Writing for Engineering and

Science Students

Writing for Engineering and Science Students is a clear and practical guide for anyone under-
taking either academic or technical writing. Drawing on the author’s extensive experi-
ence of teaching students from different fields and cultures, and designed to be accessible
to both international students and native speakers of English, this book:

• Employs analyses of hundreds of articles from engineering and science journals to

explore all the distinctive characteristics of a research paper, including organization,
length and naming of sections, and location and purpose of citations and graphics;
• Guides the student through university-level writing and beyond, covering lab reports,
research proposals, dissertations, poster presentations, industry reports, emails, and job
applications;
• Explains what to consider before and after undertaking academic or technical writing,
including focusing on differences between genres in goal, audience, and criteria for
acceptance and rewriting;
• Features tasks, hints, and tips for teachers and students at the end of each chapter, as
well as accompanying eResources offering additional exercises and answer keys.

With metaphors and anecdotes from the author’s personal experience, as well as quotes
from famous writers to make the text engaging and accessible, this book is essential reading
for all students of science and engineering who are taking a course in writing or seeking a
resource to aid their writing assignments.

Gerald Rau is Adjunct Professor in the Department of Electrical Engineering at National

Chung Cheng University, Taiwan.
Writing for Engineering and
Science Students

Staking Your Claim

Gerald Rau
First published 2020
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
 2020 Gerald Rau
The right of Gerald Rau to be identified as author of this work has been
asserted by him in accordance with sections 77 and 78 of the Copyright,
Designs and Patents Act 1988.
All rights reserved. No part of this book may be reprinted or
reproduced or utilised in any form or by any electronic, mechanical, or
other means, now known or hereafter invented, including photocopying
and recording, or in any information storage or retrieval system, without
permission in writing from the publishers.
Trademark notice: Product or corporate names may be trademarks
or registered trademarks, and are used only for identification and
explanation without intent to infringe.
British Library Cataloguing-in-Publication Data
A catalogue record for this book is available from the British Library
Library of Congress Cataloging-in-Publication Data
Names: Rau, Gerald, 1954- author.
Title: Writing for engineering and science students : staking your claim /
Gerald Rau.
Description: New York, NY : Routledge, [2019] | Includes bibliographical
references and index.
Identifiers: LCCN 2019012283| ISBN 9781138388246 (hardback) |
ISBN 9781138388253 (pbk.) | ISBN 9780429425684 (e-book)
Subjects: LCSH: Academic writing—Study and teaching. | Technical
writing—Study and teaching. | English language—Rhetoric—Study
and teaching.
Classification: LCC P301.5.A27 R38 2019 | DDC 808.06/66—dc23
LC record available at https://lccn.loc.gov/2019012283

ISBN: 978-1-138-38824-6 (hbk)

ISBN: 978-1-138-38825-3 (pbk)
ISBN: 978-0-429-42568-4 (ebk)

Typeset in Bembo
by Swales & Willis Ltd, Exeter, Devon, UK

Visit the e-resources: www.routledge.com/9781138388253

Contents

List of figures xv
List of tables xvii
Preface xix
To the student xxi
To the teacher xxii
Acknowledgments xxiv

PART 1
Getting the big picture 1

  1 General principles of writing 3

1.1 Fitting in, standing out  3
1.2 We have to start somewhere, but why research articles?  5
1.3 Academic English: a new language  6
1.4 Identifying good exemplar articles  6
1.5 Learning the fundamentals  8
1.6 Exploring different genres  8
1.7 Creating your own masterpiece  9
Evaluate your understanding  10
Homework 10

  2 Overall format of research articles 12

2.1 One format or two?  12
2.2 Research article formats: IMRD vs. IPTC  12
2.2.1 IMRD and modified IMRD (IMRaDC) formats  12
2.2.2 IPTC and modified IPTC (IPC) formats  15
2.2.3 Why a separate format?  19
2.3 Variation in research article formats  20
2.3.1 Formats in the natural sciences  22
2.3.2 Formats in engineering and mathematics  23
2.3.3 Formats in the social sciences  23
vi Contents

2.4 Determining exemplar article structure  24

Exercise 2.1 Overall structure  25
Evaluate your understanding  26
Homework 26

  3 Argument structure of research articles 29

3.1 One long argument  29
3.2 Three implicit claims  30
3.3 Ten common component claims  32
Exercise 3.1 Locating components  35
3.4 Supporting the claims  35
Exercise 3.2 Writing and supporting a claim  36
Evaluate your understanding  37
Homework 37

PART 2
Argument structure in exemplar articles 39

  4 Establishing the importance of and need

for the research 41
4.1 Joining a community  41
4.2 Claim 1: importance of and need for your work  42
4.3 Component analysis and component markers  43
4.4 Component 1: importance  43
4.5 Component 2: need  45
4.6 Component 3: research goal [main claim]  47
4.7 Component 4: framework  49
4.8 Structure of a well-written article  50
4.9 Comparison of IMRD and IPTC  51
Exercise 4.1 Introduction division components  52
Evaluate your understanding  53
Homework 53

  5 Demonstrating the continuity and

novelty of your research 56
5.1 Connected yet distinct  56
5.2 Claim 2: reliability or feasibility of your work  56
5.3 Component 4: framework  57
5.4 Component 5: research details  61
5.5 Component 6: testing methods  63
5.6 Comparison of IMRD and IPTC  66
Exercise 5.1 Methods or Process division components  67
 Contents vii

Evaluate your understanding  67

Homework 68
  6 Verifying the contribution of your research 69
6.1 Staking your claim  69
6.2 Claim 3: the contribution of your work  69
6.3 Component 6: testing methods  71
6.4 Component 7: data patterns  71
6.5 Component 8: comparisons  73
6.6 Component 9: interpretations  76
6.7 Component 10: conclusion  78
6.8 Comparison of IMRD and IPTC  80
Exercise 6.1 Results and Discussion or Testing and
Conclusion division components  80
Evaluate your understanding  81
Homework 81

  7 Evidence from past research 83

7.1 Connecting with accepted evidence  83
7.2 Evidence from past research  83
7.2.1 Categories of past research  84
7.2.2 Past research as evidence for various components  85
Exercise 7.1 Past research and components  87
7.3 Author-prominent vs. information-prominent citations  88
Exercise 7.2 Citation types  90
7.4 Citation verbs  90
Exercise 7.3 Citation verbs  93
Evaluate your understanding  93
Homework 93

  8 Evidence from current research 95

8.1 From past to present  95
Exercise 8.1 Past to present research  96
8.2 Evidence from current research  96
8.3 Examples and equations  97
Exercise 8.2 Example and equation type and location  97
8.4 Evidence from research details  97
8.5 Data in text and graphics  98
Exercise 8.3 Graphics type and location  98
8.6 Data comparison and statistics  99
8.6.1 Types of statistics  99
8.6.2 Experimental design  100
Exercise 8.4 Statistics type and location  101
viii Contents

8.7 Pointing verbs  101

Exercise 8.5 Pointing verbs  102
Evaluate your understanding  103
Homework 103

  9 Clear reasoning 105

9.1 Follow the yellow brick road  105
9.2 Logical reasoning  105
9.3 Evaluative reasoning  106
9.4 Extended vs. condensed style  107
9.4.1 Characteristics of extended and condensed style  107
9.4.2 Summary paragraphs  108
Exercise 9.1 Identifying extended and condensed styles  109
9.5 Sequencing strategies  110
9.5.1 Basic strategies  110
9.5.2 Nested or cyclical strategies  111
Exercise 9.2 Sequencing strategies  113
Evaluate your understanding  113
Homework 113

PART 3
Exploring different genres 115

10 Argument structure in other types of writing 117

10.1 Variations on a theme  117
10.2 Planning to overcome difficulties  117
10.3 Goal, audience, criteria  119
Evaluate your understanding  120
Homework 120

11 Undergraduate writing 122

11.1 So near, and yet so far  122
11.2 Lab reports  122
11.3 Essays  124
11.4 Senior project  125
11.5 Summary  125
Evaluate your understanding  126
Homework 126

12 Graduate writing 127

12.1 Apprenticeship  127
12.2 Research articles  127
 Contents ix

12.3 Research proposals  128

12.4 Conference papers  130
12.5 Industrial journal articles  131
12.6 Thesis or dissertation  132
12.7 Summary  133
Evaluate your understanding  133
Homework 133

13 Academic writing 135

13.1 Never stop learning  135
13.2 Grant proposals  135
13.3 Short papers  137
13.4 Review or survey articles  137
13.5 Book chapters  138
13.6 Popular writing  139
13.7 Summary  140
Evaluate your understanding  140
Homework 140

14 Technical writing 141

14.1 Intertextuality  141
14.2 Progress reports  141
14.3 Technical reports  142
14.4 Management or consulting reports  143
14.5 SWOT analysis  144
14.6 Summary  145
Evaluate your understanding  145
Homework 146

15 Writing about yourself and others 147

15.1 Blowing your own horn  147
15.2 Entrance essays  147
15.3 Job applications  148
15.4 Resumes and CVs  149
15.5 Reference letters  150
15.6 Summary  151
Evaluate your understanding  151
Homework 151

16 Academic and technical presentations 152

16.1 You are the expert  152
16.2 Different format, same principles  152
x Contents

16.2.1 Preparing in stages  152

16.2.2 Tell a good story  154
16.2.3 Focus  154
16.2.4 Finish strong  157
16.3 Academic oral presentations  158
16.4 Academic poster presentations  159
16.5 Technical presentations  161
Evaluate your understanding  162
Homework 162

17 Principles for successful email 163

17.1 Importance of email  163
17.2 General email etiquette  163
17.3 Position and imposition  164
17.4 Five factors affecting email success  166
17.4.1 Terms of address  166
17.4.2 Justifying your request  170
17.4.3 Amount of information  172
17.4.4 Reducing imposition  173
17.4.5 Attention to detail  174
Evaluate your understanding  175
Homework 176

PART 4
Creating your masterpiece 177

18 Writing in stages 179

18.1 The hardest lesson for beginning writers  179
18.2 Why prewrite?  179
18.3 Write as you go  180
18.4 The 7Cs of Change  182
18.5 Do it right the first time  183
Exercise 18.1 Instructions to authors  185
18.6 Avoiding plagiarism  185
Exercise 18.2 Paraphrasing practice  187
Evaluate your understanding  187
Homework 187

19 Prewriting 189
19.1 Planning your masterpiece  189
19.2 Initial sketch: argument structure outline  189
Exercise 19.1 Argument structure outline  191
 Contents xi

19.3 Adding the base layer: brainstorming  192

Exercise 19.2 Brainstorming  193
19.4 Placing the main elements: linearizing  193
Exercise 19.3 Linearizing your article  194
Evaluate your understanding  194
Homework 195

20 Writing 196
20.1 Just do it  196
20.2 Keeping track of where you are  197
20.3 Keeping track of citations  197
20.4 Keeping track of your work  198
Evaluate your understanding  199
Homework 199

21 Rewriting for organization 200

21.1 Rewriting in stages  200
21.2 Coherence  201
Exercise 21.1 Evaluating the coherence of your writing  202
21.3 Conciseness  203
Exercise 21.2 Evaluating the conciseness of your writing  204
Evaluate your understanding  204
Homework 204

22 Rewriting for clarification 206

22.1 Clear, concise, precise  206
22.2 Connection  207
22.2.1 Connection to past research  207
Exercise 22.1 Highlighting your contribution  210
22.2.2 Connection within your article  210
Exercise 22.2 Adding connectors to strengthen
your argument  215
22.3 Connotation  216
22.3.1 Making supportable claims  216
22.3.2 Proper use of statistical terms: significant  218
22.3.3 Proper use of statistical terms: causation vs. correlation  219
22.3.4 Types and uses of qualifiers  220
Exercise 22.3 Using qualifiers to make your claims
more supportable  227
Evaluate your understanding  227
Homework 228
xii Contents

23 Revision and finalization 230

23.1 Time’s up  230
23.2 Consistency  230
23.2.1 Consistency of thought  231
23.2.2 Consistency of words  231
Exercise 23.1 Consistency of words  232
23.2.3 Eliminating ambiguity  232
23.2.4 Squid technique  234
23.3 Correctness  234
23.3.1 Frequently misused words  235
23.3.2 Prepositions  238
23.3.3 Articles  239
Exercise 23.2 Articles  242
23.3.4 British vs. American usage  243
23.3.5 Spelling, punctuation, and formatting  244
23.3.6 Variation in an ELF research world  245
23.4 Collaboration  245
Evaluate your understanding  246
Homework 246

PART 5
Adding the final touches 249

24 Illustrating your manuscript 251

24.1 The purpose of graphics  251
24.2 Terms used when referring to graphics  252
24.3 Tables  253
Exercise 24.1 Evaluating tables in your exemplars  256
24.4 Figures: graphs  257
Exercise 24.2 Evaluating graphs in your exemplars  258
24.5 Figures: illustrations  258
Exercise 24.3 Evaluating illustrations in your exemplars  259
24.6 Graphics that break the rules  259
Evaluate your understanding  260
Homework 260

25 References and citations 262

25.1 When the writing is done  262
25.2 Reference and citation formats  262
Exercise 25.1 Entering references and citations  263
 Contents xiii

25.3 Reference management  263

Evaluate your understanding  265
Homework 265

26 The first shall be last 266

26.1 Attracting attention  266
26.2 Abstract  266
Exercise 26.1 Structure of exemplar abstracts  267
Exercise 26.2 Writing the final version of your abstract  269
26.3 Title  269
Exercise 26.3 Creating a title for your article  270
26.4 Keywords  270
Exercise 26.4 Deciding on keywords  270
26.5 Highlights  271
Exercise 26.5 Writing highlights  271
26.6 Acknowledgments, author information, and other bits  271
Exercise 26.6 Gathering the last bits  272
Evaluate your understanding  272
Homework 272

27 Submission and review 274

27.1 Warning: not every article is accepted (but it’s not
the end of your career)  274
27.2 Choosing the right journal  274
27.3 Submission methods  276
Exercise 27.1 Preparing for submission  277
27.4 How the review process works  277
27.5 Responding to reviewer comments  280
27.6 Borrowing from your own work  282
27.7 Closing thoughts  283
Evaluate your understanding  283
Homework 283

Appendix 1: Generalized component list 285

Component claims  285
Support: evidence  288
Support: reasoning  288

Appendix 2: Concordance, Academic Word List, and related tools 289

Definitions 289
Concordance: purpose  289
xiv Contents

Concordance: steps  289

Academic Word List  290

Appendix 3: List of supplemental material (online) 292

Teacher material  292
Student material  292
Class exercises  292

Glossary 293
Index 295
Figures

  1.1 Fitting in and standing out 4

  2.1 Overall shape and relative length of divisions in IMRD
and IPTC formats 19
  2.2 Summary of the focus of each division in IMRD and IPTC formats 21
  3.1 Implicit claims in science and engineering research articles 31
  3.2 Ten common components of science and engineering writing 32
  4.1 Common sub-components of Component 1: importance 44
  4.2 Common sub-components of Component 2: need 46
  4.3 Common sub-components of Component 3: research goal 47
  5.1 Common sub-components of Component 4: framework 58
  5.2 Common sub-components of Component 5: research details 61
  5.3 Common sub-components of Component 6: testing methods 63
  6.1 Common sub-components of Component 7: data patterns 72
  6.2 Common sub-components of Component 8: comparisons 74
  6.3 Common sub-components of Component 9: interpretations 77
  6.4 Common sub-components of Component 10: conclusion 79
  7.1 Common evidence from past research 84
  8.1 Common evidence from present research 96
  9.1 Common reasoning strategies 106
  9.2 Common sequencing strategies 110
  9.3 Nesting strategies for Materials and Methods in IMRD 111
  9.4 Nesting strategies in IMRD, IPTC, and IMRaDC,
by component and division 112
14.1 SWOT analysis chart 144
14.2 Categorical progress report 145
16.1 Stages in preparing a presentation 153
16.2 Comparison of projected images (a) as laid out in 16:9, and
(b) changed to 4:3 156
16.3 Alignment of text can (a) impede or (b) facilitate understanding 157
16.4 Possible poster designs 160
17.1 General principles of email etiquette 164
17.2 Formality required is determined by the tallest arrow 165
xvi  List of figures

17.3 Indirect speech in Chinese and English 174

18.1 English vs. Chinese punctuation and line spacing 184
22.1 Implicit claims in science and engineering research articles 217
23.1 Flowchart for deciding what article to use 240
23.2 Checklist for final proofreading 245
24.1 Figure terminology: caption (American English) or
legend (British English) 252
27.1 Flowchart of the editorial decision process 278
Tables

  2.1 Section titles and length in an IMRD article, “Effects of rainfall” 13

  2.2 Section titles and length in an IMRaDC article, “Pt-Ni nanocage” 14
  2.3 Analogous divisions in IMRaDC and IPTC formats 16
  2.4 Section titles and length in an IPTC article, “Eight-element receiver” 17
  2.5 Section titles and length in an IPTC article, “Spatial query integrity” 18
  2.6 Section titles and length in an IPC article, “On 4-ordered
3-regular graphs” 18
  2.7 Focus of different types of research 22
  3.1 Locations of support (evidence and reasoning) for implicit claims 32
  3.2 Ten common components of science and engineering
research articles 34
  3.3 Typical location of components in science and engineering
research articles 34
  3.4 Two sources of evidence in science and engineering articles 35
  3.5 Types of reasoning in science and engineering articles 36
  4.1 Components in the first section (Introduction) of “Spatial
query integrity” 50
  4.2 Components in the second section (Related Work) of “Spatial
query integrity” 51
  4.3 Introduction division of the sample article “Spatial query integrity” 51
  7.1 Relative frequency of past research as evidence for each component 88
  7.2 Citation verb frequency in different fields of science
and engineering 91
  7.3 Connotations of citation verbs common in electrical engineering 91
  8.1 Examples of different ways of referring to graphics 102
  8.2 Pointing verbs common in electrical engineering 103
  9.1 Variation in styles of scientific and engineering writing 108
11.1 Components and support in undergraduate writing 125
12.1 Components, support, and additional items in graduate writing 133
13.1 Components, support, and additional items in other
academic writing 140
14.1 Components and support in technical writing 146
xviii  List of tables

17.1 Formality and style in different types of email communication 166

17.2 Factors affecting email success or failure 166
17.3 Academic salutations suitable to different levels of formality 167
17.4 Closings, from most to least formal 168
17.5 Types of signatures, from most to least formal 168
17.6 Request formats, from most to least polite 175
18.1 The 7Cs of Change 183
18.2 Examples of plagiarism and non-plagiarism 186
19.1 Argument structure outline based on the article “Spatial
query integrity” 190
21.1 The 7Cs of Change, highlighting those covered in this chapter 201
22.1 The 7Cs of Change, highlighting those covered in this chapter 207
22.2 Connection marker types and categories 208
22.3 Types of logical connectors 212
22.4 Modals 224
23.1 The 7Cs of Change, highlighting those covered in this chapter 230
23.2 Differences in number use in British and American English 244
24.1 Table terminology: title 252
24.2 Example of a summary table showing non-numerical
categorical data 253
24.3 Example of a good numerical table 254
24.4 Sample table of operating conditions, common in engineering 259
26.1 Components in each sentence of the abstract of “Spatial
query integrity” 267
27.1 Groups of questions leading to different editorial decisions 279
Preface

Every year over 80,000 bachelor’s degrees and 50,000 graduate degrees are awarded in
engineering in the US alone, and three times that number in science. Worldwide the
number is obviously far higher. Because of increasing globalization, many of those gradu-
ates must learn to write in English.
Textbooks on scientific and technical writing frequently provide prescriptive models,
with templates for different tasks. While these make teaching and grading easy, and stu-
dents using such texts may excel in class, that does not guarantee success in the real world.
Considering the audience and goal will allow a writer to choose which common compo-
nents of academic writing need to be included in a document, and an understanding of
argument structure can guide how to place them. Thus the goal of this text is to provide
basic principles that can be applied to any writing task.
The diversity of my training and work experience has played an important role in
shaping this textbook. My doctoral degree was in natural science, but I later returned to
get a master’s degree in science education, so I saw how different writing styles are in
the natural and social sciences. While working as a science teacher I wrote lesson plans,
lab and classroom activities, various reports, SWOT analyses, and a school accreditation
self-study. When a local company was bought by an international corporation, I helped
them learn to write email and technical reports to communicate effectively with their
new headquarters in the US. My own writing experience has included journal articles,
articles for a science teacher magazine, book chapters, and a book. Moreover, my work as
a freelance academic editor for over 15 years exposed me to research journal articles from
a broad range of disciplines, from linguistics to medicine to applied mathematics, as well
as other genres such as grant proposals.
Thus when I was asked to teach academic writing I knew that the structure of articles
differed considerably across disciplines, but I believed that engineering used the same basic
IMRD (Introduction, Method, Results, Discussion) format as the natural sciences. After
all, scientists tend to think of engineering as “applied science.” When one engineering
student after another said none of their research articles followed that format, I began
my own investigation. When my analysis did not match what I had read in textbooks on
academic writing, I started to develop my own teaching material.
I have now taught academic writing for seven years, first in the learning center and cur-
rently in the electrical engineering department at National Chung Cheng University. My
students have come from many branches of engineering, including electrical, mechanical,
chemical, and computer science and information engineering, as well as various fields of
xx Preface

science and social science. Their analysis of articles in their own fields formed the basis for
the general components proposed in this textbook. Combining that knowledge with the
principles of argumentation and knowledge construction that underlie modern science
education allowed me to generalize and extend the concepts to other types of writing.
This book presents a new general framework for understanding academic and technical
writing, particularly in science and engineering, based on the premise that all such writing
is argumentation. For academic writing, I claim that although the widely accepted IMRD
model is useful in science, it is of limited value for engineering. Support for that claim will
be found in the chapters that follow. I further claim that if students learn basic principles
of argumentation, learning to think in terms of claim and support, they can become more
effective communicators in any genre, written or spoken, as supported by the testimony
of former students.
To the student

Many books on science and engineering writing follow a “one-size-fits-all” approach,

giving the impression that every field uses the same basic format, and teaching you to
write that way. Unfortunately, journal articles in engineering do not follow the textbook
format for journal articles in science, and there is a lot of variation between different
disciplines, and even subfields within a discipline. The same could be said for any other
type of writing.
Thus this book approaches writing from a genre analysis perspective, which simply
means learning to write like someone in your field by analyzing good examples of writing
in your field. It provides step-by-step instructions through various stages of the writing
process, with examples and practice exercises. The book will be most useful if you have
a competent mentor to check your work, either an instructor in a writing course or an
academic advisor (or preferably both), but can be used for self-study as well.
Since I have lived in Taiwan for most of the last 25 years and most of my students have
been native speakers of Chinese or other Asian languages, the text naturally addresses the
needs of those whose first language is not English. Additional material for those using
English as a lingua franca may be found on the publisher’s website. Nevertheless, since
no one uses academic English as their first language, any novice writer should be able to
benefit from the ideas presented here. I hope you will find the book useful.
To the teacher

There are several new ideas in this book that have never been published before, although
some will be appearing in journals and a book chapter at about the same time. The first
is the proposal of IPTC (Introduction, Process, Testing, Conclusion) as a way of nam-
ing the prototypical format for engineering research articles, based on my own analysis
of articles in engineering journals. As I will explain, the focus of an engineering article is
on developing and testing a new design to solve a problem, rather than on explanation of
data to answer a question. This is reflected in an argument structure and topic placement
that differs from the traditional IMRD format of scientific papers. I also propose the 7Cs
of Change, which evolved gradually through several years of teaching stepwise revision
of academic writing. The chapter on email likewise contains novel ideas related to well-
known sociolinguistic and pragmalinguistic principles.
Applied linguists will find that sometimes I do not follow standard terminology in
the field. Since I am not a linguist by training, the first draft of the book was written
before I realized I was in many cases reinventing the wheel. Even after reading the lin-
guistics literature, I have chosen in some cases to retain my own terminology. Because
this book is written for science and engineering students, I have tried to use terminol-
ogy they will readily understand rather than the fine distinctions teased out by linguists
over the years. In the notes you will find the standard linguistic terminology and refer-
ences to some of the seminal works, in many cases published decades before I stum-
bled on a similar concept. Where my analysis differs from previously published work,
often based on analysis of texts from engineering that do not match previous studies of
research articles in science, this is also discussed in the notes, since it is of more interest
to teachers than students.
A grant from National Chung Cheng University allowed me to hire research assis-
tants to confirm many of my preliminary findings, but there is still much to be done.
Relatively little linguistic analysis has been conducted on the structure of engineering
research articles or technical reports, so this text should by no means be considered the
final word. Rather, it is my hope that this work will stimulate further research by provid-
ing a productive framework to test and modify. Any exercise in this text, conducted on
a selection of journal articles from one field or journal, or a corpus from a certain busi-
ness or industry, would provide a suitable topic for a paper or thesis in applied linguistics.
Comments from engineers are also highly welcome, whether to correct misconceptions,
make the categories clearer or more useful, or expand the work to other areas. If you have
any suggestions on how the book could be improved, please let the publisher know and
 To the teacher  xxiii

I will very happily correct any mistakes, give better examples, and cite your research or
comments in any future editions.
This text is too long for a one-semester class. Proposed syllabi and suggestions on how
to use the text in basic and advanced classes may be found in the supplemental teaching
materials on the publisher’s website. On the website you will also find PowerPoint sum-
maries of each chapter and additional exercises and materials to assist those using English
as a lingua franca. Teaching tips in each chapter give suggestions on how to make your
class more successful, based on my experience. This book works very well with a flipped-
classroom approach, where students read the textbook at home and come to class to do
the exercises and homework. This allows you to catch problems quickly, resulting in less
frustration for the students and less correcting for you.
Finally, note that the book is primarily addressed to students, not teachers. In many
cases I will address the reader as “you,” referring to the one learning to write. It is my
intent to come alongside them, as much as is possible given the medium, acting as a
mentor and sharing anecdotes from my own experience rather than providing dry formal
pabulum that they would dread reading.
Acknowledgments

First and foremost, my appreciation goes to my wife, D. Victoria Rau, a Professor in the
Institute of Linguistics at National Chung Cheng University, who identifies herself as an
applied linguist although much of her work is in Austronesian linguistics and sociolinguis-
tics. She pointed me to the books of Swales and Feak and urged me to pursue this work.
Without her, it would never have gotten off the ground.
Students in my classes over the years have contributed to my knowledge of academic
writing in their fields and pointed out where what I said in class did not match what they
saw in their exemplar articles. From the engineers who took my first writing class (when I
still thought that engineering followed the IMRD format) to the present, their questions
and comments have caused me to question and improve my analysis and clarify both the
writing and the exercises.
Work on an early draft of this textbook was partially supported by a teaching material
development grant from the Ministry of Education, 2016 Teaching Excellence Program,
administered by National Chung Cheng University (榮獲105年度教育部獎勵大學教
學卓越計畫及國立中正大學補助). The grant allowed me to hire three graduate assis-
tants, Vithong Nguyen from the Institute of Linguistics and Wei-Young Chen and Jin
Tsung-Hsin Liu from the Department of Electrical Engineering, to confirm and refine
my preliminary analysis of various aspects of engineering articles. Jennifer Kuo, a very
capable undergraduate in linguistics at Dartmouth College, also assisted with the work
and contributed important ideas to the book and related articles that will appear. Without
them, the task would have been far more difficult.
Christine Feak provided encouragement and helpful comments on an early draft of this
work. John Blake was willing to use a draft copy of this book in his class on thesis writing
at the Japan Advanced Institute of Science and Technology and offer feedback. Yun-yin
Huang, Anita Pu, Ching-Yen Ellie Yu, and Tetyana Smotrova, writing instructors who
attended my workshop at ROCTEFL 2018, also provided important encouragement and
suggestions. Please accept my thanks.
Francesca McGowan, my editor at Taylor & Francis, was willing to support a proposal
from an unknown author, and Elizabeth Cox, her capable editorial assistant, handled
many of the details of the process. I am also indebted to the five anonymous review-
ers of the project whose comments led to its approval, as well as all the members of the
copyediting and production teams. The diagrams and flowcharts in the book were drawn
by Romer Valera.
Various works by Swales and Feak provided an introduction to genre analysis, as well
as many of the linguistic terms, as will be seen in the numerous citations. The approach
 Acknowledgments xxv

taken in this book is very similar to that of Cargill and O’Connor (2013), and I have ben-
efitted from reading their work. Although the first draft was complete before I heard of
theirs, during the revision process I have combined some insights from their work with
my own observations, and cited places where they had already published similar ideas.
Thanks to various publishers for permission to use previously published graphics,
including Oxford University Press (Table 7.2), John Wiley and Sons (Table 24.2), and
the Seismological Society of America (Table 24.3). Thanks also to Gregory T. Kleinheinz
et al. for permission to reprint the Open Access article “Effects of rainfall on E. coli con-
centrations at Door County, Wisconsin beaches” in the Sample Component Analysis,
as well as to InterVarsity Press (US and UK) for permission to quote from my previous
book, Mapping the Origins Debate, in Class Exercise 21.3 (both in the online supplemental
material).
Undoubtedly there are many other works in the field that should have been cited, ideas
that I pulled from various sources and used in my teaching before I decided to write for
publication. I have tried to track them down, but undoubtedly have missed some. Any
oversights are unintentional, and any remaining errors my own.
Many thanks go to my blind masseur at Siloam Helping Hands who saved my neck
(and back, and shoulders) multiple times along the way. Finally, special thanks to El
Shaddai and the nurses and doctors at Dalin Tzu Chi Hospital, without whom I would
not have been able to see this book to completion.

Reference
CARGILL, M. & O’CONNOR, P. 2013. Writing scientific research articles: Strategy and steps, New
York, John Wiley & Sons.
Part 1

Getting the big picture

Chapter 1

General principles of writing

1.1 Fitting in, standing out

I currently teach at a university in a rural area in southern Taiwan. There is no written
dress code, but almost everyone dresses informally. Students come to class in shorts and
tee shirts, with sneakers or flip-flops on their feet. It just seems to fit the area. Someone
wearing high heels and the latest fashion would seem out of place, perhaps a visitor from
elsewhere. They just don’t fit. But among the informally dressed students, some stand out.
They ask the right questions, work hard, and excel in class.
The same principle applies to writing in science or engineering, whether in academ-
ics or industry. For example, there may not be a written code that says what a research
article should look like in your field, but everyone who has been in it for a while knows
the unwritten rules. If you follow those rules, you will fit in. If not, they will look at you
like an outsider, and your paper is less likely to be accepted. But within those papers that
all look similar, some stand out. They ask the right questions, are well written, and wind
up being highly cited. Nevertheless, if you don’t fit in, you never get the chance to stand
out (Figure 1.1).
The problem is that when you begin to write in a new field of study, you do not know
the expectations, the unwritten code. I suggest there are three ways to learn the unwritten
expectations in your field:

(1) Gradually learn by experience;

(2) Ask someone with more experience;
(3) Purposefully study the structure of successful writing.

Many students choose the first route, often by default rather than active choice. They
finish their research and are ready to write, but then realize they have no idea how to
structure their writing. Nevertheless, they try their best, and wind up very discouraged
when their teacher or advisor says they have completely missed the mark. After a number
of attempts, they finally succeed, and after a few more attempts manage to master the
style. Nevertheless, they may not be able to describe it to others because they have never
considered what makes some writing better structurally, only in terms of content. This is
also the limitation of the second route—some professors can train their students to do the
research but have never studied or thought critically about how to write.
That brings us to the third route, and the purpose of this book. Every journal has its
own unwritten rules that define the expected structure for articles, and you must learn
4  Getting the big picture

You cannot learn the expectations

Submit
for a particular journal or situation
by reading a book that gives you a
general style of scientific writing.

Fit expectations Reject

Good content Reject

Accept

Figure 1.1  Fitting in and standing out.

what those rules are if you want your work to be accepted by that journal. Every business
has similar unwritten rules for reports in that company. You cannot learn the expecta-
tions for a particular situation by reading a book that gives you a general style for a certain
type of writing. You can do
it by reading and working
for many years and gradu-
The method taught in this book ally assimilating the style, or
allows you to write better, with you can expedite the pro-
cess by actively studying the
less revision, and easily adapt your
structure of good examples.
writing to meet the expectations In this book you will learn
of different situations. and practice how to do that.
The process initially will
seem very tedious. At first
you may wonder if it is
worthwhile, but it has two advantages. The first and most immediate is that whatever
documents you need to write as a student will require less revision and get higher marks,
 General principles of writing  5

because they meet the expectations. The second and ultimately more important is that
it will be easier to adapt your writing to different situations. After using the method to
purposefully study journal articles, you can compare any subsequent writing task with
what you already know. By noting the differences, you will be able to adjust your style
of writing accordingly. If you have learned to write in one format by trial and error, you
will have to use the same slow method for the second as well.

1.2 We have to start somewhere, but why research articles?

No matter what you want
to learn, it pays to start with
the fundamentals. Therefore, Research articles are readily
no matter what you plan to available examples of good
write first, we will begin with arguments; understanding their
an analysis of research articles. argument structure will help you
There are four good reasons
read and write better.
for this:

(1) You will learn to read more

effectively to get the infor-
mation you need to work in science or engineering;
(2) Research articles are complete arguments, thus a good example for your own writing,
but also part of an ongoing research effort;
(3) Other genres of student and technical writing are predictable modifications of research
article structure in that field;
(4) Research articles are readily available.

What the American novelist Stephen King said is no less true for academic writing than
fiction: “If you don’t have time to read, you don’t have the time (or the tools) to write.
Simple as that” (King, 2000: 147). Whether in school, academics, or industry, you will
need to read research articles to get up-to-date information on the latest advances in your
field. As we will see, research articles have a relatively predictable order of claims. Articles
include most of the components common in science and engineering writing, and certain
phrases are frequently used to mark each component. Understanding that structure and
recognizing those phrases will help you locate the information you need for your own
work more quickly and efficiently.
Although each research article stands alone as a complete argument, it is also part of the
growing body of knowledge in that field. Thus each article is one link in a long chain of
research and makes sense only in the context of that whole chain. You will need to read
many articles before you can see the overall structure of that chain and understand which
links are most important in holding the whole chain together, and which merely connect
one item dangling from that chain like a charm on a bracelet. The same is true for techni-
cal reports, email, or any other type of writing discussed in this book.1
Furthermore, as we will see in Chapters 10–17, although other genres make different
basic claims than journal articles, they include many of the same components, evidence, and
reasoning and are strongly influenced by the research article structure in that field. Thus
understanding the structure of journal articles will assist you with almost any type of writing.
6  Getting the big picture

From a practical perspective, research articles are easy to obtain. Although it is difficult
to access authentic examples of industry reports, grant applications, reference letters, and
many other genres because of confidentiality concerns, there is a virtually inexhaustible
supply of journal articles available in any field.

1.3 Academic English: a new language

If you are a graduate student, or even
an upper-level undergraduate, you have
“Is this English? Why probably read an academic journal arti-
can’t I understand it?” cle. What was your first impression?
Academic English is not Many students, even those whose first
anyone’s first language. language is English, wonder, “Is this
English? Why can’t I understand it?”
The language used in journals is very
different from the language of everyday
speech.
Linguists (those who study the structure and use of language) tell us that there are many
different genres of writing. Genre refers to a specific category of written, spoken, musical,
or artistic presentation, and each genre has its own style. Academic English is different
from everyday English, and each discipline even has its own discipline-specific English,
with its unique vocabulary and expressions. Technical writing is not the same as academic
writing in the same field. The technique we will be using in this book is based on genre
analysis. It will help you learn to both read and write academic and technical English,
particularly the discipline-specific English of your field.
Since academic English is not a first language for any of us, this book should be useful
to any beginning writer. As with learning any other new language, it will take time, but
the principles taught in this book will help you improve faster. Eventually you will learn
to not only write but also think in the patterns expected in your field.

1.4 Identifying good exemplar articles

Throughout the book you
will be examining exemplar
articles. As the name implies,
The best exemplar is one that is you can use exemplars as
well written and has a good argument examples when you do simi-
structure, not the one closest to your lar writing, so it is important
topic, most cited, or written by the to identify which documents
top researcher. will be good examples. A
good exemplar article is not
necessarily the one whose
content is the closest to your
topic, or the most cited, or
one written by the most famous researcher in the field. The most-cited papers often lay out
a new theoretical framework or methodology, and therefore may use a different format than
 General principles of writing  7

a typical research article in the field. Some of the best-known researchers in a field can get
away with poor writing because their research is respected, an advantage not shared by new
authors. Instead, look for an article that seems to be well written, more from the perspec-
tive of argument structure than
grammar.
The first thing to notice in
choosing exemplar articles is
These are other genres that may
the genre. Since our first goal provide good information for
is to analyze standard research your research but are not good
articles, do not use the following exemplars because they have a
as exemplars: different purpose and structure.
• Review or survey articles
• Overview of a standard2
• Brief reports, letters, communications
• Conference papers
• Trade publications
• Popular science publications

Reviews and surveys are usually written by experienced senior researchers. While
these may be an excellent place to start reading what has been done in your field, they
do not present any new research, and therefore lack many components. The overview of
a standard may likewise be vital to your research, but the structure and purpose are very
different, with no new testing or data. Usually the title or a text box at the top of the first
page will indicate if an article fits in any of these categories.
Brief articles and conference papers are often based on preliminary results, shorter, and
not as carefully reviewed, and they generally do not have as developed an argument struc-
ture as full articles. Brief articles may also be called letters or communications, and usually
have those words in the heading of the first page of the article, or perhaps in the name of
the journal. Any article whose reference includes Proceeding, Workshop, or Symposium
is probably a conference paper.
Trade and popular science publications
are written for people in industry or the
Recent articles generally
general public, respectively, and are usu-
ally a summary of applications of the latest make the best exemplars,
research rather than the research results of because expectations
a single group. They are easily identified change over time.
by their brighter colors and graphics.
It is also important when deciding
which articles to use as exemplars to
consider the date of publication. Recent articles are better, because both research and
writing styles change over time.
Every journal has slightly different expectations and requirements. You might expect
to find differences between, say, chemical engineering and electrical engineering, but
there are also huge differences within electrical engineering and from journal to jour-
nal within a field (Rau, In preparation). Thus, it is easier to begin by analyzing several
8  Getting the big picture

articles from one journal. After you are familiar with the process, doing it for another
journal will be easier. Eventually you will be able to quickly analyze an article’s format
based on comparison with
those you already know
and extend the knowledge
Starting with three articles by to other types of writing.
different authors from one journal In the homework for this
allows you to determine the chapter, you will identify
expectations of that journal, three exemplar articles by
different authors in your
so you can fit in. field. Why three? you may
ask. With only one or two
it is impossible to establish
what is typical of your field,
but with three it is likely that at least two will be similar, giving you a better picture of
the expectations.

1.5 Learning the fundamentals

Writing is an art. To become an artist,
a student must first learn basic elements
Learning to write is like like shape and color, then principles of
learning to paint. composition. After learning the basic
You learn a particular genre techniques, the student may be asked
to create their own work, copying
by copying successful work.
the style of a certain master. Similarly,
if you want to write like the experts
in your field, you have to study their
work and try to duplicate the style, but
with your own content. That is the fundamental principle underlying the use of exem-
plars, good examples of a genre. At the same time, as you improve, you will develop your
own unique style.
In Chapter 2 you will learn the overall format of academic writing in your field, using
research articles as a prototypical example. In Chapter 3 I will introduce the premise
of this textbook: that all writing in science and engineering is (or should be) structured
as a series of arguments, with nested claims supported by evidence from both past and
present work, linked by logical reasoning. Then, in Chapters 4–6 you will examine in
detail the claims made in different divisions of research articles. Chapters 7–9 complete
the picture, as you learn to identify the evidence and sequencing strategies used to sup-
port the claims.

1.6 Exploring different genres

In the next part of the book (Chapters 10–17), I will show how, once you have
learned to identify and emulate the format of an argument in a journal article, you
can apply the same principles to any writing. Just as it is easier to pick up a new genre
 General principles of writing  9

of art once you have mastered one, the

same is true for writing. The principles Once you master the
of using an exemplar and focusing on
argument structure can be applied to
technique, you can apply
any genre of undergraduate, graduate, the same principles to any
academic, or technical writing. They other writing.
can be extended further to job appli-
cations, oral presentations, and even
email correspondence.
As long as you focus on your audience and the underlying claims that need to be
addressed for each type of writing, you will be able to adapt what you have learned to
write successfully in any situation. Your first attempts may not be perfect, but should be
close enough that others can provide specific suggestions for improvement.

1.7 Creating your own masterpiece

One of the hardest things for new
writers to understand is the number of One of the hardest
revisions required. Some portions may things for new writers to
survive with minimal changes, but understand is how many
those involving integration of work times you will have to
from different authors or justification
of a plan of action are much harder to
rewrite your work.
write and will require extensive revi-
sion. That being the case, it does not
make sense to polish every sentence when you write it the first time, or even to write
everything in the order it will appear in the final document.
Rather than filling your paper from left to right, top to bottom, it makes sense to think
of how a painter creates a masterpiece. An experienced painter begins with a broad outline,
sketching in the overall size and shape of various components, then chooses one of the main
themes, filling in the broad brush strokes of that across the whole canvas, then another, and
finally adding all the small details. That is how I am going to encourage you to write.
In Chapter 18 I will describe in more detail the reason for writing in stages. In
Chapter 19 you will learn how to prewrite your article, brainstorming about what
should go into your document and putting those ideas in order, following your own
template derived from exemplars. In Chapter 20 I will give suggestions on keeping
track of what you have written, which is particularly essential in academic writing
where it may take over a year from the time you start until an article is published.
Although only one chapter is devoted to prewriting and one to writing, the next
three chapters (21–23) are devoted to various aspects of rewriting, highlighting its impor-
tance in creating your masterpiece. In these chapters you will learn about the 7Cs of
Change (Coherence, Conciseness, Connection, Connotation, Consistency, Correctness,
Collaboration), seven steps of rewriting that will gradually transform your writing from
very rough to a highly polished argument.
In the last part, Chapters 24–27, we will turn our attention to graphics, references,
and the title and abstract, important elements of academic writing which are normally
10  Getting the big picture

finalized after most of the writing and rewriting are finished. Once your work of art is
complete, you want it to be noticed and appreciated. To do that, you need to convince
someone to put it on display, so in the final chapter we will look at how to submit your
work and respond to reviewer comments.

Evaluate your understanding

(1) What are the advantages of the method proposed in this book?
(2) Why will we begin by analyzing the structure of research articles, no matter what you
plan to write first?
(3) What types of papers should you not use as exemplars, and why?
(4) What is the advantage of using three exemplar articles rather than only one, and three
from the same rather than different journals?

Homework
(1) Beginning with one article related to your field, find at least two other articles from
the same journal that might serve as exemplars.
Hint: Use Google Scholar or another academic search engine to find recent articles that have
cited the first.
Hint: It is preferable to use articles by three different authors, to ensure that the style is that
preferred by the field, not a single author.
(2) List the references from your three articles and check that they are all standard research
articles.
Hint: If you know how to use a reference manager (section 25.3), do so. If not, type the title
in Google Scholar. At the bottom of the listing, click on the quotation mark. Copy one of the
formats into a Word document.
Hint: Look for signs of other genres in the reference (Letters, Proceedings . . .).

Teaching tip: Show examples of each type of article and how to recognize them before asking
students to submit potential exemplars.
Teaching tip: It is essential to evaluate students’ exemplars early, to make sure all are stand-
ard research articles, not other genres. It will help students if you can tell them which of the
three articles is most prototypical or best written.
Teaching tip: Alternatively, ask students to submit five articles, and help them select three
that appear to be better exemplars, rather than finding that some are not suitable and having
to resubmit other options.
Teaching tip: Obviously, if all the students in your class are writing another genre and have
access to exemplars, you could begin with that instead of research articles, but by doing so
they do not learn how to read research articles.
Teaching tip: Sometimes authors will use the same title for a conference paper and a journal
article. This can cause confusion when downloading the reference, especially if the article is an
early access preprint version, common in engineering. Tell students to check that the reference
and header of the article match, and how to reference preprint versions.
 General principles of writing  11

Notes
1 Linguists call this intertextuality (the connection between texts). See section 14.1.
2 In engineering, new designs must conform to accepted standards. Overview articles reduce the
hundreds of pages of detail in the complete standards to tens of pages of general description.

References
KING, S. 2000. On writing: A memoir of the craft, New York, Scribner.
RAU, G. In preparation. IPTC: A prototypical format for engineering research articles.
Chapter 2

Overall format of research articles

2.1 One format or two?

Our first step in learning how to write will be learning how to read research articles. There
are several good reasons for this, explained in section 1.2. Research articles are the heart
of science and engineering, reporting the latest advances, so you need to learn to read
them. If you find them difficult to read, you are not alone. The vocabulary, grammar,
and overall structure are far more difficult than popular writing. Nevertheless, it is easier
if you understand how they are structured, and thus where to find the information you
need. After learning their structure, I will show you how to use that knowledge to write
whatever you want.
The IMRD format (Introduction, Methods, Results, and Discussion, also called
IMRaD) is a well-established tool for teaching academic writing. While this struc-
ture is easy to identify in many science articles, there are problems applying it to
many branches of engineering. In this chapter I propose an analogous format, IPTC
(Introduction, Process, Testing, and Conclusion), based on several distinctive features.
By the end of the chapter you will know how the formats differ and which is used in
your field.

2.2 Research article formats: IMRD vs. IPTC

Research articles in science and engineering exhibit obvious differences in the number,
names, and length of sections. Since the format of research articles in science has been
better studied and is well established, we will look at those first, then show how typical
engineering articles differ, followed by a justification of the need for a new designa-
tion for the prototypical format of engineering articles based on systematic differences
between science and engineering. In section 2.3 I will discuss how variation in research
methodology leads to predictable differences from the prototypical format within a field
or sub-field.

2.2.1 IMRD and modified IMRD (IMRaDC) formats

Science research articles usually have four sections and use general section titles. The
Results and Discussion comprise the longest and most important part of the article, particu-
larly if tables and figures are included. This is illustrated from two sample articles, which I
 Overall format of research articles  13

will refer to by shortened titles

rather than authors’ names, for “Section”—the numbered
ease of recognition.1
Throughout this book I will
and named portion of a
use the term “section” to refer to research article;
the actual named sections of an “Division”—a conceptual unit,
article, and “division” when refer- including one or more sections.
ring to the conceptual units of a
format. A division sometimes
contains several sections.
The section numbers, titles,
and length of each section for
an article from a microbi- In IMRD, there are usually
ology journal are shown in four sections, with general titles
Table 2.1 (Kleinheinz et al., closely related to the division names.
2009). For this article, the
section titles follow the pro-
Results and Discussion
totypical format for science are the longest divisions.
articles, matching the division
names taught in almost all
textbooks on scientific writ-
ing: Introduction, Materials and Methods, Results, and Discussion (IMRD). Subsection
titles, on the other hand, are specific to the article. Comparing the length of divisions
in this article, we find that in word count the Results division constitutes over half of
the article, although the paragraph count is only one third, because of the extensive tables.
This format may be modified, as shown in Table 2.2 for a chemical engineering arti-
cle (Peng et al., 2017). Here the Results and Discussion are combined into one section,
with a separate Conclusion. This modified IMRD format (Introduction, Materials and

Table 2.1  Section titles and length in an IMRD article, “Effects of rainfall”

Div. Section/Subsection titles Para- Word %

graphs count

I 1. Introduction 6 787 15

M 2. Materials and Methods

2.1. Water Sample Collection 2 257
2.2. Water Sample Collection After Rainfall Event 2 449 16
2.3. Sample Analysis 1 48
2.4. Statistical and Graphical Analysis 1 79

R 3. Results 9 2903 55

D 4. Discussion 6 788 15

Total 27 5311 100

14  Getting the big picture

Table 2.2  Section titles and length in an IMRaDC article, “Pt-Ni nanocage”

Div. Section/Subsection titles Para- Word %

graphs count

I 1. Introduction 4 513 11

M 2. Experimental section
2.1. Chemicals 1 69
2.2. Catalyst preparation 1 296
2.3. CCM preparation 1 118 24
2.4. Characterization 1 250
2.5. Electrochemical measurements 1 113
2.6. CCM testing 1 243

R&D 3. Results and discussion

3.1. Physical characterization 3 773 61
3.2. Ex-situ RDE characterization 2 395
3.3. CCM morphology and PEMFC performance 6 1545

C 4. Conclusions 1 160 4

Total 22 4475 100

Methods, Results and Discussion, Conclusion) we will call IMRaDC.2 In this article,
both the second and third section are divided into numbered and named subsections.
Although the name of the second section is “Experimental section,” a general title com-
mon in this journal, it clearly describes the materials and methods used, and would
also be referred to as the Materials
and Methods division.3 Over half
of the article is in the Results and
The Introduction division Discussion.
in IMRD situates the Each division has a specific pur-
current research relative pose. In IMRD (or IMRaDC) format
to previous research and the purpose of the Introduction divi-
states the research goal. sion is to demonstrate the importance
of and need for the work and to state
the research goal. This places the cur-
rent research in the context of what
has already been done. In a few fields
there may be a separate second sec-
Subdivisions are common tion for literature review within the
in the Materials and Introduction division.
Methods division, but This is followed by the Materials
specific topics vary by field and Methods division, which pro-
vides details about how the research
and research method. was done. This may be divided into
subsections to make it easier for read-
ers to find certain information, which
 Overall format of research articles  15

in natural science may include

materials purchased, material
preparation procedures, material Data (evidence) and
characterization, data collection interpretations (inference) are
procedures, and data analysis in separate sections in IMRD,
procedures (often including sta- but combined in one section in
tistics). Specific topics will vary IMRaDC.
by field and research method.
Data summaries are pro-
vided in the Results divi-
sion, often in tables or graphs,
sometimes with accompanying statistical test results. Interpretation of that data is found
in the Discussion. If the two are combined, there is almost always a separate Conclusion
to summarize the main points of the article, often by answering the research ques-
tions, if that is how the goal
was stated.
In IPTC, there are usually more
2.2.2 IPTC and modified than four sections, many with
IPTC (IPC) formats titles specific to the research.
Unlike science, engineering The second division, describing
articles usually have more than the new design, is the longest.
four sections, with several
section titles specific to the
research. Moreover, the new
design presented in the second division is the focus. Other differences will appear as we
look more closely at each division.
To highlight these and other differences that will be described later, I propose a new
format, IPTC, with division names analogous to IMRaDC, as shown in Table 2.3. To
justify why this is necessary we must examine how the analogous divisions differ. Since
neither the section number nor titles match IMRD, we will first consider how to assign
sections to divisions.
There is quite a bit of varia-
tion in the first division of IPTC The Introduction division in IPTC
format. Like IMRD, the pur-
pose of the Introduction is to
may be as short as one paragraph
establish the importance of and with no citations or several pages
need for the work and to state long with graphics. It may include
the research goal. Almost all more than one section.
engineering articles begin with
a section titled Introduction.
Some articles also have a second
section that contributes to this purpose, just as a literature review would in social science.
Other Introductions are a single section, but that section may be over a page long and contain
tables and figures, very rare in IMRD. The shortest Introductions may be only one or two
paragraphs, with no citations of previous work. Each journal tends to have Introductions
with a similar length and structure, but there is great variation between journals.
16  Getting the big picture

Table 2.3  Analogous divisions in IMRaDC and IPTC formats

IMRaDC IPTC

Introduction Introduction
Materials and Methods Product or Process
Results and Discussion Testing
Conclusion Conclusion

The second division, the

focus of the paper and
The second division in IPTC, therefore usually the long-
which is usually the longest, describes est, describes the current
the new Product or Process in one or research, whether it be
development of a new
more sections with descriptive titles Product or a new Process.
specific to the research. The analogy between these
names and Materials and
Methods should be clear,
but most IPTC articles
report the development of only one entity, as reflected in the use of the singular and
“or” rather than “and.” In natural sciences the name of the second division is often
shortened to Methods, a convention I will follow throughout this book. Similarly, I
will refer to the second division of
IPTC as Process.
The Testing division in The third division discusses the
IPTC is comparable to testing of the new solution and how it
a combined Results and compares with the current best solu-
Discussion division. It may tion or a standard. This is comparable
to a combined Results and Discussion
include several sections. division. Although many articles use
some variant of the term Experiment,
Testing is a better name because the
test rarely fits the current preferred definition of an experiment, which involves manipu-
lation of an experimental variable and measuring the effect on a response variable. This
is undoubtedly a result of lax use of the term experiment in many science textbooks and
laboratory guides to denote any sort
of data collection. It may also reflect
The one- or two- that the definition of experiment has
paragraph Conclusion changed over time (Bazerman, cited
in Swales, 1990: 113).
summarizes the
A brief Conclusion summarizes the
contribution of the contribution of the article to the field.
article to the field. This corresponds to the Conclusion
in IMRaDC or the final paragraph or
two of the Discussion in IMRD.
 Overall format of research articles  17

Now we will apply these principles to identifying the divisions of three engineering
articles, beginning with one from electrical engineering, shown in Table 2.4 (Sayginer &
Rebeiz, 2016). The section numbers and titles are written in the style used by the journal.
Section II is assigned to the Introduction because it gives background on the standard
architecture rather than details about the current work. Section III, the new method
developed and described in the article, comprises almost half of the total paper, whereas
the corresponding Method division accounted for less than 25% of the IMRD articles.
Sections IV and V describe how the new design is tested and the results of those tests,
respectively.
Next, in Table 2.5, we will look at an article from computer science and informa-
tion engineering (Hu et al., 2013). Again, there is a second section in the Introduction
that describes Related Work, a section title common in computer science. There
are three sections in the Process division, comprising more than half of the article.
The first, Preliminaries, explains the mathematical terms and procedures used in the
current research, followed by two sections describing the current research. Testing
methods and results are presented in section 6, followed by a separate Discussion, rare
in engineering articles.
Finally, another electrical engineering paper is shown in Table 2.6 (Tsai et al., 2011).
This paper differs from the previous two because the contribution of the work is a
new mathematical procedure and its proof. Since mathematical proof does not require

Table 2.4  Section titles and length in an IPTC article, “Eight-element receiver”

Div. Section/Subsection titles Para- Word %

graphs count

I I. INTRODUCTION 3 391
II. EIGHT-ELEMENT PROGRAMMABLE PPA 3 351 17
RECEIVER ARCHITECTURE

P III. TECHNOLOGY, DESIGN, AND BUILDING

BLOCK MEASUREMENTS
A. Technology 1 106
B. Reconfigurable Input Switching Network 9 402
C. Wideband Phased Array Channel 5 441 48
D. Reconfigurable Output Combining Network 4 404
E. Differential-to-Single-Ended Converter 1 140
F. Digital Beamforming Path 1 301
G. Digital and Reference Current Control 2 188
H. Layout 1 77

T IV. SYSTEM-LEVEL SIMULATIONS 6 683

V. MEASUREMENTS 1 82
4 450 33
A. S-Parameters, Linearity, and Noise
B. Isolation and Forward Coupling 3 208

C VI. CONCLUSION 1 109 3

Total 45 4333 100

18  Getting the big picture

Table 2.5  Section titles and length in an IPTC article, “Spatial query integrity”

Div. Section/Subsection titles Para- Word %

graphs count

I 1 Introduction 8 1164
19
2 Related Work 4 742
P 3 Preliminaries 8 551
3.1 Voronoi Diagrams
3.2 Signature Aggregation
4 Data Transformation 4 609
5 Authenticating Spatial Inquiries 27 4304
5.1 Query Processing at the SP 53
5.2 Signature Verification
5.3 Geometric Verification
5.3.1 kNN and Range Query Verification
5.3.2 RkNN Query Verification
5.3.3 KaNN Query Verification
5.3.4 Spatial Skyline Query Verification

T 6 Experiments 13 2219
6.1 Experimental Settings
6.2 VN-Auth and MR- and MR*-trees 26
6.3 Advanced Spatial Inquiries
7 Discussion 2 471

C 8 Conclusions 1 172 2

Total 67 10248 100

Table 2.6  Section titles and length in an IPC article, “On 4-ordered 3-regular graphs”

Div. Section/Subsection titles Para- Word %

graphs count

I 1. Introduction 6 595 13

P 2. 4-Ordered cells 17 1966

84
3. Generalized honeycomb torus 24 1819

C 4. Main result and concluding remarks 3 125 3

Total 50 4505 100

empirical testing, there is no testing division, resulting in an IPC structure. Since there is
no Testing, and only a short Introduction and Conclusion, the Process division accounts
for most of the article.
Several basic differences between IMRaDC and IPTC formats are illustrated in
Figure 2.1.4 Remember first that although the section names of the former are usually
general, section names for the middle two divisions of IPTC tend to be specific to the
 Overall format of research articles  19

Introduction Introduction

Materials and
Methods

Product or
Process
Results

Testing

Discussion
Conclusion

Figure 2.1  Overall shape and relative length of divisions in IMRD and IPTC formats.

research. Thus, whereas IMRaDC is the final structure of the paper, IPTC is a scaffold to help
you understand or build the structure.
Next, the length and shape in the figure indicate the relative length and breadth of
each division. As we saw above, whereas IMRD focuses on Results and Discussion,
in IPTC the new Product or Process frequently accounts for half the length of
the article. Furthermore, IMRD papers often have an “hourglass” configuration.5
Starting with a broad overview of the history of the field, they gradually become
more specific until they reach the focus of the current research. Then the bulk of
the paper focuses on the current research, but the Conclusion frequently broadens
again to applications of or future directions for the research. IPTC papers, although
they also focus from broad to narrow at the beginning, tend to talk less about future
extensions of the work. Again, these are general statements that may not be true
for each article. For example, if an IPTC article combines two established methods,
the Process division may be shorter, since both have been described before, and the
Testing division longer.

2.2.3 Why a separate format?

Since IPTC could be conceived as simply another modification of the well-known
IMRD structure, with more sections, more specific names, and an emphasis on the
second division, the obvious question is why a new designation is needed. Taxonomy
is always an ontological issue, generating a debate between “lumpers” and “splitters,”
whether in biological species (Rau, 2012: 125, 132) or genres of academic writing
(Swales, 2004: 63–65).
20  Getting the big picture

Unfortunately, every recent book on academic writing has presented the IMRD for-
mat from the perspective of quantitative empirical science, in which the Results section
dominates and directs the structure of the rest of the paper, with Methods relegated to a
minor status. Nothing that has been written on IMRD in the last 50 years matches the actual fea-
tures of a prototypical engineer-
ing article, to be described
in detail in the next sev-
Prototypical engineering articles do eral chapters, in which the
not match past descriptions of IMRD. Process division is the long-
Rather than trying to differentiate est and most important part
of the paper. Rather than
IMRD2, I propose a new
trying to broaden the deeply
designation, IPTC. entrenched understanding
of IMRD and then try-
ing to distinguish IMRD1
from IMRD2, it is easier
and clearer to propose a
new name for the analogous
structure, and the difference
Science asks questions and is easily understood by stu-
develops explanations based on data; dents (Rau, In preparation).
The distinction between
engineering defines problems and
the two formats can easily
designs superior solutions. be explained by considering
how the goals of science and
engineering differ. Science
asks questions (Introduction)
and seeks to develop expla-
nations (Discussion) based on an analysis of the data (Results), whereas engineering defines
problems (Introduction) and seeks to develop solutions (Product or Process) which are
compared with the current best solution (Testing), definitions based on the Next Generation
Science Standards (Rau & Antink-Meyer, 2020).
A summary of the focus of each division of the paper will serve to highlight both the
similarities and an important difference between IMRD and IPTC (Figure 2.2). Note
that the focus of the IMRD format is on the data generated and their interpretation.
Therefore, the longest and most important part of the paper is usually the Results sec-
tion, where a summary of those data is presented, or sometimes the Discussion section,
where they are explained. On the contrary, the focus in the IPTC format is on the
new design, the new Process or Product that is described, so this is the longest and most
detailed part.

2.3 Variation in research article formats

So far we have only considered the prototypical formats of science and engineering, but
even a quick look at research articles will reveal that there is substantial variation in format
within a field, often within a sub-field, and even sometimes within a single journal. In this
 Overall format of research articles  21

IMRD:
Introduction: Questions to be addressed by the data
Materials and Methods: How the data were generated
Results: Summary of the data
Discussion: Explanation of the data

IPTC:
Introduction: Problem to be solved by the design
Product or Process: Solution proposed by the design
Testing: Comparison with previous design
Conclusion: Advantage of design

Figure 2.2  Summary of the focus of each division in IMRD and IPTC formats.

section I will present a framework for understanding a likely cause of that variation. Since
this is more theoretical, if your exemplar articles fit the prototypical pattern you may skip
this section and come back to it later, after you understand the basic structure better.
The foundation of this
framework is the idea that our
traditional delineation of fields is Fields are delineated based
based on topics of study, whereas on topics studied, whereas the
the reporting format used for writ-
ing research articles is more closely
choice of reporting format is
aligned with research methodol- based on what works best for the
ogy. For example, electrical research methodology.
engineering (EE) investigates
various practical applications
of electricity and magnetism.
Prototypical EE studies design a physical product, but some develop a new algorithm,
and thus have a reporting structure almost identical to applied mathematics, while oth-
ers spend more time testing various alternatives and are similar to a typical science paper.
One important corollary of this is that although IMRD is the prototypical paradigm
for science and IPTC the prototypical paradigm for engineering, not all research in a field
will follow the prototype. For example, research in biomechanical and chemical engi-
neering and materials science frequently follows the IMRD reporting structure. On the
other hand, much of the physics literature utilizes IPTC.
Science has not always used the IMRD format, and it is by no means universal in
natural science research articles, let alone engineering or other disciplines. Even a quick
look at journals from different fields reveals at least three basic types of research, each
focusing on a different division of the paradigmatic IMRD framework, as summarized
in Table 2.7. Each type appears to be associated with a different reporting format, all of
which have previously been classified as variants of IMRD.
22  Getting the big picture

Table 2.7  Focus of different types of research

IMRD focus
Research method Research focus Reporting format

Methodological Solution Methods

Non-empirical Theoretical, mathematical, logical IPC
Design Product, process IPTC
Empirical Data Results
Quantitative Repeatable, statistical IMRD (or IMRaDC)
Qualitative Non-repeatable, situation-specific IMRaDC (or IMRD)
Analytical Interpretation Discussion
Inductive Integration, generalization IMRaDC, Recursive
Deductive Evaluation, comparison IRDC

Most previous work on genre analysis has considered only differences between various
fields of research, rather than research methodology. The framework proposed here may
help explain what were considered anomalies, such as the great difference occasionally
seen between articles within the same field or science fields such as seismology in which
many articles follow a recursive structure. Work to confirm and refine the framework,
especially for analytical research (more common in the humanities and social sciences), is
ongoing (Rau, In preparation).

2.3.1 Formats in the natural sciences

Prototypical research in the natural sciences at present uses a quantitative methodology. IMRD
is the most common reporting format in the natural sciences because it is uniquely suited to
this type of research. In quantitative studies, new data are generated, often to test a predic-
tion and thus its underlying hypothesis (see section 8.5.2). After an Introduction, the Method
of collecting the data is described, then the data are reported in the Results section and
explained in the Discussion. This is the simplest and most direct method for reporting a single
quantitative empirical study, and thus has become the standard reporting format. Some fields
use a modified structure, with a combined Results and Discussion and separate Conclusion.
However, many branches of science employ other research methods. Theoretical
research, common in many branches of physics, may be entirely non-empirical, in
which case there may be
no Results or Discussion,
only an Introduction fol-
Although the quantitative paradigm lowed by a mathematical
establishes IMRD as the prototypical proof or logical argument
format for natural science, other and a Conclusion. When
reporting formats are suitable for a new method is presented
different research methods. for gathering or analyz-
ing data in empirical fields,
the focus of the article
will obviously be on the
 Overall format of research articles  23

Methods, with a short section comparing the new method with existing methods, lead-
ing to a structure essentially identical to IPTC.
On the other end of the scale, fields such as epidemiology frequently use a qualitative
paradigm of observational research to study objects or events that cannot be controlled or
repeated. Some research uses an inductive approach, examining and trying to consolidate
the results of earlier primary research studies. This is common in environmental science,
seismology, and other fields exploring large-scale phenomena, where it is often impossible
to collect sufficient data in a single study. Since the focus is on analysis and interpreta-
tion, the traditional role of the Discussion, these fields often use the recursive reporting
structure typical of social science. Some disciplines that make extensive use of deductive
or descriptive research, such as traditional botany and taxonomy where a new type is
compared with known types, have a very different reporting structure, totally lacking a
Methods section.

2.3.2 Formats in engineering and mathematics

In engineering the goal is to find a design solution to a practical problem, so in the pro-
totypical case the focus is on how the new Product or Process was developed and what
makes it distinct. This leads to
the typical IPTC format.
However, some sub-fields Since research methods vary,
of electrical and mechanical
even within a field, some
engineering may include whole
sections in which mathemati- engineering articles are written
cal proof is used, and are thus in IPTC, others in IMRD or
largely non-empirical, with little mathematics format.
or no Testing division. On the
other hand, engineering studies
using quantitative techniques,
common in chemical engineering, will follow an IMRaDC format when reporting. Some
fields, such as civil engineering, may use either format depending on whether the research
is more oriented toward basic materials or their practical use.
Mathematics articles, apart from the Introduction, typically follow the conventions of
a non-empirical, logical deductive argument to present a new way of solving a problem or
an extension of a previous solution. As with theoretical science, since the emphasis is on
logical rather than empirical verification, there may be no Results and Discussion.

2.3.3 Formats in the social sciences

The social sciences are equally varied. Given the difficulty of collecting large data samples,
some work in social sciences is secondary research, so the research type is analytical. The
structure of these articles is similar to that found in the humanities, with extensive quoting
and comparison of different primary sources.6 The focus of the article is on analysis, either
comparison of the present study with previous studies, compilation of the results of other
studies, or other topics commonly found in the Discussion. In many cases the order within
an individual section seems to follow IMRD format. The authors will discuss literature
related to a particular method, the results that have been obtained, and their interpretation
24  Getting the big picture

of those data. The same process may

Likewise, social science be repeated recursively (cyclically) in
research uses various subsequent sections for other types or
sources of data, followed by an overall
reporting formats, final analysis.
depending on the Nevertheless, not all social sci-
research methodology. ence literature is analytical. Primary
research from a qualitative paradigm
places far more emphasis on describ-
ing the process of data collection and
interpretation. Primary research in the social sciences done from a quantitative perspective
tends to be written up in IMRD format, just as in the natural sciences. Theoretical studies
in economics and related fields may employ extensive mathematics and use a non-empirical
format like theoretical science, without a Results section. Some branches of linguistics,
such as formal semantics, employ a similar structure, given their emphasis on logical
consistency rather than empirical verification. Given the variety of reporting formats, the
technique described in Chapters 3–9 for analyzing exemplar articles in the science and
engineering literature is equally useful for the social sciences.7

2.4 Determining
exemplar
Although science articles typically article structure
use IMRD format, while engineering
In section 2.2 I introduced
uses IPTC, there are exceptions.
the prototypical formats
Exemplars will reveal the format(s) for science and engineering
used by each journal. articles, but individual fields
or articles may differ, for
reasons explained in section
2.3. In this section I will
give more details on the difference between the formats so you can decide which is used
by your exemplar articles and designate the division breaks.
First, based on the number of sections and the section titles, decide which for-
mat your article seems to follow, then assign sections to divisions. This is easy with
IMRD, since general section titles
are used by many articles. Almost
A section with a title all IPTC articles have sections enti-
like Related Work, Previous tled Introduction and Conclusion,
Work, Background, or so assigning those to the respective
Motivation is usually part of the divisions is easy, but other sections
require more explanation.
Introduction division. If there is a section immedi-
ately after the Introduction section
with a title such as Related Work,
Previous Work, Background, Motivation, or something similar, this is usually part of the
Introduction division. As we will see in the next chapter, its main purpose is to further
 Overall format of research articles  25

establish the need for the

present work. Occasionally
this will be placed at the end In IPTC, identify the Testing
of the article, as the Method division based on words
division may be in IMRD, but that indicate testing, results,
logically it is still part of the experiment, measurement,
Introduction division. simulation, or comparison.
The Testing division
addresses the claim that the
proposed solution is better
than other existing solutions,
which includes both testing the solution and comparing it with others. Section titles in
this division frequently include the words testing, results, experiment, measurement, sim-
ulation, comparison, or other words that relatively unambiguously identify the division.
Sometimes there will be more than one section, either two types of tests, a description
of tests and results, or a sepa-
rate Discussion section. Articles
where the main contribution is The Process division,
a new mathematical proof or between Introduction and
algorithm may have no Testing Testing, includes sections with
division. titles like Preliminaries, Problem
What remains between the Definition, or names specific
Introduction and the Testing
to the current research.
division is the Process division.8
This may include sections with
titles like Preliminaries or
Problem Definition that give background or vocabulary necessary to understand the pre-
sent work, in addition to titles specific to the current research. If a section includes many
sentences with “we” or “this paper,” it describes the current research rather than previous
research, the focus of the Introduction. Sometimes the Process division will contain inter-
mediate testing necessary to the development of the new design. In this case there may be
little or no Testing at the end, so the Process and Testing divisions may be combined. This
will become clearer in subsequent chapters as we look in detail at each division, so you can
change your designations later if necessary.

Exercise 2.1 Overall structure

(1) Identify the format (IMRD, IMRaDC, IPTC, or IPC) and break points between the
divisions for your first exemplar article.
(2) Create a table showing the section titles, paragraph numbers, and word count for
each division in your exemplar article, following a format similar to Tables 2.1–2.6.
Hint: List the section numbers and titles in the format used by the journal, to get used to it and
note differences between journals. If your journal does not use numbers, you will have to pay
close attention to font size and location to determine the level of the headings. If the PDF has
bookmarks, the levels can easily be seen there.
26  Getting the big picture

Hint: Word count will be useful not only to determine the relative length of each division, but
also the total length, giving you a goal for your own writing. Export the text from PDF into
Word, or copy each section into a Word document, then use the Word Count tool.
Hint: It may be hard to get an accurate word count for articles with large numbers of non-text
features like equations. Give the best estimate you can.
Hint: Should you count the words separately for each section and subsection? I did for most of
the tables in section 2.3, but not for Table 2.4 because the large number of sub-subsections was
not a common feature in that journal.Your goal is to establish the typical pattern.
(3) Note any difficulties, for example in getting an accurate word count, just as you
would any other problem with data collection.
Hint: Make sure you keep your notes from all these Exercises, as they will be helpful when you
do your own writing and rewriting.

Evaluate your understanding

(1) What is the difference between a “section” and a “division” of a research article?
(2) How do section titles differ in IMRD and IPTC?
(3) Why does the relative length of divisions differ in IMRD and IPTC?
(4) Why are there differences in reporting format within a field?

Homework
(1) Repeat Exercise 2.1 for your other two exemplar articles. Number and give titles to
your tables.
Hint: Think about how to best organize your information to make it easy to understand. The
way you collect and record the data may not be the best way to present it. This is true for all
research. For example, instead of using a table to show the total word count for each division, you
could use a graph or figure to indicate the relative proportions. What would be the advantage of
each? Which would convey the important information most clearly?
Hint: If you have more than one table or figure, how should they be numbered? Should they be,
for example, Table 1, Table 2, and Table 3, or Table 1a, 1b, and 1c, or Table 1.1, 1.2, and 1.3
(that is the format used in this book, but why)? What would make it easier for the reader to see
that they are connected? Which is more similar to your exemplar articles?
Hint:The table title should describe the contents of the table, not just list the name of the article.
For example, “Table 1a: Overall structure of exemplar article 1.”
Hint: Eliminate extra white space from your tables to make them easier to read.

Teaching tip: It may be useful to look at students’ work on one article first, before asking
them to complete the same exercises for the other two articles. Your feedback on the first will
reduce both their work and yours by minimizing errors, and gives a second chance to see if
they have understood the chapter.
 Overall format of research articles  27

Teaching tip: Have students compare their work, either in class as they do the exercises or
after completion of the homework, so they can see how much variation there is from one
journal or field to another. Also have them compare table formats, table titles, and the overall
organization of their summaries, to determine what works best and why.
Teaching tip: As students learn about their exemplar articles they will be asked to write a
description step by step. This will give them practice with writing in stages, rewriting, making
tables, and various skills that will be taught deliberately later in the book.
Teaching tip: Ideas from section 24.3 could be discussed at this point, to teach students how
to format and title a table effectively.

Notes
1 The sample articles were selected not for being prototypical, but by historical contingency. They
were submitted by my students as exemplar articles during the years I was developing the material
in this book.
2 IMRD and IMRaDC are often jointly called IMRaD.
3 In the journal Carbon, the Methods division is commonly given the section title Experimental,
Experimental Procedure, Experimental Details, or Experimental Setup and Methodology.
4 Earlier versions of Figures 2.1 and 2.2 appear in Rau and Antink-Meyer (2020).
5 The hourglass shape of the IMRD format has been noted by many authors, including Cargill and
O’Connor (2013) and Glasman-Deal (2010), but none I have seen distinguish the relative length
of the divisions.
6 This has important implications for academic writing classes. While students in the humanities
and social sciences can write a research paper summarizing other work in a format very similar to
that of a research article in the field, this is not the case in science and engineering, where such a
paper would be similar to a thesis literature review, but not to any part of an academic article.
7 I have had several social science students in my classes who have used the method presented
in this textbook for their work, although they needed additional help to identify the recursive
nature of writing in their discipline.
8 Occasionally presentation of each portion of a design is followed by testing of that portion in the
same section, leading to a combined Process and Testing division, but this seems to be rare.

References
BAZERMAN, C. 1983. Reporting the experiment: The changing account of scientific doings in
the Philosophical Transactions of the Royal Society, 1665–1800 (mimeo).
CARGILL, M. & O’CONNOR, P. 2013. Writing scientific research articles: Strategy and steps, New
York, John Wiley & Sons.
GLASMAN-DEAL, H. 2010. Science research writing for non-native speakers of English, Singapore,
World Scientific.
HU, L., KU, W.-S., BAKIRAS, S. & SHAHABI, C. 2013. Spatial query integrity with Voronoi
neighbors. IEEE Transactions on Knowledge and Data Engineering, 25, 863–876.
KLEINHEINZ, G. T., MCDERMOTT, C. M., HUGHES, S. & BROWN, A. 2009. Effects
of rainfall on E. coli concentrations at Door County, Wisconsin beaches. International Journal of
Microbiology, https://doi.org/10.1155/2009/876050.
PENG, X., ZHAO, S., OMASTA, T. J., ROLLER, J. M. & MUSTAIN, W. E. 2017. Activity
and durability of Pt-Ni nanocage electocatalysts (sic) in proton exchange membrane fuel cells.
Applied Catalysis B: Environmental, 203, 927–935.
28  Getting the big picture

RAU, G. 2012. Mapping the origins debate: Six models of the beginning of everything, Downers Grove,
IL, InterVarsity Press.
RAU, G. In preparation. IPTC: A prototypical format for engineering research articles.
RAU, G. & ANTINK-MEYER, A. 2020. Distinguishing between science, engineering and tech-
nology. In: MCCOMAS, W. & ORAMOUS, J. (eds.) Nature of science in science instruction:
Rationales and strategies, New York, Springer.
SAYGINER, M. & REBEIZ, G. M. 2016. An eight-element 2–16-GHz programmable phased
array receiver with one, two, or four simultaneous beams in SiGe BiCMOS. IEEE Transactions
on Microwave Theory Techniques, 64, 4585–4597.
SWALES, J. M. 1990. Genre analysis: English in academic and research settings, Cambridge, Cambridge
University Press.
SWALES, J. M. 2004. Research genres: Exploration and analysis, Cambridge, Cambridge University
Press.
TSAI, M., TSAI, T.-H., TAN, J. J. & HSU, L.-H. 2011. On 4-ordered 3-regular graphs.
Mathematical and Computer Modelling, 54, 1613–1619.
Chapter 3

Argument structure of research

articles

3.1 One long argument

When Charles Darwin wrote his classic, On the Origin of Species, he famously described it
as “one long argument.” In the first sentence of the final chapter (XV) he states, “As this
whole volume is one long argument, it may be convenient to the reader to have the lead-
ing facts and inferences briefly recapitulated” (Darwin, 1859).1 Actually it is a complex
series of arguments, nested one inside another, with claims, evidence, and reasoning inter-
spersed. Each chapter presents one general claim. Paragraphs generally begin with a topic
sentence that lays out the specific claim supported in the remainder of that paragraph,
unless the purpose of that paragraph is to provide general background before stating the
next claim. Part of the reason his work had such a great impact was his command of argu-
ment structure, which is essential to any scientific or engineering writing.
Broadly speaking, any research based on empirical evidence can be considered to be
scientific. In this sense both social science and engineering, along with natural science, can
be considered to be scientific research. All scientific study involves four main elements,
although the relative importance of each varies:

(1) Presuppositions;
(2) Evidence; Even though it may appear
(3) Inference; to be an objective report of
(4) Argumentation.
the data, scientific writing is in
The most important point fact an argument supporting a
for our purposes in this book particular conclusion.
is to understand that even
though it may appear to be an
objective report of the data,
scientific or engineering writing is not so much a report as it is an argument. With regard to research
articles, in a typical natural science article you are trying to demonstrate the claim that you
have collected reliable data (evidence), interpreted it in a reasonable way (inferences), and
thus have a valid explanation for a phenomenon. In an engineering article you are trying
to show that you have a better solution to a problem than any that has been presented to
date. Other genres support different claims, as we will see in Part 3 of this book.
Almost all scientific study is based on some form of empirical evidence.2 Empirical
evidence is something that can be observed by our senses, either directly or indirectly.
30  Getting the big picture

In primary research the researcher collects the evidence; in secondary research the
researcher relies on the observations of others, either contemporary or historical records.
Although philosophers of science have shown that collection of evidence and what is
accepted as evidence can be affected by the expectations of the researcher, evidence is
generally considered to be fact.
Science also involves logical inference. Inferences are interpretations of the evidence
based on reasoning. Predictions of what we ought to observe in specific situations can
be made based on general principles (deductive inference) and then tested to see if
those predictions are correct, or general conclusions made based on specific evidence
(inductive inference). Deductive inference and prediction testing are far more com-
mon in normal science activities, while inductive inference, coming up with a new
theory to explain a vast array of diverse data, can lead to major paradigm shifts. In the
absence of a clear hypothesis, abductive inference chooses the most likely of several
possible explanations.
The importance of presuppositions in science is often overlooked, but here I will only
say that science cannot be done without certain philosophical presuppositions about the
reality and orderliness of the world and our ability to understand it. Other presupposi-
tions are specific to one theoretical model. These presuppositions may affect the way we
perceive and interpret the data more than we like to admit, and incorrect presuppositions
lead to faulty conclusions (Gauch, 2003; Rau, 2012).3
Writing in science and engineering almost never mentions presuppositions, which can
cause problems in areas where the topic impinges on values, ethics, or other areas sci-
ence cannot address.4 The degree to which the writing is based on evidence or inference
depends on the field and the question studied.
No matter what, all academic and technical writing is (or should be) structured as a series of
arguments. Argument in this context does not mean disagreement, but a logical argu-
ment, where a claim is made and supported by evidence and reasoning. In the US this is
often taught as “claim, evidence, reasoning,” which is based on the classic formulation
“grounds, claim, warrant” (Toulmin, 1958), with the order of the first two terms reversed.
When analyzing your exemplar articles, it is important to remember that a research
article is not just a report, but an argument, even though in science and engineering it
may be harder to detect than in the analytical reporting style of the social sciences. Thus
it is important to look not only at the format of the article, but also the structure of the
argument. It may be hard to see the argument structure at first because there is frequently
a nested structure of claims within claims, and each claim may be stated before or after
the evidence supporting that claim. Nevertheless, recognizing what types of claims and
support to look for is an important first step.

3.2 Three implicit claims

These implicit claims Within any article, there will be sev-
direct the overall eral separate but related arguments,
argument but are each making a claim that requires
rarely stated directly. support.5 In my experience, science
and engineering articles make three
general implicit claims, as shown in
Figure 3.1.6 The claims for the two
 Argument structure of research articles  31

Implicit claims in science and engineering research articles

Science (IMRD)
1. You have identified an important unanswered question
2. You have gathered data in a valid, reliable way
3. You have a good explanation for your data

Engineering (IPTC)
1. You have identified an important unresolved problem
2. You have designed a workable solution to the problem
3. Your solution is better than existing solutions

Figure 3.1  Implicit claims in science and engineering research articles.

formats are similar, differing only because science asks questions and seeks to develop
explanations based on an analysis of the data, while engineering defines problems and
seeks to develop solutions that are better than the current best solution, as described in
section 2.3.3. These three implicit claims can be broken into ten component claims,
described in the next subsection. In the next three chapters you will look at exemplar
articles in your field to try to identify each of these claims.
The first claim, supported in the Introduction, has two parts—that the work is
important and that it has not been done. If it is unimportant or an adequate explanation
or solution is already available, the research is not needed, so the Introduction estab-
lishes why an article should be published.
The second claim is implied in the way the Methods or Process divisions are written.
Statements in the Methods about following established procedures, care in measurement,
and replication establish the implicit claim that the data were collected in a valid, reliable
way. Detailed description of the new product or process similarly supports the implicit
claim that you have designed a workable solution.
In IMRD, supporting evidence for the third claim, that you have a good explanation
of your data, is placed in the Results, but the logical argument occurs in the Discussion.7
In IPTC, the claim that your design is superior to existing designs is supported in the
Testing division.
You may have noticed that the arguments are closely related to the structure of the
article, as summarized in Table 3.1. In both formats, each claim is closely associated with
one division, although in IMRD evidence and reasoning for the third claim are split
between the Results and Discussion, respectively. Because the Results and Discussion together
support a single claim, we will treat the two of them together throughout this book. Even though
these claims are rarely stated explicitly, the fixity of the overall order of a research article
makes it clear that both scientists and engineers feel it necessary to address these issues.
32  Getting the big picture

Table 3.1  Locations of support (evidence and reasoning) for implicit claims

Implicit claim Support location

IMRD IPTC

1 I I
2 M P
3 R&D T

3.3 Ten common component claims

Within these three implicit claims are ten common component claims, which I will
henceforth refer to as components, present in most research articles (Figure 3.2). We
will introduce the components here, then expand on each in Chapters 4–6.
The first common component, Importance, is obviously related to the first part of
the first implicit claim, establishing the importance of the research, but the degree of
support necessary varies. For a new area of research, more reasoning will be necessary.
For established questions or problems in the field, citing the authors who have already
worked on it is sufficient. As a result, this tends to be less prominent in IPTC, where
each paper is likely to present one small improvement on a problem many research
groups are already addressing.
The second common component, Need, addresses the second part of the first implicit
claim in each format, that the question is unanswered or the problem unresolved. This
is often expressed as a limitation of previous studies, following a description of what has
already been done. Detailed support for the first and second components may be placed
in a second section entitled Related Work, Background, or Motivation.

Common Components
Science and engineering
1. Importance research articles
2. Need frequently (not always)
3. Research goal contain these components,
4. Framework
although the emphasis
5. Research details
6. Testing methods and location differ,
7. Data patterns with variations such as cycling.
8. Comparisons
9. Interpretations
10. Conclusion

Figure 3.2  Ten common components of science and engineering writing.

 Argument structure of research articles  33

It may seem odd to consider the third component, Research goal, as a claim, when it is
often stated as either questions or aims. Nevertheless, it is an essential component of any
research article, and makes an implied claim that the questions are answered in the paper
or the aims achieved.
Framework involves justification of the research methodology chosen. Even if not stated
explicitly, the implication is that the researchers have considered different ways of address-
ing the question or solving the problem and believe the model, theory, or framework
used to be the best for the purpose.
The fifth component, Research details, likewise should not merely be a list of what was
done but should be written in a way that addresses the second implicit claim, that the
data are valid and reliable or the solution workable. The care shown by writing a detailed
description is taken as evidence of similar attention to detail during the research. The same
is true of the sixth component, Testing methods.
Raw data are of little use in research until a Data pattern can be described, or exceptions
to a general pattern noticed. Thus in the Results division the focus is not on reporting the
data itself, but patterns in the data, which are often easier to see graphically, explaining
the abundance of figures and tables in this division. The claim that there is a pattern is often
confirmed by statistics, triangulation with other data sources, or another method.
Many different types of Comparisons are used in science and engineering research,
including comparisons with predictions, theoretical ideals, or previous results. These are
more obviously claims that, along with Interpretation of the data, must be supported by
both evidence and reasoning.
The final component, Conclusion, is usually stated in the last one or two paragraphs,
whether that is the end of the Discussion or a separate Conclusion, either as answers to
the research questions or a statement that the aims have been accomplished. Support for
this claim is spread throughout the rest of the document.
A summary of the components and
differences between IMRD and IPTC
where important is shown in Table 3.2.8
More detailed descriptions of each will The order of components
be provided in subsequent chapters and is similar in IMRD and
are summarized in the generalized com- IPTC, but the division
ponent list in Appendix 1. structure differs.
Although all ten components are
common in both IMRD and IPTC, and
tend to follow the same order, they are
not necessarily found in the same divisions, as shown in Table 3.3. This is largely due to
the differences between the three implicit claims in the two formats.
One difference between IMRD and IPTC is the location of the research frame-
work, the model or theoretical framework on which the current research is built. Since
there are so many unanswered questions in science, there is a large available research
space. It is easier to show the importance and need for your research if you can tie your
work to an established research tradition, thus the research framework is often part of
the Introduction, frequently implied based on the citation and evaluation of previous
research. In engineering the research space is much more crowded, with many people
trying to find a better solution to the same problem. Thus it is more important in the
34  Getting the big picture

Table 3.2  Ten common components of science and engineering research articles

Component Sub-components common in science and engineering

1 Importance This research is important to society

This topic is important to researchers in the field
2 Need A gap exists in current knowledge or understanding (IMRD)
The current best solution is limited or less than ideal (IPTC)
3 Research goal We answer a question or improve understanding (IMRD)
We propose a better solution to a problem (IPTC)
4 Framework The research is based on an accepted model or framework
This is the best model or method to follow for our research
5 Research details Care was taken to ensure good results (IMRD)
A workable solution was developed (IPTC)
6 Testing methods We can predict results based on the model
Testing followed verifiable procedures
7 Data patterns A pattern can be discerned in the data
Exceptions to the pattern can be identified
8 Comparisons Data [support/question] previous work
Data [conform to/differ from] expectations
9 Interpretations Data are best interpreted in a certain way
There is a reasonable cause for the data
10 Conclusion The question has been answered or the aim achieved
The solution is an improvement on the current best design

Table 3.3  Typical location of components in science and engineering research articles

IMRD (Science) IPTC (Engineering)

Section Components Division Components

Introduction Importance Introduction Importance

Need for work Need for work
Research goal Research goal
(Framework)* Product or Process Framework
Research details
Materials and Research details Testing Testing methods
Methods Testing methods Data patterns
Comparisons
Interpretations
Results Data patterns
Discussion Comparisons
Interpretations
Conclusion Conclusion Conclusion

*Often implied by the author’s attitude toward previous work cited; see section 5.3

Introduction to show the limitations of previous work—the problem that remains—

whereas the framework for the present work is frequently discussed at the beginning of
the Process division as background for the new modification.
The other main difference is in the location of the Testing method. In IMRD the
Testing method is usually more closely related to the validity of the data (claim 2) than
 Argument structure of research articles  35

the interpretation of the data (claim 3). On the other hand, in IPTC the Testing method
is associated with demonstrating superiority over existing solutions (claim 3), more than
proving the solution is workable (claim 2).

Exercise 3.1 Locating components

(1) Try to identify the main component or components in each section of your exemplar
articles.
Hint: Use Table 3.3 as a guide, but remember that your article may differ.

Teaching tip: Argument structure of each division will be discussed further in Chapters 4–6.
At this point students should get the idea that articles are written as a series of arguments and
which components tend to appear in each division.

3.4 Supporting the claims

Supporting a claim requires evidence. This
has not changed since the English philoso- Each claim must be
pher William Kingdon Clifford said, “To
supported by evidence,
sum up: it is wrong always, everywhere,
and for anyone, to believe anything upon either from past or present
insufficient evidence” (Clifford, 1877). research, and reasoning.
Thus it is essential to support every claim
with sufficient evidence, either from past
or present research. Support also requires
reasoning. Reasoning involves inference and evaluation, but also employs various rhetorical
devices. The ideas introduced here will be expanded in Chapters 7–9.
As shown in Table 3.4, there are two basic sources of evidence in scientific writ-
ing: what is already known and new evidence from the present research. Anything well
known in a field, affirmed by many previous authors, can be used as evidence without
citation. Specific research findings presented by a single author can be used as evidence
by citing the original publication. Publication implies that the results have already been
accepted by at least some portion of an academic community.
Past research is generally used to support claims at the beginning of the paper:
Importance, Need, and Framework. Research details and Testing methods usually build
on previous research as well. Commonly used methods do not require a citation, but
methods used by only a few previous researchers do.

Table 3.4  Two sources of evidence in science and engineering articles

Evidence Examples

1 Accepted (past) General knowledge, definitions (no citation)

Theoretical basis, specific research findings (citation)
2 New (present) Detailed description of current research procedures
Data (text, graphics, examples, statistics)
36  Getting the big picture

Details of the present design and testing procedures, often including illustrations in
IPTC, are necessary to support the reliability of the data in the present work and thus also
constitute evidence. Data patterns show evidence collected during the present research
and used thereafter to support the claims. Past research may be brought in again when
doing the Comparisons or Interpretation.
There are two basic types of reasoning used in scientific writing: logical and evalua-
tive, as shown in Table 3.5. Reasoning also involves ensuring the organization is clear
to the reader.
As explained in section 3.1, all scientific work involves at least one type of logical
inference, with deductive being the most common. Science also involves interpretation
of evidence, new and old. Organizational strategies can help a reader follow the author’s
reasoning. Papers in which the reasoning is clear are much easier to read than those in
which it is not. Thus the instructions for rewriting in Chapters 21–23 focus on how to
improve the argument structure and make the reasoning clearer to your audience.
In the next Part you will look at both the claims and support in more detail. In
Chapter 4 you will learn how to identify components in the Introduction division.
Chapter 5 will explore the second main division, either Methods (IMRD) or Process (IPTC),
which are very different in the two formats, followed by a chapter on the last two divisions,
which report the data summary and analysis. Chapters 7 and 8 show how evidence from
past and present research, respectively, is used to support each component claim, before con-
cluding Part 2 with a discussion of reasoning and sequencing strategies in Chapter 9.

Exercise 3.2 Writing and supporting a claim

(1) Practice writing and supporting a simple claim. Based on your homework for Chapter
2, claim which format your first exemplar follows, and the support for that. If there
are any unusual features, note them as well.
Hint: For example (based on the article in Table 2.5), you might write, “This article is written
in IPTC format. As shown in Table 1a, there are eight sections, two of which (4 and 5) have
titles specific to the research. More than half of the article is in the Process division. It has sections
named Related Work and Preliminaries, which are found in almost every paper in this journal.
The article is unusual because it has a separate Discussion section in the Testing division.”

Teaching tip: This is support from present evidence.

Table 3.5  Types of reasoning in science and engineering articles

Reasoning Examples

1 Logical Deductive (general theory tested by specific observations)

Inductive (specific observations to general conclusion)
Abductive (most likely of several possible explanations)
2 Evaluative Evaluation of previous research findings
Interpretation and explanation of current data
3 Organization Introductory or concluding summaries
Referring to prior or following arguments
 Argument structure of research articles  37

Evaluate your understanding

(1) Why is it important to understand the argument structure of research articles?
(2) What implicit claims are commonly made in science and engineering articles?
(3) What components are found in most science and engineering writing?
(4) Why are some components found in different divisions in IMRD and IPTC?
(5) How are claims supported in research articles?

Homework
(1) Repeat Exercise 3.2, combining information from all three exemplar articles into a single
summary of the general structure in your field. Do they follow IMRD or IPTC format?
How many divisions are there, and how are they named? Which division is the longest?
Refer to each table or figure as support. Note any unusual features or difficulties.
Hint:Would it be better to write this summary in one paragraph or one for each article? Which
should be placed first, the written summary or the tables, or should the two be interspersed?
What would make it easier to read?

Teaching tip: Emphasize that students should combine information from all three arti-
cles into one general description. Peer review helps students consider different formats
for the summary.

Notes
1 This sentence mentions all the elements of science except presuppositions: facts (evidence), infer-
ence, and argument.
2 The exception being theoretical work, which may be based on mathematical proof and precede
empirical confirmation, as in the case of Einstein’s laws of relativity.
3 This is closely related to what has been called motivated cognition or motivated reasoning.
4 In fact, this is quite common, e.g. artificial intelligence (AI), environmental issues, genetic engi-
neering, etc.
5 This is a broader use of claim than the “knowledge claim” discussed in linguistics literature (e.g.
Myers, 1992), which is similar to what is in this book called the first “implicit claim” and closely
associated with the “research goal.”
6 Ironically, I am making this claim, but can only back it up with reasoning and anecdotal evi-
dence. That is why I need to soften the claim by adding, “In my experience.” Whether papers
in science and engineering actually make these claims has not been documented in the applied
linguistics literature, but would make a wonderful topic, to confirm or disprove my anecdotal
observations by a systematic study.
7 Thus neither the Results (evidence) nor the Discussion (logic) by itself constitutes a complete
argument. The Discussion must frequently refer back to the Results, as well as to data or analysis
from other authors. This is why it is difficult both to describe what should go in the Discussion
and to write it.
8 Earlier versions of Tables 3.2–3.3 appear in Rau and Antink-Meyer (2020).

References
CLIFFORD, W. K. 1877. The ethics of belief, Contemporary Review, Vol. 29 (December 1876–
March 1877): 289–309. In: VELASQUEZ, M. (ed.) Philosophy: A text with readings (2017), 13
ed., Boston, MA, Cengage Learning.
38  Getting the big picture

DARWIN, C. 1859. The origin of species, Mentor Edition.

GAUCH, H. G. J. 2003. Scientific method in practice, Cambridge, Cambridge University Press.
MYERS, G. 1992. ‘In this paper we report . . .’: Speech acts and scientific facts. Journal of Pragmatics,
17, 295–313.
RAU, G. 2012. Mapping the origins debate: Six models of the beginning of everything, Downers Grove,
IL, InterVarsity Press.
RAU, G. & ANTINK-MEYER, A. 2020. Distinguishing between science, engineering and
technology. In: MCCOMAS, W. & ORAMOUS, J. (eds.) Nature of science in science instruction:
Rationales and strategies, New York, Springer.
TOULMIN, S. 1958. The uses of argument, Cambridge, Cambridge University Press.
Part 2

Argument structure in exemplar

articles
Chapter 4

Establishing the importance of and

need for the research

4.1 Joining a community

In this and the following chapters you will examine your exemplar articles to determine
the unstated expectations for an article in your field. I will give guidelines about things
that are commonly found in science and engineering research articles, and the way such
articles are typically structured, but will also show you how to decide whether your articles
actually follow that structure or some other.
The first step in building a house is to lay a solid foundation. Although the founda-
tion determines the shape of the building, it may be hard to see when the structure is
complete. In a new community many houses may be constructed on a similar basic plan
within a short period of time. In an older community some new houses built on empty
lots may imitate the older style while others choose the latest style, but sometimes the
community sets standards limiting the type of construction allowed.
Analogously, all research is built on an underlying framework. It influences every
aspect of the research, from the planning through the final writeup, but may be hard to
see in the published article. In science it determines what hypotheses are tested, what
type of evidence is sought, what methods are used to obtain it, and what types of inter-
pretations are expected. In engineering the problem identification, design methods, and
standards are largely defined by the framework chosen. Without a solid foundation the
research, like a house, will be weak and unlikely to stand long.
Over time a community grows up around each research framework. In new areas of
research much of the work will follow a few basic designs, but in older areas of research
there may be a mix of old and new frameworks. As with houses, in research there are
differences of opinion on the merit
of different frameworks, and poten-
tially animosity if the presuppositions A researcher must
differ. Thus the framework not only
1 demonstrate understanding of
guides the research, but also estab- the history and terminology
lishes a connection with an academic of a research community
community. before adding to it.
In the Introduction a researcher
begins to stake a claim in a commu-
nity. Just like a physical community,
each academic community has a common language and history, and in some cases limits on
how much variation is acceptable. Thus, to join and contribute to a community, researchers
42  Argument structure in exemplar articles

must show that they can use the terminology, are familiar with the past research, and fit the
expectations of that community.
Each component in the Introduction contributes to this goal. In the following sections
I will show you how to identify each component present in the Introduction of your
exemplar articles, but first we will look at which components are likely to be present and
the reason for inclusion of each.

4.2 Claim 1: importance of and need for your work

Each paragraph in a research article should contribute to the implicit general claim being
made in that division. Recall that the purpose of the Introduction is to show that:

1 You have identified an important unanswered question (IMRD)

or

1 You have identified an important unresolved problem (IPTC).

The current research must

add something valued by the
The components Importance (1), research community. This is
Need (2), and Research Goal (3) are why the first three components,
Importance (1), Need (2), and
found in almost every Introduction,
Research Goal (3), are nearly
from which the Framework (4) may ubiquitous in the Introduction
be inferred. to any research article, as many
previous works on academic
writing have noted, using dif-
ferent names.2 By reading the
Introduction, experienced researchers in a field will be able to discern the Framework (4),
even if it is not stated explicitly, just as the foundation of a house may be largely underground
but can be inferred from what is visible.
Most science and engineering research articles begin with a statement of the impor-
tance of the research, either to society or the field or both. In science several paragraphs
may be devoted to this component, with extensive description of past research, while in
engineering a single sentence may suffice, for reasons discussed below.
After establishing the motivation for the work, the author must show that no adequate
answer or solution has been proposed, so there is a need for the present work. To do
this, the author must demonstrate awareness of past research in the field. Limitations of
previous research point to a gap in the research that the present research goal can fill. This
component tends to be more prominent in engineering.
Every article must clearly state the research goal, which indicates the direction the
authors will take to provide a new answer or solution. As we will see in section 4.6,
this goal directs not only the remainder of the article, but also the preceding descrip-
tions of importance and need. By stating that goal, you indicate that you have identified
an important unsolved issue. Thus the first three components are found in almost all
Introductions to academic journals, as summarized in the implicit claim.
 Importance of and need for the research  43

The framework can be identified by what past research is cited and the author’s atti-
tude toward it. Sometimes it may be implied simply by where it is published, if the journal
has a very limited scope. If a journal has a broader audience it may be necessary to specify
the framework in either the first or second division.

4.3 Component analysis and component markers

Each component should be
evident in the paragraph
structure. Every paragraph in In English, each paragraph
English should have a single should have a single purpose
purpose and discuss a single that clearly contributes to
idea. It is considered good the argument. This is often stated
writing style in English for
in a topic sentence.
each paragraph to have a topic
sentence which summarizes
that main idea, usually the first
sentence of the paragraph, to
make it easier for readers to follow the argument. Thus in the following analysis we will
attempt to determine the purpose of each paragraph.3
It is important to note that paragraphs in English tend to be shorter than paragraphs
in Asian writing.4 For example, one sentence in Chinese academic writing may be
about the same length as an English paragraph, while a Chinese paragraph can be
as long as an entire section
in English. If an author does
not follow the expectations Recognizing common
for English writing, some
paragraphs may include more
component markers can help
than one component. you both read and write research
For each component I will articles more quickly
point out component markers and effectively.
that frequently occur in sci-
ence and engineering articles.5
Component markers are phrases or
groups of words used by authors to draw attention to key components of an article. Learning to
recognize these markers can help you both read and write more effectively.
Component markers help you skim an article to look for certain information or iden-
tify the focus of a paragraph. They are often found in topic sentences, making them easier
to find, but may also occur later in a paragraph.
As you write, you can add these markers to the specifics of your research.6 This helps
you write effective sentences related to each main component. Furthermore, it helps your
readers understand your article better by highlighting essential information.

4.4 Component 1: importance

Most research articles begin with a statement of importance, either a broad statement of
importance to society or the importance of the question to researchers, or both. These are
44  Argument structure in exemplar articles

1) Importance
a. General importance
Claim: This research is important to society
Support: Often general knowledge (to those in the field)
b. Specific importance
Claim: This topic is important to researchers in the field
Support: Past research (citations)

Figure 4.1  Common sub-components of Component 1: importance.

shown as sub-components in
Figure 4.1, with generalized
The more related research an statements of each subclaim
author can cite, the easier it is to and the type of support most
establish the importance of the frequently associated with it.
topic to the academic community Support for this compo-
nent comes primarily from
represented by the journal. past research. The easiest
way to support importance
to the field is to cite related
research articles. If many
researchers are working on a topic, that alone establishes its importance. For a new or
marginalized topic, the author will need to present stronger reasoning. If something is
well known or unquestioned in the field, such as general knowledge about how the
research addresses a soci-
etal need, citations may not
be necessary. We will talk
In engineering, importance may more about how to support
be assumed for recognized claims in Chapters 7–9.
problems in the field, so there In science articles, impor-
may be only one sentence tance often constitutes a large
part of the Introduction.
on it, or none. Since engineering seeks to
solve a recognized prob-
lem, less time is spent on this
component—rarely more
than one paragraph and frequently only one or two sentences. Brief articles announcing the
latest news on hot topics at the forefront of research or those in specialized journals may
not discuss importance at all. This is another difference between IMRD and IPTC formats.
The following examples show how component markers can help identify this com-
ponent.7 General importance can be shown by phrases indicating the ability to help solve
a recognized problem, the value to society, or frequency of use, as indicated by the first
three examples, respectively. Importance to the research field can be demonstrated by the
 Importance of and need for the research  45

amount of related work, either past or present, as illustrated by the last two. In engineer-
ing these may be followed by a citation with a long list of related articles, but in science
it is common to mention the contribution of each article separately, increasing the length
of this component.

• The rising incidence of . . . has emerged as a great concern . . .

• . . . is of particular value for emerging mobile applications.
• . . . is widely used in . . .
• . . . has been actively studied for decades.
• . . . is increasingly recognized as an important research area for current and future
devices.

4.5 Component 2: need

The unclaimed space in your
field that your research will
fill is sometimes called the To demonstrate need, the
research gap (Swales, 1990). author must show that the
The term research gap almost research is related to but different
never appears in an article, but from previous work—that it
is nonetheless a useful descrip- fills a research gap.
tive term. In English writ-
ing it is important to specify
both the research gap and the
research goal—what is missing
from previous work and how you propose to fill that gap. The term gap highlights the
idea that your research cannot be identical to what has been done before, but must be
related to it, filling a gap. Showing there is a gap is the purpose of the second component,
need for the work.
The importance of the gap can perhaps be best understood by means of illustration.
For six years after we moved to the university where I now teach, Google Maps satellite
view showed an empty field on the street across from the main gate of the university. I
knew it was not empty because I had eaten at several restaurants in that block, which had
been developed into a row of shops. From the outdated satellite view it seemed it would
be possible to build something on that block, but locals knew otherwise. That space was
already taken. The same thing can happen in research. From a distance, a newcomer
might think there is unclaimed space in a research field, but those in the area may know
otherwise.
In science it is common to identify the gap in previous research, but in engineering
this is frequently replaced by limitations (2b), sometimes called problem identifica-
tion, as shown in Figure 4.2. The distinction is important, because a research gap
implies that the particular question posed as the research goal has never been answered,
although there may be research on closely related questions. Limitations imply that
work has been done on that problem, but the solution is still insufficient. In science
a similar situation could exist if the author seeks to show that previous answers are
incorrect or inconclusive.
46  Argument structure in exemplar articles

2) Need
a. Research gap (IMRD)
Claim: There is a gap in current knowledge or understanding
Support: Absence of past research (nothing has been done)
b. Limitations (IPTC)
Claim: The current best solution is limited or less than ideal
Support: Past research (present best solution, limitations)
c. Benefit
Claim: This will be beneficial to society or the field
Support: General knowledge or past research

Figure 4.2  Common sub-components of Component 2: need.

In some fields, particularly those with multiple approaches to similar questions, there
may be a second section in the Introduction division devoted primarily to need. This is
standard practice in computer science and related fields, where the section is frequently
called Background or Related Work.8 Previous approaches are described, along with
their limitations. The last paragraph of the section may restate the research goal and how
it addresses the need.
It is also possible to claim there is a need based on the benefit of the research (2c). This
can be done in both science and engineering. This subcomponent is closely related to
the importance of the research and is used where there is an immediate practical benefit
to society.
As with Component 1, the main support for Component 2 comes from past research.
With a research gap, the absence of past research shows its existence, while citations of
similar research show the importance of filling it and thus are more closely associated with
the first component. On the contrary, citations in engineering more commonly show the
limitations of past research and there-
fore are directly associated with the
In engineering, sentences second component.9
indicating limitations are Component markers can also help
often found at the end of a identify the need for the current
long paragraph describing work. Current approaches are very
field-specific but are generally easy
previous research. to find, as they are usually accom-
panied by a citation. Sometimes the
need is identified as a gap in the
research that has not yet been investigated, as in the first three examples below. This
is more common in science than engineering. In engineering it is more common to
note a limitation, either by means of a contrast (however, although, nevertheless) or
terms that indicate disadvantages of current solutions, as in the other examples. These
markers often occur toward the end of a paragraph describing previous research, making them
harder to find.
 Importance of and need for the research  47

• So far, nothing has been done . . .

• What has been missing until now . . .
• . . . have not been characterized to date.
• However, a key problem with most algorithms . . .
• Nevertheless, both solutions incur . . .
• These drawbacks . . .
• Although this . . . has favorable properties, it does have some disadvantages, such as . . .
• The two primary concerns are . . .
• The standard . . . are inadequate . . .
• . . . limit the applicability of these methods to broader applications.
• . . . require a series of complicated steps and special instruments. They are high cost,
labor intensive, and time consuming.

4.6 Component 3: research goal [main claim]

Although listed as the third
component based on its typical
location in a research article, The research goal is often found
the research goal is the overall in the last paragraph of the
claim of the whole article. It
is so vital to both editors and
Introduction, or the previous
potential readers that simi- paragraph if the last describes the
lar statements appear at least organization of the article.
three times. The main claim is
usually found near the end of
the Introduction section (not
division, if there are two sections). It also appears in the Abstract, while the Conclusion
(Component 10) says that the goal has been accomplished.
Recall that science usually asks research questions and seeks explanations, whereas
engineering defines a problem and seeks a solution. Thus science may phrase the goal as
questions while engineering speaks of aims, but many different terms are used. In engi-
neering it is also common to state the proposed solution or the contribution directly, as
shown in Figure 4.3. In this case the wording of the research goal and Conclusion may
be very similar. Note that, unlike the other components, no support is listed for the research
goal, because all the other components support this main claim.

3) Research goal [main claim]

a. Specific topic (goal, aim, objective, focus, question) (IMRD or IPTC)
Claim: We add to previous research in this area
b. Proposed solution (IPTC)
Claim: We provide a better solution to this problem
c. Contribution (IPTC)
Claim: We make an important contribution in this paper

Figure 4.3  Common sub-components of Component 3: research goal.

48  Argument structure in exemplar articles

If stated as an objective, the research goal may be a single sentence. For example, in
the IMRD article, “Effects of rainfall” (Kleinheinz et al., 2009), the last sentence of the
Introduction says (emphasis added):

• “The overall objective of this project was to determine what impact rainfall had on
the E. coli concentrations at the selected beaches.”

In IPTC format, the objective is frequently followed by a brief summary of the proposed
solution, and perhaps a statement that as far as the authors know, no one has yet published
that proposed solution. All three of these elements appear in the penultimate paragraph of
the Introduction to the article “Spatial query integrity” (Hu et al., 2013) as shown by the
underlined words:

• “This paper subsumes our earlier

The research goal may work in [10] by extending . . .
be stated in one sentence, to handle more advanced spatial
one sentence with queries, . . .
numbered or bullet points, • Specifically, we propose . . .
or several sentences. • To the best of our knowledge,
this is the first paper that tackles
the query verification problem
for these types of queries.”

Many papers clearly and succinctly state the research goal, using easily recognizable
component markers. With others a problem definition may be stated as part of the larger
summary of the current state of knowledge in the field, followed by a proposed solution.
Nevertheless, in almost all cases these statements occur near the end of the Introduction
section and there is some sort of marker to indicate the transition from past to present
research, usually indicated by “we,” “our,” or “this paper,” as shown by the underlined
phrases below. Sometimes multiple goals are made more conspicuous by using numbered
or bullet points.

• This paper tests the hypothesis . . .

• This paper seeks to answer the following questions: . . . [three numbered questions]
• This paper proposes . . .
• Toward this end, this paper presents a methodology . . .
• In this paper we discuss the implementation of . . .
• In this paper, we study the following two goals . . . [two goals listed in sentences]
• In contrast to prior studies . . . here we investigate . . .
• We seek to answer the question . . .
• Here we investigate this issue using calculations . . . We in fact find . . . and provide
a plausible explanation for . . .
• Here we examined three specific predictions . . . [three predictions listed in sentences]
• We intend to demonstrate . . .
• In addition, we extended the technique to estimate . . .
• Our goal is to recognize . . . Our system . . . manipulates . . .
• The contributions in this paper are threefold. [three sentences]
 Importance of and need for the research  49

• The unique contributions made in this paper are as follows. [four numbered points]
• Our major contributions in this paper thus include: [three bullet points]
• We designed . . .
• We fabricated and tested . . .
• To our knowledge, this is the first . . .

In a well-written article the

research goal will be easy to find,
often in a topic sentence at the A research goal hidden in the
beginning of a paragraph, but unfor- middle of a paragraph is harder
tunately not all research articles are to find, and occasionally is not
well written. Sometimes it is hid- marked by these expected
den in the middle of a paragraph, component markers.
following a statement of the limi-
tations of other research, particu-
larly in Asian writing. If you cannot
find it, try to identify the goal in
the Title, Abstract, or Conclusion,
then look for a similar idea in the
The research goal states the
Introduction. In some IPTC papers
it will appear at the beginning of goal of the article, which may
the Process division. Occasionally it be different from the goal a
will be restated at the beginning of researcher had in mind when
the third division, for reasons dis- beginning the research.
cussed in section 6.2.
If you write up your own
research it is important to remem-
ber that the research goal is the goal of that article, not the overall or original goal of your
research. When pursuing research, you will have a goal in mind. When you write it up,
you may find that goal has shifted slightly. Rather than redoing the research or rewriting
the paper, it is more reasonable to rewrite the research goal. We will return to this idea in
Chapter 23 when discussing the final stage of revision.

4.7 Component 4: framework

As noted above, research is almost always related to an established framework.10 If a
certain framework guides most of the current research in a field or journal, it may not
be specified in the article. If there
are multiple ways of approaching a
question, or if a researcher uses a The framework may be
non-standard approach, it is much found in either the first
more important to explicitly state or second division, and
the framework.
In science the framework often is often inferred rather
can be inferred from the author’s than stated directly.
attitude toward past research in
the Introduction and the choice
50  Argument structure in exemplar articles

of Methods. In engineering the framework may be the last method mentioned in the
Introduction, with the fewest limitations, then discussed more extensively as the founda-
tion of the research in the Process division. Since in both formats the framework may be
identified from both the first and second division, we will leave a complete explanation
and analysis of it until the next chapter.

Teaching tip: Covering Component 4 with the second division balances the work better and is less
confusing for students, since it is logically connected with the second implicit claim.

4.8 Structure of a well-written article

Before you analyze your exemplar articles, let’s look at an example of a well-written
engineering article, “Spatial query integrity” (Hu et al., 2013). The first sentence of
each paragraph is a topic sentence that identifies the purpose of that paragraph. When
you do your own writing, I encourage you to use topic sentences, as this makes it far
easier for you to frame your argument and for your readers to find the information
they need. In Table 4.1 the main component of each paragraph in the Introduction
has been identified.
Note that the analysis is not simply based on keywords. For example, paragraph 4
mentions the word “framework,” but talks about the framework most common in other
research, not this article, and therefore has been assigned to Component 2, since it gives
information about the current best solution and its limitations.
Note also that the last paragraph in the Introduction section is labelled Organization.
This is rare in science, but common in both engineering and social science. Its purpose
is to help the reader follow the logic of the argument, as will be discussed in Chapter 9.

Table 4.1  Components in the first section (Introduction) of “Spatial query integrity”

Paragraph Topic sentence with component marker Component (purpose)

1a The amount of digital spatial information available . . . has 1a (General importance)

grown at an exceptional pace over the past decade.
1b Consequently, . . . is becoming increasingly popular 1b (Specific importance)
and has received a lot of attention in the research
community.∗
2 In this work, we focus on the . . . model, . . . However, 2b (Limitations)
there exist two major concerns with this model.
3 The general framework commonly used in the 2b (Current best solution)
literature . . . is based on . . .
4 The current state-of-the-art solution . . . 2b (Current best solution)
5 We argue that . . . suffer from several drawbacks. 2b (Limitations)
6 Motivated by the above observations, we propose . . ., a 3b (Proposed solution)
novel approach . . . based on . . .
7 This paper subsumes our earlier work . . . by extending 3a (Specific objective)
the . . . approach . . .
8 The remainder of the paper is organized as follows. [Organization]

∗This is actually the second part of the first paragraph, which discusses both levels of importance, but in other
articles the two aspects may be in separate paragraphs.
 Importance of and need for the research  51

Table 4.2  Components in the second section (Related Work) of “Spatial query integrity”

Paragraph First and last sentences of each paragraph, with component Component
markers (Purpose)

1 The idea . . . was first introduced by . . . 2b (Limitations)

Nevertheless, both solutions incur significant authentication
overhead . . .
2 . . . introduced . . . supporting fast query processing and 2b (Limitations)
verification.
Nevertheless, neither . . . handle data updates efficiently.
3 Efficient verification in the presence of frequent updates has 2b (Current best
been studied . . . solution)
The proposed mechanism achieves both correctness and . . .
4 All the aforementioned solutions require changes . . . 2b (Limitations)
The . . . method introduced in this paper is both efficient and 3c (Contribution)
deterministic, and does not require any modifications . . .

Table 4.3  Introduction division of the sample article “Spatial query integrity”

Section Paragraph Components

1 1a 1a Importance (general)
1b 1b Importance (specific)
2–5 2b Need (limitations)
6 3b Research goal (solution)
7 3a Research goal (objective)
8 Organization
2 1–4 2b Need (limitations)
4 3c Research goal (contribution)

In IPTC, the Introduction division may include a more extensive Related Work sec-
tion, as is the case with this article. There are only six citations or strings of citations
in the Introduction section of this paper, but 15 in the second section, in half as many
paragraphs.11 As shown in Table 4.2, the focus of this section is limitations (2b). Each of
the first two paragraphs state the benefits of a method but end in “Nevertheless.” While
the third paragraph accentuates the value of previous methods, the final paragraph again
points to their limitation and ends with a restatement of how the present work overcomes
this limitation.
Thus, the structure of the Introduction division of this sample article could be summa-
rized as shown in Table 4.3. Eight of the twelve paragraphs and all but two of the citations
are associated with need.

4.9 Comparison of IMRD and IPTC

To conclude, Components 1–3 are found in the Introduction of almost all research arti-
cles, usually in order. Component 4, the framework of the research, may be identified
from either the first or second division, or both, often by the attitude of the researcher
toward previous research rather than a direct statement.
52  Argument structure in exemplar articles

There is far more variation in

There is considerable the Introduction in IPTC than
variation in the structure of in IMRD. Sometimes there are
the Introduction in IPTC, in no citations at all, either because
length, components present, the Introduction is very short,
only one or two paragraphs, or
number of citations, and because most of the citations
presence of graphics. are in a Related Work section,
which is rare in science. In engi-
neering the Introduction may
also be very long, sometimes with a figure or table showing an overall schematic of the
design presented in the paper, while graphics are rare in the Introduction to science articles.
As mentioned above, importance (1) tends to be more prominent in science, but
need (2) in engineering. In engineering, different aspects of the research goal (3) may be
stated in separate paragraphs, with each paragraph discussing one characteristic that needs
to be incorporated into the new design, followed by that part of the goal. If stated as a
contribution, the wording of the research goal may be very similar to the Conclusion.
Furthermore, engineering articles frequently conclude the first section with a paragraph
showing the organization of the article, which is rare in science. Some of these differences
may be due to less editing of engineering articles, but others appear to be due to the innate
differences between science and engineering research.

Exercise 4.1 Introduction division components

(1) Identify the components in each paragraph in the Introduction division of your first
exemplar article and the component markers that help you identify them. Summarize
this in a table.
Hint: Use the list of components in Appendix 1 to avoid flipping pages.
Hint: Remember that component markers are general words that could help identify that compo-
nent in many different articles, not the specifics of that article.
Hint: Mark what components are present in each paragraph in the margins of your article. After
collecting the data, decide how to summarize and present them. Look at the tables in this chapter
and the online supplemental materials for ideas, but do not feel bound by them.
Hint: If your article contains Related Work or a similar section, remember to include it in the
Introduction division.
Hint: If you are in a writing class, your instructor may ask you to share with the class exam-
ples of component markers. If not, consider doing this with a friend. This will help you learn
to recognize the structure of the articles more quickly.
Hint: Past research may be cited to support the importance, need, or framework. We will look
more at past research in Chapter 7.

Teaching tip: Direct students’ attention to the online supplemental materials, which include
a complete component analysis of one article each in science and engineering.
 Importance of and need for the research  53

Teaching tip: Asking students to mark the components on the article (hard copy or electronic)
eliminates questions about what they consider a paragraph, which is sometimes difficult to
determine if there are many equations or other non-text items, common in engineering articles.
Teaching tip: Students may list specific content rather than general components. Help
them to identify the general components rather than the specific content of the article, which
is probably all they have focused on previously.

(2) Does the Introduction follow the order of components as listed?

Hint: Practice writing your answers as sentences that include the question, rather than using a
question-and-answer format. For example, you could say, “In this article the order of components
in the Introduction fits the expectations for IMRD.”
(3) Where is the research goal stated in your exemplar articles, and how? If they use the
phrase “to [the best of] our knowledge”12 or additional bullet points or sentences to
expand on the goal, make note of that.
Hint: If you do not find the research goal in the Introduction, it is occasionally found in the first
paragraph of the Process division in IPTC format.

Teaching tip: You might ask students to identify the research goal before the other components
commonly found in the Introduction. Since it is important and usually easy to recognize, it
provides a good starting point.

(4) Is there a paragraph at the end of the Introduction that describes the organization of
the rest of the paper?
Hint:This is common in IPTC, but not in IMRD.

Teaching tip: If this is too many questions for the level of your students, questions 1–2 are
the core questions.

Evaluate your understanding

(1) Why is it important to be part of an academic, disciplinary, or research community?
How does the Introduction help you claim your role in that community?
(2) What common components are generally present in an Introduction, and what is the
role of each in supporting the main claim?
(3) Why is the research goal the most important component of any research article?
(4) How do topic sentences and component markers help you, both as a reader and a writer?
(5) How do Introductions differ in IMRD and IPTC format?

Homework
(1) Repeat Exercise 4.1 for your other two exemplar articles.
(2) After gathering all your data, summarize them in the form of a claim and support.
Note anything special about individual articles, for example paragraphs on organiza-
tion or the location of the research goal.
54  Argument structure in exemplar articles

Hint: For example, “My first two articles follow the expected order of components, with a final
paragraph on organization, as shown in Table 2a and 2b, but the third has an unusual structure.
As shown in Table 2c, . . .”
Hint: Add this to what you wrote in the last chapter to build a more complete picture of the
general expected structure of articles in your field.

Teaching tip: It may be useful to set up a system of peer review. This has multiple
advantages:

1 As a writer, peer review gives students additional feedback.

2 As a reviewer, peer review develops the skill of reviewing others’ work.
3 Peer review instills the importance of collaboration in academic work.
4 Students will often notice things that writing teachers may not, or the process may
reveal misunderstandings or questions that the students may be unwilling to ask in
front of the whole class.
5 For students using English as a lingua franca, it also gives an opportunity to practice
spoken English and listen to different Englishes.
6 Peer review allows students to see more examples of how similar components are
expressed in other fields, as well as different organizational structures.
7 Students are actively engaged in the material and the revision process, and always
report the peer discussion is an important and enjoyable part of their learning.
8 In the process, students sometimes make new friends. Switching the groups every week
or two maximizes these advantages.

Teaching tip: Consider teaching the chapters on support at the same time as the chapters on
divisions, especially for advanced writing classes: Introduction (4) with Evidence from Past
Research (7), Methods/Process (5) with Evidence from Current Research (8), and the final
two divisions (6) with Reasoning (9).

Supplemental material: Sample component analysis.

Notes
1 Each framework has its own set of presuppositions, often about questions outside the realm of
science such as human nature, desired results, or ethical standards (see section 3.1).
2 Many linguists have proposed a similar structure for the Introduction division. For example,
Swales’ CARS model (1990: 141) and revised CARS model (2004: 230, 232) propose three
moves that parallel my three components, but Swales associates past research only with Move 1.
This is usually true in science but less so in engineering. Likewise, the 6 Stages listed by Cargill
and O’Connor (2013: 44) for the Introduction include my components 1–3 (splitting research
goal into two stages, goal and contribution), but also include past research and organization,
which I consider evidence and reasoning, respectively. Other similar models include Lewin, Fine,
and Young (2001) and the OARO model (Swales, 2004: 244). My component list was generated
independently, modified in response to student feedback, and largely completed before I saw any
of these, supporting the categories identified by previous authors, but I have not followed any of
their naming systems for several reasons. First, none have related the components to argument
structure, to an implicit general claim. Second, none have separated the three component claims
 Importance of and need for the research  55

from evidence, in this case past research, nor have they noted the distinction between science
and engineering in the primary location of citations of past research.Third, the names they have
used (moves and steps or stages) suggest to non-linguists that the components follow a certain
order, which is usually true for the Introduction, but not necessarily for the other divisions.
More details will be presented in Rau (In preparation).
3 Many linguists have split paragraphs, and even sentences, in their analysis of moves or stages.
Analysis by paragraph uses a clearly identifiable feature to attempt to reconstruct the intent of
the author.
4 This is definitely true in Chinese, and based on feedback from my students appears to be a com-
mon feature of many Asian languages.
5 “Lexical bundles” are common in engineering writing and can be used effectively to teach the
rhetorical conventions of a discipline (Hyland, 2008). The same concept is also called “sentence
patterns” (Cargill & O’Connor, 2013).
6 Since component markers are portions of sentences that are common and not content-specific,
this is not considered plagiarism.
7 Throughout the book, the component markers listed are taken from published articles, with
words specific to that article removed and the remaining clause edited for clarity if necessary.
8 Related work is very similar to the Literature Review often found as the second section in social
science articles.
9 “Reviewing items of past research” is Move 1, Step 3 of the revised CARS model (Swales, 2004:
230, 232). In engineering this is more frequently associated with Move 2, although since the
review precedes the claim (e.g. “nevertheless”), the order is the same. See Rau (In preparation).
10 Even theoretical work to establish a new framework must show how it improves on the current
one.
11 Citations can be counted either by number of occurrences or number of articles cited. The for-
mer counts the number of ideas that need to be cited, ignoring the number of articles cited in
each string and multiple citations of each article, and is easier practically. The latter counts each
citation in a string separately but eliminates multiple citations of the same source.
12 “To the best of our knowledge” is considered proper use of the set phrase in American English,
but “to our knowledge” is increasingly appearing in articles in the ELF world.

References
CARGILL, M. & O’CONNOR, P. 2013. Writing scientific research articles: Strategy and steps, New
York, John Wiley & Sons.
HU, L., KU, W.-S., BAKIRAS, S. & SHAHABI, C. 2013. Spatial query integrity with Voronoi
neighbors. IEEE Transactions on Knowledge and Data Engineering, 25, 863–876.
HYLAND, K. 2008. As can be seen: Lexical bundles and disciplinary variation. English for Specific
Purposes, 27, 4–21.
KLEINHEINZ, G. T., MCDERMOTT, C. M., HUGHES, S. & BROWN, A. 2009. Effects
of rainfall on E. coli concentrations at Door County, Wisconsin beaches. International Journal of
Microbiology, https://doi.org/10.1155/2009/876050.
LEWIN, B. A., FINE, J. & YOUNG, L. 2001. Expository discourse: A genre-based approach to social
science research texts, London, Continuum.
RAU, G. In preparation. IPTC: A prototypical format for engineering research articles.
SWALES, J. M. 1990. Genre analysis: English in academic and research settings, Cambridge, Cambridge
University Press.
SWALES, J. M. 2004. Research genres: Exploration and analysis, Cambridge, Cambridge University
Press.
Chapter 5

Demonstrating the continuity and

novelty of your research

5.1 Connected yet distinct

In some communities every house or apartment on a street is identical in construction.
Yet as soon as someone moves into one of the units, they begin to transform it to reflect
their interests. One person might put a flowerpot outside the door, another hang a flag or
other ornamentation. Inside the house the differences will be even greater. The same is
true of a research article. On the surface, the second division of many papers will appear
very similar, particularly in science, but as you look at the details, you’ll find each has its
own distinctive features.
On the other hand, houses around the world may be very different in structure, but
all fulfill the same purpose of providing shelter and all have rooms or areas that serve the
functions of living, dining, and sleeping. Similarly, the purposes of the Methods division
of IMRD and the Process division of IPTC are quite similar, despite major differences
in structure.
The second division is
not merely a description
of what you did. Like the
The second division does not Introduction, it serves the
merely report what you did, but far more important function
more importantly how it is of placing your work in the
both connected to yet distinct context of what has already
from previous research. been done. Much engineer-
ing research involves mini-
mal changes to previous
research, just as in science
one variable is changed at
a time and the results measured. Thus the Methods section in IMRD is not merely an
instruction manual, but shows how your work is similar to previous research while high-
lighting the differences. Similarly, the Process division in IPTC is a demonstration that
you have designed something new that improves on previous work.

5.2 Claim 2: reliability or feasibility of your work

Most of the second division is devoted to details of the current method or design. Recall
the claim the author is trying to support in this division:
 Continuity and novelty of your research  57

2 You have gathered data in a valid, reliable way (IMRD)

or

2 You have designed a workable solution to the problem (IPTC).

This claim is established by providing details of the present research, supported as

needed by references to past research. In IMRD, the details allow others to validate the
method, and your attention to detail in the description vouches for similar attention to
detail in carrying out the procedure and thus the reliability of the data. In IPTC, the
attention to detail allows others to verify your solution is workable. Yet the purpose of
presenting the detail is quite different, because the focus of IMRD is the data, which
will be presented in the following division, whereas the focus in IPTC is the new design
described in this division.
This difference in focus is reflected in very different division structure. In IMRD the
Methods division may be very similar to previous articles, so it tends to be written in a
very concise style and may even be placed at the end of the article in smaller font. In
contrast, in many IPTC articles the Process division is the longest division and written in
an extended style, with very little similarity between papers.1
Since the present research builds on past research, it is quite common in both formats
to find Component 4: framework in this division, but part of the extra length in IPTC
may be due to a long initial description of the framework. In both formats much of the
division will be devoted to Component 5: research details. Because of the difference in
the implicit claims, Component 6: testing methods is usually found in the Methods divi-
sion in science, but the Testing division in engineering. Statistical testing and the related
experimental design are far more important in science than engineering, as will be dis-
cussed at length in section 8.3.

5.3 Component 4: framework

Many terms could be used for this component, including framework, approach, model, and
theory. While these are not interchangeable, definitions given for them overlap. As I will
use the terms, framework is a general way of addressing a topic. Approach is often more spe-
cific, referring to tools or methods developed by previous authors. Model refers to a theo-
retical, conceptual, or mathematical
basis of the research. Widely used
models with broad explanatory abil- Previous research can be
ity are called theories. For the pur- cited to reveal its limitations
poses of this book we will use the (need) or positive aspects
general term framework, but your the current work builds
field may prefer another term.
on (framework).
Framework provides background
information about previous research
to show how it lays the foundation
for the current research. Framework (4) can be distinguished from need (2), because need
focuses on the limitations of previous research, whereas framework focuses on the advantages of one method.
58  Argument structure in exemplar articles

Experienced researchers in a field will be able to identify the framework based on

the author’s attitude toward past research in the Introduction. Since people talk posi-
tively about things important to them, more time and positive words will be devoted to
talking about research using the same framework. In science the framework can often
be inferred based on what the author thinks important and which previous research is
evaluated most positively, or the way the research goal is stated.2 In engineering the
framework may be indicated by the last of the previous methods presented, with the
fewest limitations.
The framework is often also indicated in the second division, by the choice of meth-
ods followed or the foundation of the present design. The framework can vary in length
from a single sentence to a whole section. Some of the ways it may appear are listed in
Figure 5.1.
In IMRD articles the framework can often be inferred from the methods used (4a),
since each methodology tends to be based on a certain approach or theoretical model.
Explicit discussion of the framework (4c) is less common in science than either social sci-
ence or engineering, but if different methods could have been used there may be a one
paragraph justification (4d) at the beginning of the Methods.
In IPTC, a subsection entitled problem identification or problem formulation may be
found in either the first or second division. When found in the Introduction, the focus is
usually on the limitations of previous research (2b). On the contrary, in the Process divi-
sion the focus is normally on positive aspects of a previous design that the present work
builds on (4b). There may be a section on the mathematical basis of the current work
(4c), including definitions and nomenclature. There may also be a comparison of previous
approaches, and justification of the one chosen (4d). For a very new solution, there may
be little or no framework.
The following component markers indicating the framework were all closely con-
nected with the research goal in the Introduction, either in the paragraph before, earlier

4) Framework
a. Accepted procedures (IMRD)
Claim: Accepted procedures or proven methods were followed
Support: Past research (citations or general knowledge)
b. Problem formulation (IPTC)
Claim: The research extends or builds on an established approach
Support: Past research (advantages of previous research approach)
c. Model or theoretical framework
Claim: The research is based on an accepted model or established design
Support: Past research (model, framework, mathematical basis, design)
d. Justification
Claim: This is the best method to follow for our research
Support: Past research (alternative methods)

Figure 5.1  Common sub-components of Component 4: framework.

 Continuity and novelty of your research  59

in the same paragraph, or part of it. The keywords underlined either highlight the benefit
of a method or how the current research builds on previous work.

• The excellent fluorescent properties give utility for . . . 

• The possibility of incorporating . . . makes the sensor a potential candidate for . . . 
• The proposed solution relies on . . . a useful paradigm to solve . . . 
• . . . exhibits exceptional features useful in diagnostic applications.
• In this paper we extend the approach . . . 
• In this work, a novel regularization technique is introduced that builds on . . . 

If there is a second section in the Introduction division, some paragraphs may be des-
ignated as framework if they focus on how the current method builds on the strengths of
previous methods rather than
overcomes the weaknesses of Framework presents
past work. This seems to be
relatively uncommon. background information needed
In IPTC the framework to understand the present study,
may comprise a whole sec- rather than details of the
tion at the beginning of the present study, but sometimes
Process division or the first the border is unclear.
subsection of several sections.
Rather than merely referring
readers to previous literature,
it is common in engineering to summarize either the theoretical principles underlying
the present design or ideas borrowed from previous designs. This makes it easier for the
reader to understand and evaluate the current design details, presented in the following
paragraphs or section, without looking at previous articles. The framework is distin-
guished from the research details because it is a description of background information
or previous work needed to understand the current study, rather than the new design.
Sometimes the boundary between the two is fuzzy, particularly if the current research is
an extension of previous work.
Sometimes entire sections or subsections can be identified as framework based on the
section titles, for example:

• Background information on . . . 

• Components of [the system used in previous papers]

Since the framework is very field-specific, it is hard to find general component markers.
Nonetheless, the following talk about specific procedures that would be classified as 4a.

• In order to understand . . . we briefly explain . . . 

• Our target architecture is . . . , the basic idea of which is . . . 
• In this paper, the method is transformed . . . 
• In . . ., a fast method to measure . . . was proposed. . . . This method also can be
extended to measure . . . 
60  Argument structure in exemplar articles

In IPTC the framework sometimes includes problem formulation (4b), often as a separate
section. This is common in articles developing a new algorithm or mathematical procedure,
although the connection with previous work is often stated in the previous section.

• Based on the aforementioned placement and routing models, . . . we define the

problem as follows.
• With the cost definition, we formally define the addressed problem as follows.
• In this section, the algorithm given in Section II is extended to the problem . . . 

The following are examples of 4c, because they explicitly refer to a model or theoretical
framework.

• As shown in Fig. 1, the . . . model represents . . . 

• According to the analysis of motor theory . . . 
• This section is devoted to a brief review of the theory . . . 
• Our proposed setup can be placed in the category referred to as . . . 

Justification of the choice of a frame­

work (4d) answers the question why,
Justification (4d) answers whereas the other sub-components
the question “why” this answer the question how. Since it
was chosen while the other involves a comparison and choice
sub-components answer between two approaches, general
component markers can be identified
the question “how.” more easily. Justification can take the
form of purpose (in order to), cause
or result (since, consequently), avoid-
ance or reduction of problems, or need for correct measurement or interpretation, as
illustrated by the following examples:

• In order to take advantage of . . . and lower the deployment cost, . . . 

• Since this device is meant for . . ., this frequency is a good choice.
• Consequently, . . . is required.
• Thus, using . . . as calibration parameter to correct . . . could facilitate accurate . . . 
• . . . [problems] can be theoretically avoided.
• The motivation for the . . . concept is to reduce . . . 
• To interpret the data correctly, . . . 
• Selecting the most appropriate technique is governed by . . . 

This component is supported by previous research but may not have citations.
Subsections dealing with definitions, mathematical notation, or well-known basic infor-
mation may be general knowledge and widely used by many in the field. More recent
work specific to the current research will probably have citations.
Sometimes framework and research details will alternate through the Process division,
with background information on one concept, then details of the current design that
build on that concept, followed by a second concept and its related design. This happens
frequently in papers that describe several different design elements in a single paper, for
 Continuity and novelty of your research  61

example hardware, software, and an algorithm that work together to accomplish a single
research goal, and in papers that combine previously existing methods.
To differentiate framework from other components, keep the following questions
in mind:

1 Is this describing limitations of previous research (2) or advantages the current work
builds on (4)?
2 Is this background necessary to understand the current research (4) or the current
research itself (5)?

To return to the metaphor of a house in a community, remember that need (2) talks
about existing houses, framework (4) lays the foundation, which will be very similar to
others but specific to the present work, and research details (5) are the things that clearly
separate the new structure from others.

5.4 Component 5: research details

Since there are many different research methodologies, Component 5 is extremely vari-
able. A generalized list of sub-components common in science and engineering is shown in
Figure 5.2.3 Do not feel constrained by the sub-components suggested. It may be useful to
add categories for things specific to your field. For example, in chemical engineering there
are frequently subsections on fabrication and characterization, so these would be useful
terms. A different list would be required for social science. For example, social science often
has sections on participants and ethics and may include an extensive description of qualita-
tive methods. Some areas of science and engineering that involve human subjects, such as
ergonomics, use a reporting format similar to social science and include these topics.
Different types of research details tend to be presented in science and engineering. In
IMRD, they are commonly restricted to the first two sub-components. Note that I have

5) Research details
a. Materials obtained
Claim: Appropriate materials were obtained from reputable suppliers
Support: Names and sources of materials
b. Methods followed
Claim: Care was taken to ensure good results
Support: Details of present research design and procedures
c. New design
Claim: A workable new solution, product, or process was developed
Support: Details of new design, intermediate testing of parts
d. New mathematical procedures
Claim: A mathematical theorem, procedure, or algorithm was developed
Support: Details of proof, lemmas, mathematical argument

Figure 5.2  Common sub-components of Component 5: research details.

62  Argument structure in exemplar articles

chosen to call these Materials obtained and Methods followed. This may not be necessary
in science, but in engineering it helps to distinguish Materials and Methods (5a and 5b)
developed previously by others from those developed by the current authors to fulfill the
goal of the present work (5c and 5d). I have chosen not to separate Product from Process
as sub-components because sometimes it is hard to distinguish the two, for example in
the development of a new circuit design or software, where a process is built into the
new product.
IMRD may not clearly specify which parts of the Materials and Methods follow previ-
ous research and which are unique to the present work. While there may be some modi-
fication of previously used methods, the focus of science is on the data to be collected
rather than the novelty of the method.4
On the other hand, in IPTC it is far more important to distinguish past from present
in this division. In engineering some materials will be purchased, and some methods fol-
lowed that were used in previous work, but the goal is development of a novel product
or process, so the third sub-component will dominate. In some fields of engineering new
mathematical procedures comprise a large portion of the Process division, so this is listed
as a separate sub-component. If the mathematical procedure is the main goal, the paper
may lack a Testing division, like other mathematical articles. In some fields of engineering
the design process is reported in stages, rather than just the final design. In this case there
may be intermediate testing between each design modification. Thus Components 5 and
6 may alternate within the Process division, again without a separate Testing division.
To identify the sub-components, it may be sufficient to examine the section or sub-
section titles, if they clearly designate the authors’ intent. For example, in the following
section titles from an article in chemical engineering, section 2.1 refers to materials (5a),
whereas section 2.2 describes methods (5b), as evident from the word “production.”
Section 2.3 then describes testing, Component 6, explained below.

2 Experimental
2.1 Materials
2.1.1 Raw materials
2.1.2 Characterization
2.2 Activated carbon production
2.3 Adsorption tests

Beyond section titles, it is

helpful to look for compo-
nent markers, especially in
Sub-components 5a and 5b are IPTC format. Unfortunately,
obtained from others and important since the materials and meth-
in IMRD, but 5c and 5d are developed ods used or the process and
in the current research and more mathematics developed are
important in IPTC. extremely field-specific, it
is hard to list general exam-
ples. As you develop a list of
markers used in your field,
 Continuity and novelty of your research  63

keep in mind that 5a and 5b are things obtained from others or from previous research,
whereas 5c and 5d are items developed and described for the first time in that paper.

5.5 Component 6: testing methods

Another important difference
between IMRD and IPTC
is the location of the test- In IMRD testing methods are
ing methods, Component 6 usually described in the Methods
(Figure 5.3). In IMRD the division, but in IPTC they are
testing methods are usually often found in the Testing division,
included in the Methods sec- just before the test results.
tion, often as a separate sub-
section (as in the example
above), whereas in IPTC they
are more commonly described
in the Testing division, frequently in the same section, or even paragraph, that reports
the results of the test. If the design process is reported as a series of steps, rather than
simply presenting the final design,
testing or optimization will fol-
low each design step to determine Science usually employs
which parameters to use in the next physical testing, often with
step, so Components 5 and 6 will statistics, whereas computer
alternate, and there may not be a or simulation testing is
separate Testing division.
There is a further difference in
more common in some
the types of testing used in the two engineering fields.
formats. While physical tests are

6) Testing methods
a. Physical testing
Claim: Testing was done carefully, following verifiable procedures
Support: Details of data collection method
b. Computer testing (IPTC)
Claim: Testing was done on a standard data set
Support: Details of testing method
c. Simulation testing (IPTC)
Claim: Testing was done under realistic simulated conditions
Support: Details of testing method
d. Statistical testing (IMRD)
Claim: Testing was done using standard statistical tests and tools
Support: Names of tests and software used, details of statistical design

Figure 5.3  Common sub-components of Component 6: testing methods.

64  Argument structure in exemplar articles

common in both, computer and simulation testing are rare in IMRD. Physical testing
involves measurement of data from physical objects. Computer testing uses electronic
data sets derived from real conditions. Simulation testing uses idealized data created by the
researcher. For example, in a robotic vision lab, physical testing could be done with cam-
eras and physical objects, computer testing could be done using 3D computerized images
of real objects, or simulation testing could be done on idealized grid patterns. Tests done
on a computer without mention of a data set should normally be classified as simulation.
Statistical testing may be used in conjunction with any of these three testing methods. It
is very common in science but relatively rare in engineering. When statistics are used, it is
essential to include a careful description. Far too many engineering papers that include error
bars in the figures contain no information on how those errors were calculated, whether
from 5 trials or 1000, which makes a huge difference in how the results are interpreted.
As with Component 5, section and subsection titles may state the type of testing per-
formed. Component markers can also be used to distinguish the various types of testing
methods. Often a sentence at the beginning of the description of the testing will identify
it, as in the examples below. Similar sentences are sometimes buried in the middle of a
section, making it harder for the novice reader, but experienced researchers in a field will
readily identify the testing type from the results.
Physical testing (6a), common in both science and engineering, can often be recog-
nized by reference to measurement of physical variables, as in the first two examples
below, or testing of physical samples, as in the next two. Development of a prototype can
also indicate physical testing. In chemical engineering, success is sometimes measured by
whether the desired product was obtained, so testing may consist of characterization of
that product, as in the last two examples.

• We measured the angle of the diffracted beam . . . 

• To test the effectiveness of the proposed approaches, we measured motion . . . 
• Different typical industrial pollutants were used to test . . . 
• To evaluate the adsorption capacity of the sample, . . . 
• The prototype was fabricated . . . 
• The end product is characterized for yield and [physical characteristics].
• The surface morphology of the sample was characterized using SEM.

Computer testing (6b) uses real data sets, but the testing is done and measurements
obtained on a computer. Sentences that talk about data sets, databases, or real-world data
indicate computer testing. The last example, which occurred in the Testing division,
reports three components in the same section: the testing method, results of those tests,
and comparison with previous work.

• The . . . digital data set was used . . . 

• To evaluate our method, we used three different databases: . . . 
• The proposed models have been tested on real-world traffic data.
• This section compares . . . against several state-of-the-art algorithms . . . on the . . . 
data sets.

Simulation testing (6c) can often be identified by words related to simulation, as in

the following:
 Continuity and novelty of your research  65

• We began our testing with a simulation of . . . 

• In this section we present simulation results and assess the performance . . . 
• We perform simulations to evaluate our proposed method.
• The effectiveness of the proposed distributed control algorithms will be verified by
simulating . . . 

Other types of testing carried out on a computer that do not use standard data sets would
also be considered simulation, including:

• The proposed placement method was implemented in the C++ programming lan-
guage and was applied in a design example to verify its effectiveness.
• In this section we demonstrate the use of the normal vector method . . . by applying
it to . . . 
• To evaluate the computational efficiency of different methods, we study . . . 
• Experiments have been performed to compare the proposed algorithm with a recently
proposed approach.

Statistical testing (6d) can usually be identified by words associated with statistical tests (or
their measures), such as mean (standard deviation, standard error, confidence intervals),
t-test or χ2-test (p-value), regression (r2), or correlation (r). Common with physical testing in
science, statistics can be calculated any time there are repeated measures of the same value,
including both computer and simulation testing. Often in engineering a graph will show
error bars output by a computer program, or a table will show p-values, but there is no
indication in the text how the statistical testing was done. Whenever statistical testing is con-
ducted, it is important to report not only the test done, but also the experimental design and
predictions, so if any of these terms or features occur in your exemplar articles, pay attention
to section 8.3. Examples of component markers that indicate statistical testing include:

• The 95% confidence intervals of the mean . . . are shown . . . 

• The level of interference . . . was analyzed using a χ2-test.
• . . . showed a significant increase in . . . (p < 0.01).
• . . . there was a strong correlation between . . . (r2 = 0.864).

Remember that in some cases involving development of mathematical algorithms, no

testing is performed. These articles are structured more like articles in applied mathemat-
ics, using mathematical proof rather than testing to verify the conclusions, without a
Testing division (IPC).
When doing the compo-
nent analysis, always remem- Pay attention to the overall
ber to think about the reason purpose of the paragraph rather
information is included rather than individual words. Intermediate
than focusing on individual
words. For example, you results or comparisons may be
may see the words “result” listed as part of the design process.
or “compare” in your arti-
cle and think those keywords
mean that a paragraph should
66  Argument structure in exemplar articles

be listed as Component 7 or 8, described in the next chapter, but if the verb indicates
potential, as in “we can compare,” it is probably talking about intermediate testing done
as part of the development process. Intermediate testing may also be described as results,
“using these results.”

5.6 Comparison of IMRD and IPTC

The difference between IMRD and IPTC formats is most obvious in the second divi-
sion. While two IMRD papers may have nearly identical Methods divisions, there is very
little similarity between Process divisions in IPTC. Furthermore, the Methods division
is often very short, written
in a condensed style, while
the Process division is fre-
Organizational features that help the quently more than half the
reader, such as topic sentences and paper and written in an
introductory paragraphs for each extended style.
We will talk more about
section, are common in the Process
extended and condensed
division in IPTC. style in section 9.2, but the
most obvious difference is
the presence of summary
paragraphs and topic sen-
tences in extended style. These features help the reader follow the argument, far more
important with the new design presented in IPTC than the relatively familiar method
presented in IMRD.
Where present, an introductory paragraph may serve different functions. Often it serves
as an overview, providing an organizational structure or summary of a section before giv-
ing details. It may also restate the research aim and show how the section contributes to
that aim. The only common use of an introductory paragraph in the Methods division
of IMRD is to discuss why one approach was followed rather than some alternative (4d).
For example, if we look at the IPTC article “Spatial query integrity” (Hu et al., 2013),
we find that most of the research details involve mathematics, but there is an introductory
paragraph at the beginning of each section and most subsections explaining the purpose
of that procedure. For example, the first paragraph of section 4, Data Transformation,
introduces why those calculations are needed: “Before transmitting the database to the SP,
the DO transforms each object by . . .” Similarly, section 5 has an introductory paragraph
before launching into various subsections relating to various aspects of the authentication
process. Most paragraphs have a clear topic sentence.
By contrast, in the IMRD article “Effects of rainfall” (Kleinheinz et al., 2009), there
is no introductory paragraph for any subsection in the Methods. Every sentence makes
a different point, with no topic sentences and no repetition. This contrasts sharply with
the Introduction, which has topic sentences for each paragraph. This is quite common in
IMRD papers.
As mentioned above, in IPTC articles framework tends to be far more prominent,
since the new design builds on the strength of past designs. Testing is often not mentioned
in the Process division, unless the design process is described in stages, with intermediate
testing before making the next design decision.
 Continuity and novelty of your research  67

Exercise 5.1 Methods or Process division components

(1) Identify the components in each paragraph in the second division of your first exem-
plar article and the component markers that help you identify them. Summarize this
in a table.
Hint: Remember that it is more important to identify the components than the specific sub-
components, which vary more between articles.
Hint: Both the titles of the subsections and the introductory paragraph (if present) may give hints
about what components to expect.
Hint: You may not have to identify every paragraph individually. Look for major transition
points, indicated by component markers. Particularly, where is the transition from “they” to
“we”? Nevertheless, in your table show how many sections or paragraphs are devoted to each
component, as this is an important aspect of the article structure.
Hint: Ask yourself the questions,“Is this background necessary to understand the current research
(4), rather than something actually done or developed in the current research? Does this describe
something that has been used before, perhaps with modification (5a or 5b), or is developed by the
current authors as part of the research goal (5c or 5d)?”
Hint: This division is very field-specific, so if something constantly recurs in your exemplars, for
example “fabrication conditions,” you may want to use that as a subcategory rather than the more
general ones suggested here. Just make sure it is a general category you can use, not specific to the
content of one article only, like “fabrication of [specific molecule].”
Hint: Component markers are also field-specific and will be easier to recognize after you have
read several articles.

Teaching tip: Students may find it useful to talk with other students in their field to
identify component markers.
Teaching tip: Help students determine what is specific to an article, and what are more general
recurring categories in their subfield.

(2) Does the second division follow the order of components as listed?
(3) Are there any organizational features such as an introductory paragraph? Are there any
other important features to help the reader, such as a figure showing the overall process?
Hint:These are common in engineering, but not science.

Teaching tip: Question 3 is necessary for an advanced class ready to write their own paper,
but could be eliminated for a basic class.

Evaluate your understanding

(1) How is the purpose of the second division similar yet different in IMRD and IPTC?
(2) How do the components present in the second division differ in IMRD and IPTC?
(3) Why are organizational features like summaries and topic sentences more important
in the second division of IPTC than IMRD?
68  Argument structure in exemplar articles

Homework
(1) Repeat Exercise 5.1 for your other two exemplar articles.
(2) After gathering all your data, summarize them in the form of claims about the general
structure of your articles and support for those claims. Note anything special about
individual articles that differ from the general pattern.
Hint: Use the format of a journal in your field for tables and table titles.
Hint: If several paragraphs in a section or subsection are the same component, rather than listing
each individually you could use a designation such as “Paragraphs 3–9: 5d, Math.”
Hint: Consider how to combine the information from different chapters to create a comprehensive
picture of the expectations for an article in your field, without repetition of information. Which
of the following would be most useful?
1 Article by article: Article 1 (Overview, Division 1, Division 2); Article 2; Article 3.
2 Division by division: Overview (Article 1, 2, 3); Division 1, Division 2.

Teaching tip: Help the students evaluate the two organizational strategies.
Teaching tip: Particularly in this division, students may be misled by individual words,
rather than looking at the context to identify a component.

Supplemental material: Sample component analysis.

Notes
1 Computer science normally follows IPTC format, but builds so closely on prior research that
occasionally the first part of the Process is identical to a previously published article, appending a
new portion at the end. Identical wording is considered plagiarism in most other fields.
2 The approach used to answer the research question is much more likely to be stated directly in
social science.
3 Previous authors have split Materials and Methods into distinct moves or stages, whereas sub-
components 5c and 5d are rare in science and thus in previous lists. In naming components, I have
sought to identify elements present in most research articles. Since most research articles contain
at least one sub-component but few have all, I have grouped them as a single component.
4 When a new method is presented, the article will take on a structure very similar to IPTC. Such
papers may be highly cited, but are not representative of most articles in the field.

References
HU, L., KU, W.-S., BAKIRAS, S. & SHAHABI, C. 2013. Spatial query integrity with Voronoi
neighbors. IEEE Transactions on Knowledge and Data Engineering, 25, 863–876.
KLEINHEINZ, G. T., MCDERMOTT, C. M., HUGHES, S. & BROWN, A. 2009. Effects
of rainfall on E. coli concentrations at Door County, Wisconsin beaches. International Journal of
Microbiology, https://doi.org/10.1155/2009/876050.
Chapter 6

Verifying the contribution of your

research

6.1 Staking your claim

To stake a claim originally meant to pound wooden stakes into the ground to mark off an
area of formerly unclaimed land as your own. The person making the claim had to prove
that no one else had already laid claim to the same area and that it contained a valuable
resource, perhaps a precious mineral. Staking your claim later became a figurative idiom
for asserting your right to something.
The claims we have been talking about in this book are obviously of a different sort,
but hopefully the analogy is clear. If you find something valuable and want to stake a claim
to it, you must prove that no one else discovered it first. That is done in the Introduction.
Then in the second division you mark the boundaries of the claim, separating it from the
work surrounding it. Finally, you must verify the value of what you have discovered. If an
idea is not valuable, why should it be published, even if it is different? Thus the last two
divisions of an article not only report and analyze results, but more importantly seek to
convince readers that the contri-
bution is valuable.
In this chapter we consider
We will consider the final two
the last two divisions together
for several reasons. Most impor- divisions together, as together
tantly, as explained in Chapter 3, they make and support a single
neither Results nor Discussion claim, the contribution of
by itself constitutes a complete the research.
argument. Moreover, Results
and Discussion are combined
in many science and most engi-
neering articles. Finally, although the same components tend to be present, the assign-
ment of components to divisions is different. Although IMRaDC is like IPTC in overall
structure, the Testing methods are usually in the Methods division, as in IMRD.

6.2 Claim 3: the contribution of your work

Whatever the format of the last two divisions, their purpose is to show that the work
makes an important contribution to the field. Recall the claim made in the last two
divisions:
70  Argument structure in exemplar articles

3 You have a good explanation for your data (IMRD)

or

3 Your solution is better than other existing solutions (IPTC).

To support this claim, five components

may be necessary: testing and analysis
Data patterns and methods (6), data patterns (7), com-
interpretations are essential parisons (8), interpretations (9), and
in IMRD, while comparisons conclusion (10). As with other divi-
are usually more important sions, there is a difference in emphasis
in engineering. between IMRD and IPTC.
In an IMRD paper two of these
are always required: a summary of the
data patterns (7) in the Results and
interpretations of those data (9) in the Discussion. Comparisons (8), often in the form
of statistical tests, may be included in either or both, while the last paragraph of the
Discussion usually highlights the conclusion (10). Testing methods (6) are usually found
in the Methods division.
In IPTC, testing methods (6) are usually found in the same section as the results of
the test. Comparison of the present design with the previous best design (8) is normally
required to prove that the new solution is better in some way, and some visualization of
the data patterns (7) is needed to conduct that comparison, followed by a conclusion (10)
to tie the paper together. Interpretation of the data (9) may or may not be required.
In both science and engineering these divisions frequently use an extended style, with
initial introductory paragraphs for at least some sections. An introductory paragraph may
look forward, summarizing that section, providing a general overview before specifics.
On the other hand, it is also quite common to begin the third division with an introduc-
tory paragraph that looks back to summarize the research aim and methods. After read-
ing the abstract, experienced researchers may jump straight to the results, since they are
already familiar with both the literature and methods in the field and want to see how
the results compare with established methods. A one- or two-paragraph summary of the
research aim and methods is sufficient for them to understand the goal and approach. On
the contrary, novice readers may have gotten so bogged down reading the details that
they will have lost sight of the overall goal. Including a short summary thus accommo-
dates their needs as well.
Data from the present work, often
presented in tables or figures, become
In these divisions it is far more important in these divisions
common to have more as support for the component claims.
than one component Nevertheless, past research will need
per paragraph. List all to be mentioned if there is any com-
that you find. parison with previous work.
In doing your component analysis
of this division you may find more than
one component in a paragraph. In IMRD
 The contribution of your research  71

a data summary may include statistical comparison, or comparisons may be interspersed

with interpretations. In IPTC it is common to find a description of the test, a sentence
pointing to a graph or figure depicting the results, a summary of those results, and com-
parison with previous studies, all in a single paragraph.

6.3 Component 6: testing methods

As discussed in the previous chapter, the types of tests typically conducted in science
and engineering differ. Testing methods are usually described in the Methods division in
IMRD, often as a separate subsection, but in IPTC this component is commonly found
in the Testing division.
Frequently in IPTC several tests will be performed, with the testing method, data pat-
terns, comparisons, and interpretations for the first test reported in one section before the
same sequence is followed for another test in the following section. Sometimes all four
components are described in only one or two paragraphs.1 For example, the following
sentences, from two successive paragraphs of an engineering article (Zhou & De la Torre,
2016), would be identified as components 6, 7/8, 8, and 9, respectively.

• We qualitatively evaluated our method with a leave-out scheme . . . 

• Figs. 7a, 7b, 7c, and 7d qualitatively compare our methods with . . . 
• Our method consistently outperforms . . . 
• This is not surprising because . . . 

Thus a good starting place

for your component analysis
is to identify what tests are Begin your component analysis
performed and where the by identifying the tests performed
data for each are reported. and where the data for each are
Then consider whether the reported. Then look for related
other components are in the
same section as the data or a comparisons or explanations.
separate location. This may
be easier than trying to iden-
tify the components by read-
ing through the paper in order. Remember too that sometimes the Process is reported
as a series of steps with intermediate testing before deciding on the next step. In this
case there may be little or no Testing at the end, so the two divisions may be combined.

6.4 Component 7: data patterns

Some sort of data summary is required in any research article, unless the research is purely
non-empirical (see section 2.4).2 Usually not all the data collected are reported, only those
important to the argument, and results are often reported as summary statistics (e.g., mean
or median) rather than raw data.
In naming this component, I have purposely chosen to use the term pattern, because
data are rarely useful as evidence until a pattern is discerned. I have used the plural because
usually more than one pattern is reported.
72  Argument structure in exemplar articles

A pattern may be shown in many

forms, often including tables and fig-
Patterns must usually be ures since these make it easier to visu-
identified for data to count alize patterns. Collectively, tables and
as evidence. Figures and figures are called graphics.3 Thus, any
tables make those patterns paragraph in these divisions referring
to a graphic should be identified as
easier to visualize.
Component 7, whereas figures in pre-
vious divisions may show an overview
of the process, or perhaps framework.
Data are usually measured in some way, using one or more of the testing methods
(Component 6). As shown in Figure 6.1, the results may include not only data patterns
(7a), but manipulation of that data after collection to make it easier to visualize the pattern
(7b). Exceptions to the pattern (7c) may be even more important than the pattern itself,
giving clues about how to interpret it. This could include possible sources of error, for
example possible human or machine errors during the collection process, or outliers that
do not fit the expected pattern and may lead to new insights. Sometimes the data may be
placed into groups, with a different pattern in each group (7d).
In the engineering literature it is
common for the data pattern (7) to be
One paragraph may presented primarily, if not exclusively,
in graphics. In science it is usually
include more than one
necessary to summarize these tables or
component, especially figures in the text, or point out some
in IPTC, and should be important feature, but engineering
marked accordingly. articles sometimes point to a graphic
without further comment. A graphic
may also contain a comparison (8).

7) Data patterns
a. General pattern
Claim: A pattern can be discerned in the data
Support: Graphical representation or data summary
b. Data transformation
Claim: The pattern is easier to see after appropriate procedures
Support: Description of data handling or transformation methods
c. Exceptions to pattern
Claim: Some data do not fit the overall pattern
Support: Identification of possible exceptions, outliers, or data errors
d. Classification of patterns
Claim: The patterns can be summarized into different categories
Support: Similarities in patterns

Figure 6.1  Common sub-components of Component 7: data patterns.

 The contribution of your research  73

Sometimes this is an explicit comparison between different methods described in the cur-
rent article or comparison with results from previous studies. There may also be implicit
comparison with a known goal in the field, for example how closely the results adhere to
an established standard or the degree of similarity between expected and measured values.
The text may also include an explanation of the reason for any differences (9). Thus, it is
quite possible for there to be several equally prominent components in a single paragraph,
which is far less common in science. Each component present should be marked.
In addition to mention of tables and figures, the following component markers indicate
data summaries. The first two describe something about the data directly (7a), whereas the
third describes a transformation (7b), and the fourth potential problems with the data (7c).
The last, classification of the data patterns into categories (7d), is actually an inference that
could be placed with Component 9, but is often reported in the results.

• We observe that . . . 
• In all cases the proposed method correctly detects . . . 
• Values were standardized by . . . 
• Possible sources of error in the measurements include . . . 
• Four different “profiles” emerged from the data . . . 

6.5 Component 8: comparisons

Comparisons pose a prob-
lem for traditional IMRD Comparisons involve
structure. In principle, evi-
dence should be presented in inferences, and should in
the Results but inferences, principle be in the Discussion,
including comparisons, in the but often are found in graphics
Discussion. In practice this is with the current Results.
nearly impossible, since data
are normally presented only if
a pattern is identified, but pat-
tern identification involves inference. Furthermore, statistical tests rely on inferences, but
test statistics are often shown in the same table as the data. A graphic may present not only
the current test results, but also results from previous studies for comparison.
It might seem that there would be a standard format and location for reporting such an
important category, but since articles are structured as arguments, good authors will present
things in the order that makes the strongest argument for their case rather than following a
set of rules for where things ought to go. Thus, instead of prescribing where comparisons
ought to be placed, it makes more sense to find where they are in your articles and why.
Different types of comparisons are shown in Figure 6.2. Often there is a comparison of
different treatments, designs, or parameters investigated in the current article (8a). Results
may be shown in a table or graph, with or without statistical testing, as in the following:

• Figures 14 and 15 schematically depict the proposed one-hour-ahead and one-day-

ahead forecasting model, respectively.
• Comparing the results between years, . . . 
• Fig. 7(c) verifies the reduction of training error . . . 
74  Argument structure in exemplar articles

8) Comparisons
a. Within the article
Claim: One treatment or design is better than others
Support: Graphics, data summary
b. With prediction based on hypothesis (IMRD)
Claim: Data [support/do not support] hypothesis
Support: Statistical testing
c. With expectations
Claim: Data [match/differ from/exceed …] expectations
Support: Graphics, data summary
d. With theoretical ideal (IPTC)
Claim: Results can be compared with a theoretical (mathematical) ideal
Support: Basis for ideal and comparison with it
e. With previous work
Claim: Data [support/refute/improve on …] previous work
Support: Past research compared with present, in graphics or text

Figure 6.2  Common sub-components of Component 8: comparisons.

Comparison with a hypothesis-based prediction and its accompanying use of statistical

testing (8b) is far more common in science than engineering. Studies that use statistics
should report the hypothesis and prediction, the experimental design (assignment of treat-
ments to subjects, repetitions, etc.), the test used and degrees of freedom, the results, and
the significance. If your exemplar article uses statistics, read section 8.3 and try to identify
where these various elements are in the article. Unfortunately, sometimes important ele-
ments are left out.
The location of these elements varies greatly. A prediction may be stated as part of the
experimental question in the Introduction, with the experimental design in the Methods,
or immediately before the data in the Results. It is also possible to give the results first,
leaving the prediction and statistical tests until the Discussion. Statistical comparisons are
often closely linked with the Results, and thus reported there, even though the statistics
are by nature inferential, and therefore should “properly” be placed in the Discussion. A
single graphic often reports both data and a statistical comparison, either as error bars on
a graph or standard deviation or significance tests in a table.
The following component markers indicate statistical comparison:

• The mean + standard mean error (SEM) . . . 

• The 95% confidence intervals of the mean . . . 
• . . . was significantly higher than . . .  (p < 0.001)
• The p-value for the entire data set was calculated as . . . 

Note that not every occurrence of the words “statistics” or “significant” indicates statis-
tical testing, especially in engineering, where these words are used much more loosely
 The contribution of your research  75

than in science. For example, the following sentences have nothing to do with statistical
testing. “Substantial” or “considerable” would be a better word choice in the first, since
no statistical test of significance was conducted. In the second, the intended meaning is
“values.” In the last, the meaning is “important.”

• This is a significant improvement over previous methods.

• The simulated statistics match the measured statistics very well.
• It is significant to note that . . . 

In many fields of engineering there is no hypothesis and therefore no prediction,

and no repeated measurement or statistical testing.4 Nevertheless, it is very common
to compare the results obtained with the results expected based on a model (8c).5 The
comparison may be between measured results and a simulation or between an observed
pattern and the expected pattern. Often this is shown in a table or figure rather than in
the text. The following component markers all indicate a comparison with the model
or expected values:

• The . . .  distribution agrees well with simulations.

• The maximum measured is . . . , whereas the simulated maximum is . . . 
• The calculated value from the simulation will be compared to the measured
value . . . 
• The estimated and actual values are very close . . . 
• The average error is less than . . . 
• As expected, case 1 has the largest . . . 
• . . . are compared with the performance of the reference geometry.

Again, single words cannot be used to assign a component. The following example
mentions “expected,” but in this case it is talking about a theoretical expectation used to
determine what value to give to a mathematical term, which has nothing to do with data
or comparison, and would be classified as 5d.

• As mentioned before, exponent γ in these equations is expected to have a small value

to smoothen the effect of the amount of length or resource constraint violations.
Therefore, we have set γ = 0.1 in this paper.

Comparison may also be made with some optimal standard, a mathematical or theo-
retical ideal (8d). This is more common in engineering but may occur in sciences like
physics when an optimum can be identified.

• Our technique can generate mappings that are close to the ideal performance defined
earlier.
• We tested against optimal parameters without noise.

Whereas comparison with previous work (8e) is optional in IMRD, it is nearly

obligatory in IPTC, for the obvious reason that the author claims the new design is
in some way better than previous designs. Sometimes testing is done using established
76  Argument structure in exemplar articles

methods in the field. Other

times it involves coming up
Statistical comparisons are more with a new test that better
common in science; comparison with demonstrates the advantage
expectations or an ideal are of the new design. The fol-
more common in engineering. lowing component mark-
Comparison with previous research ers identify comparison
with other research. The
is found in both.
first four express the intent
to compare; the next four
that the proposed solution
is better than others. The final two present confirmation of previous research, which is
more common in IMRD and likely to be reported in the Discussion.

• Our method is compared with the method proposed by . . . 

• We compared our algorithm to four others published in the literature.
• Our methods were compared with three baselines, . . . 
• Since previous methods reported their accuracy according to two different error
metrics, we show our results in both ways for comparison purposes.
• Our method consistently outperformed . . . 
• As expected, the common-centroid error obtained by the proposed method is smaller
than the one obtained in [11].
• The best method, however, turns out to be our transformation . . . 
• However, for [situation], this scheme would outperform the control.
• This confirms the results of previous study . . . 
• Results show . . ., which agrees with the results in . . . 

6.6 Component 9: interpretations

Items typically found in the Discussion are listed in Figure 6.3, all of which are far more
common in science than engineering. In scientific articles, discussion of the cause of the
results is mandatory. This is related to hypothesis testing and comparison of the actual
results with the predicted results, but often goes further in discussing whether there could
be other possible explanations of the same results, or how certain you are of your explana-
tion. Depending on the study, it will frequently include limitations on how broadly the
results can be applied.
In engineering the argument is more
likely to focus on the limitations—that
Discussion or the new design is better in some area,
explanation of the results just as good in some other, and per-
haps not as good in yet another. It will
is mandatory in science, include situations in which the new
but often more limited in design would be expected to have an
engineering. advantage and why. If practical appli-
cations have been tested, these may
also be discussed. It is less common to
 The contribution of your research  77

9) Interpretations
a. Analysis methods
Claim: Interpretation of the data was carried out in a reasonable way
Support: Details of analysis method, assumptions
b. Cause of results
Claim: There is a reasonable cause for the data
Support: Explanation of data pattern based on model or design goal
c. Elimination of causes (IMRD)
Claim: There is no better cause for the data
Support: Lack of support for alternative hypotheses or explanations
d. Practical applications (IMRD)
Claim: Understanding gained through this research has practical use
Support: Potential applications of the new knowledge
e. Limitations
Claim: Conclusions are valid within a limited range of conditions
Support: Limited range of testing or advantage

Figure 6.3  Common sub-components of Component 9: interpretations.

speculate on the underlying cause for the improvement, since often it was expected based
on theory.
Sometimes an author must justify the methods used for analysis (9a). This may be based
on standards, methods used by previous researchers, or argument that a certain method
yields better results. When present, this is frequently found where the test is described.

• In order to assess . . ., it is necessary to perform . . . 

• Therefore, it was necessary to compare . . .  when evaluating
• Better resolution was typically obtained . . . 
• We opted to use . . ., as this method allows . . . 

Explanation of the cause (9b) is not a direct observation, so it must be stated as the most
likely interpretation (see section 22.3.1), as in the following:

• For this reason, . . . 

• The reason our approach can . . . 
• This is on account of . . . 
• . . . is achievable because . . . 
• . . . is mainly caused by . . . 
• This phenomenon is attributed to . . . 
• A reasonable explanation of the increase . . . 
• . . . appears to have
• It appeared that . . .  due to . . . 
• It appears from these initial results that . . . 
78  Argument structure in exemplar articles

• . . . which can be interpreted as . . . 

• . . . could be explained by . . . 
• It is plausible that . . . 

Sometimes it is not enough to speculate on a cause. Other possible causes must be

eliminated (9c). Even in science, this is common only in cases where another possible
explanation seems likely.

• Therefore, we cannot attribute the increase . . . 

• . . . does not seem to be relevant.
• . . . appears to have limited capabilities . . . 
• On the other hand, it is possible that these effects were nevertheless responsible . . . 
• The alterations described in the present study do not represent . . .  and can be
interpreted as . . . 

In science it is also common to consider possible practical applications of the present

research (9d).

• One of the most important applications . . . 

• The approach can be applied to . . . 

In the following component markers, the focus is on feasibility or limitations of the

design (9e).

• Limitations of the current study . . .  include small sample size . . . 

• However, there are methodological concerns suggesting that . . . may be underestimated.
• However, it is very difficult to . . . 
• The estimated . . .  are inaccurate because . . . 
• There is a small discrepancy . . . , which may be due to . . . 
• The main reasons for this discrepancy are assumed to be . . . 
• . . . is the main drawback of the proposed solution.
• . . . the proposed algorithm is sensitive to initialization.
• Our . . .  algorithm can be confused when the assumptions made in this work are
violated.

6.7 Component 10: conclusion

The final component, shown in Figure 6.4,
will comprise the Conclusion section, if there
The Conclusion verifies is one, or the last paragraphs of the Discussion.
that the research goal It summarizes how the article achieves the
was achieved. research goal and therefore makes a valuable
contribution to the literature of the field. If
the research goal had several questions or sub-
points, each should be addressed separately.
In both engineering and science the wording of the conclusion may be nearly identical
to the research goal. While the goal states, “this is our aim,” the conclusion merely changes
 The contribution of your research  79

10) Conclusion
a. Summary
Claim: The research goal has been achieved
Support: Summary of research, answers to questions or new design
b. Achievement or advantage (more common in IPTC)
Claim: Solution improves on past work
Support: Contribution of research
c. Future work (more common in IMRD)
Claim: Remaining questions can be answered or further improvements made
Support: Planned or possible future work

Figure 6.4  Common sub-components of Component 10: conclusion.

it to “we achieved our aim.” Far greater difference is found in the social sciences, which
tend to word the research goal as a question, with the answer presented in the conclusion.
A summary of the article, demonstrating accomplishment of the research goal (10a),
is common in both IMRD and IPTC format. Perhaps because of the competitiveness
and rapid change in engineering, the second sub-component is very common, highlight-
ing the immediate advantage of the new design (10b). IMRD articles frequently include
future work (10c), which goes beyond the immediate results of the paper to suggest
related research planned by the authors or that others could contribute to. In many cases
the conclusion will include two or even all three of the sub-components.
The following component markers introduce summary statements (10a). They point
to accomplishment of the research goal, without explicit comparison with other research.
This is common in both IMRD and IPTC.

• Our results show . . . 

• In summary, we have shown that the properties . . . 
• We have presented results of . . .  experiments . . . 
• We estimated the thicknesses . . . 
• A novel system for . . .  was presented. A novel strategy for . . .  was proposed.
• A novel dual-mode control scheme . . .  has been developed.
• The novelty of this work is the design and fabrication of a complete system for meas-
urement of . . . 
• A new algorithm has been proposed . . . 
• We develop a robust, optionally fully automatic method that utilizes . . . 
• The comparisons above indicate the feasibility and accuracy of the combined method.

Achievement or advantage (10b), more common in engineering, emphasizes why the

current design is better than previous solutions. The following component markers high-
light the advantage:

• Some specific features and contributions of the paper include: [list]

• Through extensive experiments, we have demonstrated the effectiveness of our algorithm.
80  Argument structure in exemplar articles

• We have demonstrated that . . .  reduces energy cost and provides significant energy
savings.
• We have shown how to determine the minimum number of elements . . . 
• To overcome these limitations, in this paper a novel technique has been proposed . . . 
• We showed how . . .  outperforms state-of-the-art approaches to . . . 
• Unlike existing approaches, . . .  incurs no design modifications and comes at a lower
test cost.
• The main advantages over previously established methods are as follows. [list]
• The proposed design methodology can also be applied to different types . . . 
• Another potentially beneficial application . . .  is the possible use for . . . 

The following component markers talk about future work (10c). While this is more
common in IMRD, it is also found in IPTC when the work is currently underway.

• Our study poses new important questions that need to be addressed.

• Research on this . . .  can be extended further to reduce . . . 
• The optimization of this design to achieve thinner modules with higher gain is
currently under investigation.
• Future work entails making our method more efficient and supporting a greater
variety . . . 
• In the future, in order to validate the expected value of . . .  in a practical setting, we
plan to implement and evaluate . . .  on a real storage system.

6.8 Comparison of IMRD and IPTC

As has been mentioned several times, the location of some components differs in the two
formats. Although IMRaDC is structurally similar to IPTC, Testing methods still tend to
be found in the Methods division. Most of the components are present to some degree in
all research articles, but because of the difference in overall goal, engineering is more likely
to focus on Components 7 and 8, while Components 7 and 9 tend to be more impor-
tant in science. Particularly in IPTC, two or three components may be found in a single
paragraph. Some sub-components are more common in science, others in engineering.
Nevertheless, there is considerable variation between papers.
Since one article often contains multiple tests, each with their own results and discus-
sion, this almost always leads to some sort of cycling through the components in the final
two divisions. Sequencing strategies will be addressed in section 9.5.

Exercise 6.1 Results and Discussion or Testing

and Conclusion division components
(1) Identify the components in each paragraph in the last two divisions of your first
exemplar article and the component markers that help you identify them. Summarize
this in a table.
Hint: Remember that there is frequently more than one component per paragraph in these divisions.
Hint: As you start, look at the section and subsection titles. Do they give you an idea of what
you will find in each section?
 The contribution of your research  81

Hint: It may be helpful to first identify what types of testing are performed, then where the results
of each test are reported, and finally where the comparisons and interpretations related to that
data are found, which may be in the same paragraph, the same section, or a different section. If
you can identify the locations for the first test, other tests will probably follow a similar pattern.
(2) Does the order of components generally follow the order listed? Do any of them
repeat cyclically? Can you figure out a reason for the order?
Hint:The order may repeat cyclically, e.g., data1/explanation1, data2/explanation2, either because
it is easier or because the authors need the results of test 1 before they choose what to do for test
2. For further discussion of sequencing methods, see section 9.5.
(3) How close is the wording of the Conclusion to the research goal in the Introduction?
How similar is the wording of the same idea in the Abstract?
Hint: Recall that this is the main claim of the paper.
(4) Is there an introductory paragraph at the beginning of any section? If so, does it point
back to the research goal or forward to the contents of that section?

Teaching tip: Questions 3–4 are necessary for an advanced class ready to write their own
paper, but could be eliminated for a basic class.

Evaluate your understanding

(1) Why are the last two divisions discussed together in this chapter?
(2) What is the purpose of each component found in the last two divisions?
(3) How do science and engineering articles differ in the last two divisions?

Homework
(1) Repeat Exercise 6.1 for your other two exemplar articles.
(2) Write general claims about the structure of your exemplars, with support and
exceptions.
(3) Compile a complete description of the structure of a typical article in your research
field, based on your analysis of the three exemplar articles.
Hint: Number any tables and figures in your description sequentially, following the title for-
mat in your exemplars, and make sure you refer to all the tables or figures in the text of your
description, either summarizing them or pointing out certain information the reader should pay
attention to. For example, you might say, “As shown in Table 4, . . .” or “. . . as can be seen
in Table 5.”

Teaching tip: After students turn in a complete description, it is useful to have them do a
revision, based on your comments and/or peer review. They will be able to see the improve-
ment, and will understand the process when working on their own writing later.

Supplemental material: Sample component analysis.

82  Argument structure in exemplar articles

Notes
1 I have also seen this in social sciences such as economics and management, where a new method
is presented in the equivalent of a Process division, followed by a section comparing it to previous
methods.
2 In this case, the Results and Discussion will be totally absent, apart from justification, limitations,
and other comments interspersed with the mathematical proof.
3 See section 8.4 and Chapter 24.
4 These are, however, common in biomedical engineering.
5 This is essentially identical to comparison to a prediction (8d), as both compare to a theoretical
value, but for the sake of student understanding I have chosen to list it separately since the terms
hypothesis and prediction are commonly associated with statistical testing and are therefore rare
in engineering.

Reference
ZHOU, F. & DE LA TORRE, F. 2016. Spatio-temporal matching for human pose estimation in
video. IEEE Transactions on Pattern Analysis and Machine Intelligence, 38, 1492–1504.
Chapter 7

Evidence from past research

7.1 Connecting with accepted evidence

In the previous three chapters we examined the general claims commonly made in research
articles. Now we will look in more detail at the evidence and reasoning used to support
those claims. Evidence can
come from past research, the
focus of this chapter, or pre- Past research may be cited
sent research, discussed in the throughout an article to support
next. Reasoning, including various claims, since publication
ways to help readers follow
the argument, will be the sub- implies acceptance by the
ject of the third. research community.
References to past research
establish a connection not only
to a research community, but
also to the evidence presented in those papers. Publication implies a degree of acceptance
by a community, so published research can be cited by other scholars as evidence in their
own writing. Previously published work is less subject to scrutiny than new evidence from
the current research, but cannot be used to support every claim.
Before going further, we need to distinguish two terms. A citation is a marker in the
text that points to previous work. Author-date citation format, common in science
and social science, places the name of the author and the date of publication in paren-
theses, while numerical format, common in engineering, uses a numeral in superscript
or brackets. The citation points to a reference listed in the References, Works Cited, or
Bibliography at the end of the article. In Chapter 25 we will examine reference and cita-
tion formats, but for now we will consider only their use in the text used as evidence.

7.2 Evidence from past research

Past research is never mentioned for its own sake. To have value it must support the argu-
ments made in the present article. Past research is essential for supporting certain claims
and may be useful for others.
As discussed in Chapter 4, past research is used in the Introduction to establish the
importance of and need for the research which leads up to the research goal. In IMRD
this is the main function and location of citations, so textbooks on academic writing often
84  Argument structure in exemplar articles

discuss past research only in this context, with at most a passing mention in the Discussion
section. Nevertheless, citations are common in all divisions of IPTC format except the
Conclusion. They often appear in the Process as the framework underlying the current
research and in Testing to compare the current design with past research.

7.2.1 Categories of past research

Past research can be mentioned for several purposes, listed in Figure 7.1. The first three occur
in the first two divisions of an article, while the last may occur throughout, and is the only
one that always requires a citation. The first two rarely have citations, so they must be identi-
fied based on the content. A widely used framework may be mentioned without citation,
unless it is closely associated with a single researcher. As discussed earlier, the framework or
approach used is sometimes not stated, but can be inferred based on the papers cited.
Background knowledge (I a) is general information the field considers a well-established
fact. It is usually stated in the present tense, since it is considered inherently true independent
of time. It may include anything from the needs of society to the specifics of the field, and
usually meets the following four criteria:

(1) Written in present tense;

(2) Without citation;
(3) Not about new research presented in that article;
(4) Known and accepted by experienced researchers in the field.

Since it is so field-specific, there are no predictable sentence patterns, but the following
would all be considered background knowledge. All were found in the Introduction and
most are related to importance. Common in the first paragraph of engineering articles,
this category appears to be less common in science, which prefers citations even for
apparently obvious statements.

• A number of mmWave bands are currently being considered for global 5G networks.
• Power flow on overhead transmission lines is limited for a number of reasons.
• An additional challenge present in the telehealth environment is the large quantity of
data which requires analysis.
• An attacker could launch a Denial of Service (DoS) attack by generating a large num-
ber of anomalous packets with different header fields.
• The SEE characteristics of insulator materials are difficult to study because the charge
release is a gradual process over time.

I. Past research
a. General background knowledge for those in field (often without citation)
b. Definitions, terminology, or notation (with or without citation)
c. Description of theoretical frameworks, models, or approaches
d. Citation of specific methods, results, or conclusions from past research

Figure 7.1  Common evidence from past research.

 Evidence from past research  85

• The storage mechanism consists in electrostatic adsorption of electrolyte ions on the

electrode surface, which means that the capacity increases with the electrode-specific
surface area.
• During the activation step, the oxidation effect of the activating agents involves an
increase in O content and a decrease in C content.
• The electrocardiogram (ECG) is an important tool in assessing cardiac health.
• Breast cancer is the most common form of cancer among women worldwide.

Definitions (I b) may have citations if they were proposed recently, but not if they have
become common knowledge in a field. In computer science they often occur in a separate
section at the beginning of the Process division with the name Preliminaries. In other papers
definitions may appear as a list at the beginning of the Process division with one definition per
paragraph and the words defined in italics. Mathematical representations or notation, which
may appear as an unnumbered section entitled Nomenclature placed before the Introduction,
also fall in this category. Unless first defined in the current article, definitions are part of the
Framework component. The following would all be called definitions or notation:

• Here we define: . . . 
• A timing path is defined as: . . . 
• Definition 1 (Connected Graph): A graph is represented as . . . 
• We represent graph vertices as positions in the image.
• We denote binary vectors in bold lowercase characters.

Framework (I c) refers to information

other than definitions related to the theo-
retical framework of the present research.
Citations will help you
It may occur in the Introduction in identify many but not
IMRD but more commonly in the all references to past
Process division of IPTC. We will dis- research.
cuss this more in the next subsection.
Citations of specific information (I d)
are the easiest to identify, because of the
citation, but it is important to determine what purpose they serve in the argument. The
connection with various components will be described in the next subsection.

7.2.2 Past research as evidence

for various components
Past research is frequently used as evi- Past research is frequently
dence in connection with certain compo- cited in the Introduction
nents, yet rarely with others. This makes as evidence for
sense when you consider the purpose of importance or need.
each component.
Past research is frequently cited in
the Introduction in both IMRD and
IPTC. Here it shows either the importance of the research or limitations of previous
research and therefore the need for the current work.1
86  Argument structure in exemplar articles

Importance (Component 1) establishes why anyone should care about the research,
either to publish it or read it. Just like promoting a tourist destination, if many people
have gone there, pointing to the numbers is sufficient. Thus in engineering it is common
to have strings of citations, but these quickly become cumbersome in the author-date
citation format of science, as shown in the third example.

• Since then, numerous query authentication solutions have been proposed . . . [8, 14,
17, 21, 22, 29, 32, 33].
• Previous works on circuit-level modeling and simulation . . . [8]–[21] . . . 
• Rheological properties . . . have been studied extensively in the last three decades
[Doi (1980a); Osaki and Kurata (1980); Osaki et al. (1981); Venerus et al. (1990);
Isono et al. (1991); Venerus and Kahvand (1994)].

It is also common to describe at least a few studies individually. Sometimes by only

reading the description of each study it is impossible to tell whether the author intends to
use it to support the importance or need of the research, but the intent may be clarified by
other statements in the paragraph. For example, in two successive paragraphs of a science
article we find the following sentences (Wen & Hua, 2009: 781–782):

• Thus, understanding the entanglement dynamics and structure is of great importance

as the transient entangled network of the fluid affects the subsequent material defor-
mation and stress relaxation.
• Thus, a better understanding about the impact of the transient entanglement structure
is essential if one wishes to more accurately describe the chain dynamics and associ-
ated rheological properties in fast flows.

While importance and need are closely related, the first paragraph and the citations in it
stress why we care about this topic (it affects important physical properties), while the sec-
ond builds the case that more work needs to be done (to better describe how those prop-
erties affect flow). The following paragraph, which states the research gap and research
goal, contains no citations. This makes sense because the claim is that there is no research
in that area, but the current study will fill that gap.

• It appears to us, however, that no earlier experimental studies have conducted . . . 

• In the present study, . . . 

As we saw in Chapter 4,
some IPTC articles have a
section immediately follow-
Related Work and Background
ing the Introduction which
sections in IPTC frequently are more contains much of the past
related to need, but Theoretical research cited in the article.
Framework and Preliminaries This is often named Related
to framework. Work, Background, or a
similar general title, although
sometimes a section or subsec-
tion serving the same purpose
 Evidence from past research  87

will have a descriptive title. It appears to be most common in rapidly changing fields
such as computer science and information engineering, ensuring that readers are aware
of the most recent literature. Like a Literature Review in social science, this mainly
discusses past approaches and their limitations, but the framework of the present study
may be indicated as well.2
Framework is one of the hardest components for novices to identify. On the one hand,
it may not be clearly stated by the author; on the other, readers may be unaware of or not
have names for alternative frameworks. In the same way, a child may recognize a structure
as a house, but only later acquire names for different types of houses in their area.
In science the framework may be implicit based on the choice of methods and
analysis, particularly in highly specialized journals. In engineering, where there are usu-
ally several possible approaches to a problem, it is often found at the beginning of the
Process division as information necessary to understand the present work. Usually in
this case it can be identified by the presence of multiple citations, rare in the Methods
division of science articles. Sections entitled Theoretical Background or Preliminaries
may include definitions or nomenclature and a general description of the framework,
with or without citations.
Citations may also occur in engineering in connection with the fifth component,
details of the current research, when the new design incorporates a previously reported
element without modification. Fewer citations are associated with Component 6, testing
methods, since widely used standardized
methods usually do not require a citation.
For the same reason, citations are rare with
Past research is frequently
these components in science.
Citations may also be found in the cited in the Discussion,
Testing or Discussion divisions. Citations but rare in the Conclusion.
will be mandatory if there is any compari-
son with previous research (8d), for obvi-
ous reasons, but less frequent with other
types of comparisons, thus again more common in engineering. Citations will be more
common with interpretations (9) in science, either to eliminate other possible explana-
tions, confirm previous studies, or generalize the conclusions.
On the other hand, citations will rarely be found in conjunction with components
that focus on the current research. Thus we would expect to find few associated with
the research goal (3), data patterns (7), or the conclusion (10), and indeed this is the case.
If data from previous studies are presented in the same data table or figure as the present
study, this will have a citation. If the present work is an extension of prior work by the
same author, the previous work may be cited in the research goal. The relative frequency
of citations of past research in conjunction with each component in IMRD and IPTC
formats is summarized in Table 7.1.

Exercise 7.1 Past research and components

(1) In your first exemplar, is there a separate section for definitions, related work, or
theoretical framework? Is the title general or specific?
(2) Is accepted knowledge ever mentioned without citation (background, definitions,
framework), and if so, where?
88  Argument structure in exemplar articles

Table 7.1  Relative frequency of past research as evidence for each component

Component IMRD IPTC

1 Importance ++++ ++
2 Need ++ ++++
3 Research goal - -
4 Framework ++ ++
5 Research details + +
6 Testing methods + +
7 Data patterns + +
8 Comparisons + ++
9 Interpretations ++ +
10 Conclusion - -

(3) Where are the citations in your exemplars? What components do they seem to
support?
Hint: As with earlier chapters, your exemplars may differ from the common pattern, and this is
important to note.
Hint:You may find it useful to create a table showing how many citations are associated with
each component.

7.3 Author-prominent vs. information-prominent citations

As we saw above, in every field some information is so well known and widely accepted
that citations of specific past research are not needed. In your own writing you may
wonder whether you need to include citations or not. The expectations vary somewhat
between science and engineering; as always, your exemplar articles are your best guide to
your field. If your exemplar articles mention the importance to society, definitions, or the
name of some technique without citations, you probably can too. If other papers give a
citation for certain information, you should too.
When citing an article, an author needs to decide which is more important in the
situation: the author or the information. Author-prominent and information-prominent
citations3 differ in how much attention is given to the author of the article cited.
In an author-prominent citation, the
name of the author is an integral part
of the sentence, often the subject, as in
In an author-prominent
the following examples from “Spatial
citation, the name of the author query integrity” (Hu et al., 2013):
appears in the sentence, but
in an information-prominent • Papadopoulos et al. [26] designed
citation it does not. a solution for . . . 
• On the other hand, Ku et al. [13]
proposed . . . 
• The idea . . . was first introduced
by Hacigümüs et al. [9].
 Evidence from past research  89

In an information-prominent citation, the name of the author is not part of the sentence,
as in the following examples from the same work:

• The first mechanism for verifying query results in multi-dimensional databases was
proposed in [3].
• However, since [13] is not a deterministic solution, attacks may escape the auditing
process.
• Aggregate signatures are provably secure [1, 18, 19] . . . 

A variation of the information-prominent citation seems to be more common in

mechanical engineering than other fields.4 It does not use the author name, but inserts the
word “Reference,” as in the following sentences:5

• Results for stiffness design con-

sidering large displacement theory
The impact of using an
have been presented in References
[21–23], . . .  information-prominent
• In this paper, the problem for- citation is much higher with
mulation from Reference [14] is numerical citations than with
advocated . . .  author-year format.
Note that there is a far greater impact
of using information-prominent cita-
tions in the numerical citation format common in engineering than there is in the author-
year citation format common in science. Consider the following two examples:6

• Based on [3], Cheng and Tan designed a mechanism for authenticating kNN queries
on multidimensional databases, ensuring that the result set is complete, authentic, and
minimal [4, 5].
• We plotted the epicenter of the mainshock estimated by Shin et al. (2000) and that of
the aftershock estimated by the Taiwan rapid earthquake information release system
(Wu et al., 1997).

In the first example, without looking at the Reference section there is no way to
determine who the author of citation 3 is, while in the latter we still know who
both authors are, even though
the first citation in the sentence
is author-prominent and the Author-prominent citations
second information-prominent, are rare in engineering, used
appearing in parentheses rather mostly for the first paper in a
than the sentence itself. field or the model on which the
The relative frequency of current research is based.
author-prominent and information-
prominent citations varies by field,
but information-prominent citations
predominate in engineering. Author-prominent tend to be used only for the first paper in a
field or the framework on which the current paper is based.
90  Argument structure in exemplar articles

In addition to citation type, the range of dates cited also differs. In the sciences there
may be references to classic papers, sometimes 50 years old, but in engineering most of the
papers cited are within 10 years, and it is rare to cite anything over 20 years old.

Exercise 7.2 Citation types

(1) For each section of your exemplar article, count the number of author-prominent
and information-prominent citations. If the frequency differs, can you determine any
factor that seems to influence the type of citation used?
Hint: Count by section rather than division, because the frequency may be far different in, for
example, the Introduction and Related Work sections, both in the Introduction division.
(2) Do any information-prominent citations insert the word “Reference” or “Ref.”?
Hint: Insertion of “Reference” is relatively unusual. It is only important to mention this in your
description of your exemplars if it does occur.7
(3) How many references are less than 5 years old? 5–10 years? 10–20 years? Older?
Hint: Count the years between the reference and the date of publication of the article, not the
date you are reading it. For example, a citation of a 2011 paper in an article published in 2015
would be four years old, no matter when you read it.
Hint: Again, consider the best way to present the information: table, text, or a combination.

7.4 Citation verbs

Word choice is important in any writing, not only to convey concepts accurately, but
also to identify with a particular group and their shared view of reality. A change in gov-
ernment may be referred to as liberation or a rebellion. One side’s freedom fighter is the
other side’s insurgent. In science the consequences of using the wrong term may not be
as drastic as they could be in the political realm, but words are still important as a way of
indicating your connection with an academic community.
What I will call citation verbs report what a previous author did. The preferred words dif-
fer by field, as illustrated in Table 7.2.8 Nine of the words appear only once, compris-
ing almost half of the top five
words in these four fields. Also
note that many common verbs
Citation verbs, used to describe used to talk about writing, such
what a previous author did, differ as write, state, and say, do not
from journal to journal and carry appear at all.
positive, neutral, or negative All words carry connotations.
connotations. Consider the difference between
the citation verbs in the follow-
ing three statements:

• Jones (2012) discovered that A causes B.

• Jones (2012) stated that A causes B.
• Jones (2012) claimed that A causes B.
 Evidence from past research  91

Table 7.2  Citation verb frequency in different fields of science and engineering

Rank Biology Physics Electrical engineering Mechanical engineering

1 Describe Develop Propose Describe

2 Find Report Use Show
3 Report Study Describe Report
4 Show Find Show Discuss
5 Suggest Expand Publish Give

Selected and redrawn from Hyland (1999), used by permission, with additional data from Swales and Feak (2012: 213).

The first implies acceptance of the causality, while the last implies that the current author
does not think the causality had been proven. The middle statement is neutral.9 Citation
verbs tend to have positive, neutral, or negative connotations, as shown in Table 7.3.
Note that far more of the citation verbs are positive or neutral than negative, since in
engineering if an author disagrees with a previous work, they are more likely to simply
not mention it than cite it negatively, unless the purpose of the article is specifically to
challenge that previous method.
In engineering articles, most citations occur in the first quarter of the paper, in the
Introduction and beginning of the Process division, but not all citations employ a cita-
tion verb. Author-prominent
citations always contain a
citation verb, either active Author-prominent citations
or passive, as in the examples always contain a citation verb,
shown above. either active or passive.
In author-prominent cita- Active voice further highlights
tions the tense and voice the researcher as opposed
used also make a difference. to the research.
Consider the difference in
connotation of the following
statements.

Table 7.3  Connotations of citation verbs common in electrical engineering∗

Positive Neutral Negative

Demonstrate Calculate Assert

Determine Conclude Claim
Discover Describe Suggest
Establish Develop
Prove Explain
Resolve Find
Show Present
Solve Propose
Validate Report
Verify State

∗Based on unpublished data of 30 articles from six journals.

92  Argument structure in exemplar articles

Active voice:

• Wang [1] reports that . . .  (present)

• Wang [1] reported that . . .  (past)
• Wang [1] has reported that . . .  (present perfect)
• Wang [1] had reported that . . .  (past perfect)

Passive voice:

• It is reported by Wang that [1] . . .  (present)

• It was reported by Wang that [1] . . .  (past)
• It has been reported by Wang that [1] . . .  (present perfect)
• It had been reported by Wang that [1] . . .  (past perfect)

The active voice directs attention to the researcher, whereas the passive places more
emphasis on the research. Therefore, the active voice is commonly used for landmark
studies in a field to highlight the contribution of the author. Moreover, authors’ names
can be eliminated from passive constructions, but not active. Thus authors will sometimes
refer to other studies in the active, but their own in the passive. In science and engineer-
ing it is common to find over half of the verbs in the passive.
Tense is also important, carrying
a connotation about how research-
ers currently view that past research.
The tense used when citing
Past tense is often used when citing
previous studies also carries a single study, because it describes
connotations about whether that the past action of one researcher at
research is considered valid. a point in time. On the other hand,
the present perfect may be used
when referring to several studies or
a general area of inquiry because it
describes what was reported over
a wider time period. The present
Information-prominent tense can be used to indicate some-
thing that is commonly held to be
citations frequently do not use true by researchers in the field today.
any citation verb, and are far more Conversely, the past perfect is used
common than author-prominent to report something that was previ-
in engineering writing. ously believed, but no longer is.10
In information-prominent cita-
tions, passive citation verbs are more
common, as shown above. However,
unlike author-prominent citations, information-prominent do not require the use of a cita-
tion verb. In a typical engineering article, very few of the sentences citing previous work
contain a citation verb.11 This absence of a verb where there could be one is called a
null occurrence. Null occurrences are very common when generalizing the work of
many authors, in which case a string of citations may be listed at the end of a sentence
 Evidence from past research  93

or phrase, as in the first example below. Similarly, the second example below does not
talk about what the author did in that paper; rather, the citation is a shorthand way of
indicating a particular process, which is stated as an accepted truth. The third example
below refers to the problem formulation from the reference without using a citation verb.
The verb “advocated” refers to what is done in the current paper, not the reference cited,
and is therefore not a citation verb.

• Aggregate signatures are provably secure [1, 18, 19] . . . 

• However, since [13] is not a deterministic solution, attacks may escape the auditing
process.
• In this paper, the problem formulation from Reference [14] is advocated . . . 

Exercise 7.3 Citation verbs

(1) Look at each sentence in your exemplar article containing a citation. Underline all
the citation verbs and make a list of the most common. Count the number of null
occurrences, times where a citation is not accompanied by a citation verb.
(2) How many of the citation verbs carry a positive, neutral, or negative connotation?
(3) Look at the verbs you underlined. What tenses and voices are used in each section
and for each verb? Is tense or voice used to mark anything the reader should pay
special attention to?
Hint: How could you most effectively present the answers to these three questions: one table, more
than one, a text summary?

Evaluate your understanding

(1) How does the location of past research differ in IMRD and IPTC, and why?
(2) Which components tend to use past research as evidence, and why?
(3) What differences are there between author-prominent and information-prominent
citations, and when might an author choose to use each?
(4) How does the choice of citation verb, voice, or tense affect the connotation?

Homework
(1) Repeat the Exercises in this chapter for your other two exemplar articles.
(2) Write a brief summary of what you have learned about the location of past research
and the use of citations in your exemplar articles.

Teaching tip: Sections 1–2 are important for all students. Sections 3–4 are more difficult and
best left for advanced students.
Teaching tip: This chapter can be taught with the Introduction division, where past research
and citations are most important. Sections 7.3 and 7.4 could also be taught in connection
with section 22.3 (Connotation) or Chapter 25 (References).
94  Argument structure in exemplar articles

Notes
1 Components 1–3 in this text correspond closely to the three moves of Swales’ revised CARS
model (Swales, 2004: 230) Swales considers past research an essential part of the first move and
optional in the second.This does not seem to hold in IPTC, but further work is needed to docu-
ment this.
2 Unpublished observations. As far as I know, no systematic research has yet been done on this.
3 Swales and others following him use the terms integral vs. non-integral citations (Swales, 1990:
148) I have chosen the terms author- and information-prominent (Weissberg & Buker, 1990),
because the meaning seems more obvious to engineers.
4 This is called a hybrid form by Hyland (1999). Because the author’s name is not mentioned
and dropping “Reference” gives exactly the same form as information-prominent, I view it as a
variant on that form.
5 From Pedersen, Buhl, and Sigmund (2001).
6 The first from Hu et al. (2013), the second from Satoh et al. (2001).
7 Linguists call the unusual situation “marked,” which is marked or noted only when present.
8 Linguists call these reporting verbs (Hyland, 1999). Citation verbs are equivalent to reporting
verbs, author focus, which are much more common than writer focus in science and engi-
neering (Rau, unpublished data). The distinction between author focus and writer focus is not
intuitive, particularly to those from an ESL background. I have chosen a term that designates the
situation in which they are used.
9 It is also possible to insert an adverb before the citation verb to indicate the current writer’s
evaluation of the previous work, but this seems to be more common in the humanities and social
science than engineering.
10 For more details and examples see Feak and Swales (2009: 52–54).
11 According to Hyland (1999: 249), roughly 40% of all citations in engineering used report-
ing verbs, but a recent preliminary analysis of electrical engineering articles showed that the
frequency may currently be closer to 20% (Rau, unpublished data).

References
FEAK, C. B. & SWALES, J. M. 2009. Telling a research story: Writing a literature review, Ann Arbor,
MI, University of Michigan Press.
HU, L., KU, W.-S., BAKIRAS, S. & SHAHABI, C. 2013. Spatial query integrity with Voronoi
neighbors. IEEE Transactions on Knowledge and Data Engineering, 25, 863–876.
HYLAND, K. 1999. Academic attribution: Citation and the construction of disciplinary knowl-
edge. Applied Linguistics, 20, 341–367.
PEDERSEN, C. B., BUHL, T. & SIGMUND, O. 2001. Topology synthesis of large-displacement
compliant mechanisms. International Journal for Numerical Methods in Engineering, 50, 2683–2705.
SATOH, T., KAWASE, H., IWATA, T., HIGASHI, S., SATO, T., IRIKURA, K. & HUANG,
H.-C. 2001. S-wave velocity structure of the Taichung basin, Taiwan, estimated from array
and single-station records of microtremors. Bulletin of the Seismological Society of America, 91,
1267–1282.
SWALES, J. M. 1990. Genre analysis: English in academic and research settings, Cambridge, Cambridge
University Press.
SWALES, J. M. 2004. Research genres: Exploration and analysis, Cambridge, Cambridge University
Press.
SWALES, J. M. & FEAK, C. B. 2012. Academic writing for graduate students: Essential tasks and skills
(3rd edition), Ann Arbor, MI, University of Michigan Press.
WEISSBERG, R. & BUKER, S. 1990. Writing up research, Englewood Cliffs, NJ, Prentice Hall.
WEN, Y. H. & HUA, C. C. 2009. Chain stretch and relaxation in transient entangled solutions
probed by double-step strain flows. Journal of Rheology, 53, 781–798.
Chapter 8

Evidence from current research

8.1 From past to present

Presenting known information before new information assists comprehension. This is
a basic principle of communication. Thus citations of past research in the Introduction
remind us what is already known by a research community, while new data from the cur-
rent research are presented in the Results or Testing division. The transition between the
two comes somewhere in the second division.
In science there is often a sharp transition at the beginning of the Methods divi-
sion. Although the methods used may be similar to past work, the focus is the present
research. In engineering the transition point may differ and the border may not be so
clear. The position and clarity of the break seems to depend on at least two factors:
(1) the overall structure of the article, and (2) how similar the new design is to previous
solutions.
Sometimes in IPTC there is a sharp transition between past and present research at
the beginning of a section, but not necessarily the beginning of the second division.
Frequently the Process division
begins with one or more sections
or subsections on the framework In engineering there
(4) before discussing the pre- may be cycles between
sent work (5) in a new section past and present research if
marked with words like “in our the design includes several
approach.” different elements.
Other times there is more than
one transition point. Often there
are several new design elements.
In this case the previous best design for each may be summarized before describing
how it was modified in the present work, leading to a cyclical repeat of past and
present work.
Occasionally in IPTC the Introduction is very short, only one or two paragraphs,
merely establishing the importance of the work and the research goal, with all the cita-
tions of past work in the Process division. Other times the Introduction may take several
pages and include extensive detail not only on the connection between past and present
work, but also a summary of the present design.
96  Argument structure in exemplar articles

If the new design is very similar to earlier solutions, particularly in the case of extension
of previous work by the same author, the transition between past and present may not be
clearly marked, making it much harder to find.

Exercise 8.1 Past to present research

(1) Is there a clear break between past and present research? If so, at what point and what
words mark the break, if any?
Hint: In science the break frequently occurs at the beginning of the Methods division. In engi-
neering, look for markers such as “we propose,” or “in this paper.”
(2) Is there only one break point between past and present, or are there cycles?
Hint: This will help you find the break
point(s) between framework and research
Evidence from the present
details.
research includes not only the
data, but also details of how
the research was conducted 8.2 Evidence from
and the data collected. current research
When we think of scientific evi-
dence, we normally focus on the
data collected, but for the implicit
claims of the second division (that the data were collected in a reliable way or the design
is workable), the evidence is a detailed description of the procedures followed. Without
this prior evidence, the data are worthless.
Figure 8.1 lists evidence related to the current study that support the claims of a
research article. The first may occur anywhere throughout an article, especially in papers
with mathematical formulas. The next four categories support claims in the Methods or

II. Present research

a. Numbered or bulleted examples or equations
b. Details and sources of materials
c. Details of present research design and procedures
d. Details of new design, intermediate testing of parts
e. Details of proof, lemmas, mathematical argument, algorithm
f. Detailed description of data collection and testing procedures
g. Data summary or highlight in text
h. Data patterns visualized in graphics
i. Data comparison, sometimes using statistical test results

Figure 8.1  Common evidence from present research.

 Evidence from current research  97

Process division by giving a detailed description of the current research (5). The sixth sup-
ports the claims of the Testing component (6).1 The final three, related to data pattern (7)
and comparison (8), are found primarily in the last two divisions. Each will be described
in the following sections.

8.3 Examples and equations Bullets or indenting make

Numbered examples are common in the examples or equations easier
social sciences but rare in natural science. to find, while numbers make
In engineering an analogous feature is them easier to reference.
numbered equations. These can be refer-
enced by number later in the paper or in
future papers.
In other cases examples or equations are not numbered, but are formatted to stand
out from the text, on a separate line or as bullet points. In social science the examples
are frequently used as evidence in the Results or Discussion. In engineering, equations
are more common in the Process, but may also appear as explanations in the Testing
division.
Examples and equations are similar in format, if not purpose. Text is usually either
indented and bulleted or centered and numbered sequentially along the right margin, in
a format specified by the journal.

Exercise 8.2 Example and equation type and location

(1) Are there any numbered, bulleted, or otherwise highlighted examples or equations in
your exemplar article? If so, in what division and in support of which component?

8.4 Evidence from

research details Detailed descriptions
identify Components 5
A description of the present research in the
second division comprises Components 5 and 6, while the claims
and 6. While the claim of this division is themselves are
usually implicit, the detailed descriptions often implicit.
used as evidence are easy to find, so rela-
tively little needs to be said about them.
In science, to prove that the data are reliable, an author must show that the materials
were good quality, which is confirmed by listing the sources (II b). Even when follow-
ing an accepted protocol, attention to detail (II c) shows that no errors were made in the
procedure. Detail of the testing procedure used (II f) is one of the most important parts of
the paper for reviewers in science.
In engineering, an author will undoubtedly purchase materials and use some estab-
lished methods, but the details of the new design (II d) or mathematical procedure (II e)
are far more important. As discussed in Chapter 6, the Testing method (II f) is likely to
be found in the Testing division.
98  Argument structure in exemplar articles

8.5 Data in text and graphics

In the Results or Testing division,
some data are summarized or high-
Graphics includes all the lighted in the text (II g), but graphics
tables and figures in a (II h) make quantitative data patterns
document, whereas a graph easier to see. Graphics is a collective
is one type of figure. term for all the non-text information
in an article, particularly tables and
figures. It is not the same as a graph,
which is one type of figure.2
There are differences from journal
to journal in both the type of graph-
ics used and their placement. Graphics
Tables and graphs are attract attention and are one of the
frequently found only in the first things potential readers look at.
Results in IMRD, but in IPTC Reviewers also evaluate the quality and
illustrations are common in usefulness of graphics when reviewing
the Process division. a manuscript.
Numerical tables are frequently used
in both science and engineering to
summarize data, while non-numerical
tables, comparing categories or words, are rare. Graphs are the most frequent figures in
scientific literature. Tables and graphs are most common in the Results or Testing, used
to summarize data and show patterns or comparisons. Few are found in other divisions
of IMRD.
In engineering other types of illustrations are also common, including schematic
diagrams, photographs, computer-generated images, and pictorial models of abstract
concepts such as waveforms. These are often found in the Process division, or even the
Introduction, diagramming a new design, particularly if there are many distinct ele-
ments. For example, one of our sample engineering articles has two illustrations in the
Introduction and five in the Process, in addition to 11 graphs in the Testing division
(Hu et al., 2013).

Exercise 8.3 Graphics type and location

(1) Create a table showing the numbers of tables and figures in each division in the first
exemplar article, separating them by type if useful.
Hint: Show where the graphics are first mentioned in the text, not where the graphic appears in
the document.
Hint:You may find it useful to make separate categories for different types of tables and figures
in your article.
Hint:What table arrangement would show this best? How will you incorporate information from
your other two exemplar articles?
 Evidence from current research  99

(2) Write a brief description, calling attention to any important features of the table, for
example, “Over half of the graphics are in the Process division, with schematic diagrams
being the most common type, as shown in Table x.”
Hint:These descriptions will help you learn to summarize important features of data, an impor-
tant skill in scientific and technical writing.

8.6 Data comparison and statistics

Like graphics, statistical tests (II i) are associated with the data, often in the Results but
sometimes in the Discussion. They are far more common in science, explaining phe-
nomena in the natural world, than in engineering, with its goal of solving a problem,
often in a highly constrained environment where variation is assumed to be negligible.
Nevertheless, some engineers use statistical tests in comparing the present to past methods,
or when dealing with real-world applications.
While full coverage of statistics is beyond the scope of a writing textbook, their misuse
is so common that a warning is required. If you use statistics, learn how to do them and
write them up properly. Take a course, read a text, talk to experts, and understand why
they are used, not merely how to plug numbers into a program. Even if you do not use
them in your own research, if statistics are required in your research field you must know
how to interpret articles you read. The number of statistical analyses in the literature that
are either incorrect or poorly written is truly astounding. Please don’t add to it.3

8.6.1 Types of statistics

There are two basic types of statistics: descriptive and inferential. Descriptive statistics are
safe but of limited utility. They provide summaries of the current data, but no informa-
tion on how those compare with other values. Statistical testing, on the other hand, refers
to inferential statistics, in which the values measured are compared with other values or
inferences made about a larger population based on the sample measured. They are tricky
because there are many different tests based on the type of data and how they were obtained.
Descriptive statistics, common in both science and engineering, are a way of organizing
and summarizing data, either numerically or graphically. The range, mean, median, vari-
ance, and standard deviation of a data set are descriptive statistics. Data can be displayed as
bar graphs, histograms, box and
whisker plots, or scatter plots.
Correlation and least-squares
Statistical testing is common in
regression lines are also descrip-
tive, because they are mathemat- science but rare in engineering.
ical descriptions of a data set, but When conducted, it is imperative
cannot be compared with other that details of the experimental
data sets without further tests. design be included.
Inferential statistics are based
on hypothesis testing, using
probability distributions to
approximate random variation. Tests include confidence intervals, equality of means
or the difference between means, goodness of fit, and analysis of variance. Inferential
100  Argument structure in exemplar articles

statistics are used in quantitative research to allow inference from the experimental
results to a larger population or situation, and thus are far more common in science than
engineering. They allow a researcher to estimate how much of the variation in experi-
mental results is due to a real difference between treatments and how much is random
variation due to environmental conditions or errors in measurement. Whenever statisti-
cal tests are performed it is vital to explain not only what tests were done and the results,
but exactly how they were done.

8.6.2 Experimental design

How statistical tests were done is called experimental design, arguably the most important
part of reporting statistics but the most often neglected by novices. It allows reviewers and
readers to see if the methods you followed justify the claims you make about the results.
It is important to report details such as:

1 how the test subjects were chosen and how treatments were assigned to them,
2 the experimental (independent) and response (dependent) variables and how each
was manipulated or measured,
3 the number of times the test was repeated and degrees of freedom, and
4 the statistical package or procedure used to compare the results.

Many statistical terms have a tech-

The terms experiment, nical meaning which is slightly differ-
hypothesis, and prediction ent from their use in common speech,
leading to misuse in the science and
have both a common and engineering literature. Three of the
technical meaning, which most common related to experimen-
can lead to confusion. tal design are experiment, hypothesis,
and prediction.4
An experiment, by modern scien-
tific (not common) definition, involves
changing the value of an experimental (independent) variable and measuring the effect on
a response (dependent) variable, while holding all other variables constant. Usually it also
involves repetition of each test to eliminate the effect of random variation in any real-world
measurement. Many tests reported as “experiments,” particularly in engineering, are not
properly experiments. Misuse of the term perpetuates and exacerbates science illiteracy. In
your own writing I encourage you to differentiate experiments from other types of testing,
even if other authors do not.
Most experiments are designed to test a prediction, which is not the same as a
hypothesis, a common misunderstanding that originates from the term hypothesis test-
ing. Hypothesis testing involves making a prediction, a statement of what you expect
to observe if a proposed hypothesis is true. Sometimes this is stated as an “if . . . then”
expression: if my hypothesis (of the proposed explanation of the underlying mechanism)
is correct, then my prediction (of what I will observe) will be correct.5 Note that the “if”
clause should be a statement of the hypothesis, not of what you do in the experiment. Let
me illustrate by comparing the following examples:
 Evidence from current research  101

• If I add a nitrogen-containing fertilizer, then a plant will grow faster or stronger;

• If nitrogen is a required element for plant growth, then adding nitrogen-containing
fertilizer will help a plant grow faster or stronger.

The first statement makes a prediction of what will happen if I add fertilizer, but it is not
a hypothesis because it does not explain why the plant will grow. The second statement
properly pairs the hypothesis that nitrogen is a required element with the prediction
that it will help the plant grow. The hypothesis in turn is based on an underlying model
or understanding of how the factor being changed is related to the one being measured.

Exercise 8.4 Statistics type and location

(1) Are statistics used in your first exemplar article? If so, are they descriptive or inferential?
In which division(s) do they occur?
(2) If they are inferential, do they provide all the necessary experimental design details?
Hint: Normally experimental design will be reported in the Methods in science, but is often
absent or incomplete in engineering.

8.7 Pointing verbs

Since the data are often shown in graphics, the author must call the reader’s attention
to that information. Convention requires the author to refer to each graphic at least
once in the text of the arti-
cle, often using pointing verbs.
Pointing verbs can direct the There are various ways to point
readers’ attention to a graphic or to graphics, with and without
highlight certain information in a
graphic. These two functions
pointing verbs, but the present
are known as locating and passive of show (as [is] shown) is
highlighting, respectively. 6 quite common in engineering.
There are several ways of
directing your readers’ atten-
tion to graphics, depending in
part on the style of writing. A topic sentence may point to the location of the data in a table
or figure, with details the reader should observe in subsequent sentences. In many cases,
locating and highlighting functions will both be found in the same sentence. The graphic
may also be referred to in parentheses, without the use of a pointing verb. Examples are
shown in Table 8.1.
Table 8.1 also illustrates another way sentences can be varied: by using active or passive
voice. For the “locate only” category, the first example illustrates active voice. Properly,
this sentence is incorrect, because a table cannot be an active agent and therefore cannot
illustrate anything. Nevertheless, most journals accept this as a shorthand way of saying
that the author illustrated the difference by composing the table. Some authors or gram-
matically inclined editors prefer to use the passive, as illustrated in the second example.
This is a correct statement, with the agent, the author, unstated.
102  Argument structure in exemplar articles

Table 8.1  Examples of different ways of referring to graphics

Category Examples

Locate only (1) Table 8.1 illustrates the difference between the three methods. In
the simplest form only the location is given by the pointing verb.
(2) Common pointing verbs are presented in Table 8.2. For most the
author is the implied agent, the one doing the presenting.
Locate and (1) Location and highlighting can both occur in a single sentence, as
highlight shown in the second category in Table 8.1.
(2) As can be seen from Table 8.2, the reader is the agent for a few
pointing verbs, which also require a modal.
No pointing (1) In summary, it is clear that there are several different ways of
verb referring to graphics (Table 8.1).
(2) By far the most commonly used pointing verb in electrical
engineering is “show” (Table 8.2).

When location and high-

lighting occur in the same
The present tense is used when sentence the passive voice is
referring to graphics because often used, with the refer-
although the work was done in the ence to the graphic placed
in a subordinate clause, as
past, the graphic is visible to the shown in the second cat-
reader in the present. egory in Table 8.1. Note
that all pointing verbs are in
the present tense, whether
active or passive voice.7
Even though both the gathering of data and production of the graphic occurred in the
past, the data are currently being shown in the graphic, and can be seen by the reader in
the present.
It is also possible to alter the sentence by changing the agent, as can be seen in the
second example of location and highlighting in Table 8.1. In the first three examples, the
implied agent is the author. By saying, “as can be observed,” the action is shifted to the
reader. This requires a whole different set of verbs, as shown in Table 8.2. It also requires
use of a modal (usually “can”), since the reader must choose to look at what is illustrated.8
In addition to pointing to graphics, verbs can also point to information in a different
section. Different verbs tend to be used for this purpose. Since they point to text rather
than graphics, it is more common to use verbs such as present or describe, as shown in the
last row of Table 8.2.

Exercise 8.5 Pointing verbs

(1) Look at each sentence in your exemplar article containing a reference to a graphic.
Underline the pointing verbs and list the most common, separating those that point
to tables from those used for figures.
Hint: There are many ways this could be presented. Is there any difference based on the type
of graphic, or from division to division? Is number of occurrences important, or just rank order?
 Evidence from current research  103

Table 8.2  Pointing verbs common in electrical engineering*

Category Pointing verbs

Author agent (table) Show

Summarize
List
Give
Present
Author agent (figure) Show
Illustrate
Depict
Present
Plot
Reader agent (table or figure, with modal) See
Observe
Author agent (section) Present
Describe
Discuss
Show

*Rau (unpublished data).

(2) Count the number of null occurrences—times where a reference to a graphic is not
accompanied by a pointing verb.
Hint: Look for references to graphics in parentheses.
(3) Are there any places where pointing verbs are used to point the reader to a different
section of the paper? If so, what verbs are used?
Hint:You can find these by searching the article for “section.”

Evaluate your understanding

(1) Why does science tend to switch suddenly from past to present evidence at the begin-
ning of the Methods division, whereas the two may overlap extensively in the Process
division in engineering?
(2) What is the purpose of highlighting examples or equations with numbers or bullets?
(3) How do science and engineering differ in the location and type of graphics, and why?
(4) If statistical tests are conducted, what information needs to be included in describing
the experimental design?
(5) What is the difference between locating and highlighting graphics?
(6) Why is present tense always used with pointing verbs?

Homework
(1) Repeat the Exercises in this chapter for your other two exemplar articles.
Hint: Consider different ways of combining and presenting the information. What would make
it easiest for a reader to understand the general pattern?
104  Argument structure in exemplar articles

(2) Summarize what you have learned about the use of evidence from present research
in your exemplar articles.
Hint: Compare your results with other students.You may be surprised how much difference there
is from one journal to another, even within the same field.

Teaching tip: This chapter can be taught with either the second or third division, since this
evidence is important in both. Related information can be found in Chapter 24 (graphics).

Notes
1 Evidence for the four sub-components of Component 5 are listed separately because each type of
evidence is very different, while only one summarizes all the testing methods because no matter
what type of testing is conducted, the procedures that must be described are similar.
2 The two letters separating graph and graphic make a vast difference. This is difficult for those
working in English as a lingua franca, but there does not seem to be a better collective term.
3 Full disclosure: Although I am not a statistician, I have a personal stake in the proper use of
statistics. The main contribution of my PhD research was proof that a statistical method used
by 15 papers over 30 years underestimated gene number by at least an order of magnitude.
Although I realized the problem, it took an expert statistician to determine the correct meas-
ure. Since then I have detected numerous statistical errors in other articles, either during review
or after publication.
4 Two others, significant and correlation, will be discussed in section 23.3 in the context of word
choice.
5 This has been called hypothetical inferencing. “If . . . then” clauses are used for different purposes
in other fields, often with an “if (evidence) . . . then (claim)” structure, for example logical deduc-
tion in philosophy, legal consequence in law, and persuasive interpretation in English, with or
without explicit statement of the reasoning (Nesi & Gardner, 2012: 118–119).
6 See Cargill and O’Connor (2013: 34).
7 Some students seeing a past participle ending in –ed or a modified form like “shown” will auto-
matically identify it as past tense, but in the passive voice the tense is identified by the form of the
verb “to be” associated with the past participle, so “is presented” is present and passive, whereas
“was presented” is past and passive. In present passive with the conjunction “as,” the verb “to be”
may be omitted, “as (is) shown in . . .,” making it even more confusing, since there is no apparent
present tense in the construction.
8 This expression is properly called a passive modal simple.

References
CARGILL, M. & O’CONNOR, P. 2013. Writing scientific research articles: Strategy and steps, New
York, John Wiley & Sons.
HU, L., KU, W.-S., BAKIRAS, S. & SHAHABI, C. 2013. Spatial query integrity with Voronoi
neighbors. IEEE Transactions on Knowledge and Data Engineering, 25, 863–876.
NESI, H. & GARDNER, S. 2012. Genres across the disciplines: Student writing in higher education,
Cambridge, Cambridge University Press.
Chapter 9

Clear reasoning

9.1 Follow the yellow brick road

Three hundred years ago the German philosopher Gottfried Leibniz wrote, “There
are also two kinds of truths: those of reasoning and those of fact. Truths of reasoning
are necessary and their opposite is impossible; truths of fact are contingent and their
opposite is possible” (Leibniz, 1898: 235–236). Although the principle remains valid,
the meaning of words has changed so much since the English translation of 1898 that
it may require explanation. Truths of “reasoning” (those based on mathematical logic)
are “necessary” (always true). Thus the Pythagorean theorem derived 2,500 years ago,
like all other mathematical proofs, is still true today. On the other hand, with truths of
“fact” (what we now call empirical truth), the “opposite” (a totally different interpretation)
is possible. How that interpretation is done is the focus of this chapter.
Ideally, by following the reasoning in an article, all readers should be able to reach the
same conclusion about the evidence that the author has. As an analogy, when Dorothy
asks how to get to the Wizard of Oz, the Munchkins reply, “Follow the yellow brick
road.” If you stay on the road, you will reach your destination. The same should be true
in an academic article. There should be no alternative paths (as Dorothy faced when she
met the scarecrow), gaps to jump, or sharp turns that might be missed.
This chapter is by no means an exhaustive treatment of reasoning and organizational
strategies, but will point out some features commonly found in science and engineering
research articles. As shown in Figure 9.1, reasoning involves different types of logic and
evaluation. There are also often signposts to help readers follow the argument.

9.2 Logical reasoning

Three types of logical reasoning
are commonly recognized in
empirical research: deductive, Deductive logic tests a hypothesis,
inductive, and abductive. Most inductive develops a new model,
scientific research is deduc- and abductive identifies the best
tive (III a), testing predictions of several explanations. All are
made by a hypothesis or theory. common in science.
Inductive research (III b) is the
opposite, combining numerous
lines of evidence into a new
106  Argument structure in exemplar articles

Support: Reasoning
III. Logical
a. Deductive: Testing predictions made by a general theory
b. Inductive: Compiling observations to reach a general conclusion
c. Abductive: Choosing the most likely of several possible explanations
IV. Evaluative
a. Evaluation of previous research
b. Evaluation of explanations
c. Evaluation of current design
V. Organization
a. Initial summary to reveal argument structure of subsequent section(s)
b. Reminder of the research goal
c. Concluding summary of argument in preceding section(s)
d. Referring to arguments made in other sections

Figure 9.1  Common reasoning strategies.

model or theory. Abductive (III c) is sometimes called reasoning to the best explanation,
choosing the most likely of several possible explanations for the data.
When developing the IPTC model presented in this book, I used inductive reason-
ing. I combined observations from many articles in different fields of engineering, trying
to identify common components. In doing your component analysis, you are practic-
ing deductive reasoning. Based on the overall structure of your exemplars, you stated a
hypothesis that articles in your field follow either IMRD, IMRaDC, IPTC, or IPC for-
mat. Since then you have been testing that hypothesis, seeing whether the articles fit the
expectations of that model or not. They may not, depending on both the research meth-
odology and the individual author. As you try to assign components to each paragraph,
you may engage in abductive reasoning. It may not be clear, for example, whether a
paragraph at the beginning of the Process division should be called framework or research
details, and you will have to choose based on the preponderance of evidence, even if one
sentence does not seem to fit.
Generally logical reasoning is more important in science, with its focus on explanations
of the data, than engineering. Each article usually can be identified as primarily one type
of research, although a longer article or book could include more than one. Articles that
perform hypothesis testing, most science articles, follow a deductive approach. Those that
seek to develop a new model use an inductive approach. Observational studies that begin
without a hypothesis may follow an abductive approach.

9.3 Evaluative reasoning

For both science and engineering, previous research is evaluated in the Introduction (IV
a). In science the framework is often revealed by the author’s attitude toward past research.
Engineering identifies a need based on evaluation of the current best solution, noting
 Clear reasoning  107

limitations or weaknesses. Sometimes the

order of presentation reveals the evalua- Evaluation of designs in
tion as well, with the method most simi-
lar to the current research, with the least
engineering often requires less
limitations, mentioned last. discussion than evaluation of
In later divisions, whereas science uses explanations in science.
data and statistics to test a hypothesis to
determine whether an explanation is cor-
rect (IV b), engineering evaluates solutions
in different situations to determine where each has a competitive advantage (IV c). Since
there are objective measures of speed, weight, and other factors important in engineering,
the evaluation often does not require the same type or degree of discussion or reasoning as
science articles. The simple statement A<B may suffice in engineering, but science might
require a statistically significant difference, a reasonable explanation for the cause of the dif-
ference, and elimination of other possible causes.

9.4 Extended vs. condensed style

To help readers follow the argument, authors often leave signposts to point the way.
Two of the most common are introductory paragraphs and topic sentences. Using an
analogy familiar to most readers, introductory paragraphs are like a GPS unit telling you,
“In 100 meters, turn left,” whereas topic
sentences say, “Turn left.” They occur
at predictable points and help the reader
navigate the text. On the other hand, a
Extended style is often used
driver familiar with the route does not when introducing new material,
need the navigation, and may find it but condensed for material
annoying and turn it off. familiar to most readers.
The two situations are analogous to
extended and condensed styles in aca-
demic writing.1 The former guides you
through but can provide too much information if you know the territory. Thus extended
style is common in sections introducing new material, while condensed is used for mate-
rial familiar to most readers.

9.4.1 Characteristics of extended and condensed style

Differences between the two styles are summarized in Table 9.1. A plus sign (+) means an
indicator is present, while a minus sign (–) means it is not present. The two easiest features
to recognize are introductory paragraphs and topic sentences.
Although intermediate styles exist, many sections are relatively easily to categorize.
The Methods division in IMRD and sections summarizing the framework in IPTC may
be written in condensed style, since readers should be familiar with that material. Most
other sections are likely to be written in extended style. In mathematics format, there
may be a paragraph at the beginning of each section in extended style to explain why
the procedure is needed, but the rest of the section will be in the highly condensed style
typical of proofs.
108  Argument structure in exemplar articles

Table 9.1  Variation in styles of scientific and engineering writing

Indicator Extended Condensed

Introductory or final summary paragraphs + –

Topic sentences + –
Details: definitions, examples, descriptions, process + –
Purpose, justification + –
Background knowledge assumed – +
Acronyms and abbreviations – +
Listing or citation without details – +

Adapted from Swales and Feak (2004: 227)2

In extended style a sec-

tion may begin with an
Extended style uses summary introductory paragraph, and
paragraphs, topic sentences, each paragraph with a topic
repetition, and explicit statement of sentence. By reading only
the first paragraph of each
purpose or justification, making it
section and the first sen-
easier to skim and read. tence of each following par-
agraph, a reader can quickly
grasp the overall thrust of
the article. Extended style
tends to use short sentences and paragraphs, with one main point per paragraph and one
important idea per sentence, another technique that makes it easier for the reader. It also
adds non-essential words to make the argument clearer, such as examples or indicators of
purpose like “in order to.” Repetition of key words from one sentence to the next helps
tie the thoughts together.
Condensed style, on the other
hand, lacks summary statements or
Condensed style assumes repetition. The first sentence in a
a great deal of background section or paragraph plunges directly
knowledge, including into details. Ideas are presented lin-
familiarity with previous early, using long sentences with many
clauses and internal connectors (and,
research and terminology.
but), but fewer linking words or
phrases between sentences (further-
more, despite). Several main points
may be presented in each paragraph. It is impossible to skim for the overall meaning
because each sentence addresses a different topic. Extensive background knowledge is
assumed: both terminology and the content of previous papers.

9.4.2 Summary paragraphs

Summary paragraphs may be found at either the beginning or end of a section, and may
serve at least four purposes, listed as the four subpoints of Organization (V) in Figure 9.1.
 Clear reasoning  109

None are essential to the argument, but they ensure your writing is well organized and
easy to follow.3
In many IPTC articles, the last paragraph of the Introduction section describes the
organization of the rest of the paper (V a). This is also common in social science but rare
in natural science. The paragraph begins with the following sentence pattern:

• This paper is organized as follows. In section II we . . . 

It is also quite common to have a paragraph at the beginning of a section in the Process
or Testing division that provides an overview of that section (V a), sometimes as a single
paragraph before numbered subsections. This often includes the words “this section.”
It may also include words like “steps” or a series of sentences that indicate sequence, as
in the last two examples below. A similar paragraph may be found at the beginning of a
subsection. Examples of this sentence pattern include:

• In this section we present a brief introduction to this model.

• This section describes an overview of our proposed method.
• Our approach consists of three main steps.
• The two main components of our system . . . 
• Section III-A introduces . . . Section III-B describes . . . 
• In this section we present . . . We also . . . Finally . . . 

As mentioned in Chapter 6, it is not uncommon in IPTC to restate the research goal

at the beginning of the Process or Testing division (V b). This helps readers who may skip
to a section of interest. The following examples restate the research goal:

• The goal of our work is to . . . 

• Our research aims to provide . . . 

Concluding summaries are common at the end of a section (V c) in the social sciences,
but not in science and engineering. They rarely occur at the end of the Testing division
to pull together different types of evidence.
It is also common in engineering to refer to something explained in a different section
(V d), either earlier or later, as in the following examples:

• As previously stated in Section II, . . . 

• We apply the . . . models proposed in Section III . . . 
• We will discuss the impact of D in more detail in Section V.
• We discuss this issue further in Section 7.1.

Exercise 9.1 Identifying extended and condensed styles

(1) Does your first exemplar article use extended style, condensed style, or different styles
in different sections?
Hint: Look primarily for summary paragraphs and topic sentences.
110  Argument structure in exemplar articles

9.5 Sequencing strategies

Even within the expected structure of a research article, there are many ways of sequenc-
ing the same argument.4 Which should come first, claim or evidence? Which evidence
should be presented first? Present all the evidence before discussion, or alternate results and
discussion? These are the questions that make writing a continuing challenge. Sometimes
the right words come easily, but often even experienced writers will try several differ-
ent orders before settling on one that seems to make the strongest argument. As Ernest
Hemingway commented, “There is no rule on how to write. Sometimes it comes eas-
ily and perfectly; sometimes it’s like drilling rock and then blasting it out with charges”
(Hemingway, 2013).
There are several basic sequencing strategies, which can be combined either linearly or
recursively (cyclically). Unlike components, evidence, or even the summaries discussed in
the previous section, sequencing strategies are not identified by markers but by the order
of components.

9.5.1 Basic strategies

At least four different sequencing strategies are commonly used in scientific and engineer-
ing writing, which I will call physical, temporal, relational, and mathematical. The first,
physical, describes a topic in terms of its type, location, or parts. The second, temporal,
traces a topic chronologically or by following the steps in a process, which are often but
not always the same. For example, if two processes are happening at the same time, it
would be possible to describe the order of events in both (chronological) or first describe
the steps in process 1, then process 2. The third strategy, relational, involves organizing in
terms of importance or some logical relationship. The fourth, mathematical, refers to the

Physical
• Type (equipment vs. consumables, hardware vs. software)
• Location (physical arrangement)
• Elements (parts in product)
Temporal
• Chronological (time sequence)
• Order (steps in process)
Relational
• Importance (priority)
• Comparison (evaluative)
• General to specific (deductive)
• Specific to general (inductive)
Mathematical
• Mathematical proof
• Algorithm description

Figure 9.2  Common sequencing strategies.

 Clear reasoning  111

logical order of argumentation common to mathematics; for example, the order used for
a mathematical proof or description of an algorithm. The various strategies and subtypes
are illustrated in Figure 9.2.
Sometimes one ordering strategy is obviously best, but frequently several different
orders could make sense. For example, in the Materials and Methods division, methods
are a process, often calling for a temporal ordering, but materials could be listed by type,
by order of use, or by importance. There may be a de facto standard strategy in your field
that can be determined by analysis of your exemplars or the author may have freedom
to choose.

9.5.2 Nested or cyclical strategies

Often two sequencing strategies are used Sequencing strategies are
in the same section. In this case, one is frequently nested, leading
nested inside the other, leading to either
a linear arrangement of the components
to cyclical repetition of
or a cyclical repetition. The nesting strat- common components.
egy is often revealed by the section and
subsection titles.
For example, in a traditional Materials
and Methods section, Materials are described first, followed by Methods, then Testing
methods.5 If two types of tests are being performed, a second level of sequencing is
required, but two nested strategies are possible, as illustrated in Figure 9.3, where the
numbers indicate component numbers. The first leads to a linear sequence of the com-
ponents, but in the second that sequence is repeated twice.
Nested alternatives can also be found in the Results and Discussion or Testing division.
All the testing methods could be listed first, followed by all the results, then all the discus-
sion (classic IMRD format). Alternatively, the testing method, results, and discussion of
test 1 could be presented first, then the same for test 2 (common in IPTC format). The
difference between the two is illustrated in Figure 9.4, where the numbers again indicate

5a Materials Test #1
Test #1 5a Materials
Test #2 5b Methods
5b Methods 6 Testing
Test #1 Test #2
Test #2 5a Materials
6 Testing 5b Methods
Test #1 6 Testing
Test #2

Figure 9.3  Nesting strategies for Materials and Methods in IMRD.

112  Argument structure in exemplar articles

IMRD IPTC IMRaDC

6 Testing (M) Test #1 (T) 6 Testing (M)
Test #1 6 Testing Test #1
Test #2 7 Results Test #2
7 Results (R) 8,9 Discussion Test #1 (R&D)
Test #1 Test #2 (T) 7 Results
Test #2 6 Testing 8,9 Discussion
8,9 Discussion (D) 7 Results Test #2 (R&D)
Test #1 8,9 Discussion 7 Results
Test #2 8,9 Discussion

Figure 9.4  Nesting strategies in IMRD, IPTC, and IMRaDC, by component and division.

component numbers and the letters in parenthesis the divisions. The second column is
called a cyclical or recursive pattern, as the same components are repeated in a loop-
ing sequence, as opposed to the first column, where each component occurs linearly,
although two different tests are discussed for each. A more complex pattern common in
IMRaDC is shown in the third column. Like IMRD, the testing method is described
in the Method division, but like IPTC there may be a cyclical repetition of components
related to the Results and Discussion.
The sequence chosen
will depend on the format
required by the journal,
Presenting topics in the same the topic and purpose of
order in different divisions facilitates the author, and which will
reading. Results often drive make the strongest or clear-
est argument. Generally,
the organization in IRMD, but
following the same sequence
Process in IPTC. in the Process and Testing
(or Methods, Results, and
Discussion) makes it easier
for the reader to follow. In
IMRD the sequence will usually be determined by the order of presentation of data in
the Results, while in IPTC the order in the Process is more important, as is fitting given
the prominence of those respective divisions in each format.
For comparisons and interpretations it is common to use a relational strategy, because it
is necessary to integrate various types of data. Since so many different orders are possible,
this is one of the most challenging parts to write.
The Introduction is somewhat distinct from the rest of the paper. It seems to be more
standardized across fields, with the importance of the work followed by the need for the
 Clear reasoning  113

work, then the research goal. The last three divisions focus on your research, and thus
are quite variable based on both the field and the research methodology (see section 2.4).
Nevertheless, there are differences in the Introduction as well. The importance of
the work frequently moves from general to specific, but needs to show the relation-
ship between past and present research and therefore uses a relational strategy, with the
approach closest to the present work often mentioned last. As we all know, relationships
can be complicated and take many different forms, so there is far more variety in this
organizational strategy. This is one of the reasons the Introduction is so difficult to write.

Exercise 9.2 Sequencing strategies

(1) What sequencing method seems to determine the order of the Methods or Process
division in your first exemplar article? Is it linear or cyclical?
Hint: Look at the order of components you identified.
(2) Is this same sequence followed in the Results and Discussion or Testing?
Hint: Remember that more than one ordering scheme may be used, either nested (e.g., time
sequence within each part) or different schemes for different purposes (e.g., part sequence when
talking about the design, time sequence when talking about the test results).
(3) Can you speculate on why that order may have been chosen?
Hint: “No” is not an acceptable answer. Thinking about why it makes sense for this paper will
help you decide whether the same method will make sense for your own writing.

Teaching tip: See class exercise 9.1.

Teaching tip: This section is more challenging for students than the basic analysis of components.
It may be skipped without loss of continuity or covered in an advanced class only. It is also possible
to simply identify linear and cyclical strategies without considering the reason the order was chosen.

Evaluate your understanding

(1) Describe the difference between deductive, inductive, and abductive reasoning in
scientific research.
(2) How does evaluation differ in engineering and science?
(3) What are the main differences between extended and condensed style, and why
would an author choose to use different styles in different sections?
(4) Why are there sometimes cycles of the same components?

Homework
(1) Repeat the exercises in this chapter for the other two exemplar articles.
(2) Summarize the reasoning and sequencing strategies common in your field.
Hint: Consider where to place this information in your overall description of exemplar articles.
Should it be separate or combined with previous information? The order in which you discover
something may not be the best way to present it to an audience.
114  Argument structure in exemplar articles

Teaching tip: Use this as an opportunity to begin to talk about the rewriting process, organ-
izing items by where they make the argument easiest to follow rather than following either a
prescribed structure or the order in which the research was conducted.

Supplemental material: class exercise 9.1—comparison of sequencing strategies.

Notes
1 Terms from Swales and Feak (2004).
2 Swales and Feak also list “subsections” and “connectors or linking words” as evidence of extended
style, but I have found these misleading for students. IMRD format papers often have subsections
only in the Methods, with a very condensed style. The subsections contribute to that style by
stating the content as a heading rather than a topic sentence. Similarly, connectors can take the
place of repetition of words to shorten a passage.
3 Linguists call these metadiscourse, and give specific names to each type: future projections (V a),
recapitulations (V b and V c), and location statements (V d).
4 Perhaps sequencing should be called rhetoric rather than reasoning, but that word has its own
problems. One definition is positive—the art of effective or persuasive speaking or writing, but
some consider it to have a negative connotation—persuasion with the intent to mislead or lack-
ing evidence.
5 In science, Methods may refer to preparation of the study material, whereas Testing methods
would be the measurements taken and statistical procedure employed.

References
HEMINGWAY, E. 2013. The Cambridge edition of the letters of Ernest Hemingway, Cambridge,
Cambridge University Press.
LEIBNIZ, G. W. F. V. 1898. The monadology and other philosophical writings, Oxford, Clarendon
Press.
SWALES, J. M. & FEAK, C. B. 2004. Academic writing for graduate students: Essential tasks and skills
(Vol. 1), Ann Arbor, MI, University of Michigan Press.
Part 3

Exploring different genres

Chapter 10

Argument structure in other types

of writing

10.1 Variations on a theme

By now you should understand the argument structure of research articles in your field.
At the beginning I told you it would be hard work but would pay rich dividends when it
came to your own writing. Now you are ready to reap the reward.
There are many different
genres of academic writing, but
all are variations on a theme. All academic writing is
Like research articles, all aca- structured as an argument,
demic writing makes a claim so the principles learned from
or series of claims, sometimes
explicit but often implicit, sup- studying research articles can be
ported by evidence and rea- applied to any writing.
soning. Perhaps you are ready
to write a research article, a
conference paper, or your the-
sis. Maybe you are an undergraduate, working on a project for a class. At some point you
may write a grant application. All are arguments. Even a resume is an argument, not a list
of past accomplishments.
In this chapter I will lay out some general principles for academic and technical writing.
In the next five chapters we will consider how to adapt the same ideas to other types of
writing, whether exemplars are available or not. Read whichever chapter is appropriate
to what you need to write first, then skip ahead to Chapter 18 for steps in the writing
process. You can return to other chapters in this part as you are ready to write other
documents.

10.2 Planning to overcome difficulties

The same components found in research articles are found in many other types of aca-
demic writing, although not all to the same degree. Difficulties writing each component
can be overcome with proper planning.
Three of the components—importance, need, and framework—are supported by evi-
dence from past research. Extensive past research. Therein lies the difficulty. How many
articles were cited in your exemplars? Thirty, fifty, one hundred? How many have you
read? That is why it is important to quickly find the information you need from each
118  Exploring different genres

article. From articles you have read you will be able to identify others you need to cite.
You may not have to read the whole article if you know where to look for the informa-
tion you need. On the other hand, make sure that you do not cite an article you have not
looked at yourself, as citations are sometimes incorrect or misleading.
A second difficulty is deciding which past research to cite for each of those three com-
ponents and how to order them. Summarizing an article in one sentence is a challenge.
Stringing the sentences together is
another. Even remembering what
Adding information to your each article says becomes a challenge
writing as you read helps as the number increases.
you connect the article My suggestion is to begin with the
general structure of the argument, as
with known information, shown in section 19.2. Then as you
improving memory. read each article think which com-
ponent or which other research it is
associated with, write a one-sentence
summary, and put it approximately where you think it will go. For example, write both
the main advantage and limitation of a design, and place that sentence in a paragraph with
others pointing to the need for your research. If it should go in two places, for example
the limitations in the Introduction and comparison to your design in Testing, write two
sentences. When you connect the information in one article with other information you
know, it helps you remember it, something that does not happen if you take notes on
each article separately.
After you have several sentences in approximately the right place, rewrite the para-
graph to connect the ideas, reordering as necessary. You may not have written this way
before, but it is effective when dealing with a large amount of information, especially in
the age of computers that makes such revision easy.
The next component, research goal, is a far more important part of a paper than most
novice writers realize. It determines what previous research must be included to show
how your work adds to what is already known in your research community. It also deter-
mines what tests need to be conducted and how those are interpreted to achieve your
goal. In short, it determines the contents of the whole paper. Remember that the goal
needs to be stated in the format expected in your field. An author who writes research
questions when the field prefers research objectives will seem like an outsider and risk
rejection. By taking the simple action of transforming the questions into objectives, the
rest of the paper may be identical but the chance of success far higher.
Introducing the framework presents different challenges in IMRD and IPTC.
In engineering the challenge often is deciding how to distinguish past and present
work. How much of the work you did is identical to what has been done before, and
what is novel to your work? Both must be stated clearly. In science the framework
is often implied by a writer’s stance toward previous work, so care must be taken in
the specific wording.1
For the research details and testing methods it is the details that make writing difficult.
In science the language for writing these is so standardized that many phrases (but usually
not whole sentences) can be copied from one paper to another without fear of plagiarism.
Although this division is not the focus of the paper, details of the experimental design are
crucial if statistics are used. In engineering this division is the heart of your paper, so the
 Argument structure in other writing  119

details are vital. Moreover, differences in

approach limit the use of sentence pat- Writing details of the
terns from previous writing. current research as you do
For either format, because of the
importance of details I suggest writing this
the work is easier than trying
division as you do your work. It is much to remember them all later.
easier to remember and write details as
you do them than several months later.
Rewriting is easier than recalling. Begin
by writing more details than you think necessary. What seems obvious to you will not be
clear to others who cannot see your lab setup.
One of the greatest struggles for students writing in IMRD format is deciding what
to include in the Results, what to place in the Discussion, and how to connect the two.
This struggle is reflected in research and textbooks on academic writing, which show
substantial overlap in what they say should be included in each section.2 Traditionally,
data should be placed in the Results section, with comparisons and interpretations in
the Discussion section. In principle this separates the evidence, considered to be correct,
from the interpretation, which needs to be verified.3 Nevertheless, since this places the
evidence and reasoning in separate sections, neither can stand as a complete argument.
Writing a combined Results and Discussion division and laying out the basic structure of
the argument first will help (section 19.2).

10.3 Goal, audience, criteria

In any writing it is crucial to consider three things: your goal, the audience, and the crite-
ria by which it will be judged. Although a conference paper, a journal article, and a thesis
may have similar content, they are quite different in structure and detail because of those
three factors, so you need to write with a purpose in mind. As the American novelist
Carl Hiaasen said in an interview, “You
can do the best research and be making
the strongest intellectual argument, but if Who will read your work,
readers don’t get past the third paragraph why, and most importantly,
you’ve wasted your energy and valuable
ink” (in Ferguson, 2002).
what criteria will they use
Most things we do have both an to judge whether it is good?
immediate and long-term goal, or at least
short- and long-term impact. Obviously,
if you are a student your immediate pur-
pose is to graduate, and certain writing tasks will be required for that, but what is your
long-term goal? If you are looking for a job in academics, you may choose to write dif-
ferently than if you want to go into industry, focusing more on theoretical aspects instead
of practical. Writing not only showcases your research, but makes you visible, for better
or worse.
Next, how large is your audience, how much do they know about your topic, and
why do they care about it? The primary audience for a thesis or dissertation is your
advisor and committee. They already know quite a bit about your research and you
can guarantee they will read it carefully because that is their job. A conference paper or
120  Exploring different genres

journal article may be addressed to a group of specialists or a broader audience, but you
need to attract and keep their attention if you want your work to be read and cited.
This will affect how much background you need to provide and how much detail you
can or need to include.
Perhaps the most important consideration, often overlooked by first-time writers, is
the criteria by which the work will be judged. A successful thesis must attain a certain
standard set by your department. A conference paper must fit the theme of the conference
before reviewers will consider the quality of the content. A research article is judged by
many criteria including whether it fits the scope of the journal, the breadth and currency
of the references, the novelty and importance of the work, the organization and quality
of writing, and the clarity and necessity of graphics.4
Sometimes you have at least two separate audiences, reviewers and readers, with dif-
ferent criteria. If you do not convince the first audience that your work has merit, the
second audience will never see it. The former will likely be quite aware of previous work
in your specialty and only want to see how your work builds on it. On the other hand,
some of the second audience will be graduate students with a less extensive background
who need more details.
Whatever you write, do the best you can. Most work now is in electronic format,
making it available to others far beyond your immediate audience, so it is important to
make sure it represents you well.

Evaluate your understanding

(1) How can adding a one-sentence summary of an article to a document you are writing
improve your memory of that article?
(2) Why is it important to write the details of your research as you do the work?
(3) Why should you consider your goals, audience(s), and the criteria by which your work
will be judged?

Homework
(1) What do you plan to write first?
Hint: See the following chapters for various types of writing. Read whichever is appropriate
to your current writing, then skip to Chapter 18 to begin writing. Come back to the others
as needed.
(2) What is the goal of your writing, both short and long term?
(3) Who is the audience for your writing? How much do they know about your topic?
Why are they or why should they be interested in it?
(4) By what criteria will it be judged?
Hint: Some of this information will be available from the instructions to authors (journal articles),
call for papers (conference papers), department regulations (theses), or professors (undergraduate
writing). Some you may need to obtain by talking to others who have successfully completed the
task you are undertaking. Keep these questions in mind as you write, and more importantly as
you rewrite, deciding what to cut or add.
 Argument structure in other writing  121

Teaching tip: Many of the ideas presented here will be developed further in subsequent
chapters. You may find it useful to discuss how these principles relate to the situation and
type of writing students are doing in your class.

Notes
1 Hyland has written extensively on stance, e.g., Hyland (1999), Hyland and Guinda (2012).
2 Examples will be given in Rau (In preparation).
3 The distinction between evidence and inference is not always clear. See, for example, the discus-
sion of indirect evidence in Rau (2012: 60, footnote 7).
4 See Chapter 27 of this text and Cargill and O’Connor (2013: 95).

References
CARGILL, M. & O’CONNOR, P. 2013. Writing scientific research articles: Strategy and steps, New
York, John Wiley & Sons.
FERGUSON, E. 2002. He’s all the rage. The Guardian, 17 March.
HYLAND, K. 1999. Disciplinary discourses: Writer stance in research articles. In: CANDLIN, C.
& HYLAND, K. (eds.) Writing: Texts, processes and practice, Harlow, Longman.
HYLAND, K. & GUINDA, C. S. 2012. Stance and voice in written academic genres, New York,
Springer.
RAU, G. 2012. Mapping the origins debate: Six models of the beginning of everything, Downers Grove,
IL, InterVarsity Press.
RAU, G. In preparation. IPTC: A prototypical format for engineering research articles.
Chapter 11

Undergraduate writing

11.1 So near, and yet so far

When climbing a mountain you may rejoice as you surmount a peak, before looking up to
see the real summit looming above you in the distance. Similarly, having mastered under-
graduate writing you think you have arrived, only to find that there are many higher peaks
beyond. Nevertheless, if you have strengthened your skill in writing arguments, you will
be prepared for the path ahead.
We will begin our look at various genres with three types of documents common in
undergraduate courses. Naming them is a challenge, as the terms used differ by country,
institution, department, and course, but we will call them lab reports, essays, and senior
projects. For each we will consider the goals, audience, and criteria, and how those affect
the structure of the document.
On the surface each genre may appear to closely resemble research articles, but at a
deeper level differences appear. Lab reports often have the same section titles but lack many
components found in research articles. Essays are usually secondary research, more similar
in structure to survey or review articles. Experience writing either of these may contribute
to difficulties in later writing if
you are not aware of the differ-
Undergraduate writing lacks ences. Senior projects, particu-
larly those involving primary
many components of research research, are longer term, offer
articles, but practice making and more autonomy, and thus may
supporting claims will prepare be like graduate theses.
you for future writing. Undergraduate writing tasks
allow you to build confidence
as a researcher and a writer.
Attention to argument struc-
ture, making and supporting
claims, will help you in any future writing. Senior projects may be written up for submis-
sion to a conference or journal, and beneficial for graduate school or job applications.

11.2 Lab reports

Lab reports are perhaps the most common undergraduate writing task for science and
engineering students. Classroom laboratory experiences are usually demonstrations of
 Undergraduate writing  123

students’ ability to follow procedures, use apparatus, and obtain, present, and interpret
results, but not part of an ongoing program of authentic research. Although instructors
guide students to use section titles reflecting research articles in the field, the similarity is
only skin deep, because the implicit claim is that you have mastered certain methods or principles,
not that you are contributing to knowledge.
First we need to distinguish three types of laboratory experience. Many traditional lab-
oratory exercises follow a “cookbook” approach, where the exact procedure is specified.
Increasingly, science education is encouraging use of an inquiry model, where students
must formulate a question and find a way to answer it. Engineering or technical courses
may use a design challenge format (Rau & Antink-Meyer, 2020).
In the cookbook approach students must follow the procedure and can predict the
expected results. Sometimes this leads to falsification of data (I wouldn’t have any
personal experience with that!), but that is not the issue here. What is important for
our purposes is that this is very different from actual research. There is no reason to
include Component 1 (“This is important because the instructor said to do it.”) or
Component 2 (“This has already been done thousands of times with the same result,
but here we go again.”). Usually the goal is stated (“We will verify the natural law
that has been verified with far greater precision for centuries.”). There is no need to
explain the framework or the details of the method, as they are already listed in the
lab manual, although usually some summary is required. So what is left? The data and
explanation. While it is useful to produce and explain tables and graphs, that is by no
means the same as writing a research article. If you are asked to write up a cookbook
lab, you already know how to get exemplars (from former students who took the
course), and your report will be very similar to theirs (hopefully not too similar), so
there is no need to talk further about this.1
Inquiry and design challenges are more like actual research, but with differences that
can be attributed to the audience, goal, and criteria. The audience for all student work is the
professor, who also sets the goal and criteria. The professor and former students are the only
possible sources for exemplars for that specific class. Since each professor values differ-
ent things, you will have to look closely at the stated criteria (or infer the unstated from
the way they teach and grade) to
succeed.
Nonetheless, a few general Lab reports often lack
comments can be made about what Components 1, 2, and 4, modify
components should be included. 2
several others, and have little or
Some sort of introduction is nec- no support from citations
essary to establish the topic of the
of past research.
report. This may include impor-
tance but frequently only provides
background information. Citations
may be required for inquiry labs. The goal should also be stated in the introduction, but
it is the goal of the lab, not a research goal. The procedures used are stated next, but
materials supplied by the instructor may not be. Statistical procedures, if used, should
be reported. The data are shown, often as tables or graphs, with interpretation, includ-
ing comparison with predictions. A separate conclusion is common in engineering lab
reports, but not science, as in the respective research articles, but is a summary of the lab
rather than a claim of contribution.
124  Exploring different genres

The most noticeable absences are the importance, need, and framework, along with
citations to support them. Lack of experience with this may make it difficult to appre-
ciate the importance of connection with past research and write about it when doing
actual research.

11.3 Essays
The second main genre of undergraduate writing is the essay, also called a report or
research paper. It reports previous work rather than original research. Thus the format
is closer to a survey or review article, or secondary research in social science, than to
primary research articles in science or engineering. Essays seem to be more common in
science than engineering.
As with lab reports, the professor is the audience, and the assignment should state the goal
and criteria. Usually the goal is to attain a deeper understanding of one topic related to the class.
Exemplars are again specific to each class, although the structure of review articles may
provide helpful guidelines (section 13.4).
The introduction of an essay will focus only
on the importance of the topic, as there is no
Essays often lack reason to prove the need for your paper. A
Components 2, 5, and statement of purpose or research questions is
6, and modify several required. The methods you followed to collect
others. your information are not needed and since no
new testing was performed, there are no test-
ing methods. Instead, various frameworks for
approaching the topic may be discussed in detail. Specific results from the literature may
be included to illustrate or focus attention on key points. By far the most important por-
tion is the discussion, where, depending on the assignment, you may be asked to summa-
rize, compare, evaluate, or integrate, so what needs to be included will vary accordingly.
Generally, references and citations are required.
Since the requirements for an essay are quite varied and dependent on the indi-
vidual professor and goals of the course, it is hard to generalize about what to include.
Nevertheless, argument structure is crucial. What claims are you making? Your first
claim is probably the importance of the topic. Next you might claim there are three
main approaches. For each approach you might need to document, with examples,
their strengths and weaknesses. Finally, you claim one approach seems to offer the
most potential. The claims will differ depending on your topic, so think about the argument
structure and support before you begin writing (Chapter 19). With secondary research,
there are many options for sequencing, especially in the discussion (section 9.5). As
you rewrite, consider alternative orders that would strengthen the argument or make it
clearer for the reader.
Essays differ from review articles and research writing in one other key aspect. In an
essay you are providing a summary of a topic, but need not include all related work. A
sampling or selection of some of the most important work is usually sufficient. However,
to prove that there is a need for your research, you must show you are familiar with all the
work in the field, including the most recent, to show there is still a gap. All graduate writ-
ing and review articles require a comprehensive literature review, so this is an important
difference to keep in mind in later writing.
 Undergraduate writing  125

11.4 Senior project

This genre is very departmentally dependent. In some departments all seniors majoring
in a department write a final paper as part of a senior seminar. In others it is an option
for honors. Thus the document may be
called a senior capstone, senior thesis,
honors project, and undoubtedly many Senior projects requiring
other names. original research are like
Senior seminars may not require orig- graduate theses; others
inal research, in which case the paper are like essays.
will follow the essay genre (section 11.3).
Honors projects in science and engineer-
ing usually require original research, so
the writeup is more like a graduate thesis (section 12.6), but with lower expectations for
length, number of references, and depth of research.
If you are writing a senior project, the first step is to get both the requirements (criteria) and
exemplars from your department. Sometimes the project will be read only by one professor,
sometimes by several. It may or may not require an oral presentation. For a seminar you
may have complete freedom of choice in essay topic. Original research often must be
related to the work of an advisor, because of restrictions of budget and material.

11.5 Summary
Table 11.1 summarizes which components tend be found in each type of undergraduate
writing, and differences from research articles. One of the main differences is that since
the goal in lab reports and essays is not original research, it is not necessary to show the
need for the work or the connection with previous research. Furthermore, the methods
and testing methods are less important than the results and discussion. Senior projects that

Table 11.1  Components and support in undergraduate writing

Component/support Lab reports Essays Senior project

1 Importance No Yes Yes

2 Need No No Yes
3 Research goal Goal, not research Goal, not research Yes
4 Framework No Yes Yes
5 Research details Summary No Yes
6 Testing methods Yes No Yes
7 Data patterns Yes Examples Yes
8 Comparisons Yes Yes Yes
9 Interpretations Yes Yes Yes
10 Conclusion Summary Summary Yes
I Past research Little or none Yes Yes
II Present research Less detail No Yes
III Logical Usually deductive Usually inductive Yes
IV Evaluative No Yes Yes
V Organization No Yes Yes
126  Exploring different genres

involve original research will contain most if not all of the components found in research
articles. Remember that general summaries may not apply to your situation, since these
genres are often assigned and graded by a single professor.

Evaluate your understanding

(1) How is each genre of undergraduate writing like and unlike research articles?

Homework
(1) Obtain exemplars and criteria for whatever you plan to write.
(2) Before you start to write, decide what claims and components you will need to
include and how you will support the claims.
Hint: Plan and write following the steps in Part 4, beginning in Chapter 18.

Teaching tip: Help students obtain suitable exemplars for whatever you want them to write
and compare them with the structure of research articles.
Teaching tip: This chapter may be useful for graduate students as a way of contrasting their
current writing with formats they have written previously.

Notes
1 Obviously, an author’s evaluation does not need to be stated explicitly to be clear.
2 See Parkinson (2017).

References
PARKINSON, J. 2017. The student laboratory report genre: A genre analysis. English for Specific
Purposes, 45, 1–13.
RAU, G. & ANTINK-MEYER, A. 2020. Distinguishing between science, engineering and tech-
nology. In: MCCOMAS, W. & ORAMOUS, J. (eds.) Nature of science in science instruction:
Rationales and strategies, New York, Springer.
Chapter 12

Graduate writing

12.1 Apprenticeship
Graduate work is an academic apprenticeship. Working with your advisor and older
graduate students, you slowly learn to do research. Reading other research, you gradually
realize what characterizes both good research and good writing, which do not always go
together.
As the American author William Faulkner said,

Read, read, read. Read everything— trash, classics, good and bad, and see how they
do it. Just like a carpenter who works as an apprentice and studies the master. Read!
You’ll absorb it. Then write. If it’s good, you’ll find out. If it’s not, throw it out of
the window.
(in Meriwether & Millgate, 1968: 55)

As you begin your own writing you will find that some of it is trash. Throw it out the
window (being careful not to hit anyone passing by) and start over. Like a carpenter,
start with straight cuts and common nails, doing the rough framing that is hidden once
a wall is completed. Gradually move to bevel cuts and finishing nails on the surface that
shows. Eventually you may be ready to do
dovetails without nails, the neat and smooth
work of a cabinetmaker. But you cannot get
Graduate writing includes
there without practice.
As a graduate student you may continue most of the components,
to write lab reports and essays for classes, but to differing degrees.
described in the previous chapter. You will
also have to write at least three types of doc-
uments related to your research: research
articles, conference papers, and your thesis or dissertation. Engineers may write for
industrial journals. Your department may require a research proposal. All include the
same components we have studied, but with different emphasis.

12.2 Research articles

Research articles are the main publication for which you get recognition in academia.
Since we have already studied their structure, this section will briefly mention a few things
to keep in mind as you write.
128  Exploring different genres

Always consider goal, audience, and criteria (section 10.3). Each journal has a different
audience and therefore expectations. A selective, indexed journal and a more special-
ized one have different criteria, and you may have a different goal in submitting to each
(Chapter 27). Some journals require payment for publication. These often accept almost
every paper submitted, with minimal review, and therefore get a reputation for publishing
low-quality work. Steer clear of such journals.
You cannot submit the same paper
or publish the same information in
Publishing the same two places. How much difference is
information in two places enough? Some authors boost their
is considered plagiarism, number of publications by mak-
regardless of the language ing only one minor modification to
each publication. This may pad your
of publication. resume but will be annoying to other
researchers. For example, as a gradu-
ate student I found two papers, pub-
lished by the same author, the same year, in two different journals. One concluded that
a method of genetic analysis was useful, the other that it was not. Careful comparison
showed that although the wording was very different, both were based on data from the
same plants, the same years, but examining different traits. Neither article mentioned
the other, so future researchers (like me) were left searching through reams of infor-
mation to determine why the conclusions were so different. Needless to say, I do not
recommend that technique.
For those working in English as a lingua franca, it is also important to emphasize that
publishing a translation of the same paper is considered plagiarism, since the content
remains the same. A translation must be designated as such, with permission from the
original publisher, who holds the copyright. Usually it is better to do further research and
publish a subsequent paper in the other language that cites the original.

12.3 Research proposals

When required, a research proposal is the first document you will write related to your
graduate research. A formal research proposal may not be required for a Master’s degree, as
the topic is often assigned by the advisor, a small part of some larger project. It is a common
but not universal requirement for a doctoral degree program. Usually the proposal will be
read by several faculty members. It often requires an oral presentation and defense of the
proposal, at which the committee will ask questions for clarification or make suggestions for
changes. The implicit claim you need to support is that you are prepared to do the research.
The purpose of the proposal is to ensure that you are familiar with other research in
the field, have considered what topics are important but need further work, are aware of
various methods used to answer such questions or design a solution, and have the ability
to complete the research. In some fields it is like an extended Introduction, but in others
requires additional components not found in research articles. As with any writing, the
first step is to determine the requirements of your department and, if possible, obtain
exemplars of previous proposals to determine unstated expectations such as length and
degree of detail.
 Graduate writing  129

Before writing a research proposal you will need to have a general topic and read
extensively, probably at least 50 and perhaps over 100 articles for doctoral work. Not all
the articles you read will turn out to be directly related to your final project, so not all
may be listed in your final proposal, but from the beginning it will be worthwhile to use
a reference manager to keep track of everything (section 25.3). As you read, you will
find out what are currently considered important questions or problems in your field
and what has already been done. Your
proposal must build on what has already Research proposals
been done, adding to it in a unique and frequently include
substantial way. You may need to do a
Components 1, 2, 3, 4, 5, and
pilot study to demonstrate the potential
of your method. 6, and may require estimates
Departmental policies will guide your of cost and time.
overall organization. If none is specified
in the regulations, you may be able to
discern an expected structure from exemplars. The section names and order may differ,
but the following topics are commonly found in research proposals:

• Objective (research goal);

• Introduction (importance, need, previous work);
• Framework (alternative approaches, justification);
• Procedures (methods or anticipated design, testing methods);
• Resources needed (materials, cost);
• Timeframe.

Unlike a research article, the research objective may be stated at the beginning to make
the proposal easier to evaluate. If the goal is not clear, too close to current work, or unat-
tainable, there is no need to read further. Thus it is important to discuss this with your
advisor before writing.
As in a research article, the Introduction will often include both the importance of the
topic and the need for further work, each supported by citation of previous work in the
field. The further the proposal is from current work in the field, the more justification will
be required. This is frequently separated from a longer description of previous research,
again for ease of evaluation.
The third part may have many names, including literature review, background, related
work, or framework. It includes extensive evaluation of previous research or approaches,
their strengths and limitations. It frequently includes a justification for following or build-
ing on one framework.
This is followed by the anticipated methods (science) or design (engineering), and
testing methods. Although these may change during the research, a full description is
required, including complete experimental design if statistics are used. If the method is
not feasible or the testing will not produce meaningful results, it is better to find that out
before beginning. The anticipated results or predictions usually are required, along with
results of any pilot studies or prototypes.
In addition to these topics, which mirror the first six components, many proposals also
require an estimate of the resources needed, their cost, and a timeline. Although these
130  Exploring different genres

are likely to change, it helps ensure that the project can be completed given the time and
resources available, or perhaps point to the need to write a grant to cover the cost.
A research proposal could be anywhere from 10 to 50 pages, so again it is impor-
tant to determine the expectations of your department. Most of what you write will be
included, after many additions, changes, and revisions, in your thesis or dissertation and
other publications.

12.4 Conference papers

Often the first document you will write for an audience outside your department is a con-
ference paper. On the surface conference papers appear identical in structure to research
articles in the same field, but there are differences in goal, audience, and criteria. They
have a narrower scope and are less com-
Conference papers give plete and less carefully reviewed, and
thus have a lower impact than journal
you visibility in the field. articles.
Some conferences are The goal of writing a conference paper is
highly selective; others to attend the conference. Attending allows
not at all. you to meet other researchers in the
field, gain recognition for yourself and
your work, and hear about the latest
developments in the field. Usually to get funding to attend you must present either a
paper (section 16.4) or poster (section 16.6). Presenting is a good opportunity to practice
your speaking skills before your thesis defense or job interview.
There are several audiences for your conference paper. At a minimum the paper will be seen
by reviewers evaluating its suitability for the conference. If accepted, it will be available
to participants in the conference. Some conferences have graduate student paper com-
petitions, a third audience. Particularly in engineering, conference papers are frequently
published as Proceedings, making them available to a far broader audience.
When submitting a proposal to a conference, consider what information will be avail-
able to each audience and when. You want people to attend your presentation, not just
read your paper. Will they have the full paper or only the abstract available before the
conference? Your abstract should attract attention to both.
Criteria for acceptance of conference papers vary. Some conferences are very selective,
and even experienced researchers can have papers rejected. Others accept almost every
submission, leading to thousands of participants and multiple parallel sessions. Conferences
often ask that submissions be related to a certain theme or track. This facilitates screen-
ing proposals, as a different group of reviewers is assigned to each. If you submit to the
wrong track your work is less likely
to be accepted. In a non-selective
Conference papers normally conference, a paper submitted to the
have a narrow focus, present wrong track may be accepted, but
only preliminary results, and not seen by interested participants.
The scope of a conference paper
are not rewritten based on is often narrower than a research
reviewer comments. article, being a written record of
what you will present orally in less
 Graduate writing  131

than 20 minutes. Instead of the whole design, you might focus on one element, or only
one of several traits you are testing. This also allows you to present the same project at
different conferences, since it is considered improper to repeat the same information in
two venues.
Conference papers often must be submitted six months or more before the confer-
ence, and therefore report only preliminary results, with the expectation that further
results will be presented verbally at the conference. Thus the components of the
Results and Discussion are often missing or very short relative to research articles
in the same field. Nevertheless, the implicit claim is that you will have something worth
listening to.
Because the same term is used, novice writers may not recognize the difference between
review of conference papers and journal articles. Review for a conference is limited to
checking the suitability for the conference and the quality of the work, but rarely are
reviewer comments given to the author,
with the chance to revise and resubmit.
Thus conference papers frequently have Conference papers
less complete citations of past research are cited frequently in
and less detailed descriptions of the cur-
engineering, but rarely
rent work than research articles.
For all these reasons, conference in science.
papers are not considered full publica-
tion, and have less impact than research
articles. Nevertheless, conference papers are important in engineering. Since they report
the latest developments in a field, conference papers are frequently cited in engineering
research articles. In science, with its emphasis on data and analysis, citation of conference
papers is extremely rare.

12.5 Industrial journal articles

In some fields of engineering, graduate students may write articles in journals with an
industrial audience.1 Whether you choose to write for these journals or others depends
in part on whether your long-term goal is work in academia or industry. The underlying
claim is that the technique presented has immediate practical utility in industry, so although the
overall structure is similar to IPTC, some components are minimized.
The Introduction usually begins with the specific importance to industry, rather
than society at large or researchers in the field. Need is similarly limited to problems
with the current best solution, with less discussion of historical or alternative meth-
ods and therefore fewer citations. The research goal is often in a separate section or
subsection.
Framework is introduced only as it relates to the current solution, which is presented at
length. Testing emphasizes the advantage of the present method, often with less attention
to comparison with other methods and more details on specific numerical test results, so
readers can compare it with their own current method or industry standards.
Overall, the articles may be shorter than other research articles, but have as many or
more graphics. They frequently lack organizational or summary paragraphs, although
topic sentences are prominent. These generalities would of course need to be confirmed
for the journals in your field.
132  Exploring different genres

12.6 Thesis or dissertation

First it is necessary to clarify terminology. In American usage, a Master’s student writes
a thesis, while a doctoral student writes a dissertation. The opposite is true in British
English. For ease, in this section I will use the term thesis to refer to either.
The structure of a thesis is usually similar to a research article, but longer. What would
be sections in an article become chapters in the thesis. The thesis may include material
that will eventually be published in several articles, as well as some that will never get
published, including extensive appendices, data tables, or approaches tried that did not
work or produced inconclusive results.
Like a research article, a thesis begins with an Introduction. In some fields this will
follow the same structure as a research article, while in others it is a synopsis of the thesis,
beginning with the research goal, followed by a summary of the methods, results, discus-
sion, and conclusions.
Frequently the second chapter is a literature review, far more extensive than will
be published in any of the related articles. The purpose of this is to demonstrate to the
committee that the student has a clear understanding of various approaches to the topic,
not only the one used in
the thesis.
Beyond this, the thesis
The thesis may be organized will follow one of two nested
like an extended article or a designs. The first is more
compilation of several articles, like an extended research
preceded by a general Introduction article, with chapters des-
ignated Methods, Results,
and Literature Review.
Discussion, and Conclusion.
In this case within each of
the three main chapters there
will be sections for different
experiments, discussed in the same order in each chapter. This works well for closely related
experiments with similar methodology.
In other cases, the chapters will be organized more like a series of articles, each with
Methods, Results, and Discussion (or appropriate sections in IPTC format). This works
better when more diverse methods or designs have been used. It may also be chosen if
some of the material has been or is in the process of being written as articles, to minimize
the need for rewriting and make it easier for readers to relate the two.
In either case, the Conclusion is likely to be longer than the one in a research article.
Frequently it includes a summary of each chapter before drawing overall conclusions. The
abstract is also more likely to be a summary of the entire thesis rather than an advertise-
ment highlighting the main contribution only.
Criteria for the contents and length of a thesis are set by each department. The format may be
set by the university. If the thesis is to be made available online, as most are currently,
there may be additional requirements for that. All the necessary information should be
available from the department secretary or a thesis secretary in the graduate school office.
Often there are several stages of approval, even after the content has been approved by
your advisor and committee, so make sure you check the requirements and allow suf-
ficient time to meet them before the deadline.
 Graduate writing  133

12.7 Summary
The main genres associated with graduate research are all closely related to research arti-
cles but will be judged by different criteria. As shown in Table 12.1, the research proposal
is preliminary, presenting the planned methods and expected results, both of which will
undoubtedly be modified in the course of the research. A conference paper usually con-
tains at least preliminary results and may be cited in engineering before a related article
is published. Industrial journal articles are more practical, lacking or minimizing some of
the components. A thesis (dissertation) is longer than several articles and summarizes all
the work done, whether publishable in another format or not. It frequently includes more
detail on both the method and results than any of the other genres, as well as a far more
extensive literature review.

Evaluate your understanding

(1) How does each of the other genres of graduate writing differ from research articles?

Homework
(1) Obtain exemplars and criteria for whatever you plan to write.
(2) Write as you do the work, then revise, following the suggestions for research articles.

Teaching tip: Help students find appropriate exemplars for each genre.

Table 12.1  Components, support, and additional items in graduate writing

Component/Support Research Conference Industrial Thesis or

proposal papers journals dissertation

1 Importance Yes Yes Practical Yes

2 Need Yes Yes Yes Yes
3 Research goal Yes Yes Yes Yes
4 Framework Yes Short Maybe Yes
5 Research details Planned Yes Yes Yes
6 Testing methods Planned Yes Yes Yes
7 Data patterns Expected Preliminary Yes Yes
8 Comparisons Expected Preliminary Yes Yes
9 Interpretations No Preliminary Maybe Yes
10 Conclusion No Preliminary Yes Yes
I Past research Yes Some Some Yes
II Present research No Preliminary Yes Yes
III Logical Yes Yes Yes Yes
IV Evaluative Yes Maybe Yes Yes
V Organization No No No Yes
A Resources Yes - - -
B Timeframe Yes - - -
C Unpublished data - - - Yes

Research articles (not shown) contain Components 1–10 and support I–V.
134  Exploring different genres

Note
1 At our school this is common in mechanical engineering, but not electrical, chemical, or
computer science.

Reference
MERIWETHER, J. B. & MILLGATE, M. 1968. Lion in the garden: Interviews with William Faulkner,
1926–1962, New York, Random House.
Chapter 13

Academic writing

13.1 Never stop learning

If you remain in academia you never stop learning, whether about your field or about
different genres of writing. You will continue to write conference papers and research
articles but will also branch out into other genres. At a minimum you will have to write
grant proposals, and probably other types of articles, book chapters, or even books. Unlike
the genres presented in the previous chapter, these are all quite different from primary
research articles. Nonetheless, they contain many of the same components and make the
same basic claim that you are contributing to knowledge.
More time will be spent discussing grant proposals, since they are the most distinct.
You will probably discover how the others differ from research articles in the course of
your own reading.

13.2 Grant proposals

Grant applications are difficult because they require skills in management and finance,
rarely taught in science and engineering programs. Grants involve not only the initial pro-
posal, but also progress reports on the funded research (section 14.2). Doctoral students
frequently have a hand in writing a grant proposal to fund their research or the ongo-
ing research program of a lab.
Although similar to research
proposals, the stakes are much
For grant proposals it is essential
higher. A failed grant proposal
may delay your work. to follow the guidelines, and
Every granting agency has helpful to examine successful
different formats, requirements, past proposals or apply with an
and criteria for grant proposals, experienced researcher.
so it is important to both check
all relevant online documents
and examine examples of suc-
cessful proposals. Writing your first grant as a co-PI (principal investigator) with an expe-
rienced researcher can help you learn the format and increase your chances of success. In
general, the main claim is that the proposed research is important to the grantor and that your lab has
the ability to complete it successfully. Grant proposals include at least:
136  Exploring different genres

• Applicant and institution contact information;

• Justification (pursuant to grant goals);
• Framework and related work;
• Research goals;
• Experimental design or research plan;
• Expected findings;
• Participant expertise and institutional support;
• Timeline;
• Budget.

Grant proposals require most of the same components found in research articles, although
stated as plans or expectations rather than accomplished fact and not always in the same
order. Additionally, estimates of budget, timeframe, and an evaluation of the resources
available to complete the work are required. Frequently the granting agency will provide
an electronic form to fill in, often with a page or word limit for each category. Failure to
comply with those guidelines will lead to swift and certain rejection.
Grant proposals begin with several pages of forms, providing contact information of
the researchers and institution. If your research involves animal or human subjects, addi-
tional questions will ensure compliance with guidelines for ethical research and approval
from an Institutional Review Board (IRB).
In the written portion, every grant application must show the importance of the
research, particularly how it meets the goals of the granting agency. Contribution to
society or future applications are secondary, unless these are stated goals of the agency.
A review and evaluation of related work is also required, frequently in a separate sec-
tion. This need not be comprehensive but must include the most recent related work,
particularly ongoing work in your lab. Like any other research, the application must build
on an existing model or framework, except grants for creative solutions, which require
more documentation and reasoning to show their feasibility. Being able to explain the
framework clearly and concisely shows your familiarity with the concepts, thus may be
more prominent than in research articles.
Instead of a single research goal, grant applications typically state three goals: easy, achievable,
and challenging (Poe et al., 2010: 88). The first should be something that will be com-
pleted by the end of the first reporting period, often within the first six months or year of
the grant. This allows you to point to progress in your first report. The second is some-
thing you are reasonably certain you will be able to achieve by the end of the grant. This
allows you to claim success in your final report. The third is something that you expect
to work on but not necessarily complete during the term of the grant. This lays the basis
for extensions or future grants.
The experimental design or research plan is also crucial, as it allows the grant commit-
tee to evaluate the potential for success of the work. This should be as detailed as possible
within the length guidelines of the application.
Later components of a research article are included as plans or expectations. Expected
results, discussion, and conclusion should be stated confidently, but with appropriate
qualification and hedges (Chapter 22). The goal must be measurable and the method for
evaluating success clear.
Beyond the components of a research article, you will need to demonstrate com-
petence and ability to complete the research, both with personnel and facilities. Core
 Academic writing  137

continuing personnel are listed not by name but by experience, including publications,
awards, and familiarity with specific protocols. Personnel to be brought on later, includ-
ing graduate students whose research will be funded, do not have to be listed, but the cri-
teria of selection may be. You must show that the lab, department, or school has enough
basic resources to manage and complete the research, or that you will partner with other
organizations with the necessary expertise or facilities.
Budget and timeline must also be included. Underestimating or overestimating either
can lead to failure. Of course, both must be within the guidelines set for the grant. This is
one of the areas where it is most useful to look at previous grants, to ensure that you have
not forgotten any categories of expenses or severely misjudged cost or time.

13.3 Short papers

As mentioned in Chapter 2, journals may contain genres other than primary research arti-
cles. One of the more common is short papers, brief reports, letters, or correspondence.
These are common in rapidly changing areas of technology, providing an opportunity for
researchers to publish quickly, with minimal editorial review, largely in order to claim
primacy for a discovery or new design. The
practice of noting the dates a document was
received by the editor, accepted for publi- Short papers may have
cation, or published online prior to print simple organization
publication serves the same purpose. and minimize many
Generalization about the format of short
components.
papers is difficult, as there are many vari-
ations. Some are several pages long, with
a structure nearly identical to journal arti-
cles, including four sections representing IMRaDC and most if not all components, but
fewer details and references. Others may be only one or two pages with no sections and
lack several components. Minimal review means the structure is determined by what the
author finds essential to the implicit claim that this is worth publishing quickly.
The Introduction is usually shorter than a research article. There may be only a short state-
ment of importance or none. Discussion of previous research may be limited, or many papers
cited but with minimal evaluation. The research goal may be stated at length, comprising half
of the Introduction, or in a single sentence. Framework and details of the current research are
often minimized, mentioning only novel features, but may be very important in engineering.
The focus of the article is often on the Results and Discussion (science) or Testing
(engineering), with sufficient background to understand the cause or implication of those
results. Graphics are often present, but with minimal commentary.
Some journals have a relatively fixed style for short papers, while others allow far more
flexibility. Doing component analysis of a few articles from your target journal will allow
you to figure out the expectations and write accordingly.

13.4 Review or survey articles

The word “review” can be used as a verb (Chapter 18, note 2), a noun (blind review,
Chapter 27; book or product reviews), or an adjective (review article). In this section we
are talking about the latter: articles written to review or survey the status of a field.
138  Exploring different genres

Review articles may be written by an experienced researcher as a way of setting

direction for the future or assigned to a doctoral candidate to help them map the terri-
tory and find a research topic. Since
they represent the opinion of one
Review articles do researcher, they are sometimes called
not report a new a Perspective.
design or results, and Since review articles do not report new
therefore lack many research, they lack many components,
components. including the research goal (although
the goal of the review will be stated),
current research details, and test-
ing methods. While data from other
research may be reported and compared, no new data are reported, so there is no Results
or Testing division. On the other hand, there is extensive evaluation of alternative frame-
works and interpretation of various results.
While the section structure differs, the order of review articles is relatively constant. They
typically begin with a more extensive discussion of Importance than research articles, in
some cases comprising a whole section. Then a summary or grouping of various approaches
precedes a detailed evaluation of each. Finally, the various approaches are compared.
Some reviews are comprehensive, mentioning every paper published on a topic within
a specified time period, while others are broader, highlighting major papers and shifts in a
field. They may be presented chronologically or follow some other organization scheme
(section 9.5). Results obtained by different methods may be compared graphically or only
in text, but interpretation is expected.
Before writing a review article, make sure you are clear about its purpose. If the main pur-
pose is to help you find a researchable topic, it will help others do the same. This requires
extensive evaluation of the advantages and drawbacks of each approach. Whether those
can be identified from the methods themselves or the results obtained, from rational
evaluation or empirical comparison, depends on the topic. If the purpose is to point to
new directions for future research, you may wish to emphasize the diminishing returns of
current methods and the foundations and potential of new approaches.

13.5 Book chapters

Let me begin this section with a word of warning for those considering publishing a book
chapter. Maybe two words of warning. First, it is less likely to be cited than an article
(unless it becomes a hugely popular book). Journals are more likely to be available through
your university library than a book, and no one wants to track a book down through
interlibrary loan or pay the price of downloading a chapter. Second, don’t count on a book
chapter for your resume, let alone to report the latest advance, as it may take several years
to be published. On the other hand, book chapters can be useful for introducing yourself
and your research to a different group of specialists, particularly in crossover areas.
Book chapters seem to be less common in science and engineering than in social sciences
and humanities, probably largely because of the importance of speed of publication. They
may be important in areas where there is an overlap between technology and society, sci-
ence and education, or newly emerging fields on the borders of previous disciplines.
 Academic writing  139

Book chapters are usually solicited by

an editor and must fit the focus and per-
Book chapters in an edited
spective of the book. How much control
the author has over the contents depends volume may be like either
on the editor. Some books are collec- research articles or
tions of research on one topic, so each review articles.
article is very similar to a research arti-
cle. Others are intended as primers, with
each article more like a review article on one topic. Make sure you understand the pur-
pose of the book as a whole and how your chapter contributes to it.
For book chapters, contributors are usually chosen by the editor, and review is often
done internally, with each author commenting on two other chapters, with additional
comments from the editor. Since chapters are more likely to be undertaken by experi-
enced authors, I will not say more on these.

13.6 Popular writing

While many scientists and engineers are content to communicate only with other spe-
cialists, others want to get their message out to a broader audience. Different genres of
popular writing include science magazines, science teacher magazines, newspaper col-
umns, online news services, blogs, social media, and many others. Each genre has a different
audience, expectations, and format.
Some science and engineering organizations have popular journals as well as research
and industry publications. Such journals can be distinguished by prominent use of colorful
eye-catching visuals—the work of graphic designers rather than the author. Articles may
be written by science writers,
rather than research scientists,
although prominent scientists The purpose of popular
may also be asked to contrib- writing is to attract the attention
ute articles. of non-specialists, employing
Although the format of graphic design and one main point,
each varies, the implicit claim with limited details or qualifiers.
is that this is something inter-
esting and important to non-
scientists or scientists from other
fields. Instead of a general
statement of importance, the article will begin with an attention-catching example,
illustration, metaphor, or statistic. Each article normally focuses on one small topic,
sometimes not even the main point of the original research article it is drawn from,
with minimal mention of other research and few if any citations. The interpretation
of the results is often presented as truth, without either the data or qualifiers common
in research journals.
Since popular writing encompasses many different genres, it will be necessary to do an
analysis of several exemplars to determine the format and level of the audience. In many
cases the publisher exerts far more control over the product than in a research article, even
rewriting the title to make it more interesting.
140  Exploring different genres

Table 13.1  Components, support, and additional items in other academic writing

Component/support Grant Short Review/ Popular

proposal paper survey writing

1 Importance Yes Maybe Yes Interest

2 Need Yes Maybe Yes No
3 Research goal Yes Yes Goal No
4 Framework Yes Short Multiple No
5 Research details Planned Short No No
6 Testing methods Planned Short No No
7 Data patterns Expected Yes Previous Summary
8 Comparisons Expected Yes Yes No
9 Interpretations Expected Yes Yes Yes
10 Conclusion Expected Yes Yes No
I Past research Yes Limited Yes No
II Present research Planned Short No No
III Logical Yes Yes Yes No
IV Evaluative Yes Maybe Yes No
V Organization Forms No Yes No
A Resources Yes - - -
B Timeframe Yes - - -
C Budget Yes - - -

Book chapters (not shown) may be like research or review articles.

13.7 Summary
All the genres in this chapter have a different structure than research articles, although
some components overlap. To learn how to write each, obtain exemplars and do your
own component analysis. Use Table 13.1 as a guide for the types of components com-
mon in each, but realize there is extensive variation. Book chapters, not shown in the
table, may be similar either to research articles (with all the components and support) or
review articles.

Evaluate your understanding

(1) How are these genres of academic writing like and unlike research articles?

Homework
(1) Obtain exemplars and criteria for whatever you plan to write.
(2) Talk with an experienced author of the genre you are writing, to get tips specific to
your field and audience.

Reference
POE, M., LERNER, N. & CRAIG, J. 2010. Learning to communicate in science and engineering: Case
studies from MIT, Cambridge, MA, MIT Press.
Chapter 14

Technical writing

14.1 Intertextuality
All academic and technical writing is part of an ongoing process. Your work is con-
nected to other work that has been done before and currently by other workers. Thus
no science or technical writing stands on
its own, but always in reference to other
writing. This connection between writ- Intertextuality makes
ing is called intertextuality (the relation it difficult to understand
between texts). any text without
While it is important in academia to the context.
show how your work connects with previ-
ous work, by citing past research, in busi-
ness and industry it is often assumed that those
reading a document know the connection with what has gone before. This makes it hard for a
newcomer to join the conversation until they know the background.
This is especially true for the reports discussed in this chapter and email (Chapter 17),
all of which build on previous reports, meetings, or correspondence, usually without
specifically referring to it. Thus they would make no sense to outsiders, but they are not
intended to. Imagine walking into the middle of a conversation between two friends
about a recent sporting event. If you saw the event, you could join in. If you follow the
sport, you could ask relevant questions. Otherwise it would mean nothing to you.
Scientists and engineers need to write various types of reports, the names of which
differ by location, field, and industry. In the following sections I will discuss progress
reports about an ongoing project, technical reports and management reports written for
technical and non-technical audiences, respectively, and SWOT (Strengths, Weaknesses,
Opportunities, Threats) analyses, used for a variety of purposes.
These types of reports are not public, making it harder to find exemplars. Each organi-
zation has a different preferred format. Furthermore, because of intertextuality, authentic practice
is not possible in a class situation. Thus much of your learning about these genres must come
on the job. Nevertheless, understanding the goal of each report type will help.

14.2 Progress reports

Progress reports share a common central claim: that you have made progress on a project. That
project may be your graduate research, with an informal weekly report to your advisor
or lab group. It could be a monthly project report in industry. It might be an annual
progress report on a grant. In almost all cases, the main audience is someone in a supervisory
142  Exploring different genres

role, with your colleagues as the secondary audience. In all cases, those involved are
familiar with the work, so the report only needs to highlight changes since the last report, rather
than the entire project.
Several factors influence the contents of a progress report, including the frequency,
the stage of the project, and the relationship between the author and the recipient of the
report. The frequency of reports may vary during a project, perhaps being more frequent
during the early planning stage and as it nears completion but less frequent in the mid-
dle. Less frequent reports or relationships that are more distant require greater detail and
formality. The length and format depend on the formality of the report. A weekly report
to your advisor may be a list and some graphics on your computer. An annual report to
a granting agency will be many pages long and include changes to personnel, process, or
timeframe, and an accounting of funds.
Since the main claim is that you have made progress, the main support will come from
your present work. Viewed in terms of components, that means the components repre-
sented are likely to be research details (what you have done), testing (what you have tested
and how), data patterns (your results), and possibly either comparisons or interpretations
of those results, depending on the purpose of the report. The conclusion will summarize
your progress. Since the work is ongoing, the first four components are less likely to be
included. A typical progress report will contain the following elements:

• Progress since the previous report;

• Problems encountered;
• Plans for the next stage.

Each situation is different, so it is diffi-

A progress report claims cult to give specific guidelines on how to
that you have made write, particularly since much of the var-
progress toward the goal, iation depends on how much detail the
supported by evidence. boss wants. The only way to know that
(other than trial and error) is to ask your
colleagues. The most important idea to
keep in mind is that a progress report is
not just a list of what you have done, but a claim that you have made progress toward the
goal, supported by evidence.

14.3 Technical reports

Engineers frequently write reports to propose a design solution, predict potential prob-
lems, or analyze actual design failures. These are submitted to a supervisor with equal or greater
experience in the same field who is not
directly involved with the project but has
Technical reports analyze
background to understand it.
or propose a solution to Technical reports are the conclu-
a single problem, and are sion of a process of evaluation. Unlike
complete without reference progress reports, they should be com-
to other documents. plete in themselves without reference
to other documents. Like progress
 Technical writing  143

reports, the focus is on a single issue. The importance and need are normally assumed—
the reason the study and report were assigned is known to both parties. A technical report
may include:

• Summary;
• Problem statement or goal;
• Evaluation and support (section titles specific to the problem);
• Conclusion or recommendation.

The report will often begin with a summary of the goal, the findings, and the conclu-
sion. This introductory summary saves the reader time. The goal may be expanded into
a problem statement that lists the criteria, constraints, or assumptions used in evaluating
alternatives.
The majority of the report will be sections supporting the final conclusion or chief
recommendations. This may include a theoretical model, simulations, or testing data,
but citations are rare, on the assumption that the report recipient is familiar with back-
ground knowledge. Support may include evaluation of alternative designs or expla-
nations, and whatever logic is appropriate to the situation: inductive, deductive, or
abductive. More time will be given to alternative designs or explanations with more
complex or important issues.

14.4 Management or consulting reports

In contrast to technical reports, management or consulting reports are written for an audi-
ence who may not understand the technical details. Management reports are internal company
documents, whereas a consulting report is for a different organization. These reports will
be written differently depending on the goal and recipient. For example, you might need to write
different reports for the accounting
and marketing departments, or for
top management. Nevertheless,
you are supporting the implicit claim Since the recipient may not
that you have done your work well and read or understand all the
are presenting a fair assessment of the technical details, an initial
situation that will be beneficial to the summary and organizational
one receiving it. markers are crucial.
In these reports, like techni-
cal reports, there is often an initial
summary of the goal, key findings,
and conclusions or recommendations. Since the recipient may not read the entire report,
it is more important to use organizational markers such as introductory section summaries
and phrases pointing to information found in other sections.
Fewer of the common components will be present, and less technical detail. Thus
details of the framework and testing method may be reduced or eliminated. Instead of
comparing various alternatives, more time is likely to be spent justifying the recom-
mended course of action. This of course depends on what the recipient wants. Again the
report should be self-contained, with few citations apart from industry standards or similar
documents.
144  Exploring different genres

14.5 SWOT analysis

SWOT (Strength, Weakness, Opportunities, Threats) analysis is a popular strategy for
organizing claims. It considers Strengths and Weaknesses within the organization and
Opportunities and Threats outside it, relative to a goal (Figure 14.1). Each box is filled
in with claims appropriate to the category. Support
may be listed in the box as sub-points or placed
In SWOT analysis, in an accompanying explanation, where the
claims of each category are more obvious (e.g.,
each box lists one type “One of our greatest strengths is . . .”).
of claim. SWOT analysis can be used in academia or
industry for many purposes, from school accred-
itation to new product evaluation. Separation of
different categories of claims makes assessment easier.
As with other reports in this chapter, importance and need are often assumed, or
written separately from the SWOT analysis. The goal must be stated, as each category
is related to that goal. Beyond that, the expected components are totally different from
those used in academic writing, being the four categories on the chart. This illustrates that
although the common components are useful for academic writing, other categories may
be useful for other genres.
For example, some organizations develop their own progress report charts, following a
similar pattern, but focusing only on internal factors. The example shown in Figure 14.2
is modified from a form used by the technical department of a local industry. It ensures
that all areas are addressed in each report.

Positive: Negative:
Help achieve goal Hindrance to goal
Strengths Weaknesses
Internal:
Within
organization

Opportunities Threats
External:
Outside
organization

Figure 14.1  SWOT analysis chart.

 Technical writing  145

Positive Negative
Accomplishments Open issues and
since last update concerns
Current

90-day plan Issues requiring

help and support
Future

Figure 14.2  Categorical progress report.

14.6 Summary
Each business or industry uses different types of reports and different reporting formats,
but an attempt to generalize expectations is presented in Table 14.1. Just as once you
have learned to write research articles you can modify that to any other form of aca-
demic writing, once you have learned to write one type of report it is easier to learn a
second. The main thing to keep in mind is that the reports vary according to the goal,
audience, and criteria, all of which affect the claims made and the type and amount of
support needed.
Many other types of reports could be mentioned. White papers, policy papers, environ-
mental impact statements, or safety analysis may be written for government organizations.
Product specifications are written for business customers, and users’ manuals for retail custom-
ers. Space limitations preclude discussing each in detail.
By analyzing exemplars, often available from coworkers if you ask, you will be able to
write anything. Just remember that some genres will require other components, so add
whatever is appropriate, and that not every report will be a good exemplar. Evaluate the
samples you get to see whether the organization, claims, and support are clear. If not,
include the same things, but try to stand out for the quality of your argument.

Evaluate your understanding

(1) What makes learning technical writing more challenging than academic writing?
(2) What accounts for differences between the reporting formats?
146  Exploring different genres

Table 14.1  Components and support in technical writing

Component/support Progress Technical Management SWOT

reports reports reports analysis

1 Importance No No Maybe No
2 Need No No No No
3 Research goal Maybe Goal Goal Goal
4 Framework No Maybe No No
5 Research details Yes No Yes No
6 Testing methods Yes Yes Maybe No
7 Data patterns Yes Yes Yes No
8 Comparisons Maybe Maybe Maybe No
9 Interpretations Maybe Maybe Yes No
10 Conclusion Progress Conclusion Conclusion Conclusion
I Past research No No No No
II Present research Yes Yes Yes Yes
III Logical No Yes No Yes
IV Evaluative No Yes Maybe Yes
V Organization No Maybe Yes SWOT
A Timeframe Maybe Maybe Maybe Maybe
B Accounting Maybe Maybe Maybe Maybe
C Strengths - - - Yes
D Weaknesses - - - Yes
E Opportunities - - - Yes
F Threats - - - Yes

Homework
(1) Obtain exemplars for whatever you plan to write, if possible. Evaluate their argument
structure and components before attempting something similar yourself.
(2) Talk with an experienced coworker to find out what your boss expects.

Teaching tip: Class assignments to write these genres lack intertextuality and therefore are
not authentic. Nevertheless, if you have access to authentic texts, you can help students
evaluate the quality of the organization and argument structure.
Chapter 15

Writing about yourself and others

15.1 Blowing your own horn

The idiom “blowing your own horn” often implies pride, and boastful confidence in
your own ability, but I am using it in the sense of announcing your arrival, as horns used
to be blown to announce the presence of an important person, or a returning voyager
would sound his horn to announce his arrival. All the genres in this chapter have to do
with getting a new position, for you or someone else. Thus your basic claim is that you (or
they) are the best person for the position.
For any position, whether
in graduate school, academia,
or industry, there may be The person who gets the position
many qualified candidates, all is not necessarily the most
with good grades and at least
some experience. How will
qualified, but the one who makes
you stand out, getting to the the best argument showing they
shortlist, the interview, and are qualified.
finally the position? In order
to claim that you are the best
person for the position, you
will need to do some research to find out what they are looking for, then arrange your
support to show that you are that person. You need to not only be qualified, but convince
them you are qualified. That involves writing a well-supported argument.

15.2 Entrance essays

Two different types of essays may be important for getting into graduate school:
those on standardized tests and entrance exams or essays administered by the school.
Standardized tests include language proficiency tests and aptitude tests. A quick look
at the rubrics for grading the latter reveal that they look at a student’s ability to write
an argument. Since there are many specialized test-preparation programs, I will simply
remind you that you need to make clear claims and support them well with evidence
and reasoning.
In addition to standardized tests, many schools require students to write one or more
essays, either in the application or in person at the school. The former allows you time to
write and rewrite, but is subject to excessive help from other individuals, while the latter
148  Exploring different genres

tests your ability to write quickly

Find out all you can under pressure, with no assistance.
about the program and If you prepare for standardized tests,
professors, so you can it makes sense to prepare for these
ask specific questions and exams as well.
One of the main questions a school
indicate specific interest. will ask is why you want to go there
rather than elsewhere. Accordingly,
your main claim is that there is a good
match between your interests and the strengths of the program. This requires that you do some
research, looking at what the professors teach, and what they research. Just like a grant
application, it may require creative thinking to connect something you have done with
their interest. For example, when my wife saw a poster offering a scholarship to study
Austronesian linguistics, she had never considered that specialty. Nevertheless, she told
the professor that as a child she had performed Austronesian dance. She got the scholar-
ship and now has researched Austronesian linguistics for over 30 years.
As with any other argument, using specific evidence to support your claim is essential. If you
can write coherently about the concepts of a field and show why you could plausibly be
interested in the specialty of one or more professors, based on something you have done
before, you will have an edge.

15.3 Job applications

Job applications involve many elements. There may be a standard form for basic informa-
tion. They may also require a cover letter and resume. In general, you should not dupli-
cate information, but can highlight things written in another portion.
Like entrance essays for graduate work, make sure you do your research, not only
about the job itself but also the company. This shows your interest and allows you to talk
knowledgeably in an interview, but
also allows you to adjust your claims
Writing is important in and slant your evidence in the resume to
any job, so write well highlight items of interest to the employer.
on the application and Writing is an important part of
list any publications or any position in science or engineer-
ing. You have learned how to write,
presentations.
and hopefully write well. Show that
skill. Work on publishing early and
often. Even undergraduate work can
be published in some journals. If you worked with a professor and were second author,
that is still far better than nothing. Presentation at conferences shows your experience
with both public speaking and presentation software. If the conference allows it, write a
conference paper.
Other things count too. What is unique about you? If you have skill in graphic design
(illustrations), or finance (grant proposals), that is important. If you spent time in a dif-
ferent country, learning a different language and culture, you have learned to think and
communicate in creative ways. Starting your own business, even if it did not make money,
shows initiative. Lots of applicants have good grades and recommendations. What do you
 Writing about yourself and others  149

have that they don’t? To support your claim that you are the best person for the position, you must
provide specific evidence.

15.4 Resumes and CVs

The terms resume and CV (curriculum vitae) are often used interchangeably, but there
are differences in length and purpose. Most job applications require a resume. A resume
is short, one or at most two pages, highlighting your most important qualifications. A CV
is a complete list of your accomplishments, and is usually reserved for research positions,
grant applications, or personal information on your website.
Data never speak for themselves. You are trying to get a potential employer to see your
record in a certain way—that you are highly suitable for the job. You are also trying to
draw attention to anything that will make you stand out from the crowd. Thus layout is
far more important in a resume than a CV.
There are two approaches to resume writing. One suggests a simple one-page resume
that can be used with any position, with additional information placed in the cover letter.
The second suggests a longer two-page
resume tailored to each job. Arguments
could be made for each. The first is sim- Keep an ongoing CV that
pler, but cover letters may not be read lists all your accomplishments,
carefully. then create your resume
No matter which method you choose,
from that.
I suggest that you create a CV that you
update regularly (at least once a year)
with your jobs, publications, and other
information. Just like “write as you go” for the details of your research, it is easier to
add the details early than to search your files to find them later. Your CV should include
categories for:

• Name;
• Contact information;
• Education and training;
• Degrees, licenses, certifications with dates and granting agency;
• Job experience, both employment and self-employment;
• Specific job accomplishments including major reports;
• Publications and presentations;
• Awards, honors, and organizational affiliations;
• Reference contact information.

You may add categories such as military service, security clearances, travel, volunteer
experience, foreign languages, or others relevant to your field. Adding upcoming papers
or conferences will remind you to go back later to fill in the details.
When it comes time to write your resume, the first five elements listed for the CV should
be included. Others are optional. You may choose not to list every detail, particularly short-
term or part-time jobs, but beware of gaps in your school and employment history.
Organizing your resume is an art. While many books and websites give guidelines,
remember that you are trying to highlight the strongest evidence that you are the best candidate
150  Exploring different genres

for the job. Whatever that is—employment, education, or licensure—should be placed

front and center where it cannot be missed. That will change as you gain experience, so
your resume should change as well.
Material that will not fit in the resume can be added to your job application as support-
ing information. The employer will not look at this unless they are impressed with your
resume, so make sure that whatever makes you stand out is on the resume.

15.5 Reference letters

If you become either a professor or a manager, at some point you will be asked to write
a recommendation letter for a student or employee. If they have done exceptionally well,
this is easy. If they have not done well, you probably want them to move on, but have an
obligation to be honest. Either way, your goal is to help them get the position, so the implicit
claim is that they are the best person for that position.
Potential employers read between the lines of a reference letter, looking for what is
not being said as well as what is being said. Thus recommenders may try to write between
the lines so the message is clear, while avoiding direct criticism. Any direct mention of
a negative characteristic will kill a person’s chances, unless it is the flip side of something
extremely positive. For example, saying someone is extremely diligent in the lab but lacks
confidence in front of people alerts the potential employer both that they may not show
their best face in an interview and that they may not be suitable for a position requiring
high-profile presentations.
Reference letters usually have five parts:

• Self: why were you chosen to write this reference?

• Situation: how long have you known the person and in what relationship?
• Strengths: what are their greatest strengths (and weaknesses)?
• Suitability: how well suited are they to the position?
• Summary: how strongly do you recommend them?

First it is important to introduce yourself, and state why you are qualified to give a recom-
mendation. Next you need to specify the situation in which you have known the person,
and for how long. Often these are combined
in one paragraph.
Specific incidents and Next comment on the greatest strengths
skills show whether you have observed directly in your contact
someone is suitable for with the individual. Specific incidents or
a position. supporting evidence are helpful. In order
to comment about suitability, you need to
know the job requirements, so make sure
you get those from the individual when you
agree to write the recommendation. A person may be very good, with many strengths,
but not be suitable for a particular job. Honesty requires you to state or at least imply
your reservations.
The final paragraph is a summary, including a statement of how strongly you recom-
mend them for a position. Without reservation is the strongest compliment and should be
reserved for exceptional individuals.
 Writing about yourself and others  151

Reference letters for entry-level positions are usually less than one page. Those for
higher-level positions may be two pages. Anything beyond that should be added as
supplemental material.

15.6 Summary
Good writing is always an asset in getting a position, but not the only factor. Luck and
relationships play a role as well. One time while waiting for interview for a graduate assist-
antship, I was reading a book on string theory. The professor saw it and commented he
had just finished reading the same book. That gave us a good starting point that was not
related to the position itself. Connections of any kind give an edge.
The genres discussed in this chapter are very different from other academic writing.
Thus it should not surprise you that they contain different components. Nonetheless,
there are similarities in structure. Some sort of introduction says why the reader should
pay attention, followed by the goal and reason for pursuing it. Supporting evidence
must be specific and support the claim that you are suited to the position. A conclusion
summarizes the argument.

Evaluate your understanding

(1) How are the genres discussed here like other academic writing, although the compo-
nents included differ?

Homework
(1) Begin to build your own CV so it is ready when you need it.
Chapter 16

Academic and technical presentations

16.1 You are the expert

Many novice engineers and scientists are quite comfortable working alone in a lab, but
far fewer exude confidence when they step in front of a group to present their research.
Nevertheless, just like writing, the more you do it, the better you will get and the more
confident you will feel.
Remember that you are the expert.
No one knows your work better than
Remember that no you do, so prepare well and speak
one knows more confidently. Everyone in the audience
about your topic than has been or will be in your position at
you and everyone some point and, unless you have made
wants you to succeed. a lot of enemies or are in a competi-
tive job interview, wants to see you
do a good job.
Many written documents are also
presented orally. After examining some basic principles we will consider three genres:
academic oral and poster presentations and technical presentations.

16.2 Different format, same principles

Although the format of oral presentations is different from written work, similar principles
apply. Just as you examined exemplar articles before starting to write your own, analyze
similar presentations, then prepare in stages. As with writing, this will help you fit the
expectations, yet stand out for excellence.

16.2.1 Preparing in stages

As you listen to other presentations, don’t be a passive observer. Take notes, not only
on the content but on the structure. How did they begin? Which components did
they include and how many slides were devoted to each?1 Was the total number of
slides suitable for the time, or did it seem too fast or too slow? How did they con-
clude? Think about which talks caught and kept your attention. Was it the topic that
interested you, or was it a good presentation? Comparing notes with a friend can help
distinguish the two.
 Academic and technical presentations  153

Because of time constraints, some components important in written work may be

reduced or eliminated in a presentation. For example, there may be less emphasis on
showing the research gap or need for the work. Standard methodologies or tests may be
mentioned without details. In engineering you may explain only the most novel design
details. Your research goal is essential,
as is evidence showing you achieved
that goal and made a contribution. Just like writing, prepare a
Once you know what you need to
include, start preparing early. I suggest
presentation in stages, beginning
following the stages shown in Figure with the large scale and moving to
16.1. Stage 1 is the equivalent of smaller changes.
prewriting and writing. Stages 2, 3,
and 6 correspond to the three major
phases of rewriting, while 4 and 5
address special requirements of a presentation. As with writing, allow at least half the time
for several rounds of revision, and additional time to practice.
In stage 1, think about argument structure. What claims do you want to make and
how will you support them? List the components you plan to include. Begin with titles
and a few keywords on each slide. Each slide should address one main point, just like a
paragraph. Slides are free, so use them generously, without multiple layers or squeezing
too much on one slide. On the other hand, avoid large blank spaces, which the audience
will subconsciously interpret as gaps or weaknesses in the argument.
Next, organize the content, changing the order of slides and points on each slide so the
audience can follow your argument. Introduce vocabulary and abbreviations as needed,
rather than at the beginning, since the audience cannot look back as they can in a paper.
During stage 3, flesh out the slides, but keep the number of words to a minimum. Use
keywords or bullet points rather than sentences or paragraphs. Too many words on a page
will tempt you to read rather than speak. More importantly, it will distract the audience,
as they will not know whether to read or listen. Use graphics liberally, but if several are
similar, choose one as an example rather than showing all.
Once the basic order is established, think about how to attract interest. What is unique
about your work? Why should other people care about it? How could they use it or build
on it? What social relevance does it have? Is there any controversy related to it? Any unex-
pected findings? Add an introduction that will draw people in and a strong conclusion.

1. Lay out the content—titles and keywords for each slide

2. Organize the content—cut, add, and reorganize slides
3. Clarify the content—add visuals and text to each slide
4. Attract attention—strong introduction, conclusion
5. Maintain attention—animations and transitions
6. Finalize the content—proofreading and revision

Figure 16.1  Stages in preparing a presentation.

154  Exploring different genres

At this point the basic content of the talk should be ready, but you are not done. Add
animations to direct people’s attention to each item in the order you will talk about them.
Finally, go through the whole presentation, checking for spelling or formatting errors.
Then set it aside for several days or a week. When you look at it again, you will undoubt-
edly notice irrelevant or unclear portions and important points you missed. Fix them.
Then practice it with your colleagues, if possible, and modify it again based on their
feedback.

16.2.2 Tell a good story

Just as a research article is
not merely a report but an
You do not have time to present argument, so a presentation
your full argument, but everyone loves is not merely a report but a
story. Presentations are too
a good story. Make yours believable short for a full argument,
and compelling, and interesting but must present a compel-
to your audience. ling and believable story.
People get bored listening
to reports, but never tire of
a good story.
First you must catch your audience’s attention. As one author writes, “During the first
few minutes of your presentation, your job is to assure the audience members that you
are not going to waste their time and attention.”2 This begins the moment you step on
stage. If you appear and sound confident, even if you are shaking in your shoes, the audi-
ence will be alert. Don’t lose their interest. Just like a research article, first highlight the
importance of the work, leaving the outline for the end of the introduction.
The more mundane your results, the more crucial it is to present them in an interest-
ing way. The best presenters are not necessarily those with the best research, but those
who tell an exciting story. Fair or not, they are the ones that the audience will remember.

16.2.3 Focus
There is a symbiotic relationship between audience and presenter. If you are excited
about your work, let that show. Excitement is contagious. If the audience is looking at the
clock wondering when the agony will end, you will feel the same. Distraction is deadly,
so do your best to eliminate it.
The first distraction is the number
To maintain audience of slides. With 10 slides in a 20-minute
talk, each will be on the screen for
attention, the number of slides,
two long minutes. Unless you are
overall design, page setup, fonts, an excellent orator, the talk will also
positioning, and amount of seem to drag. On the other hand, try-
information are all important. ing to fit 60 slides into 20 minutes
allows only 20 seconds per slide, leav-
ing the audience overwhelmed and
 Academic and technical presentations  155

you out of time. Depending on your speaking style and the amount of information on
each slide, one or two slides per minute is about right for academic presentations. Fewer
slides may be better in business or industry if details are important.
A second distraction is slide design.3 Academic and technical presentations use simple
background designs to avoid cluttering the screen and restricting the space. Moving back-
grounds or images are particularly distracting because our eyes and minds are naturally
drawn to motion. Also, make sure the page setup (on the design tab) matches the screen
you will use. The default page setup may be “widescreen (16:9)” to fill the computer
screen, but most classroom projection screens are “standard (4:3).”4 Widescreen format will
cause the font and graphics to appear smaller. Changing the page setup alters the propor-
tion of images and text alignment, as shown in Figure 16.2. Changing the design has a
similar effect.
Make sure your slides can be read from the back of the room—at least font size 16,
even on graph axes, preferably 24 or larger. It is easier to read sentences written in serif
fonts (with small lines at the end of strokes, like Times New Roman and Cambria), while
san serif fonts (without those lines, like Arial and Calibri) are more suitable for headings
or small letters. Calligraphy, Comic Sans, and other fancy fonts are generally best
left for non-academic uses.
Another source of distraction is having too much on one slide. A presentation should
be spoken, not read. Spoken English incorporates repetition, pauses, and fillers to give
listeners time to process the message. The material on the slide should supplement the
talk, not distract from it. Alignment of text on a slide is also important. Figure 16.3,
explaining why it is better to introduce abbreviations as needed, demonstrates how
placing related text on one line allows our brain to process it more easily and thus retain
it better.
Complex graphics can also be distracting. Rather than reproducing graphics directly
from your writing, redraw them to highlight what you want your audience to focus on.
Limit complexity, using color, borders, or animations to draw attention to important
features. Frames or donuts mark items faster and more effectively than a laser pointer, and
remain on the screen.
Remember that the resolution and color of the computer projector is not as good
as your computer screen. Whenever possible, crop pictures to show only the area of
interest, then resize them, maintaining the proper aspect ratio. Correct the brightness
and contrast to make them easier to see. Add arrows to point to important features,
with or without a label. Positioning text over a portion of a figure that is hard to crop
out allows you to increase the size of the figure while simultaneously covering the
distraction.
Animations can effectively focus
attention, either by sequentially reveal- Animations and
ing lines of text or highlighting with transitions can focus
frames or arrows, but if overused or attention if they help
too dramatic they can also be distract- add meaning, or can be
ing. Setting animations to appear after
the previous one, with an appropriate
distracting.
delay, can reduce mouse clicks and
keep you on pace.
(a)

(b)

Figure 16.2  Comparison of projected images (a) as laid out in 16:9, and (b) changed to 4:3.
 Academic and technical presentations  157

(a)

(b)

Figure 16.3  Alignment of text can (a) impede or (b) facilitate understanding.

Similarly, transitions should add meaning rather than distracting.5 Fade connects
slides, while a turning or changing motion can indicate shifts between sections of a pres-
entation. Very dramatic transitions have a place in social or motivational slide shows,
but not academic presentations.6 Make sure you go through the whole presentation
in slideshow mode, preferably on the computer and projector you will use, to check
everything before presenting.

16.2.4 Finish strong

Feeling comfortable during the question and answer time will help you finish strong.
Perhaps the most important thing is anticipating questions that may be asked. There is
nothing more empowering than hearing a questioner ask a question, and feeling, “Yeah,
I got this!” Practicing with your lab mates can help you both predict the questions and
practice the answers.
You can even drop hints to get people started in a direction. For example, during
my doctoral defense on tomato breeding, one of the examiners was a plant physiolo-
gist, so I purposely mentioned that botanically a tomato is a berry, expecting him to
pick up on the clue and ask about it, even though it had no direct bearing on my
research. He did. As soon as he started down that track, I knew exactly what the series
of questions would be and had planned how to answer each, and had to suppress
a smile.
158  Exploring different genres

Make it easier for the audience

If you spend less time to ask questions by adding slide
looking for a slide, you can numbers. In slideshow mode, typ-
focus your attention on ing a number followed by “Enter”
causes that slide to appear. If the
answering the question and questioner does not remember the
have more time to do so. number, type “Control-S” (Show)7
to bring up a box listing the slide
titles, then click the appropriate
one. Both tricks save time and allow you to focus on formulating your answer (and
are impressive to those who don’t know how to do it).
You can also hide a slide you might need during the Q&A time. For example, you
might show one graph. Similar graphs can be marked with “Hide slide” and will not show
until you move there by slide number or title.
Many of these tricks, and others, I have learned from watching presentations. If you
see something that is interesting, but do not know how to do it, ask! The presenter will
undoubtedly be happy to tell you.

16.3 Academic oral presentations

There are at least four different
types of academic oral pres-
Class presentations, oral thesis entations: class presentations,
defense, conference papers, and oral thesis/dissertation defense,
job interviews have different goals conference papers, and aca-
demic job interviews. As with
and audiences, thus emphasize writing, it is important to con-
different things. sider the goal, audiences, and
criteria for success for each.
For a class presentation,
normally you must show you
have mastered the class content. Class presentations have two audiences: the professor
and the rest of the class. While the class wants an interesting presentation, the instructor
is looking for certain information or demonstration of skills specified in the assignment.
Since your immediate goal is to pass the class, obviously the professor is the primary audi-
ence, but since the class is the majority, don’t forget their needs.
Graduate degrees and some honors programs require a concluding oral presentation or
defense of your thesis. Since by that time you have completed both your research and writ-
ing, and know everyone in the audience, it should be a celebratory event. Nonetheless,
because of the high stakes, it is often feared. Remember that the audience is on your side.
They want you to succeed. Your advisor is convinced you are ready. If you can clearly
present what you have done and how it contributes to the field, you will get your degree.
Conference presentations not only present your research findings, but also showcase
you as a researcher. While your immediate goal may be to fulfill a graduation require-
ment, others are evaluating both your research and presentation ability. Unless there is a
prize for the best presentation, the criteria may not be stated, but novelty of your research
and clarity and interest of the presentation are always important. As with a research article,
 Academic and technical presentations  159

remember that your audience includes both experienced researchers who want to know
what you have added to the field and neophytes who need a broader picture.
When planning an academic job interview, focus on your immediate goal: getting the
job. You may be most comfortable talking about the main findings of your research, but if
the potential employer is interested in something else, try to identify some aspect of your
research that more closely matches the job description. In this case, your main audience is
those who have a say in the decision, which normally does not include graduate students,
so your talk should include more technical detail, particularly any theoretical aspects of
the work.

16.4 Academic poster presentations

In addition to oral papers, many academic conferences have poster sessions. There are
many differences between the two, including the format, the time you have for present-
ing your work, and the number of times you may do so.
First, let me explain how a poster session works. Your poster, usually no larger than
36 × 48 inches (90 × 120 cm), is mounted on a wall or movable divider. During the
poster session participants wander by, look at your poster, and ask questions if they are
interested. Thus it is important to have an eye-catching poster and be able to give a
quick but interesting summary of your research, so they will be willing to stay and listen
to more details.
Unlike an oral presentation, you cannot plan to speak in a fixed order but must be
able to interact with the questions posed. If you get a general question to talk about your
work, summarize the main contribution within two minutes. While speaking, be aware
of the body language and cues given by the listener. If they are interested, continue. If
they keep glancing away and do not ask follow-up questions, ask if they have any further
questions, and if not thank them for their time and let them move on. As people come
and go you may talk about your research several times, focusing on a different aspect each
time depending on their interest.
A poster makes the same argument as you would in an oral presentation on the
same topic, but because of the format the organization is not as straightforward. Make
sure that the main contribution of your research is highlighted and central, not tucked away in
the bottom corner. Visuals, not words, attract attention. Leave the details and longer
explanations as answers to questions,
rather than trying to squeeze every-
thing onto the poster. A poster need not be
A good poster can be read sequen- linear, but must have
tially, combining graphics and text a logical structure,
in a way that your eyes are naturally drawing attention to
drawn to the starting point and from the main points.
there proceed naturally in the order
intended by the author, which may
not be linear. The graphics by them-
selves should tell much of the story, supplemented by details in the text. There are many
ways to organize a poster to accomplish this. Figure 16.4 shows some alternatives.
Figure 16.4a shows a design that might be useful in chemical engineering to describe
three steps of fabrication of a new compound, followed by characterization to demonstrate
 Title Title

Picture Details Background Design

Algorithm Picture
Picture Details

Picture Details Testing

Comparisons
(Graphs)
Contribution
Testing and Conclusion

a. Chemical engineering b. Electrical engineering

Title Title

Background Method
Step 1 Step 2
Research
Picture question
and answer
Evaluate

Results
(Graphs and Discussion
Tables)
Comparison and
Contribution

c. Science d. Computer science

Figure 16.4  Possible poster designs.

 Academic and technical presentations  161

its formation. For each step, a picture and brief description are placed on each side of an
arrow with the name of the process.
Figure 16.4b illustrates a format that could be used in electrical engineering to present
a new design. The top half describes and illustrates the new design, while the bottom
describes the testing and comparison with other designs.
Figure 16.4c shows how an IMRD format paper could be laid out in four unequal
quadrants, with the unifying research question and answers in the middle. The picture
helps draw attention to the middle of the poster, and thus to the questions.
Figure 16.4d illustrates how a cyclical algorithm or other cycle could be shown, with
each step in a section. Comparison with other algorithms and the benefit of the research
are summarized at the bottom.
Although it is hard to show in such small illustrations, space is important in a poster.
It should look full, but not cluttered. More important elements should be in larger font,
so the viewer’s eyes are drawn to them. Graphics should be closely connected with the
accompanying text, separated from other blocks of graphics and text so the connection is
clear. Blank space can be as effective as a line in directing the viewer’s eyes.
You are attempting to give a quick visual representation, not all the details. If you
are considering landscape (horizontal) orientation, check the conference regulations, as
some only allow vertical. Sketch different possible layouts before creating a rough draft
using PowerPoint or a specialized program. After filling in the text you will have to
make adjustments. Use bullet points rather than sentences and paragraphs. Make sure you
reserve space for the title, author, and your department and school. Just like writing, after
you are clear what needs to be included, it is sometimes better to set the whole thing aside
and try again.

16.5 Technical presentations

There are as many types of technical presentations as there are technical reports. The
technical presentation accompanying a position paper, project proposal, progress report,
or concluding report will differ, because of the different goals. Audience is also important,
the main question being the background and interests of your target audience.
Many of the principles for academic oral presentations also apply to technical presenta-
tions in business or industry: timing and number of slides, simplifying graphics to show
the main point, highlighting information. Begin by laying out your argument, based on
the format of the related technical report. Fill in the details and revise to clarify. Catching
and maintaining interest is less important because of the inherent interest of the parties
involved.
Nevertheless, since technical presenta-
tions in business or industry involve many
considerations different from academia, I Your presentation
encourage you to consult other resources must provide the
for more details. From a single techni- information needed by
cal report you might need to prepare four your audience.
different presentations for managers from
research and development, production and
quality control, sales and marketing, or
accounting and finance.
162  Exploring different genres

Evaluate your understanding

(1) Why is it important to begin to prepare your presentation early?
(2) Why should an oral presentation be more like a story than a complete argument?
(3) How can you keep the attention of the audience?
(4) What techniques will help you feel confident while answering questions?
(5) How does presenting a poster differ from an oral presentation?
(6) How do technical presentations differ from academic presentations?

Homework
(1) Prepare a presentation, following the principles in this chapter.
Hint: Practice presenting it to at least one other person and get their feedback on how to improve
it before presenting it publicly.

Teaching tip: Try to provide good examples throughout the course in your own teaching, but
also learn from techniques your students use.
Teaching tip: Many books specialize in oral presentations, so this chapter is not comprehen-
sive, but shows how to apply the principles learned and provides practical hints from personal
experience.

Notes
1 Take a copy of the component list and note how many slides discuss each. Add other components
such as timeline or budget if appropriate.
2 Dale Ludwig and Greg Owen-Boger, The Orderly Conversation, quoted at http://thepitcher.
org/10-quotes-on-the-importance-of-mastering-public-speaking.
3 I will be using terms from PowerPoint, the most common presentation software. The same
principles apply to other programs, although the terms may differ.
4 Large lecture halls may use widescreen format. Ask if you are not sure.
5 In linguistic terms, animations and transitions should be semiotic, with each type used to convey
a meaning.
6 For this reason I do not recommend presentation programs like Prezi for novice presenters.While
I have seen some excellent talks done with it which use the ability to zoom in or out or return
to a previous section to highlight the storyline, the movement can be extremely distracting if not
semiotic.
7 Command-S on an Apple computer.
Chapter 17

Principles for successful email

17.1 Importance of email

The final genre we will consider is email. Email is currently the most common medium of
communication in academia, used for everything from departmental memos and teacher–
student interaction to corresponding with colleagues or editors in other countries, making
email literacy an essential part of academic English.1 Other electronic social media are
rising in importance, especially among young scholars, but whether they will supplement
or supplant email remains to be seen.
Writing email might initially seem
to have little to do with other academic Applying the principles
writing, but there are many similari- of argument structure and
ties. Without explicit instruction you rewriting in stages can improve
could gradually pick up the skill, just as your email communication,
you would eventually learn the expec-
tations of academic writing, but might
especially for formal email.
fail many times in the process. Like
other writing, important emails must
be worded just right, so first thinking about the argument structure, then writing and
rewriting in stages, can lead to more successful email.
Since email involves so many different goals, situations, and relationships between
sender and recipient, it could easily be broken into a dozen genres, but that awaits further
study. For now we will focus on principles that will help you achieve an appropriate level
of formality in any situation.

17.2 General email etiquette

First let me propose a few general principles of email etiquette, listed in Figure 17.1.
These hold true no matter the purpose or recipient of your email.
One of the advantages of email is that it provides almost instantaneous correspondence
anywhere in the world. The related disadvantage is that we have grown accustomed to
an almost instantaneous response. If you need more time to think or act, send a quick
acknowledgment and say when you will be able to provide a more complete answer.
Always use a subject line that summarizes the topic of the email, both to indicate that it
is a valid email and to make it easier for the recipient to find a week later. Similarly, make
sure you include at least one personally identifiable line of text in the email, especially if you
include an attachment or web link, so the recipient knows it is from you and safe to open.
164  Exploring different genres

1. Respond promptly;
2. Use the subject line and identify yourself;
3. Send it to the right person;
4. Be careful what you say.

Figure 17.1  General principles of email etiquette.

Before choosing “reply” or

Principles for email are no “reply all,” think who needs to see
different than any other the response, and before you hit
“send,” double-check the list of
communication. Consider recipients. In a series of correspond-
the situation and respond ence with a coauthor and editor of a
appropriately and politely. book chapter, I sent an email to all
and my coauthor replied, asking for
a response from the editor. A week
later I was wondering why we had
not heard anything, then noticed that the reply had gone only to me, even though the
body of the email was addressed to the editor. On the other hand, when working on a
book on email, the editors were corresponding with an author, asking for some changes.
The author sent a very polite response, with copies to the coauthors. Unfortunately, in
the subsequent exchange the authors continued to use “reply all,” forgetting that the edi-
tors were still receiving those emails, some of which were definitely not intended for our
eyes! (We still published their chapter, after revision.)
This leads to the final principle: be careful what you say. Remember that an email
is an electronic document, easily sent but impossible to contain thereafter. Always
be discreet because you never know who might see it at some point, as people have
found with images texted to friends or notes posted on what they thought was a private
conversation.

17.3 Position and imposition

Before discussing factors that can improve your chances of success in email, we need
to consider different styles of email, how formality affects style, and how position and
imposition determine the formal-
Email can be used for ity required. The rest of this chapter
purposes ranging from will then be devoted to the question
of how to write email that involves a
informal texting to very formal high level of formality, such as letters
or face-threatening letters, of request, invitation, or apology.2
leading to a range of styles. Email is a means of communica-
tion, not a genre of communication.
Just as ink on a page can be used for a
 Principles for successful email  165

quick note, a letter to a friend, or a formal business letter, email can be used for a range of
genres from texting format to informal memos to formal letters.
The first task, then, is to evaluate how formal a style is necessary. The formality required
depends on three factors: power, distance, and imposition, as depicted in Figure 17.2.
Power and distance are measures of the relative position of the sender and recipient,
while imposition is based on the contents of the email. Although there is some interaction
between the factors, in general the formality required is determined by the highest arrow.
Power refers to the degree of power the recipient has over the sender, often based on
their relative status. Letters from someone of low status to high status require greater for-
mality than from high to low. For example, a graduate student writing to a professor or
journal editor is expected to write in a more formal style than is required for the reverse.
As with any communication, what is considered appropriate is largely determined by
whoever holds power in the interaction.
Distance refers to how well the sender knows the recipient. The better you know the
person, the less formality is required. Thus letters from a graduate student to his or her
advisor may not need to be as formal as to another professor.
Imposition requires a little more explanation. Imposition in this case can be defined as
demanding someone’s attention or commitment, thus is closely related to the time or effort
necessary to respond. Some emails merely convey information, requiring reading but no
response, and thus involve minimal imposition. Others require a quick response, but
no additional time commitment—a low level of imposition. Email that requires a longer,
detailed response or additional thought or action carries a medium level of imposition.
Many emails entail an extended commitment of time on the part of the recipient, and thus
a high imposition, including requests, invitations, and many apologies, which often contain
related requests. In these cases, different devices can be used to minimize the feeling of
imposition, as will be discussed below.
Table 17.1 shows the relationship between formality and style, and examples of email
content suitable to each level of formality. Success in email, especially in high-imposition
situations, depends on an appropriate level of formality and use of techniques to reduce
the feeling of imposition.

High Formality

Power Distance Imposition

High Unknown High

Equal Known Medium

Low Close Low

Figure 17.2  Formality required is determined by the tallest arrow.

166  Exploring different genres

Table 17.1  Formality and style in different types of email communication

Formality Style Examples

Formal Formal salutation Request, invitation, apology

Body paragraphs Question—long response
Closing and signature
Informal Informal salutation Ongoing discussion
Sentences, capitalization Informative memo
Informal closing
Very informal No salutation Notes between friends
Text language, emoticons Email chatting, texting
No closing

17.4 Five factors affecting email success

The rest of the chapter is devoted to factors affecting the success of email in situations that
require a high level of formality, such as apologies, requests, or journal article submissions.
As shown in Table 17.2, each is related both to a general communication principle and a
writing principle described in this book.3
For informal email between friends, these factors may not be that important. Formal
emails, on the other hand, like any formal writing, may need to be rewritten several times,
paying attention to every word. It is often better to compose such letters offline in a text
editor, then paste it into the email once you are satisfied. More than one email has been sent
in an incomplete form because the sender hit the wrong key or clicked in the wrong place.

17.4.1 Terms of address

Terms of address, how you refer to
The salutation and closing the recipient in the salutation and
show relationship, how to yourself in the closing, follow an
you fit in the academic expected format. This is not merely
a formality, because these terms
community, which of address reflect relationship, an
changes over time. important element in any commu-
nication. Just as using the accepted

Table 17.2  Factors affecting email success or failure

Factors affecting success Principles

Communication Writing
1 Terms of address Relationship Fitting in
2 Justifying your request Cooperation Argument structure
3 Amount of information Clarity Coherence, conciseness
4 Reducing imposition Politeness Connection, connotation
5 Attention to detail Intelligibility Correctness
 Principles for successful email  167

format for academic writing shows that you belong to an academic community, using
the proper terms of address in academic communication shows how you fit into that
community, which will change over time as you become better known and established
as a researcher.
In formal academic correspondence, it is usually best to address someone based
on the highest title or position they hold. Thus, although it would be appropriate to
address someone as Mr. or Ms. [family name] in a non-academic context, in an aca-
demic context it is better to call someone Dr. [family name]. Since someone could hold
a doctorate but not be a professor, Professor is a higher term of address. If that person
holds a higher rank, for example Dean or President of an academic institution, they
should be addressed by that rank. Salutations appropriate for various levels of formality
are shown in Table 17.3. Professor may be abbreviated to Prof. if used with a family
name, but not when used by itself.
For those using English as a lingua franca, remember that titles attach to the family
name, not the given name. Furthermore, although the Chinese word for teacher is con-
sidered an appropriate term of address in Chinese, even for university professors, that is
not the case in English. This leads to various terms of address being considered inappro-
priate, as shown in the last row of Table 17.3.
Context is vital. Even if you know someone personally, and in a different context
might address them by first name, if you are writing a formal letter it is generally better
to use the appropriate title, at least in the initial email. Subsequent emails in a series are
often less formal, especially if exchanged rapidly, and may not contain either salutation or
closing. Asian students may feel this is cold and impersonal, whereas a Western professor
may be using it either as a sign of closeness or merely for efficiency, just as you might greet
someone at the beginning of a conversation but not at the beginning of every exchange
in the conversation.
Newcomers to the academic community often do not realize that not only do
terms of address reflect relationship, they can also be used to negotiate relationship.

Table 17.3  Academic salutations suitable to different levels of formality

Formality Address Title Name

Formal Dear President Family name

Dean Family name
Professor/Prof. Family name
Dr./Mr./Ms. Family name
Informal Hello Professor/Prof. Family name
Hi Professor -
- Given name
Very informal - Professor -
- Given name
- -
Inappropriate Professor/Prof. Given + Family name
Professor/Prof. Given name
- Family name
Teacher Family name
Teacher -
168  Exploring different genres

This is accomplished in the closing and

Closings and your signature lines at the end of the letter.
signature can be used Whereas at the beginning of the letter you
are showing your respect for and distance
to negotiate a new
from the other person, especially if present-
relationship. ing a request, invitation, or apology, by the
end you are often trying to establish a closer
relationship.
In Table 17.4 you can see a listing of various closings, from most to least formal.4
“Best” is a new variant on “Best wishes” or “Best regards” and seems to be increasing
in popularity as a closing for emails. “Thank you” was never considered an appropriate
closing for paper letters but is also increasing in popularity. A closing common among
Chinese is “Your student.” Like the salutation “Teacher,” this is a translation from an
acceptable and common Chinese form, but not recommended in English.
Different types of signatures are illustrated in Table 17.5. In the signature line you are
not merely identifying yourself but hinting at the relationship you would like to have
with the recipient. Thus, a very formal signature indicates a desire to maintain a distance,
at least in the present interaction. On the other hand, an informal signature indicates a
desire to reduce the distance in subsequent emails.
An illustration of this sort of negotiation of relationship can be seen in this series
of emails I exchanged with a professor, a well-known author in English for Academic

Table 17.4  Closings, from most to least formal

Closing Use

Most formal
Respectfully yours, Respectfully Legal documents
Sincerely General, emphasizing distance
Best wishes, Best regards, Best General, emphasizing relationship
Cordially, Warmly, Warm regards General, emphasizing close relationship
Thank you, Take care Informal
[no closing] Very informal
Least formal

Table 17.5  Types of signatures, from most to least formal

Signature Example

Most formal
Given name, Family name, Degree Gerald Rau, Ph.D.
Title, Given name, Family name Professor Gerald Rau
Title, Family name Professor Rau, Dr. Rau
Given name, Family name Gerald Rau
Given name Gerald
Nickname Gerry
Least formal
 Principles for successful email  169

Purposes, who was to be a keynote speaker at a conference I would attend. I began by

writing the following email.

Dear Prof. [family name],

I am writing a textbook, tentatively entitled Writing Your First Science or Engineering
Article in English, . . .  I was therefore very excited to find that you will be a keynote
speaker at [conference], and would like to ask if it might be possible to meet and talk
with you briefly while you are in [location].
. . .
Sincerely,
Gerry Rau

I began by addressing the professor by title and family name, and closed with a formal
but distant “Sincerely,” but in the signature line I did not use a title, indicating a desire to
move to a closer relationship. By using my nickname with my family name (my formal
given name was visible in the standard signature panel at the bottom of the email), I hoped
to reinforce my intent.
The reply was as follows.

Dear Gerry, if I may,

I would be happy to meet with you while I am at the conference. Your work sounds
very interesting and I would like to learn more. . . . 
Best,
[nickname]

The professor picked up on the clues, and accepted them, while still leaving both of us an
out if there had been a misunderstanding (“if I may”). By using the more relational “Best”
as a closing and signing only with a nickname, the professor was indicating a willingness
to interact as colleagues rather than maintaining a hierarchical relationship.
Happy that I had accomplished both of my purposes, setting up a meeting and moving
to a collegial relationship, my brief response was as follows.

Dear [nickname],
Thank you for the quick reply. I look forward to meeting you.
Cordially,
Gerry

By using only the professor’s nickname in the salutation, and my own in the signature,
with a relational closing, I acknowledged my acceptance of the first-name relationship.
The importance of cultivating relationships in academia cannot be overemphasized, even
for scientists and engineers who often prefer to work with things and ideas. Collaboration is
an inevitable part of modern research, and email correspondence plays a part in that.
170  Exploring different genres

17.4.2 Justifying your request

When it comes to writing the body of the email, use the same methods we will study for
other academic writing. Begin by laying out the argument structure, then rewrite to make
the argument more coherent and concise, rewrite again focusing on using words carefully
to reduce the feeling of imposition, and finally proofread to catch grammar, spelling, or
formatting errors. It would be ideal if you could model your email after successful exam-
ples. Unfortunately, such exemplars are harder to find than exemplar journal articles, but
over time you can learn from emails sent to you. Writing textbooks can help if you want
to improve more quickly.5
It may seem strange to think of an
email as an argument, but it is, with a
A formal request is request and justification instead of claim
structured like an argument, and support. Taking a letter of request
so it is important to closely as an example, the argument typically
connect the request and includes the request itself, a justifica-
tion for the recipient’s involvement,
justification. and an acknowledgment of the impo-
sition, in some order. The justification
can be made on different grounds,
either personal or institutional. Moreover, because it is an imposition there needs to be
room for negotiation and a reasonable timeframe to complete the request, with the final
choice left to the addressee.
As you begin writing an email request, think first about your argument. What is the
strongest justification you can make for their involvement? The personal request, “Since
you are my advisor, can I talk with you about what courses I should take next semester?”
sounds like you are placing them under obligation based on their position, and may not
be well received. It would be far stronger to base the request on institutional policies, for
example, “Since the university requires that I consult with my advisor before signing up
for classes, would it be possible to meet sometime next week to go over the options?”
The justification should be appropriate to both the request and the relationship
between sender and recipient. If you do not have an ongoing relationship, as with an
advisor or coworker, is there any other relationship you could mention that would
strengthen your case?
For example, when writing the first letter to the professor mentioned above to ask for
a meeting, since I had no personal connection, following the request in the first paragraph
I highlighted my connection with the school the professor teaches at and acquaintance with
her colleague.

My wife and I spent a sabbatical year at [professor’s university], where I [goal] while
she worked with [professor’s colleague].

I added that I was using a research method promoted by the professor.

Following a discourse analysis approach, as described in your various books . . . , I

discovered that engineering articles typically do not follow the standard IMRD for-
mat, but an analogous format that I call IPTC.
 Principles for successful email  171

Finally, I stated that I would be presenting at the same conference where the professor
would be giving the keynote speech, to cement my relationship with the field.

I will be presenting the format and the way I teach it during one of the concurrent
sessions at [conference], as well as participating in an invited panel . . . 

To minimize imposition, I left the timeframe open, both as to length and timing.

Please let me know if you would be able to spare a few minutes sometime during or
around the [conference] to talk with me about these questions.

Because I hoped to meet and talk with this professor, I gave more detail about myself
than I would have if I were merely requesting information. The following example, writ-
ten from a professor in Taiwan to a distinguished emeritus professor in forensic linguistics,
shows how the same elements can be combined in a single paragraph. The first sentence
shows the two are in the same field, the second talks about having taught at the same school,
while the third points to common research interests. These, along with common acquaint-
ances, are some of the most frequent means of establishing relationship. The request is
found in the last sentence.

I’m a sociolinguistic variationist, who organized the [conference] in Taiwan. I also

taught at [college where distinguished professor once taught] from [dates]. As I
have taught courses related to forensic linguistics during the past 6 years at [uni-
versity in Taiwan], I’m working on a government book project proposal with the
goal of purchasing 2,000 books in forensic linguistics to enrich Taiwan’s research
and teaching abilities in this field. Would you be willing to serve as one of our
consultants to ensure the titles of the books we are planning to purchase are really
the most important ones?

This brings up another question about

argument structure. Should the jus- The justification can be
tification be stated before or after the stated before or after the
request? 6 Like all arguments, although request, depending on
there is no fixed pattern, there are com-
mon features and, as with journal arti-
the recipient and the
cles, the expected order is different in length of the email.
Eastern and Western writing styles. In
American style, it is common to make
the request first, then build the argument to justify that request, as I did in the email just
cited. In Asian style, the opposite is the case, with the justification presented before the
request, as in the second example above.
Since the second example is a brief email, only one paragraph, placing the request at the
end is not a problem, but if a longer justification is required it is better to place the request
earlier. Busy academics may not read a long email unless the purpose is clear within the
first paragraph. By analogy, in a longer email the first paragraph serves the purpose of an
Introduction. Just as the research goal is found at the end of the Introduction of a research
article, the purpose of an email is often placed at the end of the first paragraph of an email.
172  Exploring different genres

When making a request, try to put yourself in the position of the recipient. How would
you feel if you received a similar letter? How much time will it take for the recipient to ful-
fill your request? How much lead time are you giving them? Make sure that the timeframe
is appropriate to the work involved. Emphasizing your need, as in “Since the deadline is
tomorrow, could you please look at this and return it to me with comments immediately,”
only works if you are in a position to impose on a subordinate, and is still resented. Apologies,
such as “I am so sorry to get this to you so late, but if you could find the time I would really
appreciate your comments,” are more likely to elicit a favorable response. Remember that
your recipient already has other things planned and is trying to fit one more thing into a busy
schedule. Why should they do that for you? It is important to be as diplomatic as possible, and
remember that it is a request, not a demand—the final decision is in their hands.

17.4.3 Amount of information

Many academics now receive tens if not hundreds of emails every day, so the amount of
time spent on each must be limited. Just like an article, too little information will make
it unclear, but too much wastes the reader’s time, and makes it more likely that they will
skim and perhaps miss pertinent information.
How much information does the recipient need? Enough to be able to respond appro-
priately. After you have decided what information to include, in what order, and written
a first draft, it is time to rewrite for coherence and conciseness. Do not be like Blaise
Pascal, who wrote, “I have made this letter longer than usual, only because I have not had
the time to make it shorter” (Pascal, 1657). While that may work for letters written to a
friend, it is not a good practice in academic email.
Consider first how much information to include about the relationship. If the recipient
knows you well, it is not necessary to mention the relationship at all. If you are writing to
a professor in a class you will be taking, it may be helpful to mention the connection, e.g.
“I am signed up for your Engineering writing class.” For most professors, your signature
is all that is necessary after the first few weeks of class, unless it is a very large class. If you
have met the person before, but have not seen them for a while, reminder of the context
may be sufficient, e.g. “I don’t know if you remember me, but we met when you spoke
at [conference or location] last year.” As discussed in the previous section, for an initial
contact it may be necessary to establish a basis for relationship.
Next consider how much informa-
tion they need to successfully comply
with your request. Provide a direct
You need to provide statement of the purpose of the email,
enough information for sufficient background information, and
the recipient to be able to enough detail that they can act on your
respond to the request, request immediately if they choose. If
but not too much. appropriate, additional information can
be attached as a downloadable docu-
ment rather than putting too much
detail in the email itself.
One of the most common email requests graduate students write is to be excused from
class. Which of the reasons shown below do you think would be most appropriate when
asking to be excused from class?
 Principles for successful email  173

“I cannot come to class next week . . .”

1 for personal reasons.

2 because I need to take my mother to the doctor.
3 because my mother has chronic arthritis, and can’t get around by herself, so I have to
go home to take her to see the doctor, and I won’t be able to get a bus back in time
for class.

Many students think choice 1 is sufficient, but most teachers prefer choice 2. As a
teacher, I care for my students, and would like to know whether “personal reasons”
means there is a problem, and if so whether it is likely to be short or long term, or
whether it is something positive like a conference or job interview. Details may not be
important unless I am your advisor, but knowing the general nature of the reason will
determine whether an offer of assistance in making up the work or a message of encour-
agement or congratulation is appropriate. Similarly, in other request situations the goal
is to provide the recipient with enough information that they can judge the importance
and respond appropriately.
It should go without saying that the reasons given should be truthful and credible.
If the recipient suspects either not to be the case, it is doubtful that the request will be
granted. Thus, if the true reason sounds incredible, it may be useful to provide or offer to
provide evidence to substantiate it.

17.4.4 Reducing imposition

Emails that impose on another person always run the risk of rejection. To minimize that
risk, it is important to ask as politely as possible. This involves a second stage of rewriting,
paying attention to connotation.
Compared to Asian languages, English is often viewed as being very direct. While it is
true that in a face-threatening situation an English speaker will often speak directly to the
other individual, instead of transmitting a message through a third person, as is common
in Asia, the manner of speaking in that situation is often very indirect. This is illustrated
in Figure 17.3.
Politeness in English is often achieved by indirect request. As an example, which of the
following requests do you think is the most polite?

1 Could you please look over my paper and give me comments by next week?
2 I want you to look over my paper and give me comments by next week.
3 Please look over my paper and give me comments by next week. Thank you.
4 I would appreciate it if you could look over my paper and give me comments on it
by next week.

Most students quickly eliminate choice 2, but some using English as a lingua franca
believe that choice 3 is the best, because it uses both “please” and “thank you.”
Unfortunately, this form is viewed by native speakers of English as a veiled command
rather than a polite request. Choice 1, an indirect request, is considered more polite,
but choice 4, the most indirect, is considered the most polite, even though it lacks a
“please.”
174  Exploring different genres

A B

Chinese

A
B
Indirect Speech
English

Figure 17.3  Indirect speech in Chinese and English.

Some other ways of requesting

In English, less direct are shown in Table 17.6, from most
requests are considered to least polite. Notice that all of the
most polite forms utilize a modal verb
the most polite, leading
(could or would), which has the effect
to extensive use of of making the action potential rather
modal verbs. than actual. The less polite forms use
the modal “can,” which can indicate
ability as well as potential, or the word
“hope,” which conveys lack of certainty that you will help, although it still focuses on my
need rather than your choice, as do those that are considered impolite.7
The choice of verb is also important. A Chinese student studying in the US pointed
out that

use of Want Statements along with the word “help” is likely to project a negative
image as a needy, helpless student, which is quite the opposite from the attitudes such
as independence and confidence that a graduate student is expected to demonstrate
in U.S. academic culture.
(Chen, 2006: 44)

Rephrasing your request to indicate your desire for guidance or advice conveys rather
that you respect their counsel but are capable of independent action and decision.

17.4.5 Attention to detail

The final stage of rewriting, the one most speakers of English as a second language worry
the most about, is probably the least important. In academia, more and more speakers of
 Principles for successful email  175

Table 17.6  Request formats, from most to least polite

Most polite I wonder if it would be too much bother for you to . . . 
I wonder if you could . . . 
I would appreciate it if you could . . . 
If possible, I would like you to . . . 
Would you mind . . . 
Polite Could you please be so kind as to . . . 
Could you please . . . 
Could you . . . 
Less polite Can you please . . . 
Can you . . . 
I hope you can . . . 
Impolite Please . . . 
I want you to . . . 
I need you to . . . 

English as a first language are becoming aware of world Englishes and their counterparts
in other countries who speak English as a lingua franca. As long as the email is com-
prehensible, correct grammar and spelling are far less important than the factors already
mentioned. As with writing an article, correction of these small points of English usage
should be postponed until the very end, after all the other factors have been considered.
The reason that attention
to English usage is the least
important factor comes back Grammar and spelling are the
to a theme repeated through- least important aspects of email
out this chapter: relationship.
success in English as a lingua franca
Salutation and closing are use-
ful in negotiating relationship. communication because they do
Justifying why a person should not affect relationship.
be involved is important in
developing or maintaining a
relationship. Concern for the
recipient, as shown in care for clarity, conciseness, and politeness, also contributes to
positive relationships. As long as there is mutual intelligibility, correctness of form does
not adversely impact relationship (think of all the errors in texting among friends). Since
your time is as precious as theirs, use what time you have on things that promote not only
understanding, but relationship.
Nevertheless, in some cases spelling may be important. One letter sometimes makes a
huge difference. For example, I once had a student named Chester who mistakenly signed
his name as “Cheater.” Fortunately, the email was not about a question of academic honesty.

Evaluate your understanding

(1) What are the four basic principles of email etiquette?
(2) What factors interact in determining the level of formality appropriate for an email?
(3) What is meant by terms of address, and why are these important in email success?
(4) How can terms of address be used to negotiate a closer relationship?
176  Exploring different genres

(5) What might affect your choice of where in an email to state a request?
(6) How can you decide how much information to include?
(7) What is the simplest method for reducing the imposition of your request?
(8) Why are details like grammar and spelling the least important factor in email success
when using English as a lingua franca?

Homework
(1) The next time you write a more formal email, apply the principles outlined above to
construct your argument and rewrite before sending.
Hint: If you apply these principles in writing to a professor who has taught them in a writing
class, it is even more likely you will succeed in your request, because they will know you were
paying attention.

Teaching tip: Although this is not directly related to writing an academic paper, students
often say it is one of the most useful lessons in the course.

Notes
1 One very common question from those using English as a lingua franca is whether email is count-
able, i.e. whether to say “I got a lot of email today” (collective) or “I got a lot of emails today”
(plural). Either is acceptable, depending on whether you want to talk about volume or number.
2 My interest in email began while listening to a conference paper on email given by Yuanshan
Chen at AILA 2011 in Beijing. Among other things, she said that one cause of email failure was
“insufficient lexical and syntactical mitigation devices,” a great phrase for linguists, but incom-
prehensible for others. While the following discussion is based on many pragmalinguistic and
sociopragmatic principles, it is written for scientists and engineers, thus the focus is on the
mechanics of writing rather than linguistic or communicative principles. For an excellent litera-
ture review and longitudinal study of email use by a graduate student using English as a lingua
franca, see Chen (2006).
3 These have been distilled from various literature on interpersonal communication and email,
as well as personal experience. While related to general communication principles identified by
various authors, they are not specifically based on one model. An earlier outline of this section
was presented as an appendix in Rau and Rau (2016).
4 Many other variants could be included and are easily found with a quick Internet search, includ-
ing specialized closings such as “hope this helps,” “looking forward to seeing you,” or “in haste.”
5 For example, Swales and Feak (2000: 228–256).
6 Request first is called deductive; justification first is called inductive.
7 Linguists call the first “Query preparatory” and the second “Want statements.”

References
CHEN, C.-F. E. 2006. The development of e-mail literacy: From writing to peers to writing to
authority figures. Language Learning & Technology, 10, 35–55.
PASCAL, B. 1657. Lettres Provinciales, No. 16.
RAU, D.-H. V. & RAU, G. 2016. Negotiating personal relationship through email terms of
address. In: CHEN, Y.-S., RAU, D.-H. V. & RAU, G. (eds.) Email discourse among Chinese
using English as a lingua franca, Singapore, Springer.
SWALES, J. M. & FEAK, C. B. 2000. English in today’s research world, Ann Arbor, MI, University
of Michigan Press.
Part 4

Creating your masterpiece

Chapter 18

Writing in stages

18.1 The hardest lesson for beginning writers

This chapter marks a major turning point in the book. In Part 1 you learned that argu-
ment is fundamental to academic writing. In Part 2 you examined research articles in
your field to determine the argument structure. In Part 3 you discovered how arguments
differ depending on the genre. Now you will apply all that information to your own
writing. Since most of you are students doing some sort of academic writing, that will be
the focus, but the same principles apply to technical writing. The method I am suggest-
ing involves prewriting, writing, and extensive rewriting. You may wonder if all that is
really necessary. I can sympathize.
It is somewhat ironic that I have become a writing instructor, let alone author of a text-
book on academic writing. In high school and college I wrote research papers by making
notes on notecards, sorting the cards into order (that was my prewriting), and typing the
final draft directly (the day the assignment was due), without a rough draft or any rewriting.
We called it keyboarding. Unfortunately, that did not work when I got to graduate school,
and it is not an option for you.
Will you actually have to do
extensive rewriting? In a word,
yes, and the amount of rewriting You will have to do extensive
required will continue to increase. rewriting to meet the demands
Papers written for class are read by of coauthors, reviewers, and
one person, the professor, and not editors, as well as the needs of
rewritten after submission. A thesis your audience.
will be read by several professors,
each of whom will require changes
before it is accepted. A published
article will be read by your advisor, coauthors, reviewers, and the editor, all of whom will
require changes after you think it is perfect. Moreover, it may be read by hundreds or
even thousands of people, who will form their impression of you and your research based
on that one paper. You need to make sure it is good. The amount of rewriting required
is perhaps the hardest lesson for beginning writers to accept.

18.2 Why prewrite?

You may wonder, if you have to rewrite so much anyway, why spend time prewrit-
ing? Although rewriting is inevitable, prewriting can minimize the scale. There is a big
180  Creating your masterpiece

difference between adding another reference, modifying the description of a process, and
having to totally rewrite an entire article because it was poorly organized.
Prewriting is important for another reason: it helps overcome the inevitable writer’s
block that comes when beginning a large project. According to English author Neil
Gaiman, “Being a writer is a very peculiar sort of a job: it’s always you versus a blank
sheet of paper (or a blank screen)
and quite often the blank piece
of paper wins” (Gaiman, 2004).
The goal of prewriting is not to The purpose of prewriting is to
figure everything out, but to get you reduce the chance that the blank
screen wins.
started—more like an artist’s sketch
The goal of prewriting is
than an architectural blueprint. not to figure everything out,
to know exactly what the final
product will look like, but to get
you started in the right direction.
It is not so much an architectural
blueprint as it is an artist’s sketch. This is very different from the days before personal
computers, when retyping was time-consuming and costly, so minimizing revision was
essential to success.
Perhaps it would help to give another analogy: writing is like putting together a 5000-
piece jigsaw puzzle. Although strategies differ, most people first separate the puzzle pieces
according to their main colors or other common features, while at the same time identify-
ing the edge pieces. Once the edge pieces are in place, they begin to build inward from
some prominent feature, some distinctive color. Undoubtedly, some pieces of that color
belong in a different part of the puzzle, so when you get stuck, switch to a different color
or part of the puzzle. Follow the main lines in the puzzle first, then fill in the gaps around
them. Join pieces that fit together even if you don’t yet know where the section belongs.
The more pieces you put together, the easier it becomes to see the whole picture and
decide where the rest might go. The
larger the puzzle, the more the initial
The writing steps sorting helps.
suggested in this book are Times and methods change, so
a scaffolding to help you you should feel free to adapt the pro-
get started, not a required cedures I suggest to your situation
pattern to follow forever. and work habits. Nevertheless, taking
time to prewrite will save you time in
the long run, as will following each of
the rewriting steps in order. As you
become a more experienced writer, you will be able to reduce or combine some of the
steps. They are a type of scaffolding to help you learn to write, not a fixed pattern to
follow rigidly forever.

18.3 Write as you go

As a writer myself, I have found one of the most important principles of writing is to
write as you go. Whenever an idea comes to you, capture it. Novice researchers tend to
 Writing in stages  181

think that they must have all the pieces in hand before they start writing. Actually, it will
save time if you draft portions as you go, and it may be easier not to write the paper in
its final order.
The two portions of a research article that graduate students find the hardest to
write are the Discussion and the Introduction. Those writing in IMRD often have dif-
ficulty with the Discussion because there is no single correct answer. You are trying to
compose an argument to convince your readers to see the data the way you do. That
is the basic challenge of research. On the other hand, both in my own writing and my
work as an editor, it is always the Introduction that needs the most revision. Even if
you know the literature, it is
not easy to decide how much
to include or how to connect The Introduction often requires
the various threads to create a
the most revision, because it is the
cohesive picture.
The first Introduction or only division that does not describe
literature review you write your own work, but becomes
will be the hardest. Another easier as you know the field.
analogy may help explain why
it is relatively easy for an expe-
rienced researcher to write an
Introduction, but it is often the hardest division of the paper for a novice. Before you
continue reading, think about the following question: how is writing an Introduction like
giving directions to a new house?
Consider the growth of your field of research as the growth of a town. Someone
was the first to settle in that area, a pioneer. Many pioneers wind up living and work-
ing by themselves for their whole life, but if they find something valuable, others will
start moving into the area and claiming a share of the land. Eventually a small town
may be established, in which everyone knows one another. Over time, as more and
more people move in, straight, wide roads are built to replace or connect the small
winding paths that were there at the beginning. Everyone knows the founders, but
gradually different neighborhoods take on their own character, and people generally
stay within their own neighborhood. As the town gets more crowded, newcomers
must either build on the edges or the few vacant lots in the middle. Later still, some of
the old homes may be torn down and replaced by new development.
If you want someone to find your new house, you need to give them directions. You
will not talk about every road in town; only those they need to follow. You may men-
tion famous landmarks along the way by name, but in general you will just tell them
what roads to turn on. The same is true in writing an Introduction. You do not need to
mention every paper in the field—only those that lead to your present work, with special
mention of landmark papers along the way.
The pioneer has seen all the development, seen every road as it is built, knows where
each leads and who lives on it, and has met each person as they move into town. As a
newcomer you may have no idea where the various roads lead, or who lives on them.
You may have a map with street names, but do not know where the landmarks are or
what they look like. Nevertheless, the longer you live in the town, the better you know
it. When a new building springs up, you see it being built and begin to talk about it
before it is even finished. You know various routes to any location in town. The same
182  Creating your masterpiece

is true with an academic field. The longer you are there, the easier it will be to direct
people to your work.
Next, think about what you encounter when you read a research article, and when
you would be able to write each. The first thing you see is the Abstract, but since that is a
summary of the article, it cannot be finalized until the article is complete. The first part of
the article itself is the Introduction, but as I just said, that is one of the hardest sections of the
article to write. Moreover, the purpose of the Introduction is to point to a research gap and
show how your work is going to fill it, which means you need to have done that work and
be sure you can indeed fill that gap in previous research before you can write it.
But why wait until the work is
complete to start writing? When are
Writing up your work as you going to be most clear about what
you do it is easier than methods you used—as you do them
trying to remember or or six months later? Why not write it
as you do it? Undoubtedly you will
decipher your notes six
change the wording later, but at least
months later. you won’t be struggling to remember
what those cryptic notes mean that
were once so clear.
In practice, this means writing the Methods or Process division first, as you perform
the work, followed by the Results or Testing. Since the analysis depends on the research
questions asked, it is beneficial to write the Introduction next, to make sure you are clear
what the goal is and what has already been done, before returning to write the Discussion
and Conclusion.
One other thing I find helpful when writing is to keep at least two files open: the sec-
tion I am writing at the time and another file with an outline and notes on the whole
project. As I am writing one chapter, ideas will come about how to reorganize or what
to add to another chapter. When the idea comes, I simply switch to the other file, record
the idea before I forget it, then go back to my main document.
Furthermore, when away from the
computer, I keep some way of record-
The best wording may ing notes handy. In the past, that was a
occur to you while writing pen and paper. Now the omnipresent
something else, walking across cellphone fulfills the same purpose.
Whenever the Muse speaks, I am
campus, or when you wake up. ready. Frequently the best wording
Be ready to capture it. comes at an odd moment, walking
across campus or lying in bed, but if
you do not record it then, it will be
gone forever. As best-selling author
Saul Bellow said, “You never have to change anything you got up in the middle of the night
to write” (in Bear, 1992: 93).

18.4 The 7Cs of Change

I propose a method of rewriting I call the 7Cs of Change.1 A summary of the 7Cs is shown
in Table 18.1. The first two Cs (Coherence and Conciseness) constitute rewriting at the
 Writing in stages  183

Table 18.1  The 7Cs of Change

Type of change 7Cs Purpose Level Scope

Rewriting 1 Coherence Organization Paragraph Section

2 Conciseness Sentence
3 Connection Clarification Sentence
4 Connotation Words
Revising 5 Consistency Finalization Words Whole paper
6 Correctness Format
7 Collaboration All

level of the paragraph or sentence to ensure your argument structure is clearly organized
and eliminate unnecessary material. The third and fourth (Connection and Connotation)
involve restructuring sentences and choosing words to further clarify your argument.
The last three (Consistency, Correctness, and Collaboration) comprise revision,2 looking
at the whole document again to check for consistency of word use, spelling, formatting,
and other such issues, or seeing it from a new perspective after asking for comments from
coauthors, making sure your argument is not misunderstood.
Inexperienced authors make
relatively few changes them-
selves during rewriting. As It makes sense to get the
a result, they need to make organization right before working
extensive changes at the level on details, because every change
of collaboration, after their
advisor or coauthor reads it. in order requires rewriting several
Experienced authors, on the surrounding sentences.
other hand, learn to make large
organizational changes before
working on smaller ones like
word choice. This reduces the total work since moving or adding even a single sentence
requires rewriting several sentences around it. We will examine the 7Cs individually in
Chapters 21–23.

18.5 Do it right the first time

Some things are worth doing right the first time, as it will save considerable time in later
revision. This includes formatting problems created by our dependence on computers and
some common punctuation errors.
Some publishers provide templates
for submitting articles. These templates Small formatting and
have sample sections, paragraphs, tables, punctuation errors are
and figures. Once your text is ready, easy to avoid but take time
you simply delete their text, insert your
own, and embedded codes ensure it is
to correct later.
properly formatted. Many universities
have standards for formatting theses and
184  Creating your masterpiece

dissertations, specifying how paragraphs are to be formatted or the number of spaces fol-
lowing a period, but lack such a template. By checking the requirements before you start
to write, you will avoid many tedious small changes later.
Other problems with formatting arise from computer defaults. When you copy and
paste in MS Word,3 the default is “keep source formatting,” which may differ from your
document. The difference between Times New Roman and Cambria or between size 12
and 11 font may not be obvious (did you notice that each was written in the respective
font/size?), but the editor will ask you to change it. Using “keep text only” when pasting
plain text into your document will save time later.
One formatting issue involves line
spacing for Chinese, Japanese, and
Some common problems Korean. Lines copied from such docu-
arise from pasting from ments, even if typed in English, will
other documents, others retain the wider line spacing designed
from foreign language to accommodate characters. To correct
this in MS Word, uncheck the box on
entry systems. the paragraph settings to “snap to grid
when document grid is defined.”4
Another problem for Asian users
is the difference between typing English and full English entry. In the former all the
punctuation marks, parentheses, and other marks will appear in Asian format rather than
English. The difference is illustrated in Figure 18.1.
The shading, none of which include spaces, illustrates the problem. In Chinese a punc-
tuation mark takes the same amount of space as a character, but in English it is smaller
than the average letter. Thus a Chinese parenthesis appears to have space both outside and
inside which is part of the parenthesis character itself. Similarly, periods and commas look
like they have spaces around them when they really do not. Chinese writers will often
either follow the same entry in English(not leaving a space outside the parentheses)or try
to make it look like what they are used to seeing in Chinese ( adding spaces both inside
and outside the parentheses ) . Nevertheless, when in America (or Britain) you must do
as the locals do, with no space inside the parentheses but a space before and after them
(unless followed by punctuation). The examples in the previous sentence can serve as a
model of proper English usage.

In an English font, punctuation is at the bottom of the line and takes a small amount of
space (and there is space around the parentheses only if you specifically add it). The line
spacing is noticeably narrower than Chinese single line spacing.

In a Chinese font that has defaulted to English炻punctuation may be in the middle of the
line and take as much space as a character炷like this炸so there appears to be space
around the parentheses or after punctuation when there really is noneˤThe line spacing
is also wider, to leave vertical space for Chinese charactersˤ

Figure 18.1  English vs. Chinese punctuation and line spacing.

 Writing in stages  185

Exercise 18.1 Instructions to authors

(1) Check the formatting guidelines for your writing. Ask your school for their thesis/
dissertation guidelines or look on the journal homepage for instructions to authors,
author guidelines, submission guidelines, or some similar link.
Hint: We will refer to the instructions to authors several times during the rewriting process. File
them for future use.
(2) Look through these guidelines to determine their requirements for font, font size,
paragraph formatting, spaces after periods, British or American English, and any other
general formatting features that affect the whole paper.
Hint: Do not worry about headings, tables, and figures (which will be a large part of the instruc-
tions).We will come back to these in Chapter 24.
(3) If your school or journal does not provide a template, create your own template,
incorporating the font, font size, and paragraph formatting specified.
Hint: This may seem like a waste of time, but it takes far less time to make a template and use
it consistently than it does to reformat every document.

18.6 Avoiding plagiarism

One final general issue we need to consider is plagiarism. Novice writers often want to
know what constitutes plagiarism and how it can be avoided, particularly when citing
numerous works in the Introduction.
Plagiarism is a particularly Western idea, based on the ideas of individualism and indi-
vidual ownership, and therefore hard for Eastern cultures to understand. In many Eastern
cultures the teacher is the expert. To quote their answer verbatim is a sign of respect.
Memorization is highly valued and often an answer must be quoted perfectly to receive
credit. Knowledge, like many other things, is communal property. In Western culture,
the reverse is true. The teacher is a facilitator, encouraging students to think critically and
form their own opinion. Quoting verbatim without citing the author is plagiarism, and
the value of direct quotes is minimized, as it hinders self-expression. Memorization is not
valued in the school system, with test questions worded to ensure that even if students
memorize the entire textbook, they can only get full credit if they can express the answer
in their own words. Knowledge, like other commodities, can be copyrighted and indi-
vidually owned (intellectual property rights).
Although the concept of plagiarism is culturally
determined, the need to avoid it has become an Plagiarizing one
accepted part of academic life worldwide. A student
time can ruin an
caught plagiarizing may fail a class. As an academic
or professional it can lead to censure or dismissal. academic career.
Thus when writing in English, Asian students (and
those from similar cultures) need to set aside their
own thoughts on what is appropriate (just as they must set aside their own thoughts on
what is an appropriate writing style) and follow English standards.
186  Creating your masterpiece

Plagiarism is defined as using the words or ideas of another person as if they were
your own.5 In other words, you need to either (1) quote the person directly, (2) change
it slightly and cite their work, or (3) change it enough that it is no longer recognizably
theirs. Direct quotes are common in the humanities and social sciences, where whole par-
agraphs may be quoted, but it is rare to quote even a single sentence in the natural sciences
and engineering. Since you want to relate your work to that of others, you cannot make
it unrecognizable. That leaves the second option. You must learn how to change the
wording and cite the previous work. But the question is, how much change is necessary?
To illustrate what is considered plagiarism, in Table 18.2 compare what Swales and
Feak (2004) wrote, quoted directly in the second column, with what I have written in the
last column. Which of the examples would you consider to be plagiarism? Note that each
of my rewrites attempts to illustrate the principle Swales and Feak mention.
If you think that the
first two are plagiarism, but
The amount you can copy from the latter two are not, you
would be correct in most
another work without citation varies cases, although the amount
by field and division of the paper. of similarity allowed var-
Rewriting several times helps ies somewhat by field. In
avoid plagiarism. highly technical fields like
computer science it is not
uncommon to reuse whole

Table 18.2  Examples of plagiarism and non-plagiarism

Modification Swales and Feak (2004: 173) Plagiarism or not?

Substituting “Copying a paragraph making only Copying a paragraph, but making a few
small changes, such as replacing minor changes, like changing some of
a few verbs or adjectives with the verbs or replacing adjectives with
synonyms.” other words meaning the same thing.
Reordering “Cutting and pasting a paragraph Putting one or two sentences in a different
by using the sentences of the order, or leaving one or two out, but
original but leaving one or two basically cutting and pasting a paragraph
out, or by putting one or two from the original.
sentences in a different order.”
Rewriting “Composing a paragraph by taking If all the authors in your field use the same
short standard phrases from a standard phrases, you will need to use
number of sources and putting them as well, but by putting them in
them together with some the context of your own work you
words of your own.” make them your own.
Paraphrasing or “Paraphrasing a paragraph by If you totally reorganize a paragraph,
expanding rewriting with substantial making it longer by adding examples as
changes in language and I have to the categories suggested by
organization, amount of detail, Swales and Feak (2004) or summarizing
and examples.” points to make it shorter, this is
considered paraphrasing.

Adapted from Swales and Feak (2004).6

 Writing in stages  187

sentences from previous works without citing the original source.7 Method sections or
other highly condensed writing may be essentially equivalent apart from the different
variables or materials used.
The American author Flannery O’Connor said, “There may never be anything new to
say, but there is always a new way to say it” (O’Connor, 1969: 76). Nevertheless, in prac-
tice, how can you avoid plagiarism? The best method I have found is to read your source at
least twice, to make sure you understand the content. After that, make notes in your own
words, retaining only the technical words that cannot be changed. Use phrases rather than
full sentences and try to write things in an order that makes sense to your own research,
rather than the order of the original paper. Finally, close the source and write your summary
or paraphrase, preferably at least a few hours after making the notes. Most importantly, write
in stages, as I will describe in the following chapters. If you follow this procedure, your final
version should have been rewritten enough times that plagiarism is not an issue.

Exercise 18.2 Paraphrasing practice

Read one portion of one of your exemplar articles. Set it aside and write one sentence
summarizing that portion. Do this for each portion of a division. When you read these
sentences in order, it should reveal the structure of the argument in that division.

Hint: Choose portions longer than a paragraph, perhaps related to one component or test, which
often begin with sentences indicating a major transition.

Evaluate your understanding

(1) What is the advantage of prewriting?
(2) What are two advantages of writing as you go?
(3) Why does it make sense to rewrite several times, beginning with the largest changes?
(4) Why should you pay attention to some small details such as punctuation as you write?
(5) What is the best way to avoid plagiarism, and why is it important?

Homework
(1) Set goals for finishing different stages of your writing, to make sure you will be able
to finish by the deadline.
Hint: Allow enough time for rewriting. I suggest the following as a rough guideline, based on the
percentage of time available before your deadline:
10% Prewriting (argument outline, brainstorming, linearization), Chapter 19
40% Writing, Chapter 20
40% Rewriting (Cs 1–4), Chapters 21–22
10% Finalization (Cs 5–7, graphics, references, citations, title, abstract), Chapters 23–26

Teaching tip: Setting deadlines for different stages of writing provides a scaffolding for stu-
dents to follow when they do their own writing later and ensures that the work is not all done
at the last minute.
188  Creating your masterpiece

Notes
1 I am not the first to come up with a writing framework based on 7Cs. John Blake called my
attention to at least three previous schemes:The Seven C’s of Effective Communication (www.
managementstudyguide.com/seven-cs-of-effective-communication.htm), The Seven Cs of
Business Letter Writing (http://epi.sc.edu/~alexandra_rowe/FOV3-001000D5/S002A8538),
and one presented in an old Hong Kong business writing book entitled Reader-friendly writing
(out of print). LaPlante (2018) suggested 5Cs of Technical Writing. My 7Cs of Change were
developed entirely autonomously, before hearing of any of these.
2 In American English rewriting and revision are often used synonymously, but I am separating the
ideas of rewriting (literally, writing again) from revision (literally, looking again). In British usage,
to revise means to look over material again, often in preparation for an exam, which is called
review (another way of saying looking again) in American English.
3 Throughout the book I will use MS Word as an example of a word processor, because it is by
far the most widely used. Terms differ in other programs. LaTeX is used extensively in computer
science and mathematics but is rare in other fields.
4 This option is not visible in English versions of Word unless one of the Asian language packs is
also installed. Because both are “single” line spacing, although the actual spacing differs, selecting
single line spacing does not work.
5 If a publisher holds the copyright, you cannot even copy your own words without permission.
Thus you cannot publish material that is substantially the same in two different journals, even in
a different language.
6 The original list also includes direct copying and quoting, neither of which involve modification.
7 John Blake (personal communication).

References
BEAR, J. 1992. The #1 New York Times bestseller: Intriguing facts about the 484 books that have been #1
New York Times bestsellers since the first list in 1942, Berkeley, CA, Ten Speed Press.
GAIMAN, N. 2004. On writing. Neil Gaiman’s journal [Online]. Available from: http://journal.
neilgaiman.com/2004/02/on-writing.asp.
LAPLANTE, P. A. 2018. Technical writing: A practical guide for engineers, scientists, and nontechnical
professionals, Boca Raton, FL, CRC Press.
O’CONNOR, F. 1969. The nature and aim of fiction. Mystery and manners: Occasional prose,
London, Macmillan.
SWALES, J. M. & FEAK, C. B. 2004. Academic writing for graduate students: Essential tasks and skills
(Vol. 1), Ann Arbor, MI, University of Michigan Press.
Chapter 19

Prewriting

19.1 Planning your masterpiece

In the first chapter I compared
learning to write to learning
to paint. An experienced artist Prewriting does not specify
may paint on a blank canvas, every detail, but helps you decide
but for a beginner it is useful what the focus of attention is,
to sketch out your masterpiece
before picking up the brush.
other elements to include, and
In this chapter you will learn how they relate to one another.
how to sketch the broad out-
line before beginning to write,
modeling successful exem-
plars, if available. This is not a paint-by-number. You are not trying to get every detail
in place first, then just fill it in. Rather it is a quick sketch to make sure you know what
the focus of attention is, what other elements you plan to include, whether each will be
foregrounded or backgrounded, and how they relate to one another.
This chapter is not about prewriting in general. There are many good books and
websites on academic writing, so I will not duplicate their efforts.1 Rather, I will tie the
basic techniques in with the principle of component analysis. My goal is not to explore
all possible methods of prewriting, but to encourage you to see its importance and find a
method that works for you.

19.2 Initial sketch: argument structure outline

In Chapters 11–15 we discussed many genres of writing, how they differ in goal and audi-
ence, what components they are likely to contain, and how to find exemplars. Now it
is time to lay out the overall structure of
whatever you want to write.
I suggest you begin with the expected Lay out your writing
argument structure rather than a content based on the argument
outline. At this point you understand the structure, rather than a
argument structure of research articles, topic outline by content.
the general claims of each division, the
component claims and where they are
190  Creating your masterpiece

normally found, and the evidence used to support each claim. If you are writing a research
article, you can use this information to plan your own writing. If you are writing some-
thing else, by comparing the typical argument structure of that genre with research articles
you will be able to lay out other types of writing as well.
Before writing anything, the following steps should be taken:

1 Obtain examples of successful documents of that genre, if possible;

2 Determine the overall organization and structure;
3 Identify the main claims and components;
4 Identify what is used for support;
5 Speculate on reasons for the sequencing;
6 Evaluate which document is the best pattern to follow for your own writing;
7 Outline the argument structure for your own writing based on that pattern.

In Chapters 1–9 you learned how to do steps 1–6 for research articles, so let’s proceed to
step 7. If you are writing something else, adapt the outline to the structure appropriate
to your genre.
Let’s say that we want to write an article for the same journal and on a similar topic as
our sample engineering article, “Spatial query integrity” (Hu et al., 2013). Based on what
we found in previous chapters, we might lay out the structure as shown in Table 19.1.
The section titles in quotes stand for titles specific to the new research, as opposed to the
others which are general titles common in the field.
Remember that the division indicates the general claim supported in those sections,
while the component tells us the specific types of claims and evidence needed in each. The
length and number of citations will undoubtedly change, but give us a range to aim for.

Table 19.1  Argument structure outline based on the article “Spatial query integrity”

Div. Section title Length Components Citations

(words) and support

I 1 Introduction 800–1200 1a Ia 5–10

1b Ib
2b Id
3b 
Va
2 Related Work 500–800 2b Id 20–25
P 3 Preliminaries 400–600 4a I b, c 0
4 “My New Algorithm” 3500–4500 5d II e 0
5 “Using the Algorithm” 800–1200 5c II d 0
T 6 Experiments 1500–2500
6.1 Experimental Settings 6a II f, I d 0–5
6.2 “New vs. Old Algorithm” 7a, 8e II g, h, i 5–10
6.3 “Practical Application” 7a, 8d II g, h 0
C 7 Conclusions 100–300 (10a) 0
10b
 Prewriting 191

Now it’s your turn. First do steps 1–6 for whatever you want to write, so you have a
good idea of the structure you are trying to copy. Use the information in Chapters 11–15
to guide your analysis, using exemplars from your class, school, or granting agency if
available.
Next, begin your outline by writing tentative section titles. For example, for a research
article in IMRD format, use the section titles that are standard in your field, whether
IMRD, combined R&D with C, or whatever variation is common in your journal, such
as “Experimental” in place of Methods. In IPTC format, begin with any section names
that are the same or similar for most of your exemplar articles, then give rough names for
those specific to your research. Don’t worry about getting them exact—just an idea for
now of how many sections you will need.
Next, think about how long each section should be. If this varies from one exemplar
to another, give a range. Adjust the number of words devoted to each section, based on
the focus of your article. If there is great variation in total words, is there a typical ratio
you could use as a guideline for section length?
Now list the components
and type of support expected
for your genre. Parentheses This sketch can be used as
can be used to indicate a com-
you write to check if you have
ponent present in only some
exemplars. Finally, if your included everything you need,
document will need citations, not too little or too much, and
give yourself a goal for each where you need to cite past work.
section, again based on the
exemplars.
This completes your sketch.
Remember that it is only a
sketch and will undoubtedly be modified as you add details, but it gives you a starting point,
general direction, and reminder of what is most important.

Exercise 19.1 Argument structure outline

For each step, follow the pattern of your exemplars, if available.

(1) Propose tentative section titles for your writing.

Hint: Should your titles be general, or specific to the research topic?
(2) Give a rough goal for the length of each section.
Hint: Adjust the relative length of the sections based on the focus of your writing, and remember
that these are just rough estimates, not exact numbers you have to match.
(3) List the components and support you need to include in each section.
Hint: If all your exemplars include a component, you should as well. If only one or two include
that component, put it in parentheses. Listing it will remind you to consider later whether it is
important to your research.
192  Creating your masterpiece

(4) List an approximate goal for the number of citations you should have in each section.
Hint:This will remind you how many total articles you will need to read.

Teaching tip: This outline will also be useful to you as a way of keeping track of the goals
for each student’s writing if you are not personally familiar with the format and expectations
of each genre or journal.

19.3 Adding the base layer: brainstorming

The next step of prewriting is brainstorming, getting as many ideas as possible down on
paper so you have something to organize. Some of the ideas may turn out to be irrelevant
to your topic, and later be discarded, but at this point you are trying to set forth as many
ideas as you can that might be included and begin to think about relationships between
them. Nevertheless, as Canadian writer and humorist Stephen Leacock said, “Writing is
no trouble: you just jot down ideas as they occur to you. The jotting is simplicity itself—it
is the occurring which is difficult” (in Leacock & Lynch, 2002: 194).
Thus I suggest that instead of
brainstorming randomly, think about
Next, write down as many each component. For example, for
details as you can. If you Component 1, what have previous
are not sure where to put authors said about the importance of
something, write it down similar work? Write as many things as
you can think of, then move to the
anyway and decide later. next component. If as you are work-
ing on Component 2 you think about
something else related to Component
1, go back and add it. Don’t worry at this point about getting the ideas in order or how they
relate to one another. If you are not sure which component something relates to, or if it might
relate to more than one, just write it somewhere. You will organize in the next step. You can
either set a specified time for brainstorming or just write until the ideas stop flowing.
As a visual learner, I like listing. I find it easier to do this on paper rather than com-
puter, as it is easier to add things to different lists as thoughts come, without worrying
about moving around the page on the computer, misspellings, improper auto-corrects,
saving, or other computer-related issues. On the other hand, doing it on a computer
makes reorganization afterward easier.
Auditory learners may prefer another method, sometimes called “freewriting,” writing
down all ideas as they come without trying to organize them. With modern technology,
it is easy to do this by speaking into a smartphone and using a talk-to-text app to convert
it into written text.
Kinesthetic learners may prefer to sketch out connections between various ideas, using
words with circles around them for each main concept. Different variants of this have
been called clustering, webbing, and mapping. This can also be used after listing or free-
writing to show how ideas are connected.
At the same time, list papers you need to cite for each component. List the author,
year, and key ideas from each paper that you want to include. If a paper relates to more
than one component, list it in both places.
 Prewriting 193

Next, identify which com-

ponents you can already begin
writing, which you could read Now decide where to begin
more past research on to get writing, based on what information
the necessary information, and you already have available, what
which will need to wait until you could get quickly, and what
you have done more research you will not have until later.
or analysis. This will help you
plan which sections to write
first and how to gather the
other information you need.
Currently we are considering organization of the whole paper, but you can do similar
brainstorming later as you begin writing each section. You may also realize later that some
of the items you listed should be eliminated, because they do not deal directly with the
goal of this writing.

Exercise 19.2 Brainstorming

(1) For each section, consider the components you need to include, and write down as
many ideas as you can that you think you will need to include in that section, using
whatever method of brainstorming works for you.
Hint: Include papers you need to cite and relevant information from each.
(2) Identify which sections or components you can begin writing, which must wait until
you are further along in your work, and what articles you need to read.
Hint: How many more articles do you need to read? Make a reading plan to finish them several
weeks before the deadline.

19.4 Placing the main elements: linearizing

Recall that an argument can be presented in many different orders, which can be catego-
rized as physical, temporal, and relational (section 9.5).2 Look back at your exemplars to
remind yourself what they use.
Before you decide what order to
follow, it is useful to consider various Recall the sequencing
possible orders, and different relation-
ships between the past and present
strategies used by your
research that you could pursue. Not exemplars, but also consider
only does this have the obvious other strategies that might be
advantage of forcing you to think better suited to your topic.
about the argument structure before
you start writing, it also gives you
other options you can return to if the
first proves difficult. If you have not considered alternatives first, it may be more difficult
to think of them after you start writing from one perspective, just as it may be hard to
end a relationship, even if you know it is bad, unless you have a better alternative at hand.
194  Creating your masterpiece

In Exercise 19.3 you will make a more detailed outline using topic sentences to lin-
earize your writing. Remember that this is still like a rough sketch or sorting puzzle pieces
into piles. You will undoubtedly change the order and even the number of sections or
subsections as you write, so do not worry about getting it perfect—you won’t. As your
writing ability improves, you will find that you have begun to internalize the expectations
and can simply write.

Exercise 19.3 Linearizing your article

(1) After you have finished brainstorming, group related items together.
Hint:Think about different groupings that would work.
(2) Thinking about the argument structure you are aiming for (Exercise 19.1), order the
groups of ideas within sections.
Hint:You may need to change your groupings or sections to do this. Remember that the order is
still tentative and will probably be changed during the writing and revision process.
(3) Generate a content outline, with each primary heading representing the main idea of
one section, and subheadings for paragraphs within that.
Hint: Copy the argument structure outline as a new file, eliminate the columns for length, com-
ponents, and citations, and add lines for each paragraph.
Hint: Do not number the headings; just indent to show sub-categories. This will make it easier
to reorganize.
(4) Think about how many paragraphs you will need to convey all the main points iden-
tified in your groupings. Write a topic sentence for each paragraph.
Hint: Use the component markers you found in your exemplar articles. For example, for
Component 1, it is unlikely that any of your articles use the pattern, “My topic is important
because . . . .”This may be the underlying claim, but not how it is stated.
Hint: Remember that reading the topic sentences should allow you to follow the basic argument,
but not all the details and support. Do not worry about getting the wording perfect but be sure
the order makes sense. If not, reorganize or rewrite.
(5) Now think again about your argument. If you see gaps, add topic sentences to fill
them, or placeholders to indicate what you will need to add later.
Hint: If you do not yet have the information necessary to fill a gap but know there is something
missing (e.g., citations or a test you have not done yet), use a placeholder, for example [data, test 1].

Evaluate your understanding

(1) What is the purpose of prewriting?
(2) How does creating an outline based on argument structure of exemplars differ from
a traditional content outline?
(3) What methods can be used for brainstorming?
(4) Why is it useful to consider various organizational strategies before beginning to write?
 Prewriting 195

Homework
(1) Ask a peer to look over your list of section titles and topic sentences from Exercise
19.3, to see if the logic of the argument is clear, and revise if necessary.
Hint: When you write, you can start with these topic sentences if you are writing in extended
style or replace them with the detailed contents of the paragraph if writing in condensed style.

Teaching tip: The purpose of writing topic sentences instead of a phrase outline is to help
clarify how the ideas tie together. Details can be listed as points below each.

(2) For academic writing, write a preliminary research goal, title, and abstract.
Hint:These will undoubtedly be very different when your article is complete, but like prewriting
they help ensure you know the overall goal and direction.

Supplemental material: Class exercise 19.1—Linearizing practice

Notes
1 One of the best I have seen was developed by The Writing Center of the University of North
Carolina at Chapel Hill. Several very useful handouts are available for download, including brain-
storming, argument, evidence, and thesis statements, at http://writingcenter.unc.edu/handouts.
2 These correspond to three perspectives listed in the brainstorming handout from UNC-CH:
describe it, trace it, and map it, respectively.The handout also discusses other techniques that seem
to be more useful for social sciences and humanities than for science and engineering, but may be
helpful for some writers.

References
HU, L., KU, W.-S., BAKIRAS, S. & SHAHABI, C. 2013. Spatial query integrity with Voronoi
neighbors. IEEE Transactions on Knowledge and Data Engineering, 25, 863–876.
LEACOCK, S. & LYNCH, G., ED. 2002. Leacock on life, Toronto, University of Toronto Press.
Chapter 20

Writing

20.1 Just do it
When the time comes to write your paper, the best advice is, to borrow an overworked
phrase, “Just do it.” All writers experience writers’ block at some point, not only when it
is time to start a new project, but throughout the process. The difference between those
who become good writers and those who don’t is that the former do not quit. Just like
any other task, the only way to succeed at writing is to sit down and do it, whether you
feel like it or not—every single day. Best-selling British author Philip Pullman is widely
quoted as saying,

Do plumbers get plumber’s block? What would you think of a plumber who used
that as an excuse not to do any work that day? The fact is that writing is hard work,
and sometimes you don’t want to do it, and you can’t think of what to write next,
and you’re fed up with the whole damn business. Do you think plumbers don’t feel
like that about their work from time to time?
(Pullman, 2017)

In the first part of this book

you learned to read articles in
Do not wait until you feel ready your field to see how others
to write; just begin. If it is bad, set do it. The more you read, the
it aside and start over. Working on more ideas you will pick up,
not only about the subject
different parts at the same time helps but also about how to struc-
you work effectively. ture your writing. But the
time comes when you just
need to sit down and write.
After you write, look at it,
and if it is bad don’t be afraid to throw it out and start over. After two or three tries you will
find that it looks a lot better than it did the first time.1
Another tip is to write several things at one time, whether different parts of one project
or different projects. All writers reach a point where they run out of ideas. Sometimes it is
necessary to just plow through, no matter how you feel, but effective writers will switch
to another section or a different task, rather than either sitting there frustrated or quitting.
Perhaps you will find that you are more effective at writing in the morning and rewriting
in the afternoon, or conversely that you need to do the final revision when you are fresh
 Writing 197

but can write a rough draft later in the day. Maybe you can work on details like format-
ting when you are tired of both writing and rewriting. You need to learn what works best
for you, but no matter what, keep going.

20.2 Keeping track of where you are

The next three sections offer practical hints to help you write more effectively, begin-
ning with moving between different sections of a document quickly without scrolling so
you don’t lose a thought before you find where you want to write it. In MS Word, the
navigation pane helps you do this.
To use the navigation pane, for-
mat your document using “Styles.”
Begin with your prewriting from Format your document using
Exercise 19.4. Make sure you have headings, especially for longer
topic sentences or ideas under each documents, so you can move
title as a placeholder for the plain quickly from one section to
text you will type for each sec-
another without scrolling.
tion. Then highlight each section
title and click “Heading 1.” Do the
same for any subsection titles, click-
ing “Heading 2.”
Now, on the “View” tab click the box to show the “Navigation Pane.” It should
appear to the left of your document, showing your titles. To move from one section of
the document to another, simply click a title. If you add section titles later, format them
the same way and they will immediately be added to the navigation pane.2

20.3 Keeping track of citations

One issue that is somewhat unique to engineering is the question of how to manage
citations while you are writing. In most engineering articles, citations are indicated by
number rather than the author name and date format common in science. Since those
numbers are based either on the order of appearance in the text or an alphabetical list
of authors, they will change as you
add more references.
If you already know how to Use placeholders (author-date)
use a citation manager, feel free to
use it as you write. If not, I rec-
for citations while writing,
ommend that you first insert the to become familiar with the
author name and date in parenthe- references you are citing and
ses as a placeholder, then insert the avoid confusion.
actual citation marks in the final
stage (section 25.3). This has at
least four advantages:

1 You become more familiar with the authors and dates of publication of important
works in your field. You will need to refer to them that way in presentations;
2 The reference numbers will not change while writing, eliminating confusion;
198  Creating your masterpiece

3 You will not be distracted by entering information in the citation manager;

4 Citation managers add a level of complexity to word-processing programs, making
errors more likely, so delaying entry eliminates a source of frustration.

20.4 Keeping track of your work

One final question is how much of your writing to save. In days gone by, writers kept
multiple versions of their work, from handwritten outlines and drafts to various typed and
marked-up copies, in addition to the final draft. Historians now plumb those records to
trace the development of thought of the author. Nowadays, although storage of multiple
versions is much easier, it is also far easier to simply hit “save” and overwrite the previous
version. Nonetheless, there are benefits of saving multiple versions.
Hopefully you already make a practice of regularly backing up your files and keeping
the backup in a separate location. This can avoid the tragedy of being just a few months
away from getting your doctoral degree and losing all your research and writing in a
fire.3 The rationale for saving multiple versions is the same. If one file becomes damaged
or lost, you do not have to start again from scratch. This needs to become part of your
regular routine.
A second advantage is that having multiple copies encourages you to try different ways
of writing the same thing. If after looking at your writing you think it is trash, don’t trash
it completely, but save and close it, open a new file with a new filename, and start over.
If the new version is still no good, do the same thing, reorganizing your outline or notes
to try a different approach. Eventually, within three or four tries, you will find something
that flows fairly easily. Even at this point, don’t trash the other versions. As you continue
to write you may remem-
ber that there was one line
you wrote before that really
Keep multiple versions of your sounded good and you will
writing, in multiple locations, both be able to go back to find it.
for safety and in case you change your Multiple versions can be
designated by placing a ver-
mind, but make sure you only work
sion number or date after
on the most recent. the main filename. Move
inactive files to an archive
folder, to make sure you do
not open and work on the
wrong version by mistake. Whether you delete that archive when you are done is a per-
sonal choice.
Finally, learn to use “Track Changes” in MS Word. Many editors now comment
on and change documents electronically using this feature. As the cloud becomes more
accessible, it is likely that platforms such as Google Docs that allow multiple users to com-
ment on or change the same document in real time will become more popular. Since
technology is changing so quickly, I will not dwell on the mechanics of these features, but
you will need to learn whatever is commonly used in your department and field.
This chapter has been brief, because at this point you do not need to read more, but
to begin writing. You may use any of the methods described above to assist you, but at
some point the only solution is “Just do it.” But remember, the rough draft will not be
 Writing 199

perfect. It may not even be that good. As Ernest Hemingway charmingly put it, “The
first draft of anything is shit” (in Samuelson, 1984: 11).4 In the next few chapters we will
learn how to transform that into useful and inoffensive fertilizer that will foster the growth
of new ideas.

Evaluate your understanding

(1) How does working on different parts of a large project at the same time allow you to
work more effectively?
(2) What is the advantage of using the navigation pane?
(3) What are the benefits of using placeholders for citations as you write?
(4) Why is it important to keep multiple versions and backups of your work?

Homework
(1) Write a rough draft of the first section of your writing.
(2) When you are done, ask someone to comment on it—your advisor, writing teacher,
or a fellow student.
Hint: Be prepared for criticism. Be prepared for marks all over the page and corrections on every
line (it may not be that bad, but be prepared!). Remember that you are not likely to do any
better with your first draft than Hemingway did.You will learn how to improve it step by step
in the following chapters.

Teaching tip: Do not assume that students know how to use features of Word like the navi-
gation pane and Track Changes. Learning them will save their time and yours, especially
for longer documents. Many students do not know how to use page breaks, section breaks,
or paragraph settings like “keep lines together.” Small caps may be needed in engineering.
Teaching tip: The rough draft can be written over several weeks while you begin to talk about
rewriting, from the following chapters. Writing the first section will take the longest. Allow
time for rewriting one division, at least at the level of organization, before writing a second.

Notes
1 See section 12.1.
2 This is especially useful when working on a long work like a thesis, as the section headers can be
used to automatically generate a Table of Contents. If there are blank lines in the navigation pane
you added a line return at the end of a heading. Click on the blank line in the navigation pane to
go to it, then click on Normal style instead of Heading.
3 This happened to a student at Cornell while I was there.
4 As one of my students helpfully noted, for beginning writers this is like baby poo, small and
poorly formed. Analogies about the importance of writing daily to prevent various problems I
will leave to your imagination.

References
PULLMAN, P. 2017. Daemon voices: Essays on storytelling, Oxford, David Fickling. Quoted at
https://writerswrite.co.za/philip-pullman-on-writers-block-2.
SAMUELSON, A. 1984. With Hemingway: A year in Key West and Cuba, New York, Random House.
Chapter 21

Rewriting for organization

21.1 Rewriting in stages

As I have said from the beginning, rewriting is the most important and most time-
consuming part of the writing process, even for the best writers. Ernest Hemingway
rewrote the final words of A Farewell to Arms at least 39 times before he was satisfied.
Your work may not need quite that many revisions, but it probably will not have the
same audience or impact either. Nevertheless, as the 19th-century German physician and
physicist Hermann von Helmholtz testified, “The writing aspect of scientific research is
exhausting . . . I have rewritten many parts of papers four to six times, restructuring the
entire organization, before I finally became satisfied” (in Alley, 1996: 228).
For experienced writers, much rewriting will occur during the process of writing, as
they delete and rewrite whole sentences or even paragraphs. They will not follow a dis-
crete series of steps but will change things on multiple levels at once. Nevertheless, they
will try to make large changes before small, and will go through the whole text many
times, as I am suggesting you do.
Do not feel that you need to slav-
ishly follow the order suggested.
The 7Cs of Change is a As with other ideas in this book,
scaffolding, not a method it is a scaffolding, a way of laying
to follow exactly forever. out what needs to be done, rather
Nonetheless, it is a helpful than a method to follow exactly.
device for beginning writers. On the other hand, there is a logic
to the order presented. It makes
sense to change overall organization
before worrying about specific word
choice. Therefore, for now I suggest completing the steps in order, reading about one and
practicing it before moving on to the next. That will help you see what you do well and
what you need to work on. There are online resources for each topic, once you know
what you need more help with.
As your writing improves, you will develop your own method. The method that
works for me may not work for you. You may prefer to write a rough draft of a whole
division or article before going back to rewrite or may rewrite one section at a time. Some
might go back and forth from one section to another as thoughts come, while others
stick mostly to one section, jotting down ideas relating to other sections as they occur.
Some will prefer to begin writing each day without reviewing what they wrote the day
 Rewriting for organization  201

Table 21.1  The 7Cs of Change, highlighting those covered in this chapter

Type of change 7Cs Purpose Level Scope

Rewriting 1 Coherence Organization Paragraph Section

2 Conciseness Sentence
3 Connection Clarification Sentence
4 Connotation Words
Revising 5 Consistency Finalization Words Whole paper
6 Correctness Format
7 Collaboration All

before, while others will reread and rewrite the previous day’s work before continuing.
Nevertheless, however you do it, you will have to read and reread, write and rewrite
many times, especially at graduate level and beyond.
In this chapter I will talk about coherence and conciseness, rewriting to improve the
organization of a paper, as shown in Table 21.1. Experienced writers tend to write more
than necessary and may prefer to tackle conciseness first, but novice authors generally
find it more beneficial to check first for coherence. Since both deal with the overall
structure of the argument, as you gain experience you will probably find yourself doing
both at the same time.

21.2 Coherence
Let me begin this section with an analogy: academic writing is like a tailor-made suit.
When complete, you want it to completely cover the subject, without feeling tight, but
without excess material. To accomplish this, before you begin cutting you need to plan,
take measurements, and lay out the pattern to make sure you have enough material. You
must make sure that once you cut out the pieces and sew it together, the pattern will
match. You don’t want the stripes running vertically on one half and diagonally on the
other, and the stripes on the two sides should be mirror images. In order to have strong
coherence at the end, you need to work it in from the beginning, as it is difficult if not
impossible to add it later.
Thus the first stage of rewriting is to check the coherence, making sure you have
enough material to cover your subject and checking the argument structure. This is one
of the most difficult practices to teach,
because of the various ways arguments
can be structured, but there are a few Arguments usually move
general principles. from known to new
An important general rule of aca- information, following an
demic writing, and indeed all commu- organizational sequence
nication, is to put known information predictable to readers.
first. This is the reason past research,
which your audience should be famil-
iar with, is described before showing
how your new work adds to it. Similarly, ideas you have already mentioned in a docu-
ment are generally placed before new ideas in a paragraph or sentence.
202  Creating your masterpiece

Recall that a claim may be stated either before or after the support, or both. For exam-
ple, in engineering the research goal often states the main claim as the contribution of the
research. This is supported throughout the document, then restated in a slightly different
form in the conclusion.
It is also possible to proceed either from a general theory to specific examples or from
specific examples to a more general theory or model. The former is far more common in
the academic literature, since new models are relatively rare.1
The argument often also follows a recognizable organizational scheme (section 9.5).
This can be chronological or the operational sequence of your design, but the order
should be clear and predictable to your audience.
Since there are so many factors to consider in terms of how to organize a document, it
is impossible to prescribe a general order. Choose an order that allows you to present your
argument in the clearest way possible. This frequently means rewriting it several times
until you find an order that works well.
Any document needs to be coherent at
multiple levels of organization, from sec-
When revising, begin tions to paragraphs to sentences, and needs
with the largest units and to be rewritten in that order. In longer works
move to the smaller. such as a thesis or book you may choose to
reorder the chapters after you begin writing.
Organization of sections should be straight-
forward for a journal article, so if your pre-
writing was effective you will mostly consider the order of paragraphs and sentences
within each section.
The first step, then, is to reevaluate the order of the paragraphs. First remind yourself
what claim or claims you are arguing in that section and what components you need to
include. Then ask, does each paragraph have its own purpose and topic, and does the
order make the argument easy to follow? You may need to reorder the paragraphs, split-
ting ones that do not have a single main topic or combining ones that contain similar
ideas. As a first-time author you will probably need to add more material at this stage, as
you find gaps in the argument.
The second step is to check the order of the sentences, striving for logical flow of
thought within each paragraph. In the process, you may find some sentences that should
be moved to a different paragraph, which will necessitate rewriting both paragraphs.
In summary, ask yourself these questions about every paragraph and sentence:

• Does each have a distinct purpose and topic?

• Does the order make it easy to follow the argument?

Exercise 21.1 Evaluating the coherence of your writing

(1) Beginning with the general claim appropriate to whatever you are writing, write a
related specific claim.
Hint: For example, for the Introduction of a research article, instead of the first general claim
you might say, “[My research goal] is important because [reasons] and it expands on [related
research].”
 Rewriting for organization  203

(2) Next, evaluate the coherence of the argument you have written to support that
claim. What is the supporting evidence? What is the purpose of each paragraph, each
sentence? Is the order logical, based both on some organizational scheme and the
novelty of the information?
Hint: For example, does your Introduction make both main points—importance and need? Are
they in the right locations and following the organization expected in your field? Are they sup-
ported by an appropriate number and type of citations?

21.3 Conciseness
Conciseness is an essential feature of scientific writing. Not only does it make it easier for
the reader; focusing on the main point also sharpens the understanding of the writer. As
Thomas Jefferson wrote,

This operation is doubly useful, insomuch as it obliges the student to seek out the
pith of the case, and habituates him to a condensation of thought, and to an acqui-
sition of the most valuable of all talents, that of never using two words where one
will do.
(in Mayo, 1942: 17)

Returning to our example of the tailored suit, you need to begin with more material
than is needed in the final product. It is easy to cut material out, but it is much harder to
stitch something in later without destroying the unity of the garment. In the same way,
it is far easier to begin with an excess of material as you brainstorm and write your rough
draft, but then the process of cutting begins.
You need to eliminate three things from your writing: off-topic ideas, repetition, and
wordiness. Hopefully between the time you did your brainstorming and the time you
wrote your first draft, you eliminated most of the excess material. Now that you have a
first draft, it is time to repeat the process. It is easier to do this in two stages, before and
after you check the order for coherence.
Before you begin to reorganize
for coherence, read through what
you have written and eliminate Eliminate anything that is
anything that is off-topic, does not
off-topic before checking for
address the goal of this writing task,
or is not necessary to your argument. coherence, and anything
Be ruthless. Scientific writing is con- repetitive or wordy
cise. If what you have written is not afterward.
relevant to the question addressed
in this paper, even if it refers to an
important study in the field, cut it.2
The second step is to look for repetition. If you have written similar things in two
sentences, try to eliminate one or merge the two. Often this will involve trying several
wordings before you settle on one that is as concise as possible without loss of meaning.
Sometimes rather than combining two sentences it is easier to set both aside and rewrite
the same thought from scratch.
204  Creating your masterpiece

The third step is to eliminate wordiness. As Jefferson said, “never us[e] two words
where one will do.” For example, instead of saying, “a sufficient number of tests were
done,” you could say, “sufficient tests were done.” Phrases can sometimes be replaced
by single words, for example replacing “in the vicinity of” with “near.” Sentences can
sometimes be reduced by simplifying the structure, for example “there are two things that
could be improved” is identical to “two
things could be improved.”
How concise your writing How much you cut depends in part
on the style of writing. Far more will be
needs to be depends on
cut from condensed style than extended.
the style of writing and the In condensed style, any topic sentences
expectations of the readers. used to organize your writing should be
removed at this stage, sentences com-
bined using relative clauses and other
structures, and every word weighed and
eliminated if not absolutely necessary. In expanded style, a certain amount of word repeti-
tion is expected as a way of linking sentences and paragraphs, but thoughts are not repeated.

Exercise 21.2 Evaluating the conciseness of your writing

(1) Evaluate the conciseness of your rough draft, making whatever changes are necessary.
• Does every paragraph and sentence relate to my goal?
• Is every paragraph and sentence necessary to my argument?
• Is there any unnecessary repetition of ideas?
• Could any words be eliminated without loss of meaning?
Hint: For most academic writing, the goal is stated in the research goal, but also consider the
larger overall goals discussed in Chapters 11–15.
Hint: As always, begin with larger units and move to smaller.

Evaluate your understanding

(1) How can you evaluate the coherence of your writing?
(2) Why should you always begin with large changes like the order of paragraphs or
sentences, before working on sentence structure, grammar, and word choice?
(3) What should you eliminate as you revise for conciseness?
(4) How do you know when your writing is concise enough?

Homework
(1) Rewrite your draft, concentrating on making it coherent and concise.
Hint: Save a copy before you start cutting and give your cut version a new filename.That way if
you want to go back to restore something that was cut, you will be able to find it.
 Rewriting for organization  205

Hint: If possible, show this to your instructor or advisor and get comments on it before mov-
ing to the next chapter or the next stage of rewriting. Make sure he or she knows it is a rough
draft—that you only want to check that all necessary information is present and well organized.

Teaching tip: If possible, work closely with the students’ advisors so they do not do too much
editing, but allow the students to do as much as they can in each step of the process.
Teaching tip: Reverse outline is also a good procedure to use at this stage to evaluate the
organization. Since it is a standard technique, I have not included it here.

Supplemental material:
Class exercise 21.1 Practice: coherence 1 (with answer key)
Class exercise 21.2 Practice: coherence 2
Class exercise 21.3 Practice: conciseness 1 (with answer key)
Class exercise 21.4 Practice: conciseness 2

Notes
1 Most normal science is based on deductive argumentation (general to specific), while inductive
arguments (specific to general) can lead to major paradigm shifts (see Class exercise 21.3).
2 There are, of course, exceptions to any rule. Sometimes it is expedient to mention at least one
publication of someone who is likely to be a reviewer of your article, even if that research is
marginal to your work, but that is a nuance you will figure out as you write more.

References
ALLEY, M. 1996. The craft of scientific writing, New York, Springer.
MAYO, B. 1942. Jefferson himself: The personal narrative of a many-sided American, Charlottesville,
VA, The University Press of Virginia.
Chapter 22

Rewriting for clarification

22.1 Clear, concise, precise

During the next step of rewriting your goal is to make your arguments as clear as pos-
sible. The easiest way to do this is to restate your points, using extra words to ensure your
meaning cannot be mistaken. However, writing in science and engineering must also be
concise, as we found in the last
chapter.1 Therefore you must
Any time you move one write precisely, making sure
sentence you will have to every detail, every word, is clear
rewrite the whole paragraph, and accurate. This makes sen-
tence structure and word choice
so reorganize before worrying extremely important.
about word choice. As the novelist Philip Roth
put in the mouth of one of his
characters,

“I turn sentences around. That’s my life. I write a sentence and I turn it around. Then
I look at it and I turn it around again. Then I have lunch. Then I come back in and
write another sentence. Then I have tea and turn the new sentence around. Then I
read the two sentences over and turn them both around. Then I lie down on my sofa
and think. Then I get up and throw them out and start from the beginning.”
(Roth, 2011)

Almost every change within a sentence, whether a choice of one word over another or
even changing the form of a word, for example from participle to infinitive, requires
changes in the structure of that sentence and surrounding sentences. That is what Roth
means by turning sentences around. Obviously, it is pointless to try to find exactly the
right words until your basic argument structure is clear. Otherwise, it would be like pol-
ishing a sculpture before you finish carving. Sometimes it is impossible to avoid, when
you realize you have left out some important element or need to cut to reduce your word
count, but try to avoid it as much as possible.
In this chapter I will talk about two ways of clarifying your argument: using connec-
tors to link the argument together and connotation to make your claims easier to support
(Table 22.1). I want to emphasize again that, at least in science and engineering, the
purpose at this stage is not to show off your proficiency in English but to strengthen the
 Rewriting for clarification  207

Table 22.1  The 7Cs of Change, highlighting those covered in this chapter

Type of change 7Cs Purpose Level Scope

Rewriting 1 Coherence Organization Paragraph Section

2 Conciseness Sentence
3 Connection Clarification Sentence
4 Connotation Words
Revising 5 Consistency Finalization Words Whole paper
6 Correctness Format
7 Collaboration All

argument. Neophyte wordsmiths may opt for grandiose terminology, perchance sup-
posing to impress their audience, notwithstanding that the only tangible consequence is
obfuscation of their contention. Or perhaps I should say, using big words does not make
sense if it only makes your
writing harder to understand.
This chapter is longer than oth-
ers because details are extremely Clarifying arguments is one of
important at this point. Additional the hardest but most important
online materials will help those tasks for any writer, involving
using English as a lingua franca. many details. Additional materials
Clarifying arguments and shades are available online.
of meaning is one of the hard-
est tasks for any writer, let
alone those writing in a second
language.

22.2 Connection
Returning to the analogy of a tailored suit, once you have laid out the pattern, cut the
pieces, and pinned them to one another, it is time to stitch them together. Without
that connection, the suit is not finished.
Similarly, you may have done wonder- Connectors clarify your
ful research and written each individ-
work, either by connecting
ual result well, but if it is only roughly
pinned together no one will appreciate your work with prior work
the quality of the material. or showing the logic of your
Connectors are used to improve con- argument.
tinuity of thought and thus clarify an argu-
ment. First I will show how to better
connect your work with previous research in the field, then how to improve connections
within your writing.

22.2.1 Connection to past research

Showing how your research connects with previous research can be done using
what I will call contribution markers, achievement markers, and application markers.
208  Creating your masterpiece

These are not grammatical categories, but rhetorical devices, using a word for a spe-
cific communicative purpose. All three are found in science, while the first two are
more common in engineering. Common categories of each marker type are shown
in Table 22.2 and discussed below. Examples of each category can be found in the
online supplemental materials (Supplemental Tables 22.1–3) or by using the synonym
and thesaurus functions in Word.

Contribution markers
The first claim in a research article could be restated as this work connects with and makes an
important contribution to the overall body of knowledge in the field. Beyond citation of previous
work, the basic method for supporting this claim is contribution markers.2
Contribution markers are verbs or
verb phrases that show how your work
Contribution markers supports, challenges, modifies, or
are verbs that show how the solves problems with previous work.
research supports, challenges, Since science proposes explanations
or modifies previous work, of data and phenomena, the first three
or solves a problem. categories predominate. Engineering
commonly modifies designs or solves
a problem. The four categories are
illustrated by the following examples:

• Our data corroborate . . . 

• These findings contradict . . . 
• Our study expands on . . . 
• This paper has resolved the issue . . . 

Contribution connectors appear throughout a document. Present tense is frequently used

in the first three divisions, highlighting how the current paper is contributing, as illustrated
by the first three examples. When reiterated in the Conclusion, the present perfect is com-
mon, as a final summary of what has been contributed by the paper, as illustrated in the last.

Table 22.2  Connection marker types and categories

Marker type Categories

Connection Support Modify

Challenge Solve
Achievement Excellence Efficiency
Importance Cost efficiency
Novelty Feasibility
Perfection Strength
Accuracy Clarity
Utility Emotive
Application Utilize Beneficial
Apply Potential
Facilitate
 Rewriting for clarification  209

Achievement markers
While connection with past work is important, you must also highlight your achievement; what
makes your work unique. Achievement markers are adjectives or adverbs that describe your
research or results in a positive way. Some give a general impression of excellence, others draw
attention to specifics such as low cost or efficiency, and yet others evoke positive emotion.
Achievement markers are far more
common in engineering than contribu-
tion markers, highlighting specific posi- Achievement markers are
tive characteristics of the new design. adjectives or adverbs that
Perhaps another analogy will help highlight your contribution
explain this. Contribution markers are
like the main ingredient in a recipe, and why it is worth
whereas achievement markers are like publishing.
the spices. The main ingredients in your
recipe may be almost the same as eve-
ryone else’s, but the seasoning is what makes one better than another. Your contribution
may be similar to others, but you want to highlight what makes it special, so sprinkle
achievement markers liberally throughout your writing. On the other hand, like spices,
make sure that they are not too concentrated in one place or they will ruin the dish.
Obviously, achievement markers are not a replacement for good research, but can
help strengthen the claim of your accomplishment, as shown in the following examples.
In each pair, the first is a basic statement of the contribution while the second shows how
adding markers highlights the achievement.

• The data challenge the interpretation proposed previously.

• The extensive data clearly and convincingly challenge the interpretation proposed
previously.

• Our prototype demonstrates that the fabrication method works.

• The unprecedented efficiency of our prototype demonstrates the feasibility of this
simple, low-cost fabrication method.

• This study presents a solution to the problem of circuit density outlined by Chen et al.
• This study presents a novel, cost-effective, flexible solution to the longstanding prob-
lem of circuit density outlined by Chen et al. more than ten years ago.

Note in the first example that both adjectives and adverbs can be used, in the second that
the sentence structure may need to be modified, and in the third that phrases can also be
used to highlight the achievement. In addition to the general categories of achievement
listed in the table, many others are field-specific. For example, there may be a value to
something being high-capacity, compact, or rapid-cycling.

Application markers
Engineering seeks a solution to a current problem, so the application of the research is
clear from the beginning. Although science seeks answers to basic questions, there are
210  Creating your masterpiece

usually possible future applica-

tions. Application markers draw
Your work may also have possible attention to potential applications
applications that have not yet been or their benefit.3
proven. These can be stated using When discussing possible
application markers, with modals to future applications, it is impor-
tant to use either modals or
indicate potential.
words such as “potential” that
indicate the results are possible
rather than actualized, as in the
following examples.
• This approach can be applied . . . 
• This research could lead to . . . 
• This technique might be beneficial in developing . . . 
• This information holds potential for resolving the question . . . 
• Based on these results, it should be possible to . . . 

Exercise 22.1 Highlighting your contribution

(1) Identify contribution, achievement, and application markers in your exemplars. How
many of each are there? Which components are they associated with?
(2) What sentences or words could you add to tie your work better with both past and
future research in your field while at the same time highlighting the importance of
your contribution?

Teaching tip: Have students share their findings, then discuss why those might be common
in their field.

22.2.2 Connection within your article

If a paper is hard to follow, readers may never reach the end of it. Perhaps another illustra-
tion would make the point more convincingly. Potholes and bumps in the road force a
driver to slow down. Sudden unexpected turns may cause a driver to go off the road or get
lost. Moreover, if a road is washed out, you can see the other side, but it is impossible to
get there. Similarly, if you want readers to follow your article to the end, you must make
it easy for them. There are many ways to improve the smoothness and continuity of your
article, including continuity connectors, logical connectors, and time sequence connectors.
The first is more common in extended style, while the other two are found in both.

Continuity connectors
In extended style,
sentences are often linked Continuity connectors link sentences or phrases
by repetition of words, by repeating an idea, either by repetition of
a word or use of a synonym, pronoun, or
synonyms, or relative relative pronoun. Continuity connectors are
pronouns. less common in condensed style, but some
are present in all writing.
 Rewriting for clarification  211

The simplest way to link sentences is to repeat the same word in subsequent sentences.
As shown by the example below, this is an effective way of stressing a keyword when it is
first introduced or when the second sentence defines that new word.

• The process of science involves making logical inferences. An inference is a conclu-

sion, explanation, or judgment based on the evidence.4

Instead of repeating the word, a synonym, pronoun, or relative pronoun can be used.
Synonyms are less common in science because of the desire for consistency of terms, but
could be used, as shown in the following example. Note that whereas in the first example
the end of one sentence was immediately repeated at the beginning of the next, in this
example the connection is between the initial words of each sentence.

• The difference between evidence and inference is essential to the question of origins, as
indeed it is in every area of science, so we must learn to clearly differentiate the two. The
distinction between the two will be a major emphasis of chapters three through six.5

To be more concise, a relative pronoun can be used. Relative pronouns can link either
sentences or clauses, as shown in the following two examples.

• The difference between evidence and inference is essential to the question of origins,
as indeed it is in every area of science, so we must learn to clearly differentiate the
two. This will be a major emphasis of chapters three through six.
• Each scientist is working from the perspective of one particular theory, which affects
both data collection and interpretation.

Less commonly, personal pronouns may be used when a subject is anthropomorphized,

given a human characteristic such as the ability to cause an effect, as illustrated in the fol-
lowing example.

• Models may be explicitly stated or implicitly held, but they direct our research in the
same way our worldview and philosophy subtly shape our thoughts. They have been
shown to affect both the choice of what data to collect and how the data are interpreted.

The least common continuity connector in science and engineering is the semicolon,
used to join two very closely related independent phrases. One phrase may also include
a relative pronoun, as in the following example. The author could have chosen to use a
comma and “which” instead of a semicolon and “that,” but used the semicolon for emphasis.

• Usually, however, it is relatively easy to list all possible classes of alternatives; that is
the case with the question of origins.

Logical connectors
Next we will turn to logical connectors: words or phrases that alert readers to the rela-
tionship between different clauses or sentences. Although many methods have been used to
group logical connectors, I propose eight categories, as shown in Table 22.3. The last
212  Creating your masterpiece

Table 22.3  Types of logical connectors

Type Connection Examples

Addition A and B additionally, furthermore, moreover

Contrast A vs. B however, on the contrary, whereas
Result A causes B therefore, consequently, thus
Cause B caused by A because, since, as a result of
Unexpected result A does not cause B however, nevertheless, even so
Unexpected cause B not caused by A although, in spite of, regardless of
Clarification A restated as B in other words, that is, i.e.
Illustration A exemplified by B for example, for instance, e.g.

two types are rare in natural science and engineer-

Logical connectors ing because of the value placed on conciseness, but
help readers follow common in social science. Several words fit in more
than one type, depending on the context. Further
the argument. examples of each type and how to change sentences
to incorporate them can be found in the online sup-
plemental material.
Most types are obvious, but two require explanation: unexpected result and unex-
pected cause. Both report results that are unexpected, surprising, or contrary to predictions.
Commonly grouped together as “unexpected” or “adversarial,” I have separated the two
and chosen names that reflect their relationship to result and cause connectors. More detail
on the logical connectors, and practice exercises useful for those using English as a lingua
franca, can be found in the supplemental materials online (Supplemental Tables 22.4–8).

ADDITION
Addition connectors The first type of logical connector,
take the place of the addition, is a way of joining two simi-
conjunction “and,” but add lar ideas. Unlike the simple coordinat-
emphasis to the phrase ing conjunction “and,” these connectors
after the connector. add different degrees of emphasis to the
second idea. Words or phrases with the
idea of “addition” or “more” indi-
cate the second idea is more impor-
tant. Phrases like “and in fact” tend
to introduce something unexpected.
Contrast connectors Numbering (firstly, secondly, . . . )
take the place of the makes it easier for readers to follow a
conjunction “but,” and series of steps in an argument.6
vary the strength of
the contrast implied. CONTRAST

The second type of logical connector

alerts readers to a contrast between the
 Rewriting for clarification  213

preceding idea and the following idea. The coordinating conjunction “but” is the simplest
way to express contrast.7 Some like “conversely” or “on the contrary” imply a stronger
contrast than “however.”
Contrast connectors also provide another way of challenging previous work:

• Contrary to what was reported previously, we observed that . . . 

RESULT AND CAUSE

Result and cause connectors both express

the same causal relationship, but in the oppo-
Result and cause
site order. Result connectors take the form connectors replace “so”
A causes B, whereas causal connectors and “for,” respectively,
have the basic form B is caused by A. As highlighting the causal
with addition and contrast, it is possible relationship.
to express result and cause using coordi-
nating conjunctions, in this case “so” and
“for,” respectively.
In most cases, a relationship can be expressed either as a result or a cause, with the
phrases inverted, as in the following:

• The equipment was not properly calibrated, therefore the data were inaccurate.8 (Result)
• The data were inaccurate because the equipment was not properly calibrated. (Cause)

Causal connectors always come

at the beginning of a dependent
clause or noun phrase, never an The clauses of sentences
independent clause, but in English using causal connectors
it is possible to invert the sentence can be reversed, putting
so that the dependent clause comes
first, separated by a comma. Thus,
the causal connector at the
although the basic form in terms of beginning to alert the reader.
the logical structure is B is caused by
A, the actual order in the sentence
may be inverted, as shown below.

• Because the equipment was not properly calibrated, the data were inaccurate. (Cause)

Note that each clause is identical to the previous example, but the order is reversed. On
the contrary, if the order of the clauses were reversed for the result connector, it would
make no sense. You could not say, “Therefore the data were inaccurate, the equipment
was not properly calibrated.”
This ability to invert the order is quite useful. Putting the causal connector first alerts
the reader to the causal relationship at the beginning of the sentence. By contrast, with
either result connectors or the basic order for causal connectors the reader is not alerted to
the causal relationship until the middle of the sentence, making it difficult to understand
if the sentence is long and complex.
214  Creating your masterpiece

Beyond connectors, there are other ways to represent causal relationships. Verbs such
as cause, affect, or lead to, and nouns like cause, effect, or result, can also be used. The
sheer number of ways of expressing the same concept indicates its importance.

UNEXPECTED
RESULT AND
Unexpected result and unexpected UNEXPECTED CAUSE
cause connectors show results that
are the opposite of what might have These two categories are
named for their similarity to
been expected, and can replace the result and cause connectors,
conjunction “yet.” except that something did
not cause the expected result.
Thus they express a surpris-
ing relationship between a
situation and an unexpected
result. Unexpected result
Sometimes it is difficult to can be expressed by a coor-
separate unexpected results from dinating conjunction “yet,”
but there is no conjunction
contrast, so both sets of connectors to express unexpected cause.
may be used, but with slightly Sometimes it is difficult
different connotations. to clearly separate unex-
pected results from contrast.
For example, comparing an
odd case with what usually
happens could be considered a contrast of two categories or an unexpected result. Thus,
“however” can be used for either. As illustrated in the sentences below, choosing a word
from either category will in many cases yield a sentence that makes sense, although the
emphasis is different and some sentences may need a little rewording.

• Outliers are usually ignored, (but/yet) this one led to a new discovery.
• Outliers are usually ignored. (Conversely/However), this one led to a new discovery.
• Outliers are usually ignored, (while/even though) this one led to a new discovery.
• Outliers are usually ignored, (unlike this one that led/despite the fact that this one
led) to a new discovery.

As with result and cause, the number of ways that contrasting or unexpected events can
be stated indicates their importance. Certainly this is the case in science and engineering,
where the unexpected or anomalous observation may lead to a new and deeper under-
standing of a phenomenon. Consequently, it is important to highlight such data for your
readers using appropriate connectors.

CLARIFICATION AND ILLUSTRATION

The final two types of logical connectors are used to restate or expand on a statement. These
are far more common in the humanities and social sciences than in natural science and
 Rewriting for clarification  215

engineering, but are included here

for completeness. Although the con-
nectors “that is” and “for instance” Clarification and illustration
can begin an independent clause, connectors restate or expand
the associated abbreviations i.e. (id on a statement, and are more
est) and e.g. (exempli gratia) are rare common in social science.
in sentence initial position. When
using either abbreviation, standard
American usage (but not British)
requires a comma after it.

Time sequence connectors

Your argument can also be strength- Time sequence connectors
ened by proper use of time sequence
connectors. These may denote the
denote a sequence of events,
sequence of events within the course of the either in the current research
current research or change in thought over or the history of the field.
time in the field. They may also be
9

used when discussing future possible

applications or developments of your
idea. Examples may be found in Supplemental Table 22.9. Ordinal numbers (first, second,
third) can be used for either logical or time sequence.
Although time sequence connectors can be used either at the beginning or in the mid-
dle of a sentence, it is far more common in concise scientific writing to place them at the
beginning for emphasis. However, when actual time is recorded, rather than sequence, it
often falls in the middle of the sentence. This is illustrated in the following examples, the
first two from (Hu et al., 2013), the latter two from (Kleinheinz et al., 2009).

• Initially, the DO obtains a private and a public key through a trusted key distribution
center.
• Then, it sends the signatures and the data to the SP which constructs the necessary
data structures for efficient query processing.
• By the time the eight-hour samples were collected, the difference was no longer
significant.
• Samples were analyzed within four hours of collection.

Exercise 22.2 Adding connectors to strengthen your argument

(1) Look in your exemplars for continuity, logical, and time sequence connectors. Circle
each, then list the most frequent types and words employed. Note where and how
each are used.
Hint: Are all types equally common in all divisions or with all components?
(2) How can you improve your use of connectors to clarify and strengthen your argument?
Hint: Organizational strategies (Chapter 9) are also important, of course.
216  Creating your masterpiece

22.3 Connotation
We now turn to the fourth of the 7Cs, connotation. Connotation refers to the fine shades
of meaning conveyed by the choice of words or modifiers. It is obviously pointless to worry about
fine shades of meaning until the basic meaning is well established. If the basic structure of
the paper is good, this stage will be relatively easy, a mere tweaking of the words. If the
basic structure still needs realigning, changing a few words will not make any noticeable
improvement.
Returning to our analogy of a tailor-made suit, connotation is like adding the
fasteners—the buttons, clasp, and zipper. Their absence or failure would quickly be
noted by others, so it is essential that they be as high quality as the rest of the suit.
Nevertheless, they cannot make up for lack of quality in the overall process of selecting
the material and cutting and stitching it together.
Connotation is important in all
types of writing but serves different
In science and engineering, purposes and therefore employs dif-
connotation is often used ferent techniques. In literary writing
to convey fine shades of connotation makes the material more
interesting, conveying details. In aca-
meaning related to the
demic writing it is used primarily to
strength of claims. modify claims, making them easier
to support. Thus, in this section we
will first discuss the general concept of
making supportable claims, followed by the proper use of statistical terms when making
claims. Finally, we will consider how to modify claims.

22.3.1 Making
supportable claims
You want to make the strongest You want to claim that you
claim you can support, without have a better explanation of
antagonizing readers who may a question or a better solu-
disagree. Qualifiers allow you to tion to a problem than any
that has yet been presented.
fulfill both conditions.
Yet among your reviewers
and readers are those who
have presented what you
claim to be bettering. Aye,
there’s the rub. How can you support your claim without antagonizing your readers?
Therein lies the reason for using qualifiers.
Qualifiers allow you to hedge your bets; to say what you think is the most likely expla-
nation without putting all your eggs in one basket. They allow you to claim your solution
is the best in a certain situation, if not every situation. It is hard to defend a claim that your
answer or your solution is the absolute best, so qualifiers make a claim easier to defend by
limiting its scope. Sometimes as you approach the end of your writing you will find that
you cannot fully support your initial claim. Instead of rewriting the whole paper, you can
modify the claim to make it more supportable.
 Rewriting for clarification  217

Let’s consider in general terms how this might work by returning to the list of basic
claims introduced in Chapter 3, reproduced here as Figure 22.1. I suggested that research
articles make several implicit claims.
The first claim is like advertis-
ing a new product. No advertiser
says their product is second best, No one advertises their
or almost as good as some other. product as second best; on the
On the other hand, if you deni- other hand, you cannot directly
grate similar products you will
suddenly find yourself in the mid- say similar products are bad.
dle of a nasty and costly lawsuit The same is true in research.
with your competitor. Similarly,
if you claim your work is the best
and others are junk, some of your
readers (who know their work is the best) may write a critique of your work without even
reading it.
For the first part of this claim, although you view your own work as important, you
need to convince others that it is worthwhile. (If you do not think your work is important,
I suggest you find a new field.) As Benjamin Franklin said, “If you would not be forgotten
as soon as you are dead and rotten, either write things worth reading, or do things worth
the writing” (Franklin, 2013: 78). Stating importance can be done without criticism of
others. Start with something everyone can agree on, so they are willing to read further.
The second part of the first claim is your research goal. This may be uncontroversial
in the natural sciences, where the research gap often lies at the border of current research,
unless it challenges previous findings. In engineering, with its emphasis on solving a cur-
rent problem, the field is much more crowded, and often your proposed solution is in
direct competition with a current solution. If so, you will need to be more careful about
how you state the need for your research.

Implicit claims in science and engineering research articles

IMRD (science)
1. You have identified an important unanswered question
2. You have gathered data in a valid, reliable way
3. You have a good explanation for your data
IPTC (engineering)
1. You have identified an important unresolved problem
2. You have designed a workable solution to the problem
3. Your solution is better than existing solutions

Figure 22.1  Implicit claims in science and engineering research articles.

218  Creating your masterpiece

The second claim is far different in science and engineering. In science, behind the
careful statement of the methods used is the implicit claim that the data are reliable, the
methods themselves being secondary. On the other hand, the description of your product
or process in engineering substantiates your central claim to have developed a better solu-
tion. Nevertheless, these sections are descriptive, with little if any inference. Others may
dispute whether you used the best method, but they cannot dispute what you did. If you
wrote this section as you went and revised it to make sure it is concise, coherent, and the
connections between different parts are clear, you should not have to worry about con-
notation for this claim.
Connotation is most important for the third claim. In IPTC, it may be necessary to
qualify your claim by saying your solution is only slightly better, or better in certain situ-
ations. In IMRD you are claiming that your inferences, your interpretation of the data
and explanation of the causal relationship, are correct. As you consider alternative inter-
pretations, qualifiers are used to indicate that this is the best supported explanation, but
not the only one.
Keep in mind that overall the article claims that you have made an important contri-
bution, however small. With few exceptions the contribution of any single article is very
small, so this is not just a matter of humility. Whether it turns out to be more important
depends largely on how well it stands the test of time and how many people cite it, so the
claims and support must be clear and not antagonize others.

22.3.2 Proper use of statistical terms: significant

Statistics are more important in science than engineering. Regardless of your field, I sug-
gest reading this and the following subsection to avoid misusing terms that have a specific
technical meaning. Full treatment could not be attempted apart from a presentation of the
statistical tests themselves and the assumptions behind them, so I will simply highlight a
few common mistakes.
In general, when the term “statistics” is used in the natural sciences and engineering it
refers not to data values themselves, as it often does in the social sciences (e.g. “If we look
at the statistics for the unemployment rate over time . . .”), but to inferential statistics.
Statistical tests ascertain how likely it is that parameters calculated from our sample data
represent true characteristics
of a larger set we are inter-
ested in.
“Significant” should only be used if a One of the biggest mis-
statistical test of significance has been takes made by those who do
done to show the result is not likely not use statistics regularly is
misuse of the word “signifi-
to be the result of random
cant.” In statistics, this has a
variation in measurement. specific technical meaning,
that the result obtained is
highly unlikely to have been
obtained by chance varia-
tion alone. If no statistical test has been performed, the word “significant” should not be used in
scientific or technical writing.
 Rewriting for clarification  219

If a statistical test has been done, the significance level should be specified, often 0.05 or
0.01, meaning that a result that extreme would be expected to happen only one time out of 20
or one time out of 100, respectively. Note that there should always be a number to the left of a
decimal point, even if that number is zero. Keep in mind too that some tests that are significant
do not indicate a real difference from expectations. One time out of every 20 times you do a
certain test you would expect the results to be significant at the 0.05 level. Thus, claims based
on statistics must be written carefully. A comparison of means cannot show that the two means
are identical, but rather that they are not different at a certain level of significance.
One of the most common errors made by those unfamiliar with statistics is to say some-
thing like, “A was larger than B, although the results of the test were not significant.” If
the results were not significant, you cannot say A was larger than B, even though the value you
measured was larger. Not significant means that if you measured again you might find that
the measured value of B was larger than A, because the difference between the two could
easily have been caused by sampling error. If you believe that A is truly larger than B,
even though the results do not show significance, you will have to be very cautious about
qualifying your claims and justifying why you believe it, perhaps because of multiple tests
that cannot be combined statistically but all show the same trend or triangulation of the
data with other independent measures.

22.3.3 Proper use of statistical terms: causation vs. correlation

Much of the work in science, and some in engineering, seeks to identify the cause of a
phenomenon. Sometimes, however, it is difficult to separate causation from correlation.
In a scientific sense, cause is the capacity of one variable to influence another in a predictable way,
either by bringing the second into existence or modifying it in some way. Correlation is
a relationship between two variables such that both change together in some predictable way, either
because of their action on one another or some other factor that affects both. It is essential
both to clarify what you think is going on and to state the claim properly.
Causation could be symbolized as:

A  B.

That is, A causes some change in B, or as A changes, there will be a resultant change in B.
One example would be if A is the number of times a wire is coiled around the magnetic
core of an electromagnet, and B is the strength of the electromagnet. As A increases, B
will also increase at a predictable rate, as a direct result of the number of coils. Of course,
this does not answer the question of why that is true—what the underlying mechanism
is—but it will suffice for now for the sake of illustration.
Correlation, on the other hand, simply means that there is some unknown relationship
between A and B. As A changes, there will be a concomitant change in B, but the causal
relationship could be any of the following:

A  B, B  A, or C  both A and B.

Many times, work on causality will begin by noticing a correlation. Causation can
only be proven, however, by eliminating all other possible causes. It is also possible that
220  Creating your masterpiece

something may be a cause, but not the only cause. Or perhaps you know it is correlated,
and have not yet proven causation, but want to indicate that you think it is likely. So what
is it that you want to claim?

• A caused B
• A partially caused B
• A may have caused B
• A is highly correlated with B
• A is slightly correlated with B
• A may be correlated with B

There are many ways to express both

“Correlated” should only causality and correlation (see Supplemental
Table 22.10). The relationship between the
be used if a statistical test variables can be expressed either as cause
of correlation has or result, just as relationships between ideas
been done. in your writing can be equally expressed
by cause or result connectors. The term
“correlated” should only be used if you have
done a statistical test of correlation and that test is significant. If you have not done that test,
use another expression.
Of course, even if the statistical correlation is very high, it is possible that there is no
relationship between the two at all. For the ten-year period from 2000 to 2009, there was
a 98.51% correlation between the number of computer science doctorates awarded in the
US and the annual revenue generated by arcades. Did the computer scientists’ work increase
the revenues (maybe by designing better video games?), or did the computer scientists get
doctorates because they spent so much money in the arcades (looking for ideas on how to
improve their own programming, perhaps?), or did some other social factor cause both to
rise and fall at the same rate? Even more surprising, there was almost perfect correlation,
99.79%, between the US spending on science, space, and technology and suicides by hang-
ing, strangulation, and suffocation for the period from 1999 to 2009. Perhaps both correla-
tions were the result of a bored law student going on a “fishing expedition” in big data.10

22.3.4 Types and uses of qualifiers

Qualifiers allow you to Qualifiers allow you to modify the strength
modify the strength of of a claim, making it easier to support.
They can indicate how confident you are
your claim, making it
that your inference is correct. Some quali-
easier to support. fiers cause your audience to view the data
a certain way. As with connectors, I have
grouped qualifiers according to their rhe-
torical purpose rather than their grammatical role. More examples of each category may
be found in the supplemental online materials (Supplemental Tables 22.11–15).
Rarely can you be 100% convinced that you have established causation. Qualifiers
modify the “truth value” of your statement, making it easier to support the claim. For
example, it is easier to defend the statement “the presence of A appears to have strongly
 Rewriting for clarification  221

affected B” than it is to defend the statement “the presence of A strongly affected B.” For
the former, you only need to demonstrate probable cause—the likelihood that it could
have had that effect. For the latter, you must eliminate all other possible causes, which is
often difficult since it may not even be clear what all the other possible causes are.
There are several different ways of qualifying likelihood, including likelihood quali-
fiers, hedges, and delimiters. Frequency and similarity qualifiers can be used not only to
modify claims about inferences or causality, but also claims about your evidence. Modals
allow you to discuss possibilities or make suggestions. Verb connotations can also modify
the strength of a claim. Depending on your purpose, it is often possible to combine sev-
eral types of qualifiers in one sentence. This ability to combine qualifiers and use them in
different contexts makes it difficult to assign them to a single role, so there will be some
overlap between discussions of the different types.

Likelihood qualifiers
The simplest way to qualify a claim of causation is using what I will call likelihood
qualifiers to specify how likely you think your claim is to be correct. There are many ways to
express how certain you are about your claim, including degree of certainty, possibility,
likelihood, and probability, which for most purposes are interchangeable.11 Thus, the
following expressions are nearly synonymous ways of making a very strong statement of
your claim of causation, without
making it absolute:
Likelihood qualifiers indicate
• It is almost certain that A
caused B. how likely you think your claim
• There is a very high prob- is to be correct, expressed as
ability that A caused B. degree of certainty, possibility,
• There is a very strong likeli- likelihood, or probability.
hood that A caused B.
• There is a very strong pos-
sibility that A caused B.

Likelihood qualifiers are most likely to be found in the section(s) of the paper devoted
to analysis, i.e. Testing or Discussion, where most inferences are presented.12 Unlike some
other qualifiers presented below, likelihood qualifiers can only be used with inferences or
predictions, not with data, because once something has been observed, it has a probability
of either 1 or 0—either it happened or it did not. Of course, if you can replace likelihood
qualifiers with actual probability values, that is even better.

Frequency qualifiers
Frequency qualifiers are used to draw attention Frequency qualifiers
to how frequently something occurs, without giv- can be used either
ing an exact number. For example, they can be to modify claims or
used to modify a claim of causality, as in summarize data.
• A occasionally caused B.
222  Creating your masterpiece

Frequency qualifiers can also help summarize data. Although science has a reputation
for being objective, data are always presented through the lens of the researcher. Scientists
attempt to lead their readers to a certain conclusion without appearing to do so. As John
Swales commented, “The art of the matter . . . lies in deceiving the reader into thinking
that there is no rhetoric . . . and that the facts are indeed speaking for themselves” (Swales,
1990: 112).
Thus, one of the main uses for frequency qualifiers is to direct your readers to view the
data in a particular way. Consider the difference between the following three statements:

• This occurred in 17% of the cases;

• This occurred rarely, in 17% of the cases;
• This occurred frequently, in 17% of the cases.

Each statement denotes the same thing, that the event occurred in 17% of the cases, but
the connotation is quite different. The first is the basic statement, without qualifier. If the
reader does not know the situation, it may not be clear whether 17% is higher or lower
than expected. Adding the qualifier “rarely” changes the meaning, implying that this
value is smaller than expected, while “frequently” has the opposite effect.
Whereas likelihood qualifiers are restricted to the probability of an explanation being
correct or a future event happening, and therefore largely restricted to the analysis, fre-
quency qualifiers often refer to past events and may be found in every division of an arti-
cle. Consider the examples below, which are representative of ideas that might be found
in each division, as summaries of previous work, methods, or aspects of the data:

• Introduction Generally it is thought . . . 

• Process/Method Each time we . . . 
• Testing/Results We never observed . . . 
• Testing/Discussion Sometimes there were problems with . . . 

Measurement qualifiers
Like frequency qualifiers, measurement qualifiers are used both to summarize data and
to modify claims. Commonly, they direct attention to how some measurement compared to
another or to expectations. Since
this must be done frequently
in an academic paper, it is
Measurement qualifiers direct useful to know different ways
attention to how measurements to express the same thought,
compare to one another or to to avoid repetition and make
expectations. They can be used with your paper more interesting
either claims or data. to read. There are also small
shades of difference in the
meaning of the terms, which
allows you to express your
thoughts more precisely.
Words that express degree of similarity of measurements can be divided into sev-
eral basic meanings (Supplemental Table 22.12). Some emphasize that two quantities
 Rewriting for clarification  223

are nearly equal. Others show that two quantities are close but not equal. Still others
highlight the smallness of either a single value or the comparative difference between
two values, while yet others highlight large differences. Enhancers can be used to further
qualify either smallness or largeness. Other expressions designate upper or lower limits on
the values observed. Within these there is often a distinction between words used with
countable quantities and uncountable measurements.
There is a difference in connotation in the words used to indicate upper and lower
limits. In the following examples, using “as many as” implies that 16 is a large enough
number, a situation known as permitting, whereas the connotation of “no more than”
is that 16 is a small or limiting number, and “a maximum of” is neutral, even though all
place the upper limit at 16.

• There can be as many as 16 different options. [permitting]

• There can be a maximum of 16 different options. [neutral]
• There can be no more than 16 different options. [limiting]

One term requires special mention: significant. While often used as an enhancer, for
example “significantly larger than expected,” this should be avoided in scientific and tech-
nical writing, for reasons discussed above (section 22.3.2).
Measurement qualifiers will most commonly be used in the results or analysis:

• As many as 15 occurrences were detected in a single trial, so the results were essen-
tially the same as predicted.
• No more than six occurrences were detected in a single trial, which is appreciably
fewer than formerly reported.

In both examples, the first underlined qualifier directs readers’ interpretation of the data,
while the second indicates a comparison, either with the predicted result or formerly
reported results. The adverbs “essentially” and “appreciably” modify the claims of similar-
ity in the first example and difference in
the second, making them more defensi-
ble or more specific, respectively. Modals express the
likelihood, advisability,
Modals or possibility of an action,
Another way to modify your claim or and are widely used to
suggestions is the use of modals. Modals modify claims.
can be used to express several different
concepts related to the likelihood, advis-
ability, or possibility of an action, as
shown in Table 22.4.13
It should be apparent that the modals for two of the uses, likelihood and advisability,
are nearly identical, apart from “need to.” This is logical, as if you tell someone that a
choice is necessary, it makes the likelihood of it happening virtually certain; if you tell
them that it is advisable, it is probable or likely that it will happen; but if you are saying
it is optional, the decision is up to them, that outcome is possible but neither certain nor
224  Creating your masterpiece

Table 22.4  Modals

Use Category Modals

Likelihood Virtually certain Must, have to

Probable/likely Should, ought to
Possible May, might, could, can
Advisability Necessary Must, have to, need to
Advisable Should, ought to
Optional May, might, could, can
Capability, Possibility Ability Can
Permission May, can

even likely. Capability can be thought of as a subset of the optional choice category, dif-
ferentiating between whether the choice is a matter of ability or permission.
As with the likelihood qualifiers previously discussed, modals can be used to modify
the “truth value” of a claim, as in the following examples.

• The rise in temperature was due to . . . 

• The rise in temperature must have been due to . . . 
• The rise in temperature ought to have been due to . . . 
• The rise in temperature may have been due to . . . 

The first example, without a modal, is the basic claim. Stated this way it implies that you
have the evidence to back up your claim, and that you are quite certain that this is the only
reasonable interpretation of the evidence, so it should be considered a fact. The second
implies that you are virtually certain of your conclusion, although some other interpretation
might be possible, either because the evidence could possibly be interpreted in another way
or because the evidence is limited in some way. The third example has the connotation that
you are fairly sure of your conclusion but want to leave open the possibility that there might
be other explanations. The last example indicates that you know your interpretation could
not be adequately supported by the available data, although it is not excluded.
Next, consider how modals modify the strength of a suggested course of action.

• To avoid this rise in temperature we must . . . 

• To avoid this rise in temperature we should . . . 
• To avoid this rise in temperature we could . . . 

In the first example, the author views the particular action as necessary, because that
is the only way the rise in temperature can be controlled. In the second example, the
action is stated as advisable, that it will have the desired effect, although it might not be
the only way to accomplish it. The final example gives one option but does not claim it
is the only or even necessarily the best option to accomplish the purpose. Thus, modals
can modify how strongly we advise a certain action.
The final use, capability, can be problematic. Although in common speech it is com-
mon to use “can” to indicate permission, in writing it is better to use “can” only for
 Rewriting for clarification  225

ability, and “may” for permission, to reduce ambiguity. “Can” is often used in a related
sense in IPTC papers, to indicate possibility, as in the following example.

• The output can be improved by . . . 

One of the difficulties of speaking or writing in a second language is that connotations

may differ from one language to another, even though the base meaning may be the
same. This is certainly true with modals. The Chinese word usually translated as “should”
implies suggestion but not obligation, whereas “should” in English tends to imply obliga-
tion.14 Thus if someone says, translating from the Chinese, “You should wear a sweater
in this weather,” they may mean it as a friendly suggestion, but someone whose first lan-
guage is English may hear it as critical of their decision not to wear a sweater. Thus you
should be careful when using “should.”

Delimiters note the

Delimiters
limitations of your
Although your work makes an important research and what remains
contribution to the field, it is not the final
to be done.
word. At the end it is common to note the
limitations of your study and what work remains
to be done, either by your lab or others.
What limits your study? Do you have a limited data set or have you had limited time
to analyze the data? Is further work going on in your lab to resolve those questions or do
you need the expertise of others outside your lab? Is this the starting point, or an ongoing
study, and what is the next step? All of these can be indicated by choosing appropriate
delimiters (Supplemental Table 22.14).
Since the research published in any article is only a small part of an ongoing project,
which is an even smaller part of ongoing research in your field, a brief discussion of the limi-
tations of your study helps coordinate that effort. By admitting the limitations, you concur-
rently point to what still needs to be done. By stating what work is ongoing in your lab, you
alert others to work that is likely to be published soon. By suggesting that other work needs
to, could, or should be done you highlight areas in which others could make a contribution.
All of this is usually done in one or a few sentences, as shown in the following examples.

• These conclusions are applicable within the range of our data, but further studies
would need to be done to confirm their general validity.
• Ongoing studies in our lab will continue to clarify these remaining issues.
• Similar studies should be conducted with other organisms.
• There is an urgent need for further work in this area.

Delimiters can also be used to indicate that you realize the limitations of your data,
that there are possible exceptions to the general rule, or that the results are inconclusive.
Examples of this type of delimiters are shown below.

• Based on (limited/preliminary) data, . . . 

• The results indicate that (with the exception of/apart from) . . . 
• The data (tend to/seem to) imply . . . 
226  Creating your masterpiece

Hedges
In addition to softening claims to make them easier to defend, words are sometimes used
to reduce the impact of an utterance on the hearer. Linguists call this hedging, and attribute it to
politeness or cooperation. Like physical hedges, linguistic hedges help ensure good rela-
tions with the neighbors, in this case others in your field whose work you are claiming
to surpass.
Many of the qualifiers listed above, particularly likelihood qualifiers, can serve not
only to reduce the strength of a claim but also to make it more palatable to your readers.
Similarly, some delimiters, particularly those claiming limitation of knowledge, serve a
double purpose as hedges. For example,
when claiming to be first, it is essential to
soften your claim, just in case there is a
Hedges soften the impact similar paper that you have not noticed,
of a statement on the or one comes out after yours is submit-
reader, and overlap with ted but before it is published, as in the
the qualifiers listed above. example below.

• As far as we know, this study is the

first of its kind.

One of the most common hedges in academic writing is using words or phrases that
indicate uncertainty. Other hedges include words referring to the thought of the research-
ers or the unproven nature of certain assertions. Advice for others doing similar work
is often given in the passive to avoid appearing too directive. Examples are shown in
Supplemental Table 22.15, but it is by no means a comprehensive list.

Verb connotations
Verb choice can also One of the easiest ways to change the strength
change the connotation of a claim is simply to change the verb. Just as
of a sentence. choice of a certain citation verb can indicate how
an author feels about previous research, verb
choice can indicate how strongly the author feels
about the conclusions of the present paper.15
Consider the following examples:16

• Our results demonstrate that . . . 

• The tests show that . . . 
• This finding indicates that . . . 
• We conclude that . . . 
• The outcome suggests that . . . 
• The observation implies that . . . 
• It appears that . . . 

It should be clear that the first is a much stronger statement, while the last is weaker.
 Rewriting for clarification  227

Combining qualifiers
All of the above methods of qualifying your claims can be used separately or in combina-
tion. For example, it is possible to begin with a likelihood qualifier and then add some
other qualifier to it, as in the following examples.

• It is likely that A caused B. [Likelihood]

• It is likely that A often caused B. The possibility of
[Likelihood + Frequency] combinations of qualifiers is
• It is likely that A caused a substan- almost endless, so do not
tial change in B. [Likelihood + expect to master them all
Similarity]
immediately.
• It is likely that A may have caused B.
[Likelihood + Modal]

Examples of some of the more common hedges found in technical writing are shown
below. All of these could be added to the beginning of the previous examples to further
soften the claim. However, as with achievement markers, using too many in the same
place can ruin an article.

• It (seems/appears) that . . . 
• This (suggests/points to) . . . 
• It can be (assumed/inferred) that . . . 

Exercise 22.3 Using qualifiers to make your claims more supportable

(1) What qualifiers are used in your exemplar articles?
Hint: Consider both the types of qualifiers used and where they are used in the article.
(2) Does your usage match? Are there places you should add qualifiers to soften your
claim or clarify your data or opinion of other research?

Evaluate your understanding

(1) What is the purpose of adding connectors to your writing?
(2) How are contribution markers, achievement markers, and application markers sim-
ilar? How are they different?
(3) What devices can provide connection between sentences?
(4) Why are there so many logical connectors, when the same basic ideas could be con-
veyed more simply by coordinating conjunctions?
(5) What is one of the most important uses of connotation in science and engineering
research articles?
(6) What commonly used terms should be avoided in research articles, unless used in
their technical statistical sense?
(7) What types of qualifiers can be used in summarizing data as well as clarifying claims?
(8) What is the purpose of using modals, delimiters, and hedges?
228  Creating your masterpiece

Homework
(1) Rewrite your draft again, concentrating on clarifying your argument using the mark-
ers, connectors, and qualifiers described in this chapter.
Hint: Look at the Supplemental online material for further details.
Hint: Again, at this stage you may want to show your draft to your instructor or advisor and get
comments on it before you do the final revision.

Teaching tip: This chapter is long, but has been kept together because both connection and
connotation are related to clarification. Connection and connotation can be taught as separate
classes, or some subsections covered with either the previous or following chapter, both of
which are much shorter.

Supplemental material:
Material particularly useful to ELF students
Class exercise 22.1—logical connectors (with answer key)

Notes
1 Conciseness is less important in some social sciences and humanities.
2 Glasman-Deal (2010) heightened my awareness of the importance of these connections and
her word lists were invaluable in developing the supplemental tables for markers, although my
analysis is far different.
3 The verbs and adjectives listed in the online supplemental material could be thought of as
potential contributions or achievements, respectively.
4 All examples of continuity connectors are from Rau (2012), emphasis added.
5 Modified for example.
6 Steps in an argument can be ordered “first, second, . . .” or “firstly, secondly, . . .” but not “at
first, . . .” which implies a time sequence rather than a logical sequence. “At” is not used with
ordinal numbers above first (“at second”) in American or British English, although this is com-
mon in Vietnamese English.
7 Grammarians differ on whether “but” can be used in sentence-initial position; some say that
since it is a conjunction, by definition it should not be used to begin a sentence, while others point
out that it is used that way in many written documents and can be treated as a connector as well
as a conjunction.
8 Properly the word “data” is plural (singular: datum) and takes a plural verb. Some journals accept
data as collective, with a singular verb.
9 Change in thought in the field would be common in the Introduction, while sequence of events
in the research may be found in the Process and Testing (Methods and Results).
10 I don’t know if the law student was bored, but the author was a law student. The examples are
from www.tylervigen.com/spurious-correlations.
11 Likelihood and probability are distinct mathematical concepts, but the distinction is hard to
understand for those outside the field, and the terms are used interchangeably in most disciplines
except mathematics and philosophy. The other two have no formal mathematical expression.
12 Biber, Connor, and Upton (2007: 69–72) show many ways of expressing the authors’ stance,
many of which relate to how confident they are about a statement.
13 This analysis is simplified from Glasman-Deal (2010: 158–167).
14 應該 vs. 必須 (simplified characters: 应该 vs. 必须; pinyin: ying1gai1 vs. bi4xü1).
 Rewriting for clarification  229

15 There is substantial overlap between these verbs and those used to express an evaluation of past
research, listed in Table 7.3. Both verbs and adjectives can be used in different ways to indicate
what linguists call stance, including how certain we are about our argument. See Biber, Connor,
and Upton (2007: 69–72).
16 Cargill and O’Connor (2013: 62) give more ideas on using sentence patterns employing a “that”
construction to modify the strength of a claim.

References
BIBER, D., CONNOR, U. & UPTON, T. A. 2007. Discourse on the move: Using corpus analysis to
describe discourse structure, New York, John Benjamins Publishing.
CARGILL, M. & O’CONNOR, P. 2013. Writing scientific research articles: Strategy and steps, New
York, John Wiley & Sons.
FRANKLIN, B. 2013. Poor Richard’s Almanack and other writings, Chelmsford, MA, Courier
Corporation.
GLASMAN-DEAL, H. 2010. Science research writing for non-native speakers of English, Singapore,
World Scientific.
HU, L., KU, W.-S., BAKIRAS, S. & SHAHABI, C. 2013. Spatial query integrity with Voronoi
neighbors. IEEE Transactions on Knowledge and Data Engineering, 25, 863–876.
KLEINHEINZ, G. T., MCDERMOTT, C. M., HUGHES, S. & BROWN, A. 2009. Effects
of rainfall on E. coli concentrations at Door County, Wisconsin beaches. International Journal of
Microbiology, https://doi.org/10.1155/2009/876050.
RAU, G. 2012. Mapping the origins debate: Six models of the beginning of everything, Downers Grove,
IL, InterVarsity Press.
ROTH, P. 2011. The ghost writer, New York, Random House.
SWALES, J. M. 1990. Genre analysis: English in academic and research settings, Cambridge, Cambridge
University Press.
Chapter 23

Revision and finalization

23.1 Time’s up
There is no fixed endpoint for rewriting. If you wait a little longer, work a little more,
you will always be able to improve your document. Perhaps it is counterintuitive, but
the better you get at writing, the truer that will be. So how do you know when to quit?
Sometimes you reach a point where you are making smaller and smaller changes, and
just feel enough is enough. In many cases, it is just that time’s up. The paper is due, the
deadline is here, or you have to begin work on something else. As you read it one last
time you may feel it still does not express exactly what you want to say. Nevertheless, the
time has come for the final revision and proofreading.
This chapter concludes the topic of rewriting using the 7Cs of Change, covering the
stages highlighted in Table 23.1. For consistency and correctness you will be making small
changes, mostly at the level of words or formatting, but collaboration may lead to changes
at any level of the document. Before or during this step you will also need to finalize the
graphics and add the Title, Abstract, and whatever other finishing touches are required
(Chapters 24–26).

23.2 Consistency
We will consider two areas where consistency is important: thought and words. I will also
talk in this section about removing ambiguity, by making sure all your referents are clear
and correct. The best way to see whether the document is consistent is to read quickly
through the whole thing from beginning to end, focusing on the main thoughts and

Table 23.1  The 7Cs of Change, highlighting those covered in this chapter

Type of change 7Cs Purpose Level Scope

Rewriting 1 Coherence Organization Paragraph Section

2 Conciseness Sentence
3 Connection Clarification Sentence
4 Connotation Words
Revising 5 Consistency Finalization Words Whole paper
6 Correctness Format
7 Collaboration All
 Revision and finalization  231

keywords rather than details. I conclude the section with a warning against employing
the squid technique.

23.2.1 Consistency of thought

It is very common to finish your research
and find that you have not achieved the Authors frequently restate
research aim or answered the research the research goal in the
questions you wrote at the beginning. final rewrite, but must
Instead of starting over, it is simpler to check for consistency.
reformulate or rephrase the research goal.
Remember, however, that your argu-
ment is entirely based on that goal. Thus
you also need to change any references to that goal in the Process or Testing sections and
the contribution or answers to the research questions in the Conclusion.
I vividly remember my first year in college writing a paper on memory. At that time
it was thought that some class of chemical was responsible for memory. I researched
the most recent findings, came up with my conclusion, put my note cards in order, and
started typing. When I reached the final note card, very close to the deadline, I realized
that final bit of data disproved my hypothesis. To incorporate it I would have to rewrite
and retype the entire paper. So I did the only reasonable thing—I threw the card away,
concluded the paper without including that reference, and turned it in. When I got it
back the professor had commented, “I am surprised that a freshman has sufficient back-
ground to write such a paper.” I laughed, relished my A, and went on to the next course.
Unfortunately, following my example is not an option with your research. If you find
at the very end that something is wrong, you will have to go back and change the whole
paper. Fortunately, that is much easier with personal computers than it was in the days
of typewriters. Unfortunately, the ease of making changes also means that editors tend to
require more changes than they did before.
The longer it has taken to complete a work, the more likely it is that your thoughts
have changed. You may have changed your terminology or substituted a new design
element, so it is essential to read through the whole paper carefully, checking for consist-
ency of thought. While writing this book I went through several different versions of the
originally 6Cs and later 7Cs of change. In the final revision I had to make sure they were
listed properly throughout the book. You will need to do the same.

23.2.2 Consistency of words

Problems with consistency often occur
at the level of words. By using the same
word consistently, the reader knows you Technical terms should be
are referring to the same concept. At the used consistently, but other
same time, keeping the reader’s attention words can be varied to
requires variation to avoid monotony. maintain reader attention.
When deciding whether to repeat
the same word or use a synonym, the
main question to ask yourself is, is this a
232  Creating your masterpiece

keyword or specific technical term, or is it a general concept that could be expressed in different words?
Technical terms must be used consistently, as no synonym expresses the same thought.
Apart from those, however, it is good to vary your word choice when you can.
The first step, then, is to identify the keywords. Think about the words I have used
in this book. I have been careful to distinguish division from section and have not used
either term for any other portion of an article. Nor did I use any other word for compo-
nent, or use that word with any other meaning. Such words must have a consistent mean-
ing and be used without variation for clarity. As you consider your writing, the keywords
may be physical apparatus or concepts. A few may be verbs—processes with a specific
technical name. These must be referred to the same way every time.
For other words, consider what synonyms could be used, especially if the same thought
appears several times in one paragraph. This is particularly true of the first two divisions,
where you report what others have done and what you did, both of which can be very
repetitive. Remember, however, that your primary goal is clarity. Do not use any word
if you are not 100% sure of its meaning. If you use the wrong word, or one that does not
make sense in the context, the reviewer will undoubtedly see it for what it is: someone
who made use of a thesaurus in a failed attempt to sound erudite.
Search your document for both old and new forms of any terms you have changed
to make sure they are used consistently. Similarly, if you think one word occurs too fre-
quently, search for all occurrences before deciding which would be the best to change,
rather than just changing some randomly.

Exercise 23.1 Consistency of words

(1) Check your writing to make sure key terms are used consistently. Apart from those
words, if the same word appears twice in a paragraph, consider whether it would be
possible to substitute a synonym.
Hint: If you notice any problems, check for occurrences elsewhere in the paper.

23.2.3 Eliminating ambiguity

Many different types of words or expressions can be ambiguous, usually because the refer-
ent is not clear. It is important to ensure that your reader understands every sentence the
way you intended it. While novelists
may purposely use ambiguity early in
Abbreviations, pronouns, a story to set up a surprise ending, that
antecedents, and modifiers technique is not appreciated in the sci-
entific literature. Gifted writers might
all need to be checked to be able to pull it off when writing pop-
eliminate ambiguity. ular science, but it is extremely rare in
academic journals. Common sources
of ambiguity include abbreviations,
pronouns, antecedents, and modifiers.
The general rule for abbreviations is that the full name must be stated along with the
abbreviation the first time it is used. While this works well for research papers for class, it
is not always sufficient for research articles. On the one hand, certain abbreviations are so
 Revision and finalization  233

well known in a field that they do not need explanation, like DNA in biology or LED in
electrical engineering. On the other, since readers may not read the whole paper it can
be helpful as you begin a new section to reintroduce uncommon or new abbreviations.
There is no hard and fast rule for this, but you need to be aware of what your audience
knows and does not know. Expectations for abbreviations in an Abstract differ, so be sure
to check the requirements.
Pronouns refer to something already mentioned, frequently but not always in the pre-
vious clause. In practice, relative pronouns may also be used to refer to the entire previous
sentence, as in the following examples, although prescriptive linguists may feel this is an
incorrect use:

• This is mainly due to the distance between . . . 

• This was not borne out by extensive  . . . field measurements.

Misplaced modifiers are the bane of a writer’s existence and can particularly be a prob-
lem with very concise writing, including headlines, section titles, or titles of graphics.
Consider this ambiguous headline from BBC News:

• Liquid living worms survive space (30 November 2011)

As a biologist, I know there are no liquid worms, and it is hard to imagine how solid
worms could be liquefied and survive anywhere. In this title it initially appeared that
both “liquid” and “living” were individually modifying “worms.” A hyphenated form,
“liquid-living,” would have resolved the ambiguity, making it clear the worms lived in
liquid.
Or consider this headline:

• EU waters a cemetery, Malta warns (12 October 2013)

This made me wonder why the EU would want to water a cemetery, and why Malta
would care if they did. Later it became clear that Malta, an island in the Mediterranean,
was warning that the waters around the EU are a cemetery, devoid of life. Here the prob-
lem is that “waters” could be interpreted as either a noun or a verb.
One more, this time reported by the editors of the Columbia Journalism review:1

• Cause of AIDS found—scientists

Colons and dashes can serve many functions. Make sure your intention is clear.
The best way to check whether your writing is ambiguous is to ask someone not
familiar with your research to read and comment on it. Since you know the background
and have thought about the subject extensively, you will tend to fill in information that
is not actually written in the paper. Since you wrote it, you will know whether you
intended a word to serve as a noun, adjective, or verb. As much as possible, put yourself
in the position of someone reading the paper for the first time. What information would
you need to know, what referents are not clear, or what words could be interpreted as a
different part of speech, leading to confusion?
234  Creating your masterpiece

23.2.4 Squid technique

A different type of ambiguity is often referred to as the squid technique, so named by
Doug Saville, who said,

occasionally, I recognize what I call the squid technique: the author is doubtful about
his facts or his reasoning and retreats behind a protective cloud of ink.2

Just like a squid trying to confuse

potential adversaries, authors can be
The squid technique involves purposely ambiguous in an attempt
intentional use of ambiguity to to hide the weakness of their argu-
try to hide the fact that you are ment. It may work for a squid, but
not clear about some aspect of rarely for writers. If you can clarify
your research. Don’t do it. your writing with a little additional
work, do it. If it requires further
research that has not yet been done,
say so.
For example, do not say, “It has long been known that  . . .” if what you really mean
is “I was too lazy to find the reference.” Look it up!
Do not say, “Two thirds of the samples showed the same pattern,” if you only had
three samples. Do the work required to get quantitative results you can truly stand behind
or use a qualitative approach to the analysis.
Do not say, “obviously” or “clearly” if what you really mean is, “I really don’t want to
list all the steps in my logic; in fact, I am not sure of all the steps myself.” Make your argu-
ment clear enough that your readers find it obvious themselves, without you saying so.
Finally, do not say, “Much more work will be required to provide definitive answers
to these questions” if you really mean, “I don’t have a clue how to explain the data.”
Emphasize what you have learned through the present study that others can build on, not
the questions that are not yet clear. If there is nothing, you really have nothing to publish.

23.3 Correctness
The final stage of revision involves one final check that everything is correct. What you
need to focus on at this stage depends on your background. Ironically, those who learned
English as a first language may make more mistakes with certain frequently misused words
than those using English as a lingua franca, who may have read and perhaps even memo-
rized such word lists. On the other hand, those using English as a second language will
have more difficulty with little words like prepositions and articles. Both need to be aware
of differences between British and American English and take the time to proofread care-
fully for misspellings and formatting errors.
When checking for consistency, I suggested that you read the whole article quickly.
To check for correctness I suggest the opposite: that you read it again slowly. Some peo-
ple prefer to read through the article backwards one sentence at a time to make sure they
are not distracted by the content. While experienced editors may be able to check all the
different categories in this section at the same time, you may find it helpful to focus on
one or a few at a time.
 Revision and finalization  235

23.3.1 Frequently misused words

Many websites list frequently misused
words.3 In this section I will mention
Some word pairs are
only a few words that are more likely frequently used incorrectly by
to be found in scientific, engineering, those using English as a lingua
or technical writing. franca, others by those whose
Some words that are frequently first language is English.
misused are linguistically related, with
similar-sounding words for the noun
and verb. In many cases the two can
be interchanged in your writing by changing the structure of the sentence, as shown in
the examples below.

Advice (n)/advise (v): When your advisor advises you, he or she gives you advice.
When I want your advice, I will ask for it.
When I want you to advise me, I will ask you to do so.

Device (n)/devise (v): A device is something that someone has devised.

Our solution to the problem is a workable device.
We devised a workable solution to the problem.

Effect (n)/affect (v): A variable strongly affects the results if it has a large measurable effect.
Variable A had a strong effect on the results.
Variable A strongly affected the results.

Other words have similar meanings that are easily confused, whether they are linguisti-
cally related and sound similar or not. These cannot be substituted for one another, as they
are used to refer to different categories of objects.

Continuous/continual: continuous means without interruption and has a broader

meaning, referring to numbers or space as well as time, while continual is used only
with respect to time and often implies an interruption in the continuity, so the event
is regular rather than continuous.
Variable A is continuous, while variable B is categorical.
We had to deal with continual breakdowns in the apparatus.

Each/every: each refers to separate members of a group of two or more, while every
refers to all the members of a group of three or more.
Each time we did the test we did it the same way.
Every time we did the test we got the same results.

Farther/further: farther is used only with respect to physical distance, whereas further
is used when referring to abstract or metaphorical distance.
We went farther from the shore than we did in previous studies.
We have gone further in this work than previous authors.
236  Creating your masterpiece

Fewer/less: fewer is used with countable objects, while less is used with uncountable
objects or numbers referring to measurement or time.
We collected fewer samples than we did the first time.
We collected less water than we did the first time.
We collected less than 2 ml of water this time.

Imply/infer: something implied is not specifically stated, whereas something is

inferred if it is deduced from the evidence.
We implied the previous results were wrong, but did not say so explicitly.
We inferred the previous results were wrong, based on our new evidence.

Yet another category of words that are commonly misused are homonyms: words that
sound almost the same although they have very different meanings. In some cases, only
one is commonly found in the scientific literature.

Complement/compliment: a complement makes something complete, either physi-

cally or metaphorically, while a compliment is praise given to a person.
The IPTC model complements the IMRD model, but does not replace it.
I hope to get at least a few compliments on my work.

Discrete/discreet: items that are discrete are separate, but things that are discreet are
handled carefully and confidentially.
There are 100 discrete points on the graph, each representing a person.
We have to be discreet not to reveal the identities of the individuals.

Elicit/illicit: to elicit is to draw something out of a person, whether information or

a reaction, but illicit is that which is not acceptable under the law or conventional
moral standards.
We tried to elicit responses from the subjects, but had limited success.
They were unwilling to talk about their illicit activities.

Idea/ideal: an idea indicates a suggestion, purpose, or plan of action, whereas ideal

means the best.
I have an idea how to solve this problem.
It may not be the ideal solution, but it should work.

Its/it’s: its is a possessive pronoun, meaning belonging to it, whereas it’s is a contraction
of it is.
It’s considered improper in many journals to use contractions like “it’s.”
If you are not sure about its usage, see if substituting “it is” makes sense.

Lose/loose: lose, used only as a verb, is the opposite of find, whereas loose, the oppo-
site of tighten or tight, can be used as a verb, adjective, or occasionally noun.
If you set a helium balloon loose, you will lose it.
If you lose it, it will be lost forever.
 Revision and finalization  237

Principle/principal: a principle is the fundamental foundation of a set of beliefs or a

theory, while principal means the main or most important item in a group.
One principle of science is that whoever publishes first gets credit for a discovery.
That is why time of publication is of principal interest to researchers.

Another frequently misused class of words are those that have irregular plural forms,
usually because the plural form of the original Latin or Greek is retained. In addition to
the words listed below, many terms in biology have irregular plurals.4

Datum (s)/data (pl): a single datum is frequently called a data point, while the word
data is a plural form.5
The data are presented in Figure 1.The first data point represents the initial conditions.

Analysis (s)/analyses (pl).

Previous analyses have underestimated gene number.The present analysis corrects
this problem.

Criterion (s)/criteria (pl).

The first criterion for choice of a site is accessibility, but other criteria may prove
more important in the long run.

Phenomenon (s)/phenomena (pl).

By definition, science seeks to explain natural phenomena, but data give evidence
of a phenomenon, so in practice science seeks to explain data.

One final word that is often used improperly is respectively. Respectively is used to
indicate that two lists are given in the same order, and is always set off from the rest of the
sentence with a comma or commas. It cannot be used if there is only one list, nor can it
be used if the number of objects in each list is different. It is also never used to indicate
that things happen sequentially.6
Correct uses:

The two subjects were male and female, aged 67 and 63, respectively.
The two subjects were, respectively, male and female, aged 67 and 63.

Incorrect uses:

The two subjects were aged 67 and 63, respectively. [No second list.]
The three subjects were male and female, aged 67, 63, and 62, respectively. [Unequal
numbers so it is unclear which were male, which were female.]
The two subjects were a male aged 67 and a female aged 63, respectively. [Two lists
are already merged, so there is no need to add “respectively.”]
The two subjects will be described respectively. [One after the other = sequentially,
not respectively.]
238  Creating your masterpiece

23.3.2 Prepositions
One of the greatest challenges for anyone learning a second language is to master the
little words. Whether the prepositions and articles of English or the measure words of
Chinese, rules can be stated but there always seem to be exceptions. One underlying
problem is that no two languages have an exact correspondence of how they categorize
thoughts.7 Another problem is that language is always changing, so British English may
differ from American English and there is even variation between regions, or based on
class, age, or gender. Moreover, the form used in a given area in common speech may
not be considered acceptable in academic writing, which is more formal and conservative
than speech or even popular writing.
With prepositions, the first issue is that the meaning of some verbs is incomplete or
different without the preposition. For example, in English we “get in” the car. Note that
get has a very different meaning without the preposition in and that no other preposition
is used in this context. In Chinese, however, to express the same meaning one would say
“get on” the car, undoubtedly because the Chinese word for “car” can mean any wheeled
vehicle, and people got on wagons long before there were cars.8 While such connections
between verbs and prepositions are naturally internalized by those who learn the language
as a child, those who learn a second language later in life need to memorize the patterns,
since there is no way to transfer
expectations from one language
Prepositions are a major to another.
problem since languages change A second problem is that often
over time, leading to variation the same preposition is used to
and many different meanings express various meanings, some-
thing few native speakers of a
assigned to a single word. language will be aware of.9 For
example, consider the following
sentence:

In the morning I found the information in an article in a folder while getting in the
car, so I will give it to you in an hour.

In this sentence, in is used in five ways, first referring to a time of the day, then a meta-
phorical use, next physical space, then meaning into, and finally meaning within a certain
length of time from now.
Since there are so many online resources on prepositions, many of them specific to
problems faced by speakers of a particular language, I will not dwell on them here. A few
expressions that seem to be common problems in scientific writing are worth mentioning,
however, particularly those dealing with comparisons.
Starting with the basic comparison, should you compare with or compare to? In most
cases, either is acceptable, as long as the comparison is between real objects. For exam-
ple, you could compare your data to previous data, or compare your design with existing
designs. If you are comparing in some way that is metaphorical or based on resemblance
or analogy, you must use to, for example when comparing to an ideal. Since the ideal does
not actually exist, you cannot compare with it, which often implies the ability to lay the
two side by side for comparison.10
 Revision and finalization  239

The next question is what preposition to attach to comparatives like same or similar.
Some Chinese seem to think that because you compare with, the same preposition should
be appropriate for the resulting judgment about sameness, resulting in same with or similar
with. Logically this might make sense, but unfortunately neither of these are considered
correct in either American or British English, which prefer same as and similar to.11
Even more difficult is what preposition to use with different. In both British and
American English different from is the most common, while different with is considered
improper, but in Britain different to is also acceptable, while America also uses different
than. Reasons for this are vague, but different from will work in most situations.12
Another common question is whether to use try to or try and in sentences like,
“We will try to/and demonstrate that our product is superior to any currently on
the market.” Some authors, particularly in America, unambiguously say that try and
is grammatically incorrect. Others label it an idiom or fixed expression and say it is
fine. Yet others suggest that it is fine for colloquial speech but should be avoided in
academic writing.13 Since try to is universally accepted but some people frown on try
and, stick to the former.
For those whose first language is not English, a good rule of thumb for using preposi-
tions is if referring to graphics, use “in,” but if referring to how you collected your data
or did your analysis, use “by.” This will be a good start in many cases. If you want to see
how other authors use prepositions, you can search your exemplars for a certain word and
see which prepositions are used after it.14

23.3.3 Articles
Those who have grown up using articles cannot appreciate the difficulty they pose for
those learning English as a foreign language. In general, definite article the or indefinite
article a(n) are used with singular nouns, while with plural nouns the choice is between
the and no article [Ø], since a(n) implies singularity. Changing the article will change the
meaning of the sentence, as in the following.

This may reveal the connection between the two. (a specific connection)
This may reveal a connection between the two. (some connection)
Attach the wires to the two adjacent sites. (specific adjacent sites)
Attach the wires to [Ø] two adjacent sites. (any two adjacent sites)

The flowchart in Figure 23.1 attempts to summarize how to determine which article
should be used. Proper nouns, the name of a person or location, are obviously specific,
so it seems that they should always take a definite article. Nevertheless, names of people,
states, cities, mountains, and lakes normally take no article. Other bodies of water like
oceans and rivers, along with buildings, on the other hand, do. Don’t ask me why. I don’t
make the rules. For example:

Queen Elizabeth II
Minnesota
240  Creating your masterpiece

London
Mount Everest
Lake Chargoggagoggmanchauggagoggchaubunagungamaugg15
The Atlantic Ocean
The Nile River
The Louvre Museum

For common nouns, the first question is whether it refers to some specific item known
to the reader. Things that have been previously mentioned or can be inferred based on
what has been mentioned are known, therefore the definite article the is used.

Noun Phrase
No article

Ocean, River,
Proper noun? The (definite)
Building?

Previously mentioned
Can be inferred
Known to reader? The (definite)

Countable? Singular? A/an (indef.)

Collective Plural
Abstract

No article No article

Figure 23.1  Flowchart for deciding what article to use.

 Revision and finalization  241

To illustrate the importance of what the reader knows, consider the following.

I had a bowl of beef noodles for lunch.

The noodles were very good, but the beef was tough.
The broth was also very salty.
So I did not drink the broth, but bought a glass of tea.

In the first sentence a is used because this is the first mention of the subject. In the sec-
ond sentence, the is appropriate because both the beef and the noodles have already been
mentioned. In the third sentence the is also used, even though it is the first mention of
the broth, because I can assume that the hearer will make the connection between the
beef noodles and the broth they are served in. In the last sentence, the tea is again a new
topic, so I use a.
The is also used if the hearer can easily infer which of several possible referents is meant.
On the other hand, a is used if the specific referent does not matter or is unknown to the
hearer. For example, I could say to someone in my department,

I ate lunch with the department head.

He had just come from a meeting in the administration building.

In this case, the hearer will infer I am referring to our department head, not some
other. Similarly, although there are many administration buildings, there is only one
on our campus, so the referent is clear. On the other hand, there are many different
meetings and I cannot assume the hearer knows which one I am talking about, so a is
appropriate.
If the noun phrase is not a proper noun and not known to the reader, the next question is
whether it is countable. Collective and abstract nouns never take an article. Unfortunately,
determining whether a noun is countable or not is not always easy. Most concrete nouns
that refer to physical objects are clearly countable, such as organism(s) or circuit(s). Yet
some nouns that refer to objects do not have a plural form, such as equipment, and are
considered collective.16
An even greater problem
is that many nouns can be
considered either countable Frequently whether a noun
or uncountable depending on is considered countable or
their use in the sentence. For uncountable depends on its
example, absence is an abstract
concept that is uncountable, context. Other rules for article
but absences from class are use likewise have exceptions.
countable. Paper can be used
as a collective term for a type
of material (sheets of paper),
242  Creating your masterpiece

or as a countable term (two conference papers). One book on scientific writing lists work,
freedom, and intelligence as uncountable nouns, and often that is true, but we can also
talk about how many works an author has produced, the freedoms of speech and religion,
or multiple intelligences. To make it even more confusing, American and British English
may differ. Any rule or list I could give would have exceptions, so you must consider
whether the word could be considered countable as you are using it. If English is not your
first language, just accept that it will take many years to master this.
Singular countable nouns that are unknown to the reader take the indefinite article a or
some other “determiner” like my, this, some, or a number. For plurals, the article a, being
singular, cannot be used, but other words can be. For example,

We submitted a/the/my/this/one paper to the journal Nature.

We submitted the/our/these/two papers to the journal Nature.

One special case important in technical writing is when the adjectives few or little are
used to modify the noun. Since these are used with plural or uncountable nouns, respec-
tively, we would expect no article. Nonetheless, a can be inserted to modify the meaning
of the sentence. Without an article, the connotation of the sentence is the rarity, whereas
with the article the emphasis is on the presence of the noun.

We found few exceptions to this rule.

We found a few exceptions to this rule.
This causes little trouble for native speakers.
This causes a little trouble for native speakers.

One final rule refers to when to use an instead of a. The rule is sometimes stated in
terms of the following letter but is better expressed in terms of the following sound:
Use a if the following sound is a consonant but use an if the following sound is a
vowel. Thus words like university and unique, which are considered to begin with
a consonantal y sound, take a. This is also true with abbreviations. Again, the only
odd case among the vowels is U, while consonants that take an include F, H, L, M,
N, R, S, and X.

I teach at a university.
We used a UV light.
We performed an MRI scan.

Exercise 23.2 Articles

(1) In your exemplar article, circle all the a, the, or places in front of a noun that could
have an article but do not [Ø], and try to determine which rule is being followed.
Hint: Consider the background of the authors when deciding which exemplar to examine for
this exercise.
 Revision and finalization  243

23.3.4 British vs. American usage

Linguists speaking of English as a lingua franca talk about multiple Englishes and try to
describe the distinctive characteristics of each, not only in countries where it is spoken as
a first language, like Singaporean English, but also places where it is spoken as a second
language, like Chinese English. When teaching oral communication, the emphasis in
many cases is on communica-
tive competence rather than
seeking native fluency and British and American English
accuracy. Nevertheless, when differ in word choice, spelling,
it comes to written academic punctuation, and numbers, but
communication, British and most journals now allow the
American standards still dom-
inate. In the past (not the dis-
author to choose either system.
tant past—I still remember it)
authors were expected to use
one or the other depending
on where a journal was published. Now most journals will accept either, as long as
you are consistent throughout the paper, but you will need to check the Instructions to
Authors to be sure.
A surprising number of words differ in British and American usage, most being
nouns referring to common objects in daily life. Examples include motor vehicles
(lorry/truck), parts of cars (bonnet/hood, boot/trunk), words related to trash (dustbin/
wastebasket), food, clothes, holidays, and others. There are a few names of chemicals,
drugs, or supplies that differ (aluminium/aluminum, cotton wool/absorbent cotton),
but fortunately most words used in scientific or technical writing are identical.17
There are also a number of spelling differences between British and American
English.18 Just as the presence of c or s distinguishes noun from verb in advice (n)/advise
(v) and device (n)/devise (v), Britain distinguishes practice (n)/practise (v) and licence
(n)/license (v), while America has merged the two parts of speech in different directions
to practice and license. Another spelling difference important in science and engineer-
ing is that in Britain you analyse while in America you analyze. There are also many
words where the –re in British English has been replaced by –er in American English,
including litre, metre, and centre. In Britain many words contain double vowels which
in America have been replaced by a single vowel, such as colour, behaviour, amoeba,
foetus, and paediatric.
Moreover, some words are considered to be collective in American English, but not in
British English. For example, in America math is considered a singular or collective noun,
while Britain speaks of maths (which makes far more sense, since geometry is as distinct
from calculus as biology is from physics). Similarly, in America research is exclusively a
collective noun and used with a singular verb, “research has shown,” whereas in Britain it
is acceptable to say, “researches have shown.” Many speakers of English as a lingua franca
also speak of researches, so it will be interesting to see whether this becomes acceptable in
American journals in the future.
There are also differences when it comes to numbers, as shown in Table 23.2. Due to
differences in the meaning of one billion and the date format, the meaning should always
be stated unambiguously in international English to avoid confusion. It is also worth
244  Creating your masterpiece

Table 23.2  Differences in number use in British and American English

British American

One billion 1,000,000,000,000 1,000,000,000

(1 million million, 1012) (1 thousand million, 109)
12/1/16 12 January 2016 December 1, 2016
(day/month/year) (month/day/year)

noting that some countries, including most European countries and some others, use the
comma as a decimal mark and either a space or period as a thousands separator, for exam-
ple, writing ten thousand dollars as $10.000,00.
One final difference between British and American English is the placement of
quotation marks relative to punctuation. In British English, punctuation is placed
within quotation marks only in cases such as a sentence, where the punctuation occurs
in the quoted material, whereas in American English punctuation is always placed
within the quote.

In British English, the rule is “punctuation after”.

In American English, the rule is “punctuation within.”

23.3.5 Spelling, punctuation, and formatting

Finally we come to the stage of revision commonly called proofreading. The term proof-
reading originated in the days before computers, when a typed manuscript was physically
mailed to an editor, and the journal then had to typeset it on the machine that would
be used to print the actual pages. Before final publication a “proof” copy was sent back
to the author to make sure that there were no errors in the typesetting process. This
still happens with books but is rare
for journal articles. Nevertheless, the
Errors due to autocorrect same process has to be done on your
are harder to detect than manuscript before you send it to the
editor. English editors can do much
misspellings, as they are not
of this for you, if you are willing and
marked, so it is important able to pay them, but remember that
to proofread carefully. your name is on the paper, not theirs.
The final responsibility is yours.
Make sure you check your docu-
ment not just for misspelled words,
but also incorrect ones. I had one student who was quiet confident of his analysis. Another
purchased regents for his research. Yet another throughout his whole paper talked about
addictive manufacturing rather than additive. In resent years I have also had papers on witch
frequency resulted in a lager capacity. Such unintentional humor is best avoided.
In section 18.6 I suggested things worth doing right the first time to reduce changes
during proofreading. At this point it is still necessary to check all those items, and more.
 Revision and finalization  245

Similarly, templates help but will not correct all errors. Figure 23.2 lists items you will
need to check. Changing the physical appearance of a document by changing the mar-
gins, font, or font size can sometimes reveal problems you would otherwise miss. Using
the “Show all” command in Word to reveal hidden characters such as paragraph marks,
breaks, and spaces is also very helpful.19
In addition to whatever is listed in the Instructions to Authors, there are frequently
unstated expectations you can learn from your exemplar articles. Hopefully you have
already confirmed that the length is appropriate, but at this point you will need to
double-check the overall structure, numbering, headings, number of references, and
other broader features.

23.3.6 Variation in an ELF research world

It is worth noting that as more and more scientists from non-English-speaking countries
write using English as a lingua franca (ELF), expectations are changing. While some jour-
nals still expect all authors to conform to the standards of American or British English,
other journals are showing greater acceptance of other variants.
Some journals, especially in engineering, are accepting papers with what previously
would have been called non-standard English. If you skim some articles quickly you will
find that it is fairly easy to understand the thread of the argument, although individual
sentences are not formally grammatically correct. This is not always the case, however,
and many journals still require that a paper be edited by a “native speaker,” preferably one
familiar with your field.

23.4 Collaboration
In the collaboration process you ask others to comment on your writing. Although this
is placed at the end of the list of 7Cs, in practice the timing depends on what you are
writing and how much experience you have had with academic writing. When you are
writing your thesis, many advisors suggest that you show it to them chapter by chapter.
For your first article, which probably will be written with your advisor as coauthor,

Checklist for final proofreading

Spelling
Double spaces
Foreign punctuation or other marks
Font and font size
Margins, indents
Format of paragraphs—extra line or space before/after, indented or not
Format of headings, titles, captions, footnotes
Format of reference and citations

Figure 23.2  Checklist for final proofreading.

246  Creating your masterpiece

you may also collaborate as you

Collaboration may occur at write. For articles beyond the first,
any stage, not only at the end. you may not show it to your coau-
thor until you think it is finished.
For beginning writers it is good No matter when that collaboration
to get feedback from your takes place, plan on revising your
advisor or coauthor early. work again afterwards, unless your
coauthor does the rewriting for
you. In other words, just when you
thought you were finished, you
have to start all over again. Nevertheless, if you have followed all the suggested stages
of rewriting and revision, you should have less work to do after collaboration than you
would otherwise.
Remember that revision means looking again, which is what happens when a second
set of eyes looks over your document, and those eyes will see things that you missed,
whether the eyes are your advisor, another coauthor, or a reviewer. If as you advance
in your career you do not have a second author for your paper, but still want comments
before you send the paper to a journal, the Internet provides opportunities not available
to past generations. Several websites now allow you to exchange peer review with others
in your field.20
The most important point I can make about collaboration is to learn from it. Look
carefully at the changes your collaborators or reviewers suggest and learn from them. Ask
for clarification if you do not understand why they made certain changes. Pay attention
to common problems identified and try to figure out which of the 7Cs they relate to, so
you know where you need to focus your effort next time.

Evaluate your understanding

(1) Why do many writers restate the research goal when the paper is nearly complete?
(2) What techniques can be used to check for consistency?
(3) Which of the many points of correctness mentioned in this chapter is most difficult
for you?
(4) As a first-time author, why is it important to seek advice from your collaborators
early in the writing process?

Homework
(1) Do a third rewrite of your draft, concentrating on consistency and correctness of the
words.
Hint: If you are working on this for a class, this will be the final revision. When you write a
journal article, you will have to revise it at least one more time after collaboration before it is
ready to submit.

Teaching tip: This chapter may or may not be the focus of your class, depending on how
concerned you are with structure and content vs. English usage.
 Revision and finalization  247

Supplemental material:
Class exercise 23.1: Eliminating ambiguity
Class exercise 23.2: Articles (with answer key)

Notes
1 www.cjr.org/the_lower_case/headlines_editors_probably_wish_they_could_take_back_21.
php (accessed 2 September 2016).
2 Quoted in Day (1998).
3 One particularly good site for misused words is www.oxforddictionaries.com/words/usage.
4 For a list of common Greek and Latin endings and their plural forms, see www.biomedical
editor.com/spelling-tip-latin.html. A longer list of terms common in biology can be found at
write-site.athabascau.ca/documentation/Irregular-plurals-Biology.pdf.
5 As a result of language change, in some fields it is becoming acceptable and even common prac-
tice to use data as a singular noun.
6 For a more detailed explanation and more examples, see www.scienceeditingexperts.com/
using-respectively-correctly.
7 Their ontology.
8 More specifically, “Get in the car” in English is the same as “On car” (上車) in Chinese, as
Chinese has no articles and “get in” is expressed by the single character 上 which means “on.”
9 See, for example, www.ego4u.com/en/cram-up/grammar/prepositions.
10 For more details, see https://en.oxforddictionaries.com/usage/compare-with-or-compare-to.
11 Even more interesting is identical. The preferred form is identical to, but identical with is also
found frequently in a corpus, and (unlike compare) seems to be more common with metaphori-
cal usage. See https://lingohelp.me/preposition-after-adjective/identical-to-or-identical-with.
12 See https://en.oxforddictionaries.com/definition/different.
13 For a good summary of the debate, see www.dailywritingtips.com/try-to-vs-try-and.
14 This can also be done on a large group of documents at once (a corpus) using a concordance
program. See Appendix 2.
15 Also known as Lake Webster.
16 Equipment is often considered countable and pluralized in Asian English, but not in American
or British English.
17 www.oxforddictionaries.com/words/british-and-american-terms.
18 Wikipedia is not usually cited as a reference in academic work, but their article on the spelling
differences between British and American English is actually quite a good summary: https://
en.wikipedia.org/wiki/American_and_British_English_spelling_differences.
19 Unfortunately “Show all” is not on the menu by default in the current version of Word, but can
be accessed by the “Tell me what you want to do” search. I find it useful to add it to the menu,
using File/Options/Customize ribbon.
20 The sites will undoubtedly change rapidly, but at the time of publication include (alphabetically)
Academic Karma, Paper Critic, Peerage of Science, Publons, Pub Med Commons, Pub Peer, and
Rubiq. Particular services, requirements, and costs differ.

Reference
DAY, R. A. 1998. How to write and publish a scientific paper, Phoenix, AZ, Oryx Press.
Part 5

Adding the final touches

Chapter 24

Illustrating your manuscript

24.1 The purpose of graphics

In a speech to leaders of the communication industry, Rupert Murdoch said, “Great
journalism will always attract readers. The words, pictures and graphics that are the
stuff of journalism have to be brilliantly packaged; they must feed the mind and move
the heart” (La Monica, 2009).
Whether in journalism or
engineering, it is the combina- Like a preview, graphics
tion of good writing and good
graphics that attracts readers. should attract attention,
Graphics is a collective term for showing some of the most
all the tables and figures in an interesting portions without
article.1 The title and abstract giving away the whole story.
may lead people to look at the
graphics, but often it is the
graphics in turn that will draw
them to read the whole thing. To use an analogy, graphics are the 60-second trailer
that entices people to sit down for the whole 2-hour show. Thus it is important
not only to have graphics, but to ensure that they are high quality and tell the main
story of your research, while leaving enough questions unanswered to make reading
worthwhile.
Graphics and text should supplement one another, without duplication of information.
Graphics must be able to stand alone, so an experienced reader in your field can under-
stand the main point of the graphic without looking at the text. At the same time they
should complement the text so that both are required to get the full message, just as both
the soundtrack and video are necessary to fully appreciate a movie. As you design your
own graphics, consider the purpose of each. What does it add to the paper that makes it
essential? Just like text, if it does not serve a purpose, eliminate it. Remember also that
many journals now have supplemental materials online, providing another option for
graphics of interest to a smaller audience.
Graphics, both tables and figures, can be used either to summarize or show detail. For
example, a table may give the details, while the text summarizes the main point. Or a
graph may show a trend, while the text explains possible causes of that trend. The basic
principle behind choosing whether to express something using text, table, or figure is to
252  Adding the final touches

consider what will be easiest for your reader to understand. Sometimes that basic principle
trumps any rules presented in writing books.
The maxim “writing is an art” is true for all aspects of writing, but nowhere is it clearer
than when producing graphics. The ideas presented in the following sections are not rigid
rules, but general principles that may help you convey your ideas more effectively.

24.2 Terms used when referring to graphics

Before talking about the characteristics of a good graphic, I need to clarify terminology,
because each kind of graphic uses a different set of terms and those for figures are differ-
ent in British and American English. Table 24.1 and Figure 24.1 show the terms used for
tables and figures, respectively, and the location for each. I will be using the American
designations, in line with the rest of the textbook.
Note that the table title goes at the top.2 This is easy to remember because all three words
begin with T. On the other hand, the figure caption follows the illustration. The Writing

Table 24.1  Table terminology: title

Column 1 heading Column 2 heading

(unit) (unit)

Row 1 heading (unit) Data Data

Row 2 heading (unit) Data Data
Summary row heading Summary statistic Summary statistic

4
y-axis label (unit)

2
Legend (American English)
or Key (British English)
1

0
1 2 3 4
x-axis label (unit)

Figure 24.1  Figure terminology: caption (American English) or legend (British English).
 Illustrating your manuscript  253

Center at UNC-Chapel Hill

helpfully suggests that this is
Titles go at the top of a table
because we tend to read a table
from the top down, whereas (Table Title Top, T3), but captions
we read a graph (the most com- go below a figure.
mon type of figure in science Units are placed in parentheses
literature) from the bottom up, below or after a heading
from the caption to the x-axis or label.
to the y-axis.3 Note also that
for column headings, the unit
is often placed in parentheses
on a separate line, as shown. For row headings or axis labels, the unit may be placed either
in a separate line or after the name. While some writing books will state requirements for
the placement of units, many journals now do not seem to care, and placement seems to
be related to saving space rather than comprehensibility. Unless your journal lists specific
requirements, think what would be easiest for your readers to understand.

24.3 Tables
Academic writing textbooks often state that tables should be used when the details are
important, to show exact numbers and their precision, but sometimes fail to mention that
tables are also commonly used to summarize data, whether numerical or non-numerical
(categorical). For example, the authors of one recent excellent book on scientific writing
(Cargill & O’Connor, 2013: 26) say that tables are
“most useful for

• recording data (raw or processed data);

• explaining calculations or showing components of calculated data;
• showing the actual data values and their precision; and
• allowing multiple comparisons between elements in many directions”.

Immediately after that, the authors present a table, reproduced below as Table 24.2,
summarizing the difference between tables and figures. It is ironic that this table does
not perform any of the functions they just listed for a useful table! Nevertheless, it
is useful as a summary of categorical differences in how tables and figures show
numerical data.

Table 24.2  Example of a summary table showing non-numerical categorical data

The choice between tables or figures for data display.

Most useful Table Figure

When working with number shape

When concentrating on individual data values overall pattern
When accurate or precise values are more important less important

Redrawn from Cargill and O’Connor (2013: 26, Table 5.1), used by permission.
254  Adding the final touches

It is important to distinguish between tables that are primarily numerical and those that are
primarily non-numerical, because there are differences in preferred layout based on the way
our brains process informa-
tion. In English we read words
as whole units and normally
It is easier to compare numbers read across a line of text, so
in a column, but words written it is more natural to compare
in a row, so numerical and words across a row. On the
non-numerical tables should be other hand, it is easier to com-
arranged differently. pare numbers in a column, as
we do not need to process the
meaning of each number but
can quickly compare their rela-
tive size based on place value.
The difference between the two arrangements can be seen by comparing Table 24.2 with
Table 24.3. In Table 24.2, the categories the authors want us to compare are in columns,
and we naturally read across each line to compare tables with figures. On the other hand, in
Table 24.3, the categories to be compared are listed in rows and we can quickly compare
the difference between the different arrays by scanning down each column of numbers,
looking only at the position and value of the first numeral to compare their relative size.
You may have noticed that there are no vertical lines in either table. Traditionally,
vertical dividing lines were not allowed because of difficulties in typesetting prior to com-
puters, and some journals retain this restriction. Whether permitted or not, reducing the

Table 24.3  Example of a good numerical table

Table 1
Array size and number of stations

Smallest station Largest station

separation separation Number
Name (m) (m) of stations

TCU062 S array 12.5 43.3 6

M array 25.0 86.6 6
6 No vertical lines.
L array 106.7 449.4
Horizontal lines between:
TAL001 S array 12.5 43.3 7 • Table title/
M array 25.0 141.6 7 column headings
L array 69.5 237.5 6 • Column headings/
X array 317.4 1192.1 6 table body
• Table body/footnotes
TAL002 S array 10.0 34.9 7 (or bottom)
M array 19.8 61.2 7
L array 84.0 227.6 7
TCU067 S array 21.7 40.4 5
M array 35.0 84.0 6

Redrawn from Satoh et al. (2001), used by permission.

 Illustrating your manuscript  255

number of lines in a table makes it easier to read, just like eliminating any other unnec-
essary information. In many cases, differences in spacing between rows or columns can
guide which way the table should be read and which material should be considered as a
single block, even more effectively than lines could.
Horizontal lines are common between:

• Table title/column headings

• Column headings/table body
• [Table data/summary statistic: totals or means] (optional, or white space, as in
Table 24.1)
• Table body/footnotes (or at the bottom of a table)

There are many details to consider when creating a table. First, check the instructions
to authors to find the requirements:

• Table size: single or double column, or no restriction?

• Font size: minimum? (make sure everything will be readable as printed)

There are additional things you can do to make numerical data easier to read:

• Necessity: delete any data unrelated to the research goal.

• Importance: most important comparison first.
• Order: logical based on the most important comparison, not testing order.
• Alignment: align numerical data on the decimal, right align if no decimal.
• Precision: identical for all data of one type, only enough to allow comparison.
• Summary: report a mean or total to no more precision than the data.
• Exponents: place exponents in the unit rather than within the table.

There are often specific requirements for the format of titles:

• Justification: centered or left?

• Placement: title on the same line as the table number or the following line?
• Font: Arabic or Roman numeral, capitalization, bold, etc.?
• Periods: period after the number? Period at the end of the title?

Here are some further suggestions for the title and headings:

• Short titles: table titles in many fields of engineering are less than one line. Additional
explanation may be placed as a caption at the bottom of the table.
• Column headings: center headings over data. By default, heading text is left justi-
fied and numbers are right justified, leading to misalignment.
• No split words: do not split words in headings. Make the column wide enough to
fit the words, hyphenate by syllable, or abbreviate.
• Units: indicate units as a second line after the column headings, in parentheses, not
in the table body.
• Minimize white space: eliminate extra space. Make each column wide enough for
the headings and data, not wider, unless space is used to guide reading.
256  Adding the final touches

The default settings for tables

in MS Word do not produce a
Journals and theses have publication-quality table. By
minimum requirements for default, columns are the same
table format, but you should width, text is left aligned,
do more, to make them and cells have complete bor-
ders. Fortunately, it is easy
easy to read.
to modify most of these set-
tings. Since you want your
paper to stand out for excel-
lence, it is worth taking the
time to find out how to use the features on the Tables tab. Here I will only note two
procedures that are less well documented: aligning on a decimal and table footnotes.
To align on a decimal, open the ruler by clicking the box at the top of the left slider.
Next, click on the tab marker until you see an inverted T with a dot next to it, meaning
align on decimal. Then highlight the cells in the table that you want to align. Finally, click
on the ruler where you want the decimal to be. You can adjust the position by sliding
the marker.
Explanatory footnotes for your table should appear immediately below the table, not
at the bottom of the page. In Word this can be accomplished by placing section marks
before and after the table, using endnotes for the notes within the table, placing endnotes
at the end of the section, and restarting numbering each section. Of course, this only
works if you have no other endnotes in your document. Alternatively, you can use nor-
mal superscripts.

Exercise 24.1 Evaluating tables in your exemplars

(1) What types of tables are used in your exemplars (numerical or non-numerical)?
Where are they (which division)? What is the purpose of each (what components are
they associated with)?
Hint: Many will have two purposes. All summarize data patterns (Component 7), but many
also use those data in a comparison (Component 8) or interpretation (Component 9).
(2) How long are the titles (number of words)? What is the format of the titles (justifica-
tion, numbers, capitals or bold, periods)?
Hint: If your journal has an electronic template, the template should do the formatting
automatically.
(3) Could any be improved?
Hint: Just because a graphic is published does not mean it is good.You need to fit in as far as the
type of information presented but can stand out with appearance.

Teaching tip: Exercises 24.1, 24.2, and 24.3 expand on Exercise 8.1.
Teaching tip: Have students find poor examples from their exemplars and share them with
the rest of the class, saying how they could be improved.
 Illustrating your manuscript  257

Teaching tip: Do not assume students understand how to use the features on the Tables tab
in Word. Many do not.

24.4 Figures: graphs

Two distinct types of graphics are classified as figures in academic writing: graphs and
illustrations. Most of the discussion in writing books is about the former, as illustrated
by Table 24.1, in which “figure” means a graph displaying the same data as the table.
Illustrations usually are mentioned only in passing, almost a footnote. Nevertheless,
since illustrations are important in engineering, I will discuss them separately in the
next section.
Graphs are common in the Results or Testing division as a way of summarizing data.
Line graphs are used to show change over time or across a range of values of a con-
tinuous variable. Bar graphs are
required when the variable is
discontinuous. Pie graphs and Do not be satisfied with a
radar graphs are rare in science
and engineering. default graph, which almost
As recently as 30 years ago, always contains unnecessary
graphs were drawn by hand by or poorly ordered output.
an author and reproduced by
graphic artists. Every unimpor-
tant point or line was eliminated.
Now graphs can be generated by
a computer with almost no human intervention, but as a result they often contain far too
much information, making them hard to read.
Here are some ways to make it easier to read your graphs:

• Highlight: most prominent line = heavy line weight, dark solid color.
• Importance: highlight the current standard, or your work?
• Simplify: not too many lines on one graph.
• Scale: same scale on similar graphs facilitates comparison.
• Comparison: emphasize similarity or difference by choice of scale.
• Axes: time or experimental variable = x-axis; response variable = y-axis.
• Error bars: only if necessary; consider showing fewer, or a cone of variance.
• Color: still distinct if printed in black and white?
• Boxed: few points = box around graph; complex = no box.

Many of the points mentioned with reference to table titles apply to figure captions as
well. The following are more specific to figure captions:

• Captions: often longer and more descriptive than titles, sometimes sentences rather
than phrases; may contain explanation of axes or units in a second sentence.
• Axis labels: begin with a capital letter, but following words in small letters.
• Units: in parentheses, following label.
258  Adding the final touches

• Consistency: maintain order of lines and patterns for all similar graphs.
• Legend: order of legend should match order of lines on the graph, top to bottom.

Exercise 24.2 Evaluating graphs in your exemplars

(1) What types of graphs are used in your exemplars (line, bar, or other)? Where are they
(division)? What is the purpose of each (component)?
Hint: Many will have two purposes (see Exercise 24.1).
(2) Are the captions phrases or sentences? How long are they? What is the format?
(3) Could any be improved?
Hint: Discuss your suggestions with a partner. Different people will have different ideas on what
is good or could be improved.

24.5 Figures: illustrations

In sharp contrast to science articles, illustrations are by far the most common type
of graphic in engineering, in many journals surpassing the number of tables and
graphs combined. As men-
tioned in Chapter 8, in
IPTC they frequently are
Illustrations are far more found in the Process divi-
common in engineering sion, or even Introduction.
than science, frequently In some cases they may
depicting some aspect depict a physical arrange-
of the current design. ment, either schematically
or photographically, that
is hard to describe in text.
They may also represent
an abstraction such as a
waveform. Occasionally they show a flowchart or other process. As with graphs, one
of the most common problems is too much information.
Illustrations are very field-specific, so it is hard to give general guidelines. What would
be considered too cluttered in one field might be considered just right in another. Some
journals prefer a smaller number of figures, even if that means several very loosely related
figures are grouped together.
Nevertheless, some general suggestions can be made about illustrations:

• Necessity: remove unnecessary or repetitive information.

• Photographs: retain proper aspect ratio (length to width) when resizing.
• Background: declutter photographic images by cropping.
• Resolution: clear, generally minimum 300 dpi.
• Text: large enough to read in the final print version, not on your monitor.
 Illustrating your manuscript  259

Exercise 24.3 Evaluating illustrations in your exemplars

(1) What types of illustrations are used in your exemplars (photograph, diagram, schematic,
other)? Where are they? What is the purpose of each?
Hint: Illustrations are frequently used to depict details of the current study, but some show results
or comparisons.
Hint:You may want to consider different types separately.
(2) Are the captions phrases or sentences? How long are they? What is the format?
(3) Could any be improved?

24.6 Graphics that break the rules

Many books on writing present a list of rules for what should be included in a table,
what should be presented in a figure, and what should be mentioned only in the text.
Unfortunately, such rules are made to be broken. The main consideration is what will be
easiest for the reader.
For example, some authors say that a list of conditions should be reported only in
text. Yet Table 24.4 shows an example of a type of table that is very common in engi-
neering. Since it simply lists the operating conditions of the system and could easily be
presented as a single sentence of text, we need to ask why it is so common. The obvi-
ous explanation is that it is easier both to find and to read, just like bullet points in a
conference presentation.
Sometimes rules are broken for the sake of ease for the author rather than the reader.
For example, in engineering it is common to compare several dozen characteristics of
two to four designs. Ideally, according to what I said above, each characteristic should
be one column, so readers can compare similar results vertically. Yet in many papers
characteristics are listed as rows. This saves space and formatting time, since fitting the
names of all those characteristics (sometimes several words) is easier in row headings
than column headings. As long as the number of designs compared is not too great, this
seems to be acceptable.
For any rule that is given, including the suggestions I made above, there will be
exceptions. If a certain type of graphic is common in your field, feel free to use it even if
it seems to go against rules set by writing instructors who do not know your field. You

Table 24.4  Sample table of operating conditions, common in engineering

Table I
Specifications of the PV Array

Rated power 900W

Open circuit voltage 150V
PV array
Short circuit current 7A
Connection 6 (each 150W)
260  Adding the final touches

need to fit in. On the other

hand, think about whether it
The only two inviolable really is a useful way of pre-
rules are that each graphic senting the information or if
must be mentioned in the there might be a better way.
text and each should be Even if journals in your field
helpful to the reader. publish tables using the MS
Word default showing all the
gridlines, could you make
your graphics more visually
appealing? You want to stand
out for excellence. The only truly inviolable rule is that every graphic must be men-
tioned in the text.

Evaluate your understanding

(1) How do table titles and figure captions differ?
(2) How should the arrangement differ for numerical and non-numerical tables?
(3) Why is it beneficial to minimize the number of horizontal and vertical lines in a table?
(4) Why should the instructions to authors be considered minimum requirements rather
than the goal?
(5) How can you improve the default output of your graphing program?
(6) Why are illustrations more common in engineering than science?
(7) Why is it impossible for a writing textbook to give absolute rules on what makes a
good graphic?

Homework
(1) Format your graphics according to the instructions to authors for your journal.
Consider how to improve them using the suggestions in this chapter.
Hint: Just like your text, graphics should be revised several times. Do not simply accept the preset
output format from whatever program or device you are using.

Teaching tip: Students doing peer review sometimes forget that standards differ by field.
Encourage them to ask their partners why they chose to present a graphic a certain way or
use modal-based questions (Would it be possible . . .?, Could you . . .?) to make their sug-
gestions more polite.

Supplemental material:
Class exercise 24.1: Evaluating sample tables
Class exercise 24.2: Evaluating sample graphs
Class exercise 24.3: Evaluating sample illustrations
 Illustrating your manuscript  261

Notes
1 As mentioned in section 8.4, students using English as a lingua franca may have trouble distin-
guishing graph from graphic, but there does not seem to be a better general term.
2 This is almost universal, but I have seen one computer science journal (Journal of Machine Learning
Research) where both table titles and figure captions are placed under the graphic.
3 http://writingcenter.unc.edu/handouts/figures-and-charts.

References
CARGILL, M. & O’CONNOR, P. 2013. Writing scientific research articles: Strategy and steps,
New York, John Wiley & Sons.
LA MONICA, P. R. 2009. Inside Rupert’s brain, London, Penguin.
SATOH, T., KAWASE, H., IWATA, T., HIGASHI, S., SATO, T., IRIKURA, K. & HUANG,
H.-C. 2001. S-wave velocity structure of the Taichung basin, Taiwan, estimated from array
and single-station records of microtremors. Bulletin of the Seismological Society of America, 91,
1267–1282.
Chapter 25

References and citations

25.1 When the writing is done

I suggested in section 20.3 that you use placeholders for citations while writing and postpone
adding the actual citations and references until the end. This allows you to become more
familiar with the authors and dates of important papers in your field and to focus on writing
instead of bookkeeping details, while avoiding technical glitches.
In this chapter I will first talk about two different reference and citation formats. This
will be followed by a discussion of the benefit of using reference managers to avoid exces-
sive retyping and reformatting.

25.2 Reference and citation formats

There are two basic citation-reference formats, each with many different variations, the
numerical system common in engineering and the author-year system common in science.
Recall that citation refers to the marker in the body of the article that directs readers to the
reference listed at the end.
The easiest way to choose the right format for your journal is to use one of the
reference managers described in the next section. This section simply provides a lit-
tle background so you can
appreciate how much those
The two basic formats are tools do for you, and things
numerical and author-year, to check to make sure that
but the many minor variations your reference manager has
actually done the format-
from journal to journal make it
ting correctly.
very worthwhile to use Almost all engineering
a reference manager. journals use some variant
of the numerical system,
in which citations are indi-
cated by a number in the
text. The citation number may be shown in brackets, parentheses, as a superscript, or in
superscripted brackets, and thus will appear as [1], (1), 1 or [1], respectively. Multiple papers
within a single citation are usually listed in numerical order, e.g., [1, 2]. The citation
marker may occur within the sentence or after the punctuation.
There is even more variation when it comes to the references. References may be
listed numerically in order of appearance in the article (citation order) or alphabetically
 References and citations  263

by family name1 of the first author (reference order). The date of publication appears
near the end of the reference. Depending on the journal, either full names or initials may
be used for authors’ given names, and either given names or family names may be listed
first. If initials are used, some journals require use of a period after the initial while others
forbid it. All authors (or at least ten) are listed. References are often but not always listed
in block style, with the number to the left and space between each reference. This list of
differences is not exhaustive, but gives some idea of the amount of variation from one
journal to another.
Still more variation exists in the author-year format common in science journals,
because of the variation in citations. The basic format for the citation is the family name
and date of publication placed in parentheses in the text, e.g., (Alpha 1993). Sometimes
the two are separated by a comma, but not always. If there is a second author, that author’s
family name is added, with either “and” or “&”2 between the two names, e.g., (Alpha and
Beta 1998) or (Alpha & Beta 1998). Sometimes a third author’s name may be added, but
usually additional authors are indicated by “et al.”3 If multiple papers are mentioned in a
citation, some journals require that they be listed in alphabetical order, and others by date
of publication, e.g., (Alpha 1993, Beta 1990) or (Beta 1990, Alpha 1993). Citations are
usually within the sentence, before the period, but not always.
The variations in the references in author-year format are similar to those listed for the
numerical system above. The main differences between the two systems are that the year
of publication is placed immediately after the authors’ names (always listed family name
first) and a hanging indent is common, often without space between entries.

Exercise 25.1 Entering references and citations

(1) Examine the citations and references in several different journals. List the differences
between them. What system and which variations are used in the journal you are
writing for?
Hint:The instructions to authors for your journal will contain specifics on the format they require,
along with examples for various types of sources.

Teaching tip: Instead of trying to teach the differences, allow students to compare the
format of references in several different journals, noting as many differences as they can.
See Class exercise 25.1.

25.3 Reference management

Less than a generation ago authors
would spend hours typing references
Reference managers
in the format required by each jour-
nal, but those days are fortunately are changing rapidly,
behind us. Instead, the question now with many younger
is which of the multitude of reference researchers turning to
management tools to use. Since the cloud-based platforms.
tools available are changing quickly, I
will only comment on the advantages
264  Adding the final touches

and disadvantages of different types of tools and some of the better known examples rather
than all of the individual programs. There are currently three types of reference manage-
ment tools available—those built into word-processing software, separate reference man-
agers that run on your computer, and online cloud-based accounts.
Word-processing software has the advantage of integration. Since the reference
manager is an integral part of the program, compatibility is not an issue. Those writing
in LaTeX will want to use the integrated BiblioTeX reference manager. In MS Word
the main disadvantage is the limited number of formats supported. Initially only three
were offered—APA, Chicago, and MLA—which was not very useful for scientists and
engineers. Recently several others have been added, including IEEE, and the choices
will undoubtedly expand in the future, but for now it is still quite limited. Currently
sources must be typed or copied into the master list manually, another severe limita-
tion compared to the other two types.
Computer-based reference managers, such as Endnote, support formats for thou-
sands of journals.4 Just as important, it is possible to download references from many
online sites directly into the database, eliminating the need to type in the information.
On the negative side, the software is not free, takes space on your disk, and your data-
base may be lost if your computer crashes. Some schools or libraries offer a site license
on one of the reference managers so faculty and students can use it without paying
individually, but that license expires if you move to a different school without such
an agreement, requiring you to purchase an individual license or export your data to
a different program.
Online reference managers are the latest addition to the market, and the choice of many
younger researchers who grew up on computers. Some of the larger, better known sites
include Mendeley, RefWorks, and Zotero, but there are many others as well, catering to vari-
ous needs. These also allow direct import of references from journals or search engines and
have many export formats available. If you are not familiar with reference managers, I suggest
you ask your colleagues what program they are using, as they will be the best resource on
current options.
No matter what management system you choose, the input method is similar. Sources
are added to the master list, those needed for a particular paper are selected from that
master list, marks are placed in the text to indicate the location of citations, and the style
of output is set. The program does the rest. Well, not quite.
It is still incumbent on the
researcher to check the printed ref-
It is still up to you to do erences. Errors in the reference list usu-
ally reflect errors in the database that
the final check to ensure need to be corrected in the master
that the reference manager source list. Odd sources, such as
has output the results articles in an edited volume that is
correctly. part of a series, can be a particular
problem, and may require trial and
error to find out how to enter them
so they display correctly.
 References and citations  265

Evaluate your understanding

(1) What are the two basic types of citation/reference formats, and some of the variations
within each?
(2) What types of reference managers are available, and what are the advantages and dis-
advantages of each?
(3) How can errors in references arise, even when using a reference manager?

Homework
(1) Using whatever method you choose, add all the necessary references and citations to
your document in the proper format.

Teaching tip: Prepare a sample document with a few references. Show how it is possible with
a few mouse clicks to change the format of both citations and references. Show the number of
formats available in other reference managers compared to Word.
Teaching tip: This chapter and the following chapter may be taught in one class, as none of
the exercises takes long to explain or complete.

Supplemental material: Class exercise 25.1—Differences in reference formats.

Notes
1 I use “family name” and “given name” instead of the more common American designations “last
name” and “first name” because in many Asian languages the family name is written first.
2 This mark is called an ampersand, derived from the Latin word et, meaning “and.”
3 The abbreviation “et al.” stands for the Latin et aliquid, meaning “and the rest.” Since et is a full
word, but aliquid is abbreviated al., there is a period only after the latter.
4 More common formats are included in the Endnote software; others can be downloaded from
their website.
Chapter 26

The first shall be last

26.1 Attracting attention

The title and abstract are the first parts of your paper that a reader will see, but the
last parts written, a new twist on a classic paradox. After completing a 10,000-word
article, writing another 300 words might seem like a small task, but these parts are
vitally important to the success of your paper and should be taken seriously. If poorly
done, all your hard work in the paper will be for naught, because these final bits you
write are what attract readers. If graphics are the 60-second trailer that gets people
hooked, the title and abstract are the 30-second preview that catches their attention
in the first place.1
I can’t vouch for the accuracy of the numbers, but it has been estimated that the
number of people reading an article follows a law of powers of 10. Of those who read
the title, only 10% will read the abstract. Of those who read the abstract, only 10%
will skim the article, looking at the graphics. Of those, only 10% will read the whole
article carefully. If this is true, only one out of a thousand people who read your
title will read the whole article. The numbers are certainly not exact, but the princi-
ple holds true—if you want people to read the article, they have to find it, and find
it interesting. I will cover the various attention-catching parts in the order I suggest
you write them.

26.2 Abstract
Although you may have written a preliminary abstract before beginning to write,2 you
will need to completely rewrite it once the paper is complete. Not only will your
thoughts be clearer now, but you have undoubtedly made changes in your aim, proce-
dures, or conclusions in the course of writing. You may have included things you did
not plan, while eliminating others.
To understand the structure of an abstract, let us return one more time to our exem-
plar articles. The abstract is a summary of the entire article, thus it frequently contains
something from each of the divisions, often in the same order but not necessarily the same
proportion as the article.
Although in science the focus of the abstract tends to mirror the focus of the arti-
cles themselves, with much of the abstract devoted to the Results and Discussion, this
is not necessarily the case in engineering. In IPTC more attention may be focused on
 The first shall be last  267

either the need for the work

or the contribution rather
In science most of the abstract than the design itself, per-
is devoted to Results and haps because it is hard to
Discussion, just like the article, briefly summarize so many
but in engineering little space design details. Half of the
may be devoted to the new abstract may be devoted
Product or Process. to the Introduction and
Conclusion.
To demonstrate, let’s
return to our sample article,
“Spatial query integrity” (Hu et al., 2013). Table 26.1 shows key phrases that allow us to
recognize the components represented by each sentence, and the divisions where each
component is found. Half of this abstract is related to the Introduction, defining the prob-
lem and stating the proposed solution. On the other hand, only one sentence summarizes
the Process division, and even that refers to the framework rather than the design details,
which comprise more than half of the article. This is followed by a two-sentence summary
of the testing, including how the new system compares to the current state-of-the-art
solution, and therefore the contribution of the new design.
It is also important to pay attention to the length of the abstract. In some journals,
the abstract may be as short as 4 sentences or 100 words. In others it may be 12 or
more sentences and over 300 words. Even longer abstracts are common for theses.
Conference abstracts vary considerably. Some journals with medical applications
have a required abstract structure that allows practitioners to quickly find the infor-
mation they need.3 Needless to say, your abstract should match the requirements of
your target.

Exercise 26.1 Structure of exemplar abstracts

(1) How many sentences do the abstracts of your exemplars contain? What is the average
number of words per sentence?
Hint: Use exemplars of whatever genre you are writing, if available.

Table 26.1  Components in each sentence of the abstract of “Spatial query integrity”

# Component markers Comp. Div.

1 . . . has grown rapidly over the past few years. 1a I

2 . . . has provided a flexible and cost-effective platform . . . 2b I
3 Nevertheless, . . . remains a challenging problem. 2b I
4 In this paper, we focus on . . . and propose an efficient scheme . . . 3b I
5 Our approach is based on . . . 4d P
6 We evaluated . . . on real-world datasets using mobile devices . . . 6c T
7a Compared to the current state-of-the-art . . . 8d T
7b . . . produces significantly smaller verification objects and is more 10b C
computationally efficient . . .
268  Adding the final touches

(2) Based on your knowledge of the components, assign each sentence to a component
and to the division where that component usually occurs in an article. Do the sen-
tences appear in the same order as the order of the divisions? How many sentences
are devoted to each division?
Hint: Sometimes a sentence may include more than one component, for example results (7) and
comparison (8).
(3) Is there a required structure, as in some medical journals?
Hint:This is relatively uncommon.
Hint: Could you present all the information in this Exercise in one table, or would it be better
to create more than one, or should some be stated only in the text?

Teaching tip: It is very worthwhile to have students compare the tables they create for the
abstract. With so many different types of data, there will be many options, and it will help
them evaluate what works well and what does not.

Now let’s consider how to write your own abstract. An abstract will almost always
include a statement of the research goal, and something about the framework to indicate
the connection to previous research. It will also include a summary of any comparisons
made, whether to expectations or previous research. It may include an introduction of
either the background or importance and a one-sentence conclusion discussing implica-
tions, applications, or limitations of the study.
Present tense and active voice are common in abstracts. While it would be possible to
use the past tense to describe the research you did (“in this study we proposed a new solu-
tion”), it is more common to use the present tense to describe the paper that immediately
follows (“in this paper we propose a
new solution”). Similarly, although
Copy ideas but not it would be possible to write in the
whole sentences from passive voice (“in this paper a new
your paper, then solution is proposed”), it is usually
rewrite for conciseness, clearer and more concise to write in
the active voice.
coherence, and clarity. Students often wonder whether
they can copy sentences from the
body of the paper into the abstract.
While this does not constitute plagiarism, it is generally not a good idea, as the abstract is
much more concise than the body of the paper.
I recommend following the 7Cs of Change when writing your final abstract. Begin by
selecting important sentences from your paper and copying them into a new document. Next
eliminate any unnecessary or unimportant words or phrases (conciseness). Then rewrite the
remaining phrases as new sentences to create a logical flow of thought (coherence). Finally,
consider how to link the sentences together or modify your claims to make your argument
clearer and stronger (connection and connotation). Particularly for beginning writers, this
ensures that all the main ideas from your paper are present, but nothing else.
 The first shall be last  269

There are different types of abstracts, including conference abstracts and the abstract of
your thesis in addition to the abstract of an article. Each is slightly different, but in some
situations parts of one can be reused for other purposes, as we will see in section 27.6.

Exercise 26.2 Writing the final version of your abstract

(1) Using the information on what is expected in an abstract in your field from Exercise
26.1, rewrite your abstract.
Hint: The key arguments in your article should be clear from the abstract alone. Do not simply
narrate what you did, but claim and support that your work makes an important contribution.

26.3 Title
The title is the only part of your document that most people will see. Its purpose is to
allow researchers in the field to decide how relevant your work is to theirs. Thus a good
title will be self-explanatory to your target audience, containing words that identify both
the purpose and contribution of your research.
Research article titles may be noun phrases, statements, or questions. Noun phrases are
by far the most common. Statements may be used in science in situations where a simple
answer to a research question can be stated (Cargill & O’Connor, 2013: 66). Questions
are far more common in the social sciences, where the research goal may be framed as a
question. The length and amount of information contained vary greatly by field. While
titles of fewer than 10 words are common in electrical engineering, titles may be up to
30 words in some chemical engineering journals. Colons are common in some fields, and
may indicate various relationships, including:

• Topic: Method;
• Problem: Solution;
• General: Specific;
• Major: Minor.

Titles often indicate both the gen-

eral topic of the research and what
It may be possible to turn
makes the paper unique—its con-
tribution. One of the easiest ways your research goal into a title,
to make sure your title summarizes with the contribution first to
your article is to turn your research foreground it, followed by the
goal into a title. Try to make the general research area.
key contribution the first part of the
title, with the research area second,
to foreground what is unique about
the current research. The two parts of the title are usually linked with a word that indicates
the nature of the connection, for example with, for, by, using, or based on. Your title may
also include words that identify the nature of the study, such as preliminary, case study,
framework, or simulation. While it is important to be concise in all scientific writing, it is
even more important in the title, where the importance of every word should be weighed.
270  Adding the final touches

Exercise 26.3 Creating a title for your article

(1) How long are the titles of your exemplars? Are the titles statements or questions
(containing a verb) or noun phrases (no verb)? Do any contain a colon? If so, what
is the relationship between the first and second parts?
(2) Do the titles include the research goal or main contribution of the research, the
general topic or research area, or anything else? In what order?
Hint:To get a broader sample, examine the titles of all the articles in your references.
(3) Write several different versions of a title before choosing the best.
Hint:Try to be as concise as possible without losing clarity, placing the most important information
near the front.

26.4 Keywords
In order to choose good keywords, it is essential to understand how they are used, which
requires a little historical context. Keywords, also called key words or index terms, are
a holdover from the pre-Internet
era, when they were used to compile
Choosing keywords printed indices to help people search
not in your title can for articles. When I was a graduate
help extend the appeal student, the only way to find articles
of your article to a was to go to the library and pore over
different audience. the small print in those huge volumes.
Since titles often convey only the spe-
cific topic, keywords were used to
place an article within its wider con-
text, allowing more people to find it. With the rapid rise in computing power, modern
search engines search not only the title and keywords, not even just the words in the
abstract, but all the words in the article. Thus keywords no longer fulfill the purpose they
were created for. Now they are more likely to be scanned visually by a researcher, along
with the title and abstract, as part of the decision whether to read the article or not. Thus
there is little or no benefit to listing words in your title as keywords.
Some journals require or suggest that you choose from a predetermined list of keywords
or index terms.4 Where possible, specific words in your title should be supplemented by
keywords representing the broader field of study, without duplication. On the other hand,
depending on how many keywords the journal requires, it may be better to duplicate words
in your title rather than choosing words that do not fit.

Exercise 26.4 Deciding on keywords

(1) Check the instructions to authors to find how many keywords are required, and
whether they must be selected from a list.
(2) Look at your exemplar articles to see which keywords were chosen by others study-
ing a similar topic.
(3) Choose suitable keywords for your article.
Hint:Think outside the box. Are there certain techniques used in your article that did not make
it into the title, or perhaps an application?
 The first shall be last  271

26.5 Highlights
A few years ago Elsevier, a major academic publisher, began requiring highlights for all
journals they publish, adding yet another category for authors to write. Highlights are
intermediate in length between the title and abstract.
According to the Elsevier website:5

Highlights are a short collection of bullet points that convey the core findings and
provide readers with a quick textual overview of the article. These three to five
bullet points describe the essence
of the research (e.g. results or
conclusions) and highlight what Highlights should list the
is distinctive about it. core results of the article,
rather than importance, and
Thus the author is asked to write three may thus have a different
to five highlights, each no more than
85 characters including spaces, list- focus than the abstract.
ing the core results of the article. By
asking for only the core results, other
statements that might be in the abstract, such as background and application, are specifi-
cally excluded.
A quick analysis of their examples shows that each highlight is a simple declarative sen-
tence containing about ten words, including one verb in the present or past active form.
The content depends on the article, but generally includes material that would be found in
the Results and Discussion or Process and Testing divisions. They may include the material
studied or method used, but no details.

Exercise 26.5 Writing highlights

(1) Check the instructions to authors to find out whether highlights are required. If they
are, look at examples in the online version of your journal. Do they fit the general
pattern suggested above for length and sentence structure? What divisions of the
paper are represented?
(2) Write your own highlights following the same pattern.
Hint: As with the title, write several versions before deciding on the best.

26.6 Acknowledgments, author information, and other bits

Depending on the genre, there may be various other short parts to be written before submis-
sion. These may include acknowledgments, author information, and a note of online material.
If you (or your advisor) received
a grant that paid for any part of the
research (including your graduate assist- Funding sources or
antship), you must acknowledge it. previous publication in
There is a standard format for this that conference proceedings
can be copied verbatim from other should be noted in the
journal articles, apart from the grant acknowledgments.
source and number.6 If the material in
the article was previously presented at a
272  Adding the final touches

conference, this should also be noted, along with previous publication of any portion of
the material. Occasionally authors will acknowledge the assistance of someone who is not
an author. This is less common in engineering articles, which tend to have more authors,
than science, where technicians may be recognized as having done some of the labor
although not contributing to the writing.
Some publications print information on the authors’ affiliations and contact informa-
tion, while others require a brief bio. These may be part of the article template provided
by a journal or may be a separate section of the submission process if a template is not
provided. Similarly, there is a standard procedure for referring the reader to online
material, if it exists.

Exercise 26.6 Gathering the last bits

Write the final parts, following the requirements of the journal, department, conference,
or funding agency.
Hint:You may need to gather information from your advisor or other authors.

Evaluate your understanding

(1) Why is it essential to rewrite your abstract after the paper is complete?
(2) Why do abstracts in science tend to reflect the structure of the article, while those in
engineering may not?
(3) What principles can help you create a better title?
(4) Should keywords duplicate words in the title?
(5) How do highlights differ from the abstract?
(6) What types of information should go in the acknowledgments?

Homework
(1) Ask a colleague to comment on the clarity of your abstract, and which of several pos-
sible titles sounds best or most interesting.
Hint: Ask someone not in your lab, as others in your lab will be too familiar with the terms and
procedures used to evaluate the writing as an outsider would.

Notes
1 The general pattern of researchers scanning the title and author, then abstract, followed by
graphics when deciding what to read has been documented by Berkenkotter and Huckin
(2016: 3), although they note that it differs somewhat depending on how central an article is
to the reader’s research area.
2 Chapter 3 homework.
3 For example, an IMRD oncology journal requires the abstract to discuss separately the
Background, Methods, Findings, Interpretation, and Funding.
4 Including all those published by IEEE.
5 www.elsevier.com/authors/journal-authors/highlights.
6 This work was supported [in part] by [grant agency] [grant number].
 The first shall be last  273

References
BERKENKOTTER, C. & HUCKIN, T. N. 2016. Genre knowledge in disciplinary communication:
Cognition/culture/power, London, Routledge.
CARGILL, M. & O’CONNOR, P. 2013. Writing scientific research articles: Strategy and steps, New
York, John Wiley & Sons.
HU, L., KU, W.-S., BAKIRAS, S. & SHAHABI, C. 2013. Spatial query integrity with Voronoi
neighbors. IEEE Transactions on Knowledge and Data Engineering, 25, 863–876.
Chapter 27

Submission and review

27.1 Warning: not every article is accepted

(but it’s not the end of your career)
Papers written for a class are accepted. What grade you get depends on the quality of your
work, but acceptance is not an issue. Many conferences also accept almost every submis-
sion, but when it comes to more prestigious conferences or journals, perhaps the submis-
sion site should come with a warning label—Warning: not every article submitted will
be accepted. Like most warning labels, it tells us something we already know, or should
know, but don’t want to think about. On the other hand, perhaps it should also include
an encouragement—but a rejection is not the end of your career as a researcher. Almost
every author has experienced rejection at some point.
In this final chapter I will focus on submission of a research article, but the same princi-
ples apply to conferences, grants, and other academic writing. There are four possible fates
for every document submitted—acceptance, conditional acceptance, request for revision, and
rejection. Outright acceptance without change is extremely rare. The best an author can
realistically hope for is conditional acceptance: a promise of acceptance if certain concerns are
addressed or changes made to the satisfaction of the editor. Requests for revision are quite
common, where the author is asked to make certain changes, then resubmit, and acceptance
or rejection is dependent on those changes. The rejection rate depends on the journal, with
more prestigious journals having a higher rejection rate.
Getting a rejection does not necessarily mean your paper is bad or unpublishable.
Frequently the problem is that it does not meet the requirements of that journal. I edited
one paper for a seismology professor twice, but the editor rejected it twice, saying each
time that it needed to be edited by a native speaker of English, without mentioning any
specific problems. Although the grammar and spelling were correct, perhaps the word
choice was different from other articles in the journal, or perhaps the argument structure
did not match what the editor wanted. At my suggestion, the professor submitted the same
article to a different journal and it was accepted the first time, with only minor revisions.

27.2 Choosing the right journal

Submitting your paper to the right journal will greatly increase your chance of accept-
ance, while submitting to the wrong journal dooms your effort. Two factors are impor-
tant: the type of papers the journal accepts and the quality of your research relative to the
impact factor of the journal.
 Submission and review  275

It is easy these days to find out what type of papers a journal accepts, by going to the
homepage for the journal and looking for one of the following, or some related phrase:

• Instructions to authors;
• Author guidelines;
• Guidelines for manuscript submission;
• Submission requirements.

Somewhere on that page is likely to be a statement of “Scope” or “What types of papers we

accept.” For example, the IEEE Transactions on Power Electronics Guidelines for Manuscript
Submission include the following:1

IEEE Transactions on Power Electronics’ scope includes new results in the field
of power electronics. For example, papers that treat original component, device,
module, circuit, control, system, or application issues are of interest. Papers that do
not show sufficient overlap with interest in the power electronics field (for example,
papers on topics such as device physics, component manufacturing technologies, and
circuit theory) should be submitted to other IEEE Transactions that cover these other
fields. Authors should pay particular attention to documenting their work in relation-
ship to the known literature. Papers of a historical or tutorial nature within the scope
of this publication will also be considered.

This statement of scope makes it clear that certain papers marginally related to power
electronics are not likely to be accepted. It also gives a hint that in order to be accepted,
authors need to cite papers that document the connection between the current work and
past work published in that journal. This is a good reminder, and true for all journals,
whether specifically stated or not.
The impact factor of a journal is harder to find. Journals that are highly rated tend to
publicize their high impact factor, but those are not likely to be the journals you will submit
your first paper to, unless you happen to be lucky enough to discover something as important
as the structure of DNA.2 Impact factors measure how frequently articles in that journal are
cited, which does not directly translate into importance, as some fields tend to have higher
numbers of citations per article than others, which automatically boosts the impact factor of
every journal in that field. Complicating the matter, there are now several competing systems
for rating journals.
Unfortunately, there is sometimes a tradeoff in journal choice. The scope and readership
of a journal is the most important factor in getting your work read by the target audience.
Thus smaller, more specialized journals
may give you better access to researchers in
your specialty who are likely to read your
Many factors influence
work. On the other hand, the promotion
process often gives extra weight to papers choice of journal, including
published in journals with a higher impact scope of the journal, impact
factor. Thus, you may eventually want to factor, and stage of career.
publish in both—the first to get your name
known in the field, the second to look good
on your resume or promotion application.
276  Adding the final touches

If you are relatively new in your field, the best way to determine which journal will
be most suitable is to ask your advisor. He or she will know the journals in your field and
what type and quality of papers each accepts, and thus which is most likely to be willing
to accept your work.

27.3 Submission methods

Submission methods have changed drastically over the last two decades, and will undoubt-
edly continue to change rapidly with future advances in technology, so in this section I
will focus more on basic principles rather than actual methods. Twenty years ago the
researcher had to print and send three complete copies of the manuscript by what is now
known as snail mail (in those days it was just mail) to the editor, who then forwarded one
copy to each reviewer. Mailing back and forth took time, so an author would often hear
nothing for months. Email submission became common as researchers got connected to
the web and email systems were developed that could handle attachments. Now almost
every journal has some sort of electronic submission system to standardize the process,
although it may still take just as long to get a response. Nevertheless, the basic principles
of submission remain the same.
Before beginning the actual submission, it is important to find out what will be required.
In some cases the entire article must be uploaded, print-ready, on a template provided
by the journal. For others, differ-
ent sections of the article must be
Before beginning an online uploaded separately in different
submission, look through blocks on the submission form or
what will be required and as separate files. For example, the
gather all the materials you title, abstract, and keywords may
have to fit within blocks of a cer-
will need in the proper tain length to ensure they meet
format. length requirements, or tables and
figures may have to be attached
separately. Some journals require
items that are not part of the article itself, such as author bios or a cover letter. You may
also have to provide affiliations, physical and email addresses, phone numbers, or other
information on the authors, and should have those available in electronic format before
you begin the actual submission process. Some forms may require phone numbers or
addresses in a format that does not match those used in your area, and you will have to
find a way around this. This problem seems to be more common in submission to a con-
ference than to a journal. If files must be uploaded, make sure you check the requirements
for file type, size, and naming requirements. I suggest that you gather all the documents
you need for each submission in one folder, so it is easy to find and access them while
doing the submission, and months later after you get a response.
If you are asked to provide a cover letter in an electronic submission, do not duplicate
information found elsewhere on the electronic form such as author names and institutions,
the journal you are submitting to, or a statement that the research is original and not submit-
ted to or published in any other journal, as was common in hard copy and email submis-
sion. What can be included are clarifications of anything that might be unclear, such as any
problems you had entering data on the electronic form, or mention of supplemental material.
 Submission and review  277

Other things that should not be included in the cover letter include accomplishments
of the authors or advocacy for the paper. The paper should speak for itself. It is also not
necessary to include a request for a response, invitation for the editor to revise, or indi-
cation of a willingness to revise. The editor knows the expectations better than you do.
Any additional comments of this nature will undoubtedly be interpreted negatively, and
should be avoided.
If separate files are required, make it easy for your editor. Do not use the authors’
names, as review is often intended to be blind, but rather give a short but distinctive
version of the title. If there are additional files, such as graphics, give them the same title
followed by a part designator, e.g., Article title Fig 1.

Exercise 27.1 Preparing for submission

Look through the submission process for your journal. What files will you need, in
what format? What additional information will you need?
Hint: Gather all the information you need for the submission in one folder. Once you have everything
ready, submit—but not before checking everything one last time with all your coauthors.

27.4 How the review process works

Wolfgang Pauli, the famous theoretical physicist, stated, “I don’t mind your think-
ing slowly; I mind your publishing faster than you think” (in Alley, 1996: 178). The
purpose of review is to ensure that you have thought through everything before you
publish, and that your work is of high enough importance and quality to merit pub-
lication. Reviewers may point out weakness in citations, description of the current
research, data, or analysis—in other words, all the elements of evidence and inference
we have talked about.
After a paper is received by a journal, the editor will usually take a quick look to
make sure the submission is complete and within the general scope of the journal, then
send the submission to at least two reviewers. They read the paper, answer a set of
questions, write comments, and often give an overall assessment of the quality of the
paper or a recommendation on whether to accept or reject. After they complete their
review, the editor will summarize their feedback and comments, make a decision, and
send the results to you.
First-time authors often wonder what the editor and reviewers are looking for. The
types of questions reviewers evaluate could be summarized in three groups, leading to
four different fates, as shown in Figure 27.1. This is obviously an oversimplification, as
many different factors affect the final decision, including the personal preferences of the
editor, but gives a rough approximation of the process.
Table 27.1 shows examples of the types of questions in each group. Although the spe-
cific questions vary, they tend to fall into the same categories.3 As shown in the last two
columns, all the questions are related to topics covered in this text.
The questions are listed in general order of importance, so outright rejection may
occur for failure of any question in the first group. A paper may be rejected if the topic
is considered unimportant or the solution not novel enough to merit publication, or if
the argument is irreparably flawed or incomprehensible, making it impossible to assess
278  Adding the final touches

Submit article

Group 1:
Rejection
Contribution

Group 2:
Revise and resubmit
Argument

Group 3:
Conditional acceptance
Content

Acceptance

Figure 27.1  Flowchart of the editorial decision process.

the contribution. This is why

Your research must Components 1, 2, and 3 are virtu-
be important, provide ally ubiquitous in an Introduction.
a novel solution, and Perhaps the most common reason
for rejection, however, is that the
be within the scope of paper is not suitable for the journal
the journal, or it will be it was submitted to, as discussed in
rejected. section 27.2.
The second group of questions
looks at the details of the current
work. If changes are suggested at this level, it is unlikely that a judgment will be made to
accept or reject the paper until those changes are made, and you will be asked to resubmit
it for review after major revision. As shown in group 2 of Table 27.1, these questions focus
on the overall length and organization of the paper, as well as whether you have clearly
shown that you have a reasonable answer to your research question or a workable design
that is better than the current standard in at least one aspect. They will also check whether
your works cited are up to date, particularly whether you cite relevant recent work in
 Submission and review  279

Table 27.1  Groups of questions leading to different editorial decisions

Question Treatment in Location

this text in this text

Group 1—Negative review leads to rejection

1 Suitability for scope of journal Journal scope 27.2
2 Importance of problem or question Claim 1a 4.2
3 Contribution of design or answer Claim 1b 4.2
Group 2—Negative review leads to revise and resubmit
1 Acceptable organization and length Structure 3.4
2 Citations current, include journal of submission Citations 7.2
3 Acceptable statement of methods or design Claim 2 5.2
4 Acceptable results and conclusions Claim 3 6.2
Group 3—Negative review leads to conditional acceptance
1 Suitable title and abstract Title, abstract 26.2–3
2 Tables and figures necessary, clear Graphics 24.3–5
3 Appropriate amount of detail Organization 21.2–3
4 Claims justified Clarification 22.2–3
5 English usage and formatting Revision 23.2–3

the journal you are submitting to. In sci-

ence one of the most common requests at If any of the basic
this level is for clarification of the experi- components are missing
mental design or statistics. Specific things or poorly written, this will
reviewers will look for and comment on
lead to a decision to revise
are related to components other than
importance (1) (addressed in group 1), and resubmit.
including:4

• The connection between past and present research is clear (2);

• The theoretical framework chosen is justified (4);
• The design process or experimental setup is described clearly (5);
• The analysis or testing method is fully documented (6);
• The summary of the results is adequate (7);
• The claims about comparisons do not exceed what is justified by the data (8);
• The explanations are based on sound logic and consider alternatives (9);
• The aims of the paper as well as the conclusion are clearly stated and match the actual
design and testing (3, 10).

If the changes required are all at the level of group 3, the paper may be conditionally
accepted, subject to making the required changes. This could include:

• Rewriting the title or abstract to better represent the contents of the article;
• Revision of graphics;
• Adding more data or discussion, or a single reference;
280  Adding the final touches

• Eliminating unnecessary detail or unsupported claims;

• Modifying the strength of claims.

If there are extensive grammar, typographical, or formatting errors that do not affect con-
tent, an editor may ask that the paper be revised by a native speaker. If so, this should be
done after you have made all other required changes.

27.5 Responding to reviewer comments

As I have said many times, beginning in Chapter 1, it is difficult for a novice author to
understand how many times they will have to revise their paper before it is published.
Even after submission, the process is not over, as a paper is almost never accepted for
publication without further revision.
Students often wonder whether it is necessary to make all the changes suggested
by the editor or reviewers. It is not. There are three options for addressing reviewer
comments:

• Make the suggested changes;

• Make other changes to address the concern;
• Make no change, but justify your reason.

In addition to making the changes in the document itself, you must also write a let-
ter specifying how you addressed each of the reviewers’ comments. You must respond in
some way to each comment, although it
is not necessary to mention every com-
You do not have to make ment individually. It is important to
all the changes suggested be respectful in your writing, not
by the reviewers but only because it is good professional
must respond to every practice or because they are col-
leagues, but also because they will
comment. make the final decision on whether
your paper is published. There is no
fixed format for the response, but
there are at least three general formats the response may follow, and three basic ways of
ordering the individual responses.5
Just as there is a difference in the writing style of different authors, there is a difference
in the style of reviewers. Some reviewers are very terse, listing suggested changes without
additional explanation. Others will write extensively why they think certain changes are
necessary. Depending on how long the reviewer comments are, you may choose to use
one or more of the following methods when you respond:

• Highlight comments and respond;

• Extract comments and respond;
• Summarize comments and respond.
 Submission and review  281

Some reviewers take time to order their comments, while others make comments as
they go through the paper. Sometimes the order chosen by the two reviewers is similar, but
not always. As you respond to the comments, you may choose one of several basic orders:

• Respond to each reviewer separately, in the order listed;

• Respond to both reviewers in the order changes appear in the paper;
• Respond to more important changes first, then smaller changes.

The method you choose will depend on your personal preference as well as what will
make it easiest and clearest to respond to all the comments. Since most reviewers include
some explanation, questions, or quotes from your writing, it is generally good to either
highlight the main points you are responding to or extract them. The order chosen may
depend on how much similarity there is between the reviewer comments. If there is lit-
tle overlap, it may be easiest to respond to each reviewer separately. If there is a lot of
overlap, it may be easiest to
group similar comments and
respond to them together,
particularly if the two review-
Different formats can be used
ers make opposing sugges- when responding to reviewer
tions. Frequently a mixed comments. As with your
order is used, first dealing article, consider both what
with large changes in order will be easy to read and also
of importance or appearance, make a strong argument.
with small changes such as
typographical or formatting
errors summarized at the end
rather than listed individually.
If you are making the suggested changes, less comment is required. If the error would
have caused a major problem for readers, it is good to mention it separately and thank the
reviewer, while typographical or formatting errors can usually be summarized, as in the
following examples:

• Equation 1.3 has been corrected. We thank reviewer 1 for catching this error.
• All typographical errors noted by the reviewers have been corrected.

If you think the reviewers misunderstood what you wrote, try to think why they
interpreted it that way and how you could rewrite it to make the meaning clear.
This may involve rewriting it differently than suggested, including comments like
the following:

• Section 2.1 has been rewritten to clarify the main point of the argument.
• This sentence has been rewritten. We thank reviewer 2 for pointing out the ambiguity.

If you feel your wording is correct and essential, you can try to justify why it is correct.
For example, in an article on how to help science students learn to distinguish evidence,
282  Adding the final touches

inference, and argumentation (Rau, 2009), I wrote the following, but the editor struck
out certain words to make it more concise.

• “Thus, it is necessary not only to collect data, but also to make and test inferences and
convince other scientists that your interpretation is correct.”

I responded:

• “The wording is important in this sentence to emphasize that data do not speak by
themselves, but are both interpreted and used as part of an argument, reinforcing
the subtitle [An activity to help students learn about observation, interpretation, and
argumentation].”

The editor accepted my justification and reinserted the words.

One final thing related to responding to the editor is that changes made should be
highlighted to make it easier for the editor to focus on those changes. In MS Word, it is
common for coauthors to make changes using the Track Changes feature. Unfortunately,
many journals consider it inappropriate to send a document marked with Track Changes
back to the editor. Rather, items should be changed using Track Changes, then the
changes should be highlighted, and finally all changes accepted so only the highlighting
shows. It is not necessary to mark small things like typographical errors.

27.6 Borrowing from your own work

A common question in academic writing is how much you can borrow from something
you have written previously. It is tempting to think that, because it is your own writing,
it should be acceptable to reuse sentences or even whole paragraphs elsewhere. Whether
this is true or not depends largely on where the material was published.
Once something has been published in an academic journal it cannot be reused in
the same form anywhere else without proper attribution, because that printed material
is copyrighted and the owner of the copyright is the journal, not the author. This is
why publishing the same article elsewhere, or even substantially the same informa-
tion (in the same language or translation), is considered plagiarism and subject to
severe penalties, including termination of employment. Short portions may generally
be used if the original is cited, but
larger portions can only be repro-
Conference proceedings do duced with permission from the
not have the same weight as publisher.6
a journal article, so you can Articles printed in conference
borrow more from them than proceedings are a different story.
Conference papers are not considered
from other written work. to have the same weight as academic
articles, because the review process
is usually restricted to review of the
abstract, not the full article. Such papers will frequently be published later as journal articles
and may contain the same material and even much of the same wording, but always require
rewriting based on feedback received.
 Submission and review  283

A thesis or dissertation is intermediate between a journal article and a conference

paper. Unlike a conference paper, it is not considered preliminary, and is considered to
be published, but the author is usually the copyright holder and therefore can choose to
reuse some of the material. Thus, it is expected that many of the ideas from the paper,
including some of the graphics, may be reused in journal articles, but usually with exten-
sive rewriting.

27.7 Closing thoughts

There are three basic steps to any successful writing. First, consider your goal and audi-
ence. What is your purpose, and how will this affect your claims? Who will be reading
your work, why will they read it, and how will they read it? Second, look at successful
exemplars. Think about the structure, the words used, the formatting. What made those
exemplars successful? How do they fit the general pattern, yet stand out as excellent?
Finally, begin writing. Write in stages, prewriting, then writing, then rewriting over and
over until you cannot find a way to improve it—then ask someone else for comments
and revise it again.
Learning to write well is a never-ending process. The more you write, the better you
will get. If you take the time to think about what makes some writing more successful,
you will improve faster and be better able to help others, either as a reviewer or an advisor.
Enjoy the journey.

Evaluate your understanding

(1) Why might you choose to submit to different journals at different stages in your career?
(2) How does online submission differ from older methods?
(3) What factors are most likely to lead to outright rejection of a manuscript?
(4) What factors are likely to lead to a request to revise and resubmit?
(5) Do you need to make all the changes suggested by an editor?
(6) How do you write a response to the reviewer comments?

Homework
(1) Ask someone who has successfully published a paper if you could look at their response
to reviewer comments. What method and order did they use in responding? Did they
make all the suggested changes? If not, how did they change or justify their wording?
Hint: Some people will be more willing than others to help you in this way, as it is somewhat
face-threatening. Alternatively, if you coauthor a paper with a more experienced writer, notice how
they respond or modify your responses to the reviewers, and learn from that model.

Notes
1 www.ieee-pels.org/images/documents/New%20Guidelines%20for%20Manuscript%20
Submission%20to%20IEEE%20TPEL_2016.pdf (downloaded 12 September 2016).
2 As graduate students, Watson and Crick published the structure of DNA in Nature, the most
prestigious science journal in the world, but they are a rare exception.
3 One particular set of questions used for science can be found in Cargill and O’Connor (2013: 18).
284  Adding the final touches

4 For an excellent discussion of how to respond to comments on these areas, see Cargill and
O’Connor (2013: 102–103).
5 Cargill and O’Connor (2013: 99–107) provide excellent practical suggestions and examples for
responding to editors and reviewers. For a more theoretical perspective of peer review as an argu-
mentative discussion, see Berkenkotter and Huckin (2016: 61–77).
6 Short portions can be copied under “fair use,” usually fewer than 500 words of continuous text or
800 words of intermittent text, with proper citation. Reuse of figures and tables requires permis-
sion, unless they are substantially modified, in which case a citation including “adapted from . . .”
is sufficient.

References
ALLEY, M. 1996. The craft of scientific writing, New York, Springer.
BERKENKOTTER, C. & HUCKIN, T. N. 2016. Genre knowledge in disciplinary communication:
Cognition/culture/power, London, Routledge.
CARGILL, M. & O’CONNOR, P. 2013. Writing scientific research articles: Strategy and steps, New
York, John Wiley & Sons.
RAU, G. 2009. A new twist on ‘mystery boxes’: An activity to help students learn about observation,
interpretation, and argumentation. The Science Teacher, 76, 30–35.
Appendix 1

Generalized component list

Component claims
  1 Importance (I)
a General importance
Claim: This research is important to society
Support: Often general knowledge (to those in the field)
b Specific importance
Claim: This topic is important to researchers in the field
Support: Past research (citations)
  2 Need (I)
a Research gap (IMRD)
Claim: There is a gap in current knowledge or understanding
Support: Absence of past research (nothing has been done)
b Limitations (IPTC)
Claim: The current best solution is limited or less than ideal
Support: Past research (present best solution, limitations)
c Benefit
Claim: This will be beneficial to society or the field
Support: General knowledge or past research
  3 Research goal (I) [main claim]
a Specific topic (goal, aim, objective, focus, question) (IMRD)
Claim: We add to previous research in this area
b Proposed solution (IPTC)
Claim: We provide a better solution to this problem
c Contribution (IPTC)
Claim: We make an important contribution in this paper
  4 Framework (I or M/P)
a Accepted procedures (IMRD)
Claim: Accepted procedures or proven methods were followed
Support: Past research (citations or general knowledge)
b Problem formulation (IPTC)
Claim: The research extends or builds on an established approach
Support: Past research (advantages of previous research approach)
286  Appendix 1

c Model or theoretical framework

Claim: The research is based on an accepted model or established design
Support: Past research (model, framework, mathematical basis, design)
d Justification
Claim: This is the best method to follow for our research
Support: Past research (alternative methods)
  5 Research details (M/P)
a Materials obtained
Claim: Appropriate materials were obtained from reputable suppliers
Support: Names and sources of materials
b Methods followed
Claim: Care was taken to ensure good results
Support: Details of present research design and procedures
c New design
Claim: A workable new solution, product, or process was developed
Support: Details of new design, intermediate testing of parts
d New mathematical procedures
Claim: A mathematical theorem, procedure, or algorithm was developed
Support: Details of proof, lemmas, mathematical argument, algorithm

 6 Testing methods (M/T)

a Physical testing
Claim: Testing was done carefully, following verifiable procedures
Support: Details of data collection method
b Computer testing (IPTC)
Claim: Testing was done on a standard data set
Support: Details of testing method
c Simulation testing (IPTC)
Claim: Testing was done under realistic simulated conditions
Support: Details of testing method
d Statistical testing (IMRD)
Claim: Testing was done using standard statistical tests and tools
Support: Names of tests and software used, details of statistical design

 7 Data patterns (R/T)

a General pattern
Claim: A pattern can be discerned in the data
Support: Graphical representation or data summary
b Data transformation
Claim: The pattern is easier to see after appropriate procedures
Support: Description of data handling or transformation methods
c Exceptions to pattern
Claim: Some data do not fit the overall pattern
Support: Identification of possible exceptions, outliers, or data errors
 Appendix 1  287

d Classification of patterns
Claim: The patterns can be summarized into different categories
Support: Similarities in patterns

 8 Comparisons (R&D/T)

a Within the article

Claim: One treatment or design is different than others
Support: Graphics, data summary
b With prediction based on hypothesis (IMRD)
Claim: Data [support/do not support] hypothesis
Support: Statistical testing
c With expectations
Claim: Data [match/differ from/exceed . . .] expectations
Support: Graphics, data summary
d With theoretical ideal (IPTC)
Claim: Results can be compared with a theoretical (mathematical) ideal
Support: Basis for ideal and comparison with it
e With previous work
Claim: Data [support/refute/improve on . . .] previous work
Support: Past research compared with present, in graphics or text
 9 Interpretations (D/T)

a Analysis methods (IMRD)

Claim: Interpretation of the data was carried out in a reasonable way
Support: Details of analysis method, assumptions
b Cause of results
Claim: There is a reasonable cause for the data
Support: Explanation of data pattern based on model or design goal
c Elimination of causes (IMRD)
Claim: There is no better cause for the data
Support: Lack of support for alternative hypotheses or explanations
d Practical applications (IMRD)
Claim: Understanding gained through this research has practical use
Support: Potential applications of the new knowledge
e Limitations
Claim: Conclusions are valid within a limited range of conditions
Support: Limited range of testing or advantage

10 Conclusion (D/C)
a Summary
Claim: The research goal has been achieved
Support: Summary of research, answers to questions or new design
b Achievement or advantage (more common in IPTC)
Claim: Solution improves on past work
Support: Contribution of research
288  Appendix 1

c Future work (more common in IMRD)

Claim: Remaining questions can be answered or further improvements made
Support: Planned or possible future work

Support: evidence
  I Past research
a General background knowledge for those in field (often without citation)
b Definitions, terminology, or notation (with or without citation)
c Description of theoretical frameworks, models, or approaches
d Citation of specific methods, results, or conclusions from past research
  II Present research
a Numbered or bulleted examples or equations
b Details and sources of materials
c Details of present research design and procedures
d Details of new design, intermediate testing of parts
e Details of proof, lemmas, mathematical argument, algorithm
f Detailed description of data collection and testing procedures
g Data summary or highlight in text
h Data patterns visualized in graphics
i Data comparison, sometimes using statistical test results

Support: reasoning
III Logical
a Deductive: Testing predictions made by a general theory
b Inductive: Compiling observations to reach a general conclusion
c Abductive: Choosing the most likely of several possible explanations
IV Evaluative
a Evaluation of previous research
b Evaluation of explanations
c Evaluation of current design
  V Organization
a Initial summary to reveal argument structure of subsequent section(s)
b Reminder of the research goal
c Concluding summary of argument in preceding section(s)
d Referring to arguments made in other sections
Appendix 2

Concordance, Academic Word List,

and related tools

Definitions
Concordance program: Software that produces a concordance from a corpus.
Concordance: A list showing all the words in a document or corpus, with their imme-
diate context.
Corpus: A body of documents that are related to one another in some way.

Concordance: purpose
Concordance programs are useful for looking at words in context. For example, they
could be used with the exercises in this book to:

• Identify components, based on common component markers (4.1, 5.1, 6.1)

• Identify uses of statistical terminology (8.3)
• Find citation verbs and pointing verbs (7.3, 8.5)
• Find contribution, achievement, and application markers (22.1)
• Find logical and time sequence connectors (22.2)
• Find likelihood, frequency, and measurement qualifiers, modals, delimiters, and
hedges (22.3)
• See how prepositions and articles are used (23.2)

Concordance programs may be more suitable for linguists doing research than engineers
and scientists learning how to write. For the former they are a tool they could continue
to use, whereas the latter are unlikely to use them outside a writing class.

Concordance: steps
I will briefly introduce a family of programs developed by Lawrence Anthony, available online
with excellent simple instructions. The basic concordance program is AntConc (Anthony,
2018). It and the related programs mentioned below can be downloaded free from www.
laurenceanthony.net/software.html. The most current versions of the instructions are listed
under “Help” for each program. Many similar programs are available elsewhere.

(1) Create a corpus.

Science: Use AntCorGen to identify articles from the PLOS ONE database of
science and medicine. Once the articles are chosen, the program will automati-
cally generate a corpus.
290  Appendix 2

Engineering: Use AntFileConverter to convert your PDF or Word files to text

files. Open the files you want to convert in the online tool, click Start, and they
will be converted into text files in the same directory. Note that converter pro-
grams have difficulty with equations and similar features common in engineering
texts, so there will be gaps and not all lines will display properly.
(2) Create a concordance, searching for words of interest.
Use AntConc to open the text files you want to examine. Create the concordance
following page 3 of the instructions. The results can be sorted alphabetically by
words to the left or right of the target word. Currently the latest instructions are
available at www.laurenceanthony.net/software/antconc/releases/AntConc357/
help.pdf.

For example, if you want to see what prepositions are used with “correlation” in
(Kleinheinz et al., 2009), the results are shown in Figure A2.1, sorted by the words
following the search term.

Academic Word List

Another tool teachers may find handy is the AntQuickTools (www.laurenceanthony.
net/software/antquicktools), which will highlight words from the Academic Word List
(Coxhead, 2000) in any text entered, for example an exemplar article. While this will
not point out specialized words specific to the discipline, students can identify those
easily. Instead, it will point out more advanced general English words they may need
to learn.

Figure A2.1  Sample AntConc output, search for “correlation” in one journal article.
 Appendix 2  291

References
ANTHONY, L. 2018. Software. www.laurenceanthony.net/software.html.
COXHEAD, A. 2000. A new academic word list. TESOL Quarterly, 34, 213–238.
KLEINHEINZ, G. T., MCDERMOTT, C. M., HUGHES, S. & BROWN, A. 2009. Effects of rain-
fall on E. coli concentrations at Door County, Wisconsin beaches. International Journal of Microbiology,
https://doi.org/10.1155/2009/876050.
Appendix 3

List of supplemental material (online)

Teacher material
Complete contents of textbook, showing all levels of subheadings
Teaching from the textbook: schedules, assignments, and rubrics

Student material

Audio PowerPoint introductions of each chapter (about 15 minutes each)

Chapters 4–6: sample component analysis of two articles
Chapter 22: supplemental material (particularly useful to ELF students)

Class exercises
  9.1 Sequencing strategies
19.1 Linearizing
21.1 Coherence (with answer key)
21.2 Coherence 2
21.3 Conciseness (with answer key)
21.4 Conciseness 2
22.1 Logical connectors (with answer key)
23.1 Eliminating ambiguity (with answer key)
23.2 Articles (with answer key)
24.1 Evaluating graphics
25.1 Reference formats
Glossary

7Cs of Change  A stepwise rewriting process including coherence, conciseness, connec-

tion, connotation, consistency, correctness, and collaboration.
Argument  A claim supported by evidence and reasoning.
Argument structure  A summary of the claims, component claims, and types of sup-
port used in a document, related to the division structure.
Author-prominent citation  Citation of previous work in which the author’s name
appears as part of the sentence, also called integral citation.
Citation verb  Verb referring to what a previous author did.
Citation  Mark in the text indicating the source of information used to support a state-
ment. See author-prominent citation and information-prominent citation.
Claim  An assertion or proposition, which may include one or more component claims,
requiring support.
Component  Component claim, a subclaim commonly found in science and engineer-
ing research writing, often in a predictable order, with or without recursive cycles.
Component analysis  Identification of the components and support present in an
exemplar.
Component markers  Common phrases or groups of words without specific content
that help readers identify components and that can be reused without fear of plagiarism.
Condensed style  Writing style that assumes the reader has specialized knowledge of the
subject, often with many ideas in one sentence.
Division  Conceptual units of a research article (either IMRD or IPTC) that can be dis-
tinguished by their basic claims, and may include one or more sections.
English as a lingua franca (ELF)  The use of English as a common language of com-
munication among speakers of various first languages.
Evidence  Details about or data from the present work or citations of previous work,
used to give an indication whether a claim is true or false.
Exemplar  A representative work of a particular genre that can be used as an example of
the style expected in that genre.
Experimental design  Details of the subjects, variables, measurement, and testing meth-
ods used in quantitative research.
Extended style  Writing style that is easy for a reader to follow, often with an initial
summary followed by details and one main idea per paragraph.
Framework  The theoretical framework, model, or accepted procedures followed or
modified in the present research.
294 Glossary

Genre  A distinct style of writing or speaking. Defined by linguists as a class of communicative

events within a discourse community to accomplish a shared set of communicative purposes.
Graphics  All the tables and figures in a document. Figures include graphs and
illustrations.
IMRD  Introduction, (Materials and) Methods, Results, Discussion. The four divisions
of a prototypical science research article.
IMRaDC  Introduction, (Materials and) Methods, Results and Discussion, Conclusion.
A modification of IMRD. Common in some sciences and branches of engineering
dealing with more quantitative research, such as chemical engineering and some fields
of civil engineering.
Inference  A logical conclusion, either inductive or deductive, based on the evidence,
often incorporating background knowledge or assumptions.
Information-prominent citation  Citation of previous work in which the author’s
name does not appear in the sentence, also called non-integral citation.
IPC  Introduction, Proof, Conclusion. A modification of IPTC. Common in math-
ematics, theoretical physics, and some fields of electrical engineering that focus on
mathematical proof and therefore do not require empirical testing.
IPTC  Introduction, (Product or) Process, Testing, Conclusion. The four divisions of a
prototypical engineering research article. Also used in some branches of physics dealing
with design or theory rather than quantitative research.
Plagiarism  Using ideas and words of another person as if they were your own, or using
copyrighted material without proper citation or permission.
Pointing verb  Verb used to point a reader to graphics or to information found in other
sections of the article.
Prewriting  Gathering (brainstorming) and ordering (linearizing) of ideas before writing.
Recursive  Occurring in cycles, as when a series of components is repeated, e.g. repeat-
ing Components 6, 7, 8 in the Testing division to test and compare different parts of
a new design.
References  Listing of complete bibliographic information of all works cited in the doc-
ument, usually found at the end without a section number.
Research article  A report of research conducted by one or more authors, published in
a refereed journal, also called article, journal article, or academic paper.
Research gap  Identification of an important question or problem not sufficiently cov-
ered by previous research, associated with Component 2: need for current work.
Research goal  Identification of the goal of the present research, whether expressed as
research question, aim, objective, purpose, focus, or problem definition.
Section  Portion of a research article designated by a section number (usually) and title,
often equivalent to the divisions in IMRD.
Significant  In science, observations that are unlikely to occur by chance, as determined
by statistical testing, thus likely indicative of a causal relationship.
Statistics  Use of mathematical inference to generalize results from a sample to a larger
population. Common in science but not in engineering.
Style  See extended style, condensed style.
Support  Evidence and reasoning used to show the veracity of a claim.
Index

7Cs of Change 182–183, 268 component 32–35, 50, 117–119; analysis 43,
65, 70; markers 43; see also comparisons,
abstract 266–269; thesis 132 conclusion, data patterns, framework,
academic community 35, 41–42, 44, 83, 90, 95, importance, interpretations, need, research
118, 166–167 details, research goal, testing methods
academic oral presentations 158–159 conciseness 203–204, 212, 268, 269
academic poster presentation 159–161 conclusion (component 10) 78–80
acknowledgements 271 condensed style see style
active see voice conference see academic oral presentations,
ambiguity 232–234 academic poster presentations, conference
anecdotes: area knowledge 45; connections 148, papers
151; teaching 175, 244, 274; writing, graduate conference papers 7, 120, 130–131, 158, 282
128, 157; writing, undergraduate 123, 179, 231 connection 141, 161, 192, 207–215; past research
argument 29, 179; clarifying 201, 206, 277; 207–210; present research 210–215; research
exemplar articles 5–7; order 73, 105, 107, community 41, 83, 170
110, 153, 171; non-empirical 22, 23; research connotation 90–91, 194, 216–218
article 43; structure 29–31, 117, 122, 163, 170; consistency 230–232, 258
support 83, 96, 148 correctness 234; see also articles, misused words,
argument structure outline see outline prepositions, proofreading
article see exemplar articles, research articles correlation 219–220
articles 239–242 CV see resumes and CVs

background see related work Darwin, C. 29

Bellow, S. 182 data patterns (component 7) 71–73
book chapters 138–139 delimiters 225
British vs. American English 243–244, 252 dissertation see thesis
division 13–15, 19–22, 33–34, 95; vs. section 13,
CARS model 54n2, 55n9, 94n1 15, 24
causation 219–220
citation 46, 83–88; author vs. information email 163–164; address 166–169; information
prominent 88–90; format 262–263; 172–173; requests 170–172; see also imposition
management 197–198, 263–264; number of English as a lingua franca (ELF) 243, 245;
117–118; in various genres 124, 131, 139, 143; additional materials 207, 212; cultural
verbs 90–93 differences 167, 173, 225; difficulties 174, 234,
claims 8, 24–31, 190–191; implicit 30–32, 42, 238; translation 128
56–57, 70–71; main 47; qualifying 216–218, entrance essays 147–148
220–227; supporting 35, 170; see also equations 114
component essays 124
Clifford, W. K. 35 evidence 29–30, 35, 71, 96, 110; location 73,
coherence 201–203 119; from past research 83–84; from present
collaboration 245–246 research 96–97; from research details 97–98;
comparisons (component 8) 73–76 from statistics 99
296 Index

examples 114 lab reports 122–124

exemplar articles 6–8, 24–25, 31, 41, 266 Leacock, S. 192
experiment: definition 16 Leibniz, G. 105
experimental design 74, 100–101, 118, 129, 136 lexical bundles 55n5
experimental section 14, 16, 27n3, 191 literature review see related work
extended style see style
management reports 143
Faulkner, W. 127 metadiscourse 114n3
fitting in, standing out 3–5, 166–167, 260, 274 metaphors of writing: artist 8–9, 180, 189, 206;
framework (component 4) 49–50, 57–61; in building 41, 56, 61; community 41, 90,
abstract 268; citations 84–85, 87, 117–118; 141, 181, 226; connection 5, 105, 180, 210;
identifying 50, 106; IMRD vs. IPTC 51, 95; perseverance 122, 127, 196; preview 154, 251,
research community 41–42; in various genres 266; relationships 113, 193; staking a claim 69;
124, 129, 131, 136–138 tailor-made suit 201, 203, 207, 216
Franklin, B. 217 misused words 235–238
modals 223–225
Gaiman, N. 180 Murdoch, R. 251
genre 6–9, 19, 139, 164 Muse 182
goal, audience, criteria 119–120, 283; in email
163; in presentations 158, 161, 163 need (component 2) 45–47, 117–118
grant proposals 135–137; in various genres nomenclature 58, 85, 87
122–124, 130, 132, 135, 141–143
graphics 72–73, 98, 251–253, 259–260; pointing O’Connor, F. 187
to 101–102; in presentations 155, 159; see also outline: argument structure 189–191; topic
graphs, illustrations, tables sentence 194
graphs 257–258
paragraph 29, 43, 50, 66, 70, 202; Asian 43;
hedges 226 summary 107–109
Hemingway, E. 110, 199, 200 Pascal, B. 172
Hiaasen, C. 119 passive see voice
highlights 271 Pauli, W. 277
plagiarism 118, 128, 185–187, 268, 282
illustrations 258–259 pointing verbs 101–102
importance (component 1) 43–45, 117–118 popular writing 7, 139
imposition 164–166, 173–174 preliminaries 87
IMRaDC 14–15, 18–19, 22–23 prepositions 238–239
IMRD 12–15, 22–24; see also IPTC vs. IMRD presupposition 30, 41
industrial journal articles 131 prewriting 153, 179–180, 189–195
inference 30, 35–36, 73, 221; deductive, problem definition 25, 48, 60
inductive, abductive 30, 105; statistical 99–100 problem formulation see problem definition
interpretations (component 9) 76–78 progress reports 141–142
intertextuality 141 proofreading 244–245
IPC 18, 65 Pullman, P. 196
IPTC 15–19, 22–23
IPTC vs. IMRD 15, 18–22, 258, 266–267; qualifiers 220–223, 227
claims 30–35; first division 44, 51–52, 83–84,
86–88, 109; second division 57–58, 62, 63, reasoning 30, 32, 36, 105–107
66, 95; third division 69–71, 75–76, 79, 80, recursive 22–23, 110–113
111–112 reference letters 150–151
references see citations
Jefferson, T. 203, 204 related work 24, 32, 46, 51, 52, 55n8, 86
job application 148–149 reporting verbs see citation verbs
research articles 5, 12, 127–128; as arguments 30,
keywords 270 32; citing 83; review of 277–282; submitting
King, S. 5 120, 274; variation in 20–22, 56; writing 181,
knowledge claim 37n5 190–191
 Index 297

research community see academic community stance 118, 121n2, 228n12, 229n15
research details (component 5) 61–66, 118–119 statistical testing 64, 65, 73, 74, 99–100, 129,
research gap 45, 86, 153, 182, 217 218, 220
research goal (component 3) 47–49, 52, 118, 217; statistics 99–100, 218–220
abstract 47, 49, 268; change in 231; citations style 57, 66, 70, 107–109, 210
86; conclusion 47, 49, 78; title 274; various summary see paragraph
genres 136, 138 support 30, 35–36, 47, 83, 216, 220
research methodology 21–22 survey articles see review articles
research proposal 128–130 Swales, J. 222
resumes and CVs 149–150 SWOT analysis 144–145
review articles 7, 137–138
rewriting 179, 182–183, 200, 281; abstract 266; tables 253–256
email 163, 166, 170, 172, 173, 179; plagiarism technical presentations 161
187; presentations 153 technical reports 142–143
Roth, P. 206 tense 84, 91–93, 208; in abstract 268; with
graphics 102
Saville, D. 234 testing methods (component 6) 63–66, 71,
scaffold 19, 187, 200 118–119
science vs. engineering 20 theoretical background see preliminaries
section: in first division 47, 50–51, 59; number of thesis 132
12–15; in second division 62; in third division title 269–270
70, 71, 78; style 107; summary 66, 108–109; topic sentence 43, 49, 50–51, 66, 101, 107–108;
title 12–15, 122, 190; various genres 132, 138; see also outline
see also division
senior project 125 voice 91–93, 101–102, 268
short papers 7, 137 von Helmholtz, H. 200
significant 74, 107, 218–219, 223
squid technique 234 write as you go 180–182