Grabe Kaplan (1996)

Theory and Practice of Writing
APPLIED LINGUISTICS AND LANGUAGE STUDY
General Editor
Professor C hristopher N. Candlin, M acquarie University, Sydney
For a com plete list of books in this series see pages vii-viii
Theory and Practice of
Writing
An Applied Linguistic Perspective
William Grabe
and
Robert B. Kaplan
Routledge
Taylor &. Francis Group
LONDON AND NEW YORK
First published 1996 by Pearson Education Limited
Published 2014 by Routledge

2 Park Square, M ilton Park, Abingdon, O xon 0 X 1 4 4RN
711 T hird Avenue, New York, NY 10017, USA
Routledge is an imprint o f the Taylor & Francis Group, an informa business
Copyright © 1996, Taylor & Francis.
All rights reserved. N o 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 perm ission in writing from the publishers.
Notices
Knowledge and best practice in this held are constantly changing. As new research and
experience broaden our understanding, changes in research methods, professional
practices, or medical treatm ent may becom e necessary.
Practitioners and researchers m ust always rely on their own experience and knowledge
in evaluating and using any information, methods, com pounds, or experiments
described herein. In using such inform ation or m ethods they should be m indful of
their own safety and the safety of others, including parties for w hom they have a
professional responsibility.
T o the fullest extent of the law, neither the Publisher nor the authors, contributors, or
editors, assume any liability for any injury and/or damage to persons or property as a
m atter of products liability, negligence or otherwise, or from any use or operation of
any m ethods, products, instructions, or ideas contained in the material herein.
ISBN 13: 978-0-582-55383-5 (pbk)
British Library Cataloguing-in-Publication Data

A catalogue record for this book is
available from the British Library
Library o f Congress Cataloging-in-Publication Data

Grabe, William.
Theory and practice of writing: an applied linguistic perspective
/William Grabe and Robert B. Kaplan.
p. cm. - (Applied linguistics and language study)
Includes bibliographical references and index.
ISBN 0-582-55383-0 (ppr)
1. English language-Rhetoric-Study and teaching-Theory, etc.
2. English language-Study and teaching-Foreign speakers.
3. Academic writing-Study and teaching. 4. Applied linguistics.
I. Kaplan, Robert B. II. Title. III. Series.
PE1404.G65 1996
808’. 042’07-DC20 96-24745
CIP
Set by 8H in 10/12pt Baskerville
Contents
Series List vii

Foreword ix
General Editor’s Preface xiii
1 Issues in writing research and instruction 1

2 Textlinguistic research 36
3 Towards a model of text construction 60
4 Writing process approaches 84
5 Writing process research and recent extensions 113
6 Writing for professional purposes 147
7 Writing across cultures: contrastive rhetoric 176
8 Towards a theory of writing 202
9 From theory to practice 237
10 Teaching writing at beginning levels 266
11 Teaching writing at intermediate levels 303
12 Teaching writing at advanced levels 341
13 Responding to writing and writing assessment 377
14 Conclusions: Writing in English 422
v
vi Contents
Appendix: Seventy-five themes for xvriting instruction 427
Bibliography 431
Subject Index 477
Author Index 480

APPLIED LINGUISTICS AND LANGUAGE STUDY
GENERAL EDITOR
PROFESSOR CHRISTOPHER N. CANDLIN
Macquarie University, Sydney
Second Language Learning: T heory and Practice o f Writing:

Theoretical Foundations An A pplied Linguistic Perspective
MICHAEL SHARWOOD SMITH WILLIAM GRABE and ROBERT B. KAPLAN
Analysing Genre - Language Use Strategies in Learning and Using a

in Professional Setting Second Language
V.K. BHATIA ANDREW D. COHEN
From Testing to Assessment: Teaching an d Language C orpora

English as an International ANNE WICHMANN, STEVEN 6IGELSTONE,
Language GERRY KNOWLES and
CLIFFORD HILL and KATE PARRY (EDS) TONY MCENERY (EDS)
Phonology in English Language

Language and Development: Teachers
Teaching:
in a Changing World
An International Approach
BRIAN k e n n y and
MARTHA C. PENNINGTON
WILLIAM SAVAGE (EDS)
Interaction in the Language

Curriculum: Awareness, Autonomy A utonom y and In d ep en d en ce in
and Authenticity Language L earning
PHIL BENSON and PETER VOLLER (EDS)
LEO VAN LIER
M easuring Second Language Literacy in Society

P erform ance RUQAIYA HASAN and
T.F. MCNAMARA GEOFFREY WILLIAMS (EDS)
Page Intentionally Left Blank
Foreword
This volume presents an applied linguistics perspective on the

theory and practice of writing. An examination of writing and
writing instruction from an applied linguistics orientation, how-
ever, is a daunting task. First, a comprehensive study of writing in
first-language contexts is itself a difficult interdisciplinary under-
taking. One must consider perspectives from English, education,
linguistics, psychology, and sociology. In addition, there are
radically different issues and concerns which emerge when dis-
cussing the teaching of writing to children at elementary schools
as opposed to tertiary (undergaduate and post-graduate) stu-
dents.
An applied linguistics perspective complicates the situation fur-
ther. As an interdisciplinary field in itself, applied linguistics adds
new dimensions to the study of writing. The varieties of writing to
be accounted for, the increased complexity of purposes for learn-
ing to write, and the added cross-cultural variations created by
different groups of second-language learners all force the study of
writing into a larger framework. We have attempted to survey the
nature of writing and the teaching of writing from this larger
framework, and, in doing so, have inevitably remained faithful to
an applied linguistics inquiry. As applied linguists, we may also
have made assumptions about the nature and development of lan-
guage abilities that may be unfamiliar to some teachers and
teacher trainers. We sincerely believe that such potential obstacles
to understanding can be rectified with an introductory text in lin-
guistics.
Adopting a broad interdisciplinary orientation, the book
discusses major issues in writing research, current directions in
the teaching of writing, and the somewhat ill-defined set of
ix
x Foreword
connections between theory and practice in many instructional

contexts. As applied linguists, we intend this volume to contribute
especially to an understanding of those connections between the-
ory and practice in writing and, at the same time, to reveal the
many different and complex settings in which such connections
must take place. To the extent that these connections are clari-
fied, attendant problems raised, and directions for resolution
suggested, the volume will have achieved its primary goal. Of
course, it would be unreasonable to think that any volume could
be fully successful in achieving this goal. The undertaking of this
volume has certainly revealed to us the limitations under which we
have laboured, particularly with respect to the demands of an
applied linguistics framework.
This volume is expected by its publisher and by its audience to
have genuinely international interest; that is, the discussion is
expected to contain information that transcends national bound-
aries and has universal applicability. We believe that in the
discussion of theory we have succeeded in providing notions that
do indeed transcend the local - that do reflect the cutting edge of
research everywhere. However, to the extent that we have dealt
with practice, we have had to imagine an idealized classroom,
some specific context of teacher/student interaction, and some
real characteristics of teachers and students who enter into such
an interaction. It is, we believe, impossible to create these ideal-
ized situations without starting from some specific context.
We confess to some bias in creating the framework within which
our theoretical discussions occur. We, as authors, both work pri-
marily in the USA and, as a consequence, are more familiar with
US-based publications and US-based scholars. We primarily attend
conferences that are convened in the USA; we exploit the
resources of libraries in the USA; we communicate largely with
other scholars who share our geographic (and philosophical)
constraints, and - because we work in the USA - we have concen-
trated our individual and collective research on the solutions of
applied linguistic problems which derive from, and apply to, con-
ditions in that country.
This is not to suggest that our experience is entirely myopic. We
have both, at various times and for various intervals, worked in
other environments in other geographical areas and, together, we
have had at least some experience in approximately a score of
Foreword xi
educational systems. But as this exposure is certainly not universal

and covers a great many years, our knowledge of other educa-
tional systems is often dated.
In compiling this volume we have tried to incorporate a wider
international perspective, drawing upon our collective experience
of other educational systems, and have looked carefully at the
applied linguistics research deriving from other literature that was
available to us in international journals and books. On occasion,
we have been randomly assisted by colleagues in other countries
who have made more local research materials available, and the
extent to which we have succeeded in drawing together such inter-
national, or perhaps more accurately multinational, resources, will
determine the broad applicability of this volume and will reflect
perspectives from at least those parts of the world in which
applied linguistic research is done on the teaching and learning
of English.
If we are to any extent guilty of an academic myopia, we apolo-
gize to our colleagues and readers. We trust that readers who are
interested in any larger issues that have been omitted will bear
with us, as we believe that we have covered the various facets of
most of our research paradigms fairly and fairly exhaustively.
Some of the issues may appear to have pertinence only in the US
context, but the linguistic and cultural diversity of that country is
so great that notions deriving from that base will certainly have
wider implications.
With respect to the matter of practice, many of our ideas and
illustrations have admittedly been drawn from a US context, but
many of the suggestions we offer will be of use to teachers in any
geographic location. Nevertheless, we cannot pretend that all of
the suggestions have universal applicability, nor can we pretend
that, given our conceptual constraints, we have exhausted the pos-
sibilities for practice in all possible contexts. We hope that the
readers of this volume understand that we are not offering a per-
fect set of pedagogical practices - or that our notions of things to
do exhaust the possible or the useful - but rather we hope that
this volume will be regarded as an ideas supermarket in which
readers are welcome to shop.
In closing this foreword, we would like to thank a num ber of
individuals who have been willing to provide us with feedback on
various parts of the volume - Angela Barker, Jena Burgess, Ulla
xii Foreword
Connor, Tom Miller and Fredricka Stoller. We would also like to

thank Chris Candlin and Longman for their guidance and
patience as we managed to work through a project that extended
us far beyond out initial expectations.
April 1996
William Grabe
Northern Arizona University
Robert B. Kaplan
University of Southern California
This volume is respectfully dedicated to

Kathe Rositzke and Audrey Kaplan
General Editor’s Preface
‘Good soup needs no salt’ strikes one as an apt maxim for General
Editors to bear in mind with any volume such as this that seeks to
offer an extensive account of a central topic of applied linguistics.
The scope of this latest contribution to the Applied Linguistics &
Language Sudy Series is necessarily so extensive not only because of
the obvious pervasiveness of writing but because of the signal sig-
nificance of its plural roles; as a means of measuring cognitive
abilities, as a central skill area in the design of educational cur-
ricula and in the patterns of their delivery in teaching and
assessment, as a way of understanding some of the occupational
and social demands of daily communication in living and working,
in revealing its key gatekeeping role in enabling or disabling the
sheer accessibility of life chances for all in contemporary indus-
trial and post industrial societies, and its ethnographic significance
in exploring and explaining cultural variation and relativities. And
yet, despite the extensiveness of the treatm ent here, there is room
and a place to highlight why this book is needed now. The reason
lies partly in the need to highlight and unpack these ever-differen-
tiating roles and multiple purposes, and, more than that, even in
the face of apparently major distinctions of first and second lan-
guage performances, in so doing to reassert the essential nature
of writing and the writing process as overarching, despite that vari-
ation and that fragmentation.
Appropriately enough, then, the authors begin with an explor-
ation of the nature of writing, its authorships, their individual
ande collective purposes, and the rationales that motivate these
writing processes and their varied output texts. They argue that
xiv General Editor ’s Preface
the facets of writing sketched above cannot at one level be treated

individually, nor can one facet be asserted as necessarily prime:
their discussion of composing, for example, is at once cognitive,
social, personal and educational. Nor, in their view, can writing be
appropriately addressed without reference to its histories. Not
only the genesis of writing per se, (though that is interesting in itself
in just such a complex way) but in the varied histories of writing
development in different socio-cultural and socio-political con-
texts - the relative conditions of production and reception of
writing at different times and in different circumstances - in
short, the socially and culturally embedded and contested mul-
tiple literacies of writing. Nor again, as they demonstrate, can
writing now be usefully discussed as a domain entirely distinct
from that of speech. Research on text structure, into the social
practices of literacy and oracy in different communities, into the
interconnectedness of speech and writing in public discourse, into
the intertextuality and interdiscursivity of many if not most con-
temporary genres, all point to a needecl multidimensionality in
the study and practice of writing. ..
From this point, then, Bill Grabe and Bob Kaplan embark, and
provide in their first introductory Chapter a cogent and defining
account of writing research in English in first and second lan-
guage contexts, drawing on the facets of the writing process
identified above, and in the context of varied responses to a plu-
rality of writing needs. Such an account permits them a second
overlay of interdisciplinarity - that of the methodologies of writing
research. Here they show how some asserted distinctions between
LI and L2 writing, already problematic cognitively and also
socially in many contexts, similarly blend and become fuzzy in
research terms as experimental, linguistic/textual, ethnographic
and social psychological techniques form aspects and tools of a
common repertoire.
We should, however, be properly cautious in celebrating this
unity in diversity. Seen from the point of view of the history of
writing practice and writing pedagogy, especially in English, there
is now and has been at once much more overt sectarianism and at
the same time masked differences of purpose, than one might
assume. ‘Genre-based’, ‘process-focused’, ‘whole language’,
‘product-orientated’ may be less descriptions of academic or
educational value than they are manipulated slogans for the
credulous, yet they do signify importantly distinguishable
General Editor’s Preface xv
positions, if as I say, often exploited. Approaches to textual

description are not uncontested, nor are their ‘applications’ to
instructional materials. Teaching strategies ought not to be simply
read off from some linguistic analysis, however successful a blend
of the formal and functional. Students’ writing practices should
not be assessable simply as solo performances in terms of a belief
that some targetted cognitive strategy can mutate into acceptable
linguistic form. Nor in the context of writing development seen as
a not too covert recruitm ent into' some disciplinary or profes-
sional specialism, should such individual writing performances
merely be seen as momentary personal and individual responses
to some given instructional task. The value of this book is that
while acknowledging this variability, it is still bold enough to assert
and document the core issues that writing research and writing
practice has addressed and still must address.
It does so in thorough detail in its following chapters. Begin-
ning appropriately enough with the text as product, it locates it,
equally convincingly, as a tangible, analysable artefact deriving
from two related inferrable processes of discourse: the discourse
of the writer and the discourse of the reader. Such discourses are
at once cognitive and socially conditioned, personal and func-
tional, to a degree accessible through experimental, ethnographic
and social psychological accounting, but always to be seen as
mutually influencing through the potentially explanatory evi-
dence of the text. Writing and reading processes are thus
interconnected in the text. It is this position which permits the
authors to warrant their other characteristic close interconnection
in this book, that of research and practice. Indeed, one might see
writing and reading as their own research laboratories, their own
research methodologies, their own research agendas. After all,
what is a writer (or a reader for that matter) doing except to test
out theories about the realisation of some mental discursive
process within the general and actual constraints of the personally
available formal system of the lexicogrammar?
Nonetheless, despite this mutuality, the relationship between
theory and practice in writing research and instruction has not
been self-evident. In part, as the authors argue, this has been a
consequence of a more general separation of powers between the
academy and the school (let alone the home or the workplace,
one might add, in respect of both). In part this has been due to a
lack of clarity within each institution about the range of writing
skills, knowledge bases and writing processes and how these may be
xvi General editor s preface
integrated in a descriptive model of language processing in a man-

ner such as that which Grabe and Kaplan outline in their keystone
Chapter 8 ‘Towards a theory of writing’. In part this has been due
to an understandable if perverse unwillingness to grasp fully in
either teaching or research practice the verity of writing: that inter-
dependence of writer, reader, text and topic I focus on above.
In the sections of this notable chapter, who writes what, to
whom, for what purpose, why, when, where and how is, of course,
at one level simply to transpose and extend Fishman’s famous dic-
tum about sociolinguistics. At another, it defines the questions
and conditions for research and for practice. You could build a
curriculum on addressing those questions, and as the authors
show in their subsequent chapters, some educators have done so,
though few, if any, with all the questions in the air at the same
time. The conditions, as Brecht wrote, were not right. In a similar
way, you could construct a research agenda from those questions,
but here, disappointingly and with less excuse, the need for a con-
current consideration of these mutually influencing questions has
also been less evident. One might speculate why. My view is that
the particularity of the research response derives from the particu-
larity of applied linguistics training. A focus on text without a
focus as well on discourse, a focus on the individual rather than as
well on the cultural or institutional collective, a focus on the
actual rather than as well on the historically conditioned, a focus
on the cognitive in the absence of a focus as well on the social, a
focus on the instruction rather than as well on the learning, a
focus on the linguistic rather than as well as on the subject-matter,
all these particularities work to create the conditions which make
such an interdisciplinarity of the questions impossible to conceive
for researchers, and thus to turn into relevant and authenticated
research practice.
Seen in this way, this book about the theory and practice of writ-
ing turns out to have a wider implication. Read this way, it
becomes a book about the state of the art of applied linguistics,
and how the problem-focused nature of the discipline, identified
by the authors at the outset, is shown to be relevant not just for its
‘applications’ but for the discipline of applied linguistics itself.
Professor Christopher N Candlin

General Editor
Macquarie University,
Sydney, Australia
1
Issues in writing research and

instruction
1.1 Introduction
Applied linguistics has concerned itself with the development of

writing skills for at least the past 50 years, and that it has done so is
entirely appropriate. If one is to take seriously the relatively
straightforward definition of applied linguistics as the attempt to
resolve real-world language-based problems, then the develop-
m ent of writing abilities, whether for learners of English as a first
language (LI), or as a second language (L2), or learners of any
other language, surely falls well within the domain of applied lin-
guistics.
There are, however, significant differences between the two
groups of learners, since there are wide variations in learner issues
within each of these major groups. These differences and their
consequences for writing theory and instruction will be explored
throughout the book. The treatment of both groups within a
single volume is, however, the only logical applied linguistics per-
spective to adopt since both groups subsume learning and
instructional problems which are language-based, and there is sig-
nificant overlap in their historical evolution over the past 20 years.
The decision to ignore English first-language research and prac-
tice would not only lead to a badly distorted view of L2 writing
approaches, it would also misinterpret the true scope of applied
linguistics inquiry with respect to issues in writing development.
This chapter outlines many of the larger issues and problems
implicit in the theory and practice of writing instruction, adopting
a broad applied linguistics perspective. In exploring issues in writ-
ing, basic assumptions about the nature of writing must t>e
considered. For example, why do people write? That is to say, what
1
2 Theory and Practice of Writing
different sorts of writing are done by which different groups of people,

and for what different purposes? More fundamentally, one must ask:
What constitutes writing? Such basic questions cannot be discussed
in a vacuum but must also consider the larger issues raised by liter-
acy skills development and literacy demands in various contexts.
Literacy, incorporating specific writing issues with a related set of
reading issues, highlights the necessary connections between
reading and writing as complementary com prehension/produc-
tion processes. It also introduces the distinctions between spoken
and written language forms, and the specific constraints of the
written medium. Thus, a brief overview of literacy provides an
im portant background for understanding the recent develop-
ments in writing theory and instruction.
Any discussion of basic foundations must necessarily incorpo-
rate outlines of research on both writing in English as the first
language and writing in a second language. From the LI perspec-
tive, many theoretical issues and concerns in LI contexts also
affect writing approaches in L2 situations. From the L2 perspec-
tive, research in L2 writing also highlights differences between the
two contexts. The many additional variables introduced in L2 con-
texts - not only cognitive but also social, cultural and educational
- make considerations of writing in a second language substan-
tially different in certain respects.
This chapter also explores the gap between research and
instruction and considers how that gap may be bridged. The trans-
lation from theory to practice in LI writing contexts has changed
considerably over the past 20 years. Such a translation has also had
a profound impact on L2 applications from theory to practice;
some of these applications have been appropriate, while others
have been much less appropriate, given the distinct L2 context.
For example, L2 instruction may:
■ place writing demands on EFL students, and for some of them,
English may not be perceived as a very im portant subject;
■ place distinct writing demands on English for Special Purposes
(ESP) students, or on English for Occupational Purposes
(EOP) students - demands which may be very different from
those on English for Academic Purposes (HAP) students plan-
ning to enter English medium universities;
■ include writing demands on adult literacy and immigrant survi-
val English students - both groups experiencing very different
demands from those which occur in academic contexts;
Issues in writing research and instruction 3
■ include academic writing demands in which a sophisticated

level of writing is not a critical concern.
All of these issues form parts of an overview of writing theory and
practice from an applied linguistics perspective.
1.2 On the nature o f writing
The need for writing in m odern literate societies - societies

marked by pervasive print media - is much more extensive than is
generally realized. When one examines the everyday world, one
finds people engaged in many varieties of writing, some of which
may be overlooked as being routine, or commonplace, or unim-
portant. These varieties, however, all represent the ability to
control the written medium of language to some extent. It is fair
to say that most people, on a typical day, practice some forms of
writing. And virtually everyone in every walk of life completes an
enormous num ber of forms. In addition, many people write for
reasons unrelated to their work: letters, diaries, messages, shop-
ping lists, budgets, etc.
Describing the various tasks performed every day by writers
offers one way of classifying what people write, but a slightly more
abstract taxonomy of writing types will prove more descriptively
useful. A list of actual writing tasks does not provide a way to group
these tasks according to similar function - a goal in understanding
what gets written and why. In fact, many different functional sorts
of writing constitute common occurrences. These sorts of writing,
depending on the context, task, and audience, may be classified
functionally in numerous ways, including writing to identify, to
communicate, to call to action, to remember, to satisfy require-
ments, to introspect, or to create, either in terms of recombining
existing information or in terms of aesthetic form. Thus:
■ writing down one’s name identifies
■ writing a shopping list may identify, communicate, a n d /o r
remind
■ writing a memo may communicate and remind
■ writing a student essay may at least satisfy a requirem ent
■ writing a diary may promote introspection
■ writing a professional article may communicate, recombine,
and allow introspection
■ writing a novel or a poem may exemplify what is known as aes-

thetic creativity.
At yet another development level, one may distinguish writing
which involves composing from writing which does not; this dis-
tinction is useful because most of what is referred to academically
as writing assumes composing. Composing involves the combining
of structural sentence units into a more-or-less unique, cohesive
and coherent larger structure (as opposed to lists, forms, etc.). A
piece of writing which implicates composing contains surface fea-
tures which connect the discourse and an underlying logic of
organization which is more than simply the sum of the meanings
of the individual sentences. Figure 1.1 illustrates the compos-
in g / non-composing dichotomy in terms of audience. The matrix
suggests the possible options that are available for writing with or
without composing.
Composing, further, may be divided into writing which is, in
essence, telling or retelling and writing which is transforming.
Retelling signifies the sort of writing that is, to a large extent,
already known to the author, such as narratives and descriptions.
Audience Writing without Writing with

composing composing
For knowledge For knowledge

telling transforming
Self Shopping list Personal diary ‘J o u rn al’ notes
O ne known Note to milkman Personal letter

other
O ne unknown Business letter
other
Small group Lesson plan
known sermon
Small group Questionnaire Newsletter item Proposal
unknown
Large group Tax form Poem, dram a
Driver’s licence novel
application form Short story
Figure 1.1 Patterns of com posing with differing audiences

Issues in uniting research and instruction 5
The planning involves recalling and reiterating. Transforming, on

the other hand, signifies that sort of writing for which no blue-
print is readily available. The planning involves the complex
juxtaposition of many pieces of information as well as the weigh-
ing of various rhetorical options and constraints (Bereiter and
Scardamalia 1987). In this type of writing, the author is not cer-
tain of the final product; on the contrary, the writing act
constitutes a heuristic through which an information-transfer
problem is solved both for the author and for his or her intended
audience. This notion of composing is much more comprehensive
than the idea of drafting or ‘shaping at the point of utterance’
(Britton 1983), since it takes in the ‘final’ product. Many sorts
of what traditionally have been labelled expository and argu-
mentative/persuasive texts, as well as ‘creative’ writing, involve
transforming. In Figure 1.1, an attempt has been made to distin-
guish between retelling and transforming, even though both
organizing strategies are available for many sorts of composing.
In most academic settings where students are learning to write,
the educational system assumes that students will learn to com-
pose with the ability to transform information. In fact, many
students learning to write before they enter the tertiary level have
little consistent exposure to writing demands beyond retelling. In
some cases, students, both in LI and L2, have minimal practice
even with simple retelling. The problems created by these
students as they enter the academic environment certainly
deserve the attention of applied linguists. Moreover, writing places
constraints on student learning that are distinct from the develop-
m ent of spoken language abilities.
To understand these developmental constraints on students,
and the more complex demands made by academic institutions, it
will be useful to examine briefly the historical development of
writing and the changing writing/literacy expectations which have
arisen over the last two centuries.
Writing is a rather recent invention, historically speaking.
Unlike spoken language - coterminous with the history of the spe-
cies - written language has a documented history of little more
than 6000 years. And while it is generally accepted by linguists that
certain aspects of spoken language may be biologically deter-
mined, the same cannot be said of writing. While all normally
developing people learn to speak a first language, perhaps half of
the world’s current population does not know how to read or
write to a functionally adequate level, and one-fifth of the world’s

population is totally non-literate. It seems a bit absurd to suggest
that this difference is accidental, due to the inaccessibility of writ-
ing instruments or material to read. Nor does it seem appropriate
to label this one-fifth of humankind as somehow ‘abnorm al’.
The distinction between spoken and written media calls atten-
tion to a significant constraint on the development of writing
abilities. Writing abilities are not naturally acquired; they must be
culturally (rather than biologically) transmitted in every genera-
tion, whether in schools or in other assisting environments. While
there are many distinctions between the two media in terms of lex-
ical and structural use, the acquired/learned distinction deserves
particular attention. The logical conclusion to draw from this dis-
tinction is that writing is a technology, a set of skills which must be
practised and learned through experience. Defining writing in
this way helps to explain why writing of the more complex sorts
causes great problems for students; the skills required do not
come naturally, but rather are gained through conscious effort
and much practice. It is also very likely, for this reason, that num -
bers of students may never develop the more sophisticated
composing skills which transform information into new texts.
The crucial notion is not that writing subsumes a set body of
techniques to master, as might be claimed, for example, in learn-
ing to swim; rather, the crucial notion is that writing is not a
natural ability that automatically accompanies maturation
(Liberman and Liberman 1990). Writing - particularly the more
complex composing skill valued in the academy - involves train-
ing, instruction, practice, experience, and purpose. Saying that
writing is a technology implies only that the way people learn to
write is essentially different from the way they learn to speak, and
there is no guarantee that any person will read or write without
some assistance.
1.3 Literacy and writing
The history of literacy development supports such a writing-as-

technology perspective for the nature of writing. Indeed, a
num ber of literacy scholars have argued for this view strongly; any
other definition of literacy does not stand up to the historical evid-
ence (Goody 1987, Graff 1987). Since there are, in fact, many
types of literacy which have developed historically under very dif-

ferent contexts and for very different uses, any more complex
definition tends not to hold up equally well in all contexts (cf.
Cressy 1980, Graff 1987, Houston 1988, Purves 1991). Moreover,
the history of literacy demonstrates that reading and writing skills
were developed and passed on to following generations only in
response to cultural and social contexts; these skills were not
maintained when appropriate social and cultural supports were
removed.
In fact, the definition of writing-as-technology fits extremely
well with historical perspectives on literacy because many literacy
movements and developments were little more than the wider dis-
semination of very basic skills, such as the ability to write one’s
name or fill out a ledger or a form. Such literacy developments
hardly count as the sort of writing-as-composing discussed earlier
or as the type that is valued academically; yet, these literacy devel-
opments undeniably reflect aspects of writing abilities.
The history of literacy development is both enlightening and
commonly misunderstood. It is enlightening because the many
and varied literacy movements and contexts of literacy develop-
m ent provide a better understanding of the current use of, and
expectations for, students’ writing abilities; it is often misunder-
stood because many assumptions about literacy have been widely
promoted and accepted without careful documentation and analy-
sis (cf. Graff 1987). One significant point that the study of literacy
has demonstrated, both synchronically and diachronically, is that
there are many different sorts of literacy skills just as there are
many different sorts of writing abilities. Most students who display
writing problems in educational contexts do, in fact, have writing
skills; they are just not the skills which educational institutions
value (Barton and Ivanic 1991, Street 1993). This is particularly
true for L2 students in EAP contexts; they clearly come to higher
academic institutions with many different literacy practices and
many different views on the purposes of reading and writing.
The history of Western literacy begins (from c. 3100 b c ) with
the early uses of writing for recording events, traditions, and trans-
actions by scribal specialists who were able to translate orally for
the masses as was necessary. The powers of priesthood can be
attributed to the apparently mystical properties associated with
the ability to read and write. The rise of the Greek city-states sig-
nalled a greater dissemination of literacy skills among the
populace. However, literacy among the classical Greek citizenry

was less widespread and less sophisticated than has commonly
been assumed. Perhaps only 15-20 per cent of the Greek popula-
tion consisted of ‘citizenry’, and oral traditions were still the
trusted and preferred means of communicating. Similarly, the
Roman period was marked by a limited literacy among the popu-
lace; in this context, literacy was due in large part to the rise of
public schooling, the need for civil servants to do government
business in the far-flung regions of the empire, and the rise of
commercial literacy needs. Nevertheless, it was a very small intel-
lectual, political, and religious élite that possessed literacy skills
(cf. Goody 1987, Graff 1987).
The decline of Rome saw the role of literacy relegated primarily
to the religious infrastructure which emerged across Europe from
the fourth to the eighth century. While many schooling traditions
of the Roman era persisted throughout Western Europe, most
were taken over by the church for the training of priests, clerics,
and other functionaries. The development of literacy for other
than religious uses began during the eighth to tenth centuries;
and contrary to popular belief, the ‘Dark Ages’ was not a period of
complete illiteracy beyond the monastery walls. The tenth and
eleventh centuries marked the beginnings of a commercial liter-
acy and set the stage for literacy practices across Europe from the
twelfth to the fifteenth century. It is evident from the research of
such scholars as Harvey Graff that literacy has a continuous his-
tory in Europe from the Greek city-states to the present. What
must also be understood is that literacy was still restricted, tied to
church, state, or economic necessity, and to particular practices in
particular contexts. The mass literacy to which we are accustomed
simply did not exist and, during most of this time, the ability to
compose was extremely limited.
As noted above, religious institutions played a critical role in the
history of Western literacy. It was not until the evolution of
Protestantism in the sixteenth century that popular literacy
became necessary. Unlike the Roman Catholic church, Protestant
theology took the view that personal salvation could be achieved
through direct access to the biblical gospels. Protestant sects have
contributed importantly to world literacy through missionary
activities which supported the translation of the Bible into hun-
dreds of non-Indo-European languages as well as the teaching of
literacy in the languages of the missionaries (e.g. English, French,
German, Spanish) for purposes of access to the Bible in European

languages (e.g. the work of the Summer Institute of Linguistics
and the Wycliff Bible Translators).
The first well-defined popular literacy movement in Europe
may arguably be traced to the English revolution and the rise of
Oliver Cromwell or to Martin Luther and the Protestant reforma-
tion. (The earlier development of Hangul in Korea should not be
overlooked, though it is unlikely that literacy beyond the aristoc-
racy was a serious objective.) The clearest early case of a successful
mass literacy movement, however, is attributable to the Swedish
movement of the seventeenth century to require of all Swedish cit-
izens the ability to read the Bible (Arnove and Graff 1987, Graff
1987). This movement was remarkably successful since reading
the Bible was a prerequisite for religious confirmations, and con-
firmation in the church was a requirem ent for marriage. For the
first time, women as well as men were trained in literacy skills.
This literacy campaign, however, as in much of the previous his-
tory of literacy, concentrated almost exclusively on reading, and in
particular on reading the Bible. The role of writing in common
literacy development can only be seen as an innovation of the last
200 or so years. Even rudim entary writing skills (beyond signing
and recording) among the populace were unknown until very
recently.1
In the m odern historical era, the rise of popular literacy -
including the uses of writing for secular purposes beyond govern-
ment and business - emerged in the late eighteenth century,
primarily in England, France and the USA. As traced by Cook-
Gumperz (1986; see also Resnick and Resnick 1977), the rise of
m odern literacy (including writing skills) can be seen as occurring
in three stages. Roughly, the period between 1750 and 1850 can
be considered to mark the rise of common literacy (not school-
directed or taught). In the mid-nineteenth century, parts of conti-
nental Europe, the UK, and the USA supported the rise of
schooled literacy along with the beginnings of compulsory educa-
tion. Literacy then became a means whereby the ruling classes
could train the larger populace to be more efficient workers - a
tool in support of the status quo, a gateway for upward mobility
for a chosen few, and a means to control the spread of informa-
tion to the larger society. It is also the case that the quantifiable
level of skills needed to be considered literate were subject to defi-
nition by the élite. As the pressure on the élite increased, in part
through the spread of literacy, the élite redefined the degree of

literacy necessary for entry - thus, a constantly increasing literacy
demand evolved.
In the twentieth century, literacy has undergone at least one
profound transformation. The growth of positivism and the belief
that scientific progress would resolve all problems led to the use of
psychometric testing and evaluation procedures in European and
American educational systems. This pervasive influence on educa-
tional systems led to gradually increasing expectations of literacy
uses in society and of literacy expectations in academic training. It
also served to compartmentalize and stratify students at many
levels of skills and abilities; finally, it introduced to education the
notion of failure and remediation. Students could be classified as
deficient in particular skills and the notion that literacy could be a
problem rather than an opportunity took hold.
After the Second World War, the increasing demands for profes-
sional uses of literacy concomitantly increased demands on
reading and writing training. Professional literacy now is the goal
of educational systems throughout much of the literate world. The
rise of professional literacy has also led to an increasing demand
on the sort of writing ability that involved composing (Kaestle et
al. 1991, Purves 1991, Stedman and Kaestle 1987). It is also
through the growth of compulsory schooling during the twentieth
century that writing has come to be emphasized both in school
and in white-collar employment. It remains to be seen what
impact computers will eventually have on the conception of liter-
acy. While there have been a num ber of extravagant assertions
that the computer will transform literacy, and the ‘word’, it may be
quite some time before such a transformation is realized. The
Greeks took centuries to make the alphabetic system a tool of
wider literacy for its citizenry (Graff 1987), and the printing press
(a technological innovation of enormous importance in the
increasing dissemination of texts) as an agency of change took
centuries to have a major impact on mass literacy (Eisenstein
1979, 1985). It may therefore be generations before the capabil-
ities of the computer cause significant transformation in literacy
practices.
The last two centuries in the history of Western literacy are par-
alleled by changes in the demands and expectations in students’
writing abilities. Most school rhetorics in the eighteenth and nine-
teenth centuries amounted to rudim entary guides to usage and
etiquette. At the tertiary level, however, writing instruction and

composition became im portant aspects of a university education
as part of rhetorical studies, coinciding with the influential rheto-
rics of Campbell (1776), Blair (1783), and Priestley (1777). As one
outcome of this trend, Freshman Composition courses were
initiated at Harvard in 1874 (Berlin 1984); instruction in fresh-
man composition in the USA has never looked back (Berlin 1984,
1987, Crowley 1990, 1995). Why this trend should take hold so
strongly in US but not European institutions is a question which
deserves study. Whatever the motivations, they led to a rapid insti-
tutional growth in composition instruction in the USA which was
not paralleled elsewhere. And while the recent history of rhetoric
and composition is long and complex, the still accepted ‘current
traditional’ approach to writing instruction has at least a 100-year
tradition going back to the inception of freshman composition
courses in the USA. Such approaches are recognizable through
the emphasis on handbooks, the use of model texts, and theme
writing in the various modes of discourse (which, even until 20
years ago, was uniformly popular).
The current academic interest in literacy and writing is readily
traceable to the recent historical evolution of professional literacy.
Over the past 15-20 years, educators have complained of a ‘crisis
in the schools’ because of the observed decline in national test
scores in the USA (and elsewhere in the English-speaking world -
e.g. Australia, Canada, New Zealand - as well as in new nations
which rely on English as a national or official language - e.g.
Hong Kong). In the USA, arguments have been advanced by poli-
ticians, educators, and parents that literacy standards have been
declining at the time when political and economical skills are in
greatest need. These concerns, of more importance to administra-
tors and politicians, do not seem well-founded in the light of a
sober history of literacy, nor do they reflect the realities of an ever
increasing school population. As more lower socio-economic class
and minority students were added to the population taking these
tests, it should have been expected that a temporary decline
would occur (Coe 1986). It is only symptomatic of the effect of
larger numbers of people aspiring to higher education and social
mobility; it is not an inevitable sign of cultural decay or of educa-
tional literacy failure (cf. Bracey 1995, Stedman 1994).
The current popular concern over literacy has generated a
num ber of theories concerning its nature and its consequences on
both the individual and society. The consequences of literacy on

society are well attested and quite profound, as literacy is a socially
motivated phenomenon; its purposes, uses, and values are deter-
mined by the society in which such literacy skills are practised.
Literacy is not a universally uniform set of skills, but rather a set of
highly contextualized skills, determ ined by the society in which it
functions (Street 1984, 1993). An opposing view focuses more
directly on the cognitive influence of literacy on the individual:
the theory, commonly referred to as the ‘Great Divide’ or the
‘Great Leap’ theory of individual cognitive development, claims
great individual cognitive changes as a consequence of the acquisi-
tion of literacy (Goody 1977, Olson 1977; cf. Gee 1986, Goody
1987, Olson 1994). In brief, this theory holds that learning to read
and write alters the cognitive thought processes of the individual;
literacy transforms the way in which the mind functions.
Since the mid-1960s, various scholars have argued for this
‘Great Divide’ position, primarily on logical rather than historical
grounds. The seminal article triggering this view of literacy and its
consequences for the individual was published by Goody and Watt
(‘The consequences of literacy’) in 1963. This article argued for
the ‘Great Divide’ theory as a result of the Greek adoption of the
alphabetic principle. The alphabetic principle led to mass literacy
among the Greek citizenry - a change in Greek society, it was
argued, which subsequently led to the (asserted) inventive genius
of classical Greek civilization. Goody and Watt argued more gener-
ally that the rise of literacy led to a num ber of dramatic changes
(both social and individual):
■ the domination of history over myth
■ the distinction between the natural and the supernatural
■ the ability to store and access greater amounts of knowledge
■ the creation of abstract logical deduction
■ the ability to analyse language itself as an object
■ the awareness of the individual as distinct from the group and
consequently the need for private introspection
■ the rise of critical scepticism towards previously reified knowl-
edge and beliefs
■ the rise of democratic institutions.
This long list of consequences reads remarkably like a set of edu-
cational goals in m odern Western school systems. Thus, it is easy
to see how literacy has come to be equated with the essence of aca-
demie education. The ‘Great Divide’ view has received support

from a variety of sources over the last 30 years. Scholars such as
Havelock (1976), Olson (1977), Ong (1982), and Vygotsky (1962,
1978) have raised arguments of a similar nature with respect to
the acquisition of literacy.
In the last 15 years, a num ber of strong refutations of this
theory of individual cognitive differences have emerged. Even
earlier, Gough (1968) compared the rise of literacy in Greece with
similar historical evolutions in India and China. Her research indi-
cated that different societies do not necessarily follow the same
paths of developments as a result of literacy dissemination among
the population. In particular, the notions of history-as-concept,
deductive logic, and democratic institutions as consequences of
literacy were not supported. Her research led later scholars to
qualify early claims for individual differences since she demon-
strated that societal contexts play a large role in the uses and
consequences of literacy.
The line of reasoning that literacy is a social phenom enon (in
service to the social milieu) rather than a universally consistent
cognitive phenom enon was strongly supported by the research of
Scribner and Cole (1981). In a series of studies on the acquisition
of literacy among the Vai people of Liberia, they showed that liter-
acy alone did not account for any generalized cognitive differences
among the subjects; rather, they found that specific skill differ-
ences among subjects were a consequence of specific skills and
tasks emphasized by each type of literacy (Vai, Arabic, or English).
Schooled education rather than literacy appeared to predict the
sorts of individual cognitive skills - such as abstract deductive rea-
soning, drawing logical inference, and connecting unrelated
information - which are assumed to result from learning to read
and write. Rather than literacy being the cause of the cognitive
skills valued by academic institutions, it would seem that the skills
are inculcated by the institutions, which then attribute to these
skills the idea that they represent basic cognitive development.
The cognitive impact that may be attributed to literacy is more
probably the result of the gradual extension of specific literacy-
related skills learned in particular situations (cf. Goody 1987,
Olson 1991, Scholes 1993) which, themselves, are socially embed-
ded. Further support for the socially embedded nature of literacy
skills has appeared in a range of educational research (e.g. Boggs
1985, Cook-Gumperz 1986, Heath 1983, 1986b).
This current research, in combination with the historical

research of Graff, demonstrates that literacy skills and literacy
acquisition do not constitute a single unitary cognitive process
across cultures and social groups; rather, there are different sorts
of literacy skills which develop to serve the needs of each
social/cultural group. School children, when they participate in
schools, are typically not literate or illiterate; they bring the prac-
tices of their community or social group or family to the
classroom (cf. Bernstein 1972a, 1990, Halliday 1978, Hasan 1989,
Heath 1983, Wells 1986). Problems occur when the sorts of liter-
acy practices of certain students do not conform to the expected
literacy practices of the school community. The central issue in
literacy development is not the development of uniform cogni-
tive skills, but the recognition that there are many different
literacy practices, of which only a few are likely to be valued by a
given educational system. At the same time, it is likely that the
availability of the technology of literacy modifies the way in which
a society behaves; no specific theory of cognitive change is neces-
sary to support such a notion (cf. Goody 1987, Olson 1991,
1994).
Akin to this history of literacy, the study of writing - its history,
uses, purposes, and consequences - must be seen as socially con-
texted. There is not a unique set of writing skills which universally
defines all of writing, apart from particular social contexts. There
are many different ways to practise writing, and an educational
system typically values one particular set of writing practices, and
not every student - particularly not every immigrant student - will
come into the educational context having practised that set of
writing skills valued by the educational structure.2 This does not
necessarily mean that such students ‘don’t know how to write’ or
that they are cognitively deficient; but only that they are not well
versed in the practices of writing which are valued institutionally.
Following this line of reasoning, it is not enough simply to teach
the valued practices of the academic discourse community; it is
also necessary for all members of the institution to understand
that students may come with sets of writing practices that have
been valued and purposeful in other contexts. Such an awareness
should change the way in which the teaching of writing and the
writing practices of students are approached.3
The long history of literacy, and the current debates which
beset it, provide critical insights about the nature of writing devel-
opment, and provide lessons which should be heeded if the

simple erroneous assumptions of earlier literacy research are not
to be perpetuated.
1.4 Oral and written language
The study of literacy and its relation to writing shows that oral lan-
guage and written language are not contraries, nor are they exact
reflections of each other. The history of literacy has demonstrated
that oral and written language coexist in many complex patterns
of use. Even when literacy became widespread in European soci-
eties, most reading was performed aloud as a social activity; the
written form was closely linked with oral presentation. Silent
extensive reading as a deviation from oral interaction is only a rel-
atively recent practice. Present-day uses of language often go
hand-in-hand with various oral practices such as lecturing,
recounting, and debating. Moreover, certain oral practices can
either reinforce or obstruct the writing practices that students are
expected to master. Recent educational research, for example,
demonstrates that methods of oral interaction have a strong influ-
ence on the later development of reading and writing skills
(Bloome and Green 1992, Cook-Gumperz 1986, Purves 1991,
Wells and Chang-Wells 1992).
Throughout the history of m odern linguistics, most linguists
have taken the position that oral language is primary and written
language is simply a reflection of oral language (cf. Biber 1988,
Halliday 1989). Most education researchers have taken the oppo-
site position - i.e. the written language is a true representation of
the correct forms of language and should be valued and practised.
The past ten years have seen these two positions coalesce in soci-
olinguistic research on the ethnography of education and on
register variation in different types of oral and written texts. This
research has tried to come to terms with the conflicting assump-
tions of two major disciplines involved in language and language
teaching. Research by Biber (1988, 1994), Chafe (1982, 1985),
Halliday (1989), Kress (1989), and Tannen (1982, 1985) has
pointed out the various ways in which oral and written language
vary and overlap. Findings from this research have greatly
extended the knowledge of properties of both media.
The work of these scholars indicates that patterns of variation
across oral and written language texts differ on a num ber of

dimensions: structural and organizational differences; different
frequencies of use of various language features, different produc-
tion constraints; and different uses for the varieties of oral and
written language texts. Chafe (1985) and Tannen (1982) in par-
ticular have argued that any understanding of oral and written
text variation cannot be determined along one dimension of vari-
ation; rather, they have claimed, texts vary and overlap with other
texts along a num ber of functional dimensions. From this perspec-
tive, it is not enough to consider texts as simply oral or written;
texts vary in ways that cannot be captured by such a single contin-
uum.
In a recent series of studies, Biber (1988, 1992, 1994, 1995) has
clarified the complex set of relationships that hold among various
types of oral and written texts. In his research, Biber has
developed a multidimensional model of textual variation that
compares relations among 23 different genres of spoken and writ-
ten texts. By incorporating a large num ber of linguistic features, a
large num ber of texts, and a sophisticated research design, Biber
(1988) sought to overcome the difficulties of much of the earlier
research on spoken and written textual variation which often led
to confusing if not contradictory findings. Text features were ana-
lysed by an automated counting program. The resulting frequency
counts were then used in factor analyses to determine the co-
occurrence patterns among the linguistic features for all of the
texts in the corpus (approximately one million words). His factor
analysis model identified six interpretable dimensions; that is to
say, the linguistic features co-occurred together on six different
factors, each of which could be interpreted as a communicative/
functional dimension of textual variation. The communicative
interpretations were based on the grouping of linguistic features
which have been seen as having compatible discourse functions in
previous discourse analysis literature.
The results of his research indicated that three of his six dimen-
sions are related in some way to orate-literate variation (involved
versus informational production, explicit versus situation-depen-
dent reference, and abstract versus non-abstract information),
though none of the six dimensions interpreted provided an abso-
lute distinction between oral and written genres; that is, no
dimension clearly distinguished all types of written texts from all
types of spoken texts. Thus, the traditional dichotomy between
oral and written texts that is so often assumed does not appear to
be represented by any single dimension of textual variation in a
strict interpretation of his results. Biber’s results also provide
strong evidence for the multidimensional nature of textual struc-
ture. No small set of feature counts and no single notion of a
communicative dimension will offer a satisfactory interpretation
of textual variation or textual structure.
A major conclusion to be drawn is that the spoken-written con-
tinuum does not exist in any strict sense as a single dimension of
textual comparison. The implications of this line of research for
the study of writing and composition is that all texts are complex
multidimensional structures, including texts written by students;
claims made in writing research about distinctions between oral
and written language, as well as oral features in student composi-
tion, are, for the most part, greatly oversimplified, requiring
caution with respect to many of the assertions based on these
premises.
Many of the issues addressed in the comparison of oral and writ-
ten language and in the discussion of literacy point to basic
assumptions about the nature of written language which should
be clarified so that the interpretations of current research on writ-
ing are given an appropriate context. It has been argued to this
point that:
(1) the study of writing is an appropriate domain for applied lin-
guistic inquiry (including language-based problems with
English LI and L2 students);
(2) writing is a technology insofar as it is a culturally transmitted
set of practices;
(3) writing involves many different uses and functions, not all of
which are valued academically;
(4) the study of literacy demonstrates that writing should be
viewed as a set of practices which are socially contextualized -
academic writing is simply one valued set of practices appro-
priate to that context - rather than as a single universal set of
cognitive skills;
(5) academically valued writing requires composing skills which
transform information or transform the language itself;
(6) the uses of oral and written language interact and reinforce
each other as sets of practices that serve social functions;
(7) research on spoken and written language demonstrates that
all texts are highly complex multidimensional structures, and

the actual linguistic-functional nature of many of these
dimensions is just beginning to be explored;
(8) oral and written language variation points out that different
types of written texts vary greatly according to task, audience,
and purpose;
(9) assertions about sweeping differences between oral and writ-
ten language made by composition theorists must be
regarded with some scepticism based on recent sociolinguis-
tic research.
These and other related assumptions guide the following discus-
sion of research in writing, both in LI and L2 contexts.
1.5 Research on writing in English LI contexts
Generally speaking, writing research in LI contexts has been con-

ducted along four distinct but interacting dimensions which can
be represented, though only somewhat loosely, by four respective
disciplines: education, psychology, linguistics, and rhetoric/com -
position. While there are a num ber of analyses of writing research
which divide the territory somewhat differently (cf. Hillocks 1986,
North 1987, Phelps 1988, Witte 1992), this particular synthesis is
not m eant to be definitive but only to provide an organizing guide
to the somewhat confusing array of extant research. The first of
these research strands is the study of literacy development or the
acquisition of writing, particularly at the early stages. Researchers
in education, applied sociolinguistics, and educational psychology
explore ways in which children learn to write, as well as why some
students experience difficulties in learning to write.
Educationists and educational psychologists have been primarily
concerned with the socio-educational contexts for learning to
write, the need to express meaning in writing, the need for
students to view writing as a purposeful activity, and the various
stages of learning through which young writers are nurtured.
Much of this research has centred on case studies of individual
children, learning experience approaches (LEA) and whole-lan-
guage approaches (Calkins 1986, Dyson 1989, 1993, Graves 1983,
1984). Many of the practical approaches are also in line with
Vygotskean perspectives on literacy development (Moll 1990,
Tharp and Gallimore 1988, Vygotsky 1983, Wertsch 1985, 1991). In

addition, current research in Australia addresses the role of lan-
guage form and genre knowledge as critical aspects of learning to
write from early grades (Christie 1992, Cope and Kalantzis 1993).
The research is also paralleled by the ethnographic research in
educational contexts undertaken by sociolinguists. The ethnogra-
phy of education explores ways in which children from divergent
social and linguistic backgrounds learn in educational contexts
that are designed for mainstream students. Such research has
examined different groups of high-risk students and the differ-
ences between school expectations and home environment (Boggs
1985, Heath 1986b, Philips 1983, Poole 1991, Snow et al 1991), the
impact of conversational style and interactional discourse on learn-
ing (Bloome and Green 1992, Cazden 1988, Cook-Gumperz 1986,
Wells and Chang-Wells 1992), and the attitudes of students and
teachers in the classroom context (Williams 1976). In particular,
researchers have found that exposure to literacy events, attitudes
towards school literacy, and the teaching of meaningful literacy
tasks are important conditions for writing development.
A second strand of research involves study of the cognitive
aspects of writing. Researchers in cognitive psychology, education,
and composition represent those interested in modelling and
explaining the mental processes used in the act of writing. Much
of the research is empirical in nature. While earlier cognitive
research on writing principally studied the written products and
used experimental research on text recall (Britton and Black
1985b, van Dijk and Kintsch 1983), more recent research on cog-
nitive processes has included studies of the actual process of
writing on-task (that is, as it is occurring in real time), as well as
protocol analyses, and task intervention (e.g. Bereiter and
Scardamalia 1987, Flower 1994, Flower et al 1990, Hillocks 1986,
Kellogg 1994, Smagorinsky 1994).
O ther approaches involve the use of retrospective case studies
and observational research as ways to examine qualitatively the
nature of the composing processes and their development.
Im portant results from this research include the findings that writ-
ers constantly shift among pre-writing, writing, and revising tasks.
Writing is not a linear process; instead, it involves the complex
combination of content information, rhetorical demands, and
reader interpretation. Good writers and poor writers also appear
to make use of processing skills in different ways.
A third strand of research is the study of the text itself, or text

construction. This research typically falls under the heading of
‘text linguistics’ or ‘discourse analysis’. Researchers in cognitive
psychology, linguistics, applied linguistics, and rhetoric/com posi-
tion are interested in how texts are constructed and organized in
ways which allow appropriate reader interpretation. Central topics
for this research field are the study of contributing lexico-gram-
matical structures, cohesion, coherence, inferences-making
processes, and text modelling (Coulthard 1994, Halliday and
Hasan 1989, Hoey 1991, Mann and Thompson 1992, Singer 1990,
van Dijk and Kintsch 1983).
Research on these topics is both quantitative and qualitative.
Much work relating to main-idea comprehension and recall, and
comparing good and poor writers, involves the counting of sur-
face features, propositional units, or inference chains together
with appropriate statistical analyses of the resulting data (Beck et
al. 1991, Cox et al. 1990, 1991, Speigel and Fitzgerald 1991). O ther
research exploring the coherence of texts has focused on macro-
structures (or text themes), logical relations among clauses and
text units, and information structuring in texts (given-new infor-
mation, topic-comment, them e-rhem e, focus-presupposition
structures; Singer 1990, vande Kopple 1986, 1990). All three sub-
sets of coherence notions have proved im portant for research on
writing as well as for writing instruction.
The fourth and final strand of research on writing is the rhetori-
cal study of writing. Researchers from rhetoric, composition,
applied linguistics, and literary criticism examine variation in writ-
ing skill or writing interpretation as a function of writing purpose,
topic, genre, audience, intertextuality, and the social construction
of writing, as well as the larger sociopolitical, historical, and edu-
cational contexts. Rhetorical research traces its origins back to the
five canons of Aristotle: invention, arrangement, style, memory,
and delivery (Corbett 1971, H orner 1983). Modern rhetorical
research, after a long history of declines and emergences, com-
prises three branches of inquiry: historical, philosophical, and
critical/herm eneutic research (North 1987). While many rhetori-
cians may object to this classification and indeed combine
research in the three areas, this tripartite scheme provides a useful
way to encompass rhetorical inquiry.
Current research on both classical and nineteenth-twentieth-
century rhetorical history, rhetorical criticism of (literary) texts,
and social and political contexts of intertextuality represent

im portant concerns. The revival of rhetoric in the 1970s and
1980s in US universities has led to a much greater influence of
rhetoric on composition studies and, in general, on the teaching
of composition in educational institutions, particularly at second-
ary and tertiary levels. Perhaps the most enduring influence of the
m odern rhetorical revival will be the increasing emphasis on dis-
course communities and the role of social construction in writing,
both of which are having a significant impact on theories of writ-
ing and writing instruction, particularly in academic and
professional contexts. Writing instruction across the curriculum as
well as content-based writing instruction have also been influ-
enced by these rhetorical developments.
In general, research on writing in LI contexts has greatly
changed writing instruction at almost all levels - from elementary
to post-university professional contexts. While it is true that con-
servative instruction (the term used in the literature is
‘current-traditional’) still exists and dominates in some insti-
tutions, the ongoing research on writing in all four research
strands has changed writing instruction for the better. These
research strands have also had a strong impact on writing assess-
m ent and evaluation, though the demands of testing theory and
the institutional requirements for grading impose conservative
constraints on changes in assessment.
A new area in writing research which also stands to benefit from
research is computer-assisted writing instruction, with computers
being used both as a research tool and as an instructional resource
(Bangert-Drowns 1993, Cochran-Smith 1991, Pennington 1993,
Snyder 1993). At present, most computer-assisted instruction has
simply transferred the more traditional writing instruction to a new
medium. This particular field will make significant contributions
when it learns to use the capacities of the computer in innovative
ways beyond current conceptions of its utility for writing (Bruce et
al 1993, Bruce and Rubin 1993, Selfe and Hilligoss 1994, Tuman
1992)
This review of LI writing research is dominated by research in
the American educational context, but the work of Britton, Carter,
and Wells in England, of Bereiter, Freedman and Jones in Canada,
and of Christie, Halliday, Hasan, and Martin in Australia all point
to important changes in writing theory and instruction. Overall,
however, the contributions of these researchers have not yet greatly
altered the picture of writing research as it has developed in the

USA. Perhaps the fact that the USA is the only country with uni-
formly compulsory freshman composition requirements at the
tertiary level helps to explain differences in focus.
The visibility of US LI writing research has also not been
matched in research on second language writing theory and
instruction. Reasons for this difference are readily explicable:
1. The US composition research community is insulated from
much of the rest of world in terms of other students’ writing
development, writing contexts, and writing needs in educa-
tional institutions.
2. The ethnocentric attitudes among US composition researchers
towards English as the only language worthy of study meant
that, for many years, immigrant students and international
students were not given any consideration with respect to their
unique backgrounds and needs. There was an unspoken
assumption that the research on LI writers would apply equally
to these second language learners.
3. The ‘crisis’ in writing instruction in the USA has allowed
researchers to focus their interests on secondary and tertiary
LI student writing problems to the exclusion of other learners.
4. The large amounts of grant funding, the lucrative textbook
markets, and the likelihood of professional advancement
biased research in the USA, until very recently, towards main-
stream composition research.
5. The discipline most open to the concerns and problems of L2
learners - i.e. applied linguistics - emerged from linguistics
and foreign language departments rather than from English
departments. This situation greatly reduced interchange
between applied linguists and composition researchers (many
with literary studies backgrounds) as well as between applied
linguists and language arts education researchers.
6. The concerns and problems encountered by the various L2
learners needing to improve their writing abilities has only
recently received the attention it deserves. This concern is
primarily due to the growing awareness of the needs of lan-
guage minority populations within the USA rather than a
concern for English uses in international contexts.
7. Research on writing for L2 students has been pursued more
vigorously in other English-speaking countries than in the
USA.
All of these factors have contributed to a much later emergence of

research on L2 writing abilities and writing instruction, particu-
larly in the USA. This network of causes is reflective, as well, of
social attitudes in some multilingual nations. Obviously, while the
seven points raised above are true and do explain the failure of L2
research to emerge earlier in a more powerful form, they evade
the major issue that such distinctions as ESL/EFL or as L1/L2
inevitably break down. The educational systems of such countries
as Australia and the USA must deal with all students in their multi-
linguality. Socially, there should be continuity, rather than
polarity, which tends to create second-class citizenship for some
group (s) of children.
This later development of L2 writing research in the USA has
not been paralleled in the UK or Australia. Perhaps the greater
awareness of immigrant and international student needs, coupled
with a greater awareness of English language needs in other coun-
tries, created more interest in writing in a second language. At the
same time, however, recent work in the USA in L2 writing
research is again emerging as significant, due to the large and
ever-increasing numbers of international students entering US
education - particularly tertiary education - and to efforts to sup-
port language minority students’ learning needs in elementary
and secondary schools. Increased federal funding to support
research on language minority student problems has also assisted
this change in US educational research on writing.
The emergence of interest in the USA in the writing problems
and needs of L2 learners has led to an awareness that these
students cannot be judged according to the same research
assumptions guiding LI writing instruction. While there are simi-
larities between LI and L2 students of writing, there are also
differences which must be recognized if research implications and
practical instruction are to be most appropriate for the needs of
L2 learners (Leki 1992, Silva 1993).
1.6 Second language student needs and writing

instruction
Any discussion of L2 students’ writing needs must first take into

account the wide diversity among L2 learners as distinct groups
each with its own uses of, and needs for, writing. The broadest dis-
tinction that can be drawn is a two-part division between English

as a Foreign Language (EFL) and English as a Second Language
(ESL) learners. More reasonably, the two lie at opposite ends of a
single continuum but are treated separately for non-linguistic
(administrative/budgetary) reasons. EFL courses include those
students who need to learn English (or in this case need to learn
to write in English), who live in countries in which English is not
regularly spoken or written as a language of the community (e.g.
the People’s Republic of China, Indonesia, France). ESL courses
include those students needing to learn English who live in coun-
tries where English is a language, or the language, of the community.
This division between ESL and EFL includes more-specific varia-
tions: students in countries where English shares a role as an
official language (e.g. Singapore, India, Nigeria), students in
countries where English is an im portant educational language
though not recognized officially (e.g. Malaysia, Malta), students in
countries where English is a major language of the community
(the USA, the UK, Canada, Australia, New Zealand). Students in
these latter countries may be subdivided further into students who
reside permanently and students who will remain only for the
duration of their academic a n d /o r professional training. Students
who reside permanently in an English dom inant community may
be divided into those groups which seek to learn English from an
early age, typically in the public school system, and those groups
who are adults needing writing skills for survival literacy or job
training (English for Occupational Purposes [EOP]). Students in
the public school systems may be further divided into students
who have recently arrived, or whose parents have recently arrived,
in the community, as opposed to those students who are indige-
nous to the community (in the USA: Native Americans, Puerto
Ricans, Cubans, Indochinese, New Mexico Spanish speakers,
Cajun speakers, New England French speakers). All of these
students share a common designation in US public schools as
Limited English Proficiency (LEP) students. Additional issues in
writing instruction also depend on whether these student are
studying at the elementary or secondary levels.
This wide array of L2 student groups requiring writing profi-
ciency to varying degrees creates a set of research and
instructional issues that are not fully responded to by research
focusing on LI students of writing. The various L2 groups will dif-
fer markedly in terms of the need for writing abilities. Students in
EFL contexts will need English writing skills ranging from simple
paragraph writing and summary skills to the ability to write essays
and professional articles (depending on students’ educational
levels, academic majors, and institutional demands). Even within
an EFL country, it is not possible to generalize beyond local edu-
cational institutional expectations. Kachru (1985, 1992) has
looked at the uses of English as an alternative language for speak-
ers of various other languages in South Asia and Africa and has
shown that English may serve literary and social functions even
where it is not a majority language. Such variation in these coun-
tries makes textbooks and appropriate advice on writing
instruction difficult to provide.
In ESL contexts the range of writing needs is equally diverse,
although the needs will, for the most part, be more academically
oriented. Survival literacy and low-level occupational needs for
writing are typically handled not by writing specialists but by basic
adult education teachers, and they may not stress writing as com-
posing.
The reticulated structure created by different student groups
and different student writing needs is only at the margin of the set
of factors which make L2 writing theory and practice so complex.
Any appropriate instruction must take into account the influence
from students’ various LI life and cultural experiences. The need
for this information is amply illustrated by the recent research on
the ethnography of education and the sociolinguistics of literacy.
Students from different backgrounds approach educational tasks
in ways at variance with the approach adopted by students from
the mainstream culture. It has further been found that educa-
tional instruction that has been modified so as to be more
compatible with, or simply to account for, students’ LI socializa-
tion practices leads students to higher literacy rates in the
classroom (Au and Jordan 1981, Heath 1986b, 1993, Tharp and
Gallimore 1988).
In addition to the LI experiences of these students, researchers
and practitioners need to consider the level and types of language
interference that may influence learning as well as students’ edu-
cational experiences in, and attitudes towards, their LI. Problems
related to language transfer and interlanguage development are
well known to applied linguists and are central to second language
acquisition research. The problems created by this phenom enon
for L2 writing instruction are less well defined overall, though a
major thrust of research in this area is the study of contrastive

rhetoric. On the level of the social contexts of language use, seri-
ous if typically unexplored issues include:
■ Careful study of students’ educational attainment in their LI
educational system (and what it means to have reached differ-
ent levels of education in various school systems - in other
words, the ways in which literacy is defined in their LI commu-
nities) .
■ Study of students’ reading motivations and abilities in their first
language.
■ Study of students’ writing motivations and abilities in their first
language.
■ Study of students’ attitudes towards their first language and
literacy in that language.
■ Study of students’ attitudes towards English and English literacy.
In addition, processing issues beyond language interference and
language transfer include such issues as the impact of learning
through two languages; i.e. how processing information in one
language affects vocabulary development in a second language.
This issue also influences curriculum design choices in situations
in which two languages are provided in the curriculum, and the
various ways in which each of the two languages might be
accounted for instructionally:
■ Bilingual transition programmes
■ Bilingual maintenance programmes
■ L2 immersion programmes
■ L2 submersion
■ SL pull-out classes
■ Some other option.
Thus, there are a num ber of factors relating to the theory and
practice of writing instruction in a second language which go well
beyond the concerns and issues central to writing practices in LI
contexts. Given an imperfect understanding of many of these fac-
tors, it is not surprising that L2 writing research has generally not
advanced beyond the practices of LI writing research, even if the
actual instructional practices are more informed about the L2 stu-
dent than LI research might suggest. It should be clear from this
review of L2 students and factors affecting their learning that a
wholesale borrowing of research and practice from LI contexts
will not be entirely appropriate. At the same time, L2 researchers

studying writing have examined and adapted many of the research
projects and instructional practices first investigated with LI
students. In the following overview of writing research on L2
students, it is im portant not only to be aware that parallels with LI
research exist, but also to recognize the very different learning cir-
cumstances of many L2 students.
1.7 Research on writing in a second language
Research in L2 writing began in the late 1960s and early 1970s in

the USA and the UK, accompanying the increasing numbers of
international L2 students coming to tertiary-level academic insti-
tutions and an increasing awareness of domestic second language
students who are failing in the public educational systems. Writing
in second language contexts in countries where English is not
dom inant has received relatively little research treatment. This
may be more a case of research neglect than of any lack of educa-
tional difficulties to address.
The pattern of research on writing in a second language does
not follow exactly the pattern of research in LI writing. In the L2
research, applied linguistics has been the academic discipline giv-
ing most attention to the writing needs and problems of students.
Applied linguists have drawn on the work of cognitive psycholo-
gists and linguists on the one hand to study the organization of
discourse and text construction processes, and on the work of
sociolinguists and ethnomethodologists on the other to study the
social contexts in which students learn to write. The research
effort in L2 writing is quite recent. Apart from work in contrastive
rhetoric and ESP, relatively little research on writing was under-
taken before 1980. Most professional articles which appeared
prior to 1980 centred on techniques for teaching writing rather
than on the nature of writing in various contexts.
In addition to applied linguistics, research on L2 writing has
emerged recently in education and in composition studies. In
education, the concern for language minority students and other
high-risk language arts students in elementary education has led
to greater interest in and awareness of writing in the public school
in the USA and Australia. The education research in the USA has
drawn extensively on Goodman’s whole language philosophy and
on ethnographic classroom research for ways to develop students’

writing abilities (Edelsky et al 1991, Freeman and Freeman 1992,
Goldman and Trueba 1987, Hudelson 1989a, 1989b). Research
arising from composition studies has followed recently from LI
composition research, employing similar composition research
approaches on groups of L2 students (Kroll 1990, Johnson and
Roen 1989, Raimes 1991, Reid 1993). Many L2 researchers,
although they may specialize in child writing development or com-
position studies, consider themselves as belonging to the applied
linguistics community as much, if not more, than to the related
discipline in which they are working.
Writing research in applied linguistics first centred around the
study of rhetorical variation noticed in the writing of L2 students
from various LI origins (Kaplan 1966, 1972). This line of
research, commonly referred to as contrastive rhetoric, drew on dis-
course analysis and textlinguistic research to explore how student
writing could be analysed at the discourse level as a way to under-
stand the varying patterns of organizational preferences in
student writing. Work still continues and has contributed to a
more general expansion of discourse analysis studies on L2 stu-
dent writing. Since the appearance of Kaplan’s original work, a
num ber of applied linguists have drawn on research in cognitive
psychology, text linguistics, and discourse analysis to explore ways
in which L2 texts are constructed, how these texts differ from
texts constructed by LI students, and how and why these texts dif-
fer among L2 students representing different linguistic
communities. Recent research along these lines includes the study
of cohesion and coherence, the macrostructure of texts, the
arrangem ent of information, using propositional analysis, dis-
course bloc analysis, inferencing and schema theory, story
structure, the comparative study of text corpora, and the textual
co-occurrence patterns of lexico-syntactic features (cf. Chapter 2).
One of the larger subspecializations to have grown out of the
developments of discourse analysis and text linguistics in applied
linguistics is the study of the special writing needs of various
groups of students. The best known of these research directions is
the study of the writing needs of English for Special Purposes
(ESP) students. Following the early research in contrastive rheto-
ric, this field was the most active area of writing research through
the 1970s, and it has continued to be popular to the present time.
O ther special-purpose groups who need writing skills of a special
sort include future doctors, lawyers, civil service workers, etc.

(Hutchinson and Waters 1987, Johns and Dudley-Evans 1991,
Maher and Rokosz 1992, Swales 1990, Trimble 1985; see the
extended discussion in Chapter 6).
Research in the L2 writing of students in the public education
sector owes its beginnings to the concerns of informed language-
arts and bilingual-education teachers that language minority
students were facing serious difficulties developing expected liter-
acy skills. Drawing on many of the findings of whole language
research, L2 specialists have explored the applications of many
new language-arts approaches to writing instruction for children
(Enright and McCloskey 1988, Faltis 1992, Peregoy and Boyle
1993). Results of this recent research are now emerging and
should affect considerably the literacy instruction of younger chil-
dren. The needs of L2 students in secondary education have yet,
however, to emerge as an area of L2 writing research (cf. Wald
1987). One can only hope that researchers will turn to this area
and study the writing problems of L2 adolescents (e.g. Montano-
Harmon 1991).
Research on writing in a second language in contexts other
than the USA, Australia, Canada, and the UK is minimal. It is clear
that different countries will have different needs and expectations
for students learning to write in English. It is not clear what
research has been carried out in various countries to establish
these needs and expectations, or to explore the successful instruc-
tional implementation of these expectations (see Kaplan 1995).
Certainly countries such as Japan, Hong Kong, India, Malaysia,
etc., which send many students to the USA and the UK for tertiary
education, give some consideration to the writing needs of these
students. In some cases, the needs are being addressed by the
development of US university programmes in-country which claim
to provide the equivalent of the first two years of tertiary-level aca-
demic credit (Japan, Malaysia). It remains to be seen how
successful these programmes will prove to be.
The current emphasis in English L2 writing research strongly
follows English LI writing research. There are, however, certain
criticisms which question the apparently extensive adaptation of
LI research into L2 contexts that is so prevalent today. It is
important, in this regard, to distinguish research methods from
interpretations of research results. It is intuitively reasonable that
research methods useful in LI contexts should also be applicable
to L2 contexts. In this respect, distinctions in LI and L2 writing

research have become much fuzzier in the past ten years; e.g. both
LI and L2 writing research includes studies employing experi-
mental treatments, text analyses, protocol analyses, ethnographic
observations, etc. This situation does not, as we argue, mean that
L2 writing research is purely derivative; indeed, it addresses some
very different questions about some very different students.
1.8 Moving from theory to practice
While research on writing in LI and L2 contexts has developed

remarkably over the last 20 years, the teaching of writing is only
now beginning to reap the benefits of this research. In the 1950s
and 1960s, most teachers of writing, whether for LI or L2
students, felt that writing instruction had a well-established and
effective methodology. This methodology centred around what is
now term ed a ‘current traditional’ approach. As described by
Britton (1983: 2):
T here had arisen that very lim ited tradition o f com position teaching
which is so familiar to all of us in first and second-language situa-
tions, with its em phasis on ‘correct usage, correct gramm ar, and
correct spelling,’ and its focus on the topic sentence, the various
m ethods of developing the paragraph ... and the holy trinity of
unity, coherence, and em phasis ... R ichard Young (1978) calls this
tradition ‘current-traditional rh eto ric’ and he argues th at its
practitioners’ em phases and pedagogical techniques were all d eter-
m ined by certain tacit b u t shared assum ptions concerning the
n ature of the com posing process. C hief am ong these was the
Rom antic conviction that the creative aspects of the process are mys-
terious, inscrutable, and hence unteachable. W hat can be taught
and discussed are the lesser m atters of style, organization, and usage.
Similar accounts of writing practices are presented in Berlin

(1984, 1987), acknowledging a long twentieth-century tradition of
writing instruction. The parallel tradition in L2 contexts, to the
extent that it was emphasized, was reinforced by a feeling that L2
students needed to focus on correct grammar, spelling, and usage
to the point at which they were then ready for the same ‘current
traditional’ writing instruction that LI students were given
(Raimes 1983a, 1991). The mid-1960s constituted the turning
point in writing instruction. Dissatisfaction with (a) the slow devel-
Issues in nrriting research and instruction 31
opm ent of students’ writing abilities, and (b) the increasing num -
bers of international and domestic L2 minority students in
academic institutions, led to new theories of writing instruction
based on more successful teaching practices. Many of these new
approaches to writing instruction were less informed by as yet
undeveloped or underdeveloped theoretical research than by
‘techniques that worked’ (North 1987).
Raimes (1983a) noted that L2 writing instruction was typically
uninformed by writing theory through most of the 1970s, provid-
ing, for the most part, ‘proven’ techniques for teaching writing. As
Raimes (1983a: 261-2) states:
... we have stressed the ESL p art o f ESL com position at the expense
of the com position part, an d we have d one so because we have
thought th at students n eed m astery over the sentence before p ro-
ceeding to the paragraph, and m astery over the paragraph before
proceeding to the essay. So we have provided controls and limits
which make the task easier for us ... Many o f o ur students ... cry out
for rules, for som ething concrete to m o nitor their writing perfo r-
m ance with. So we give them gram m atical Band-Aids and doses of
paragraph models. We m ust th en realize th at we are teaching edit-
ing and imitation. We’re n o t teaching com posing.
Raimes goes on to assert the importance of grammar for ESL writ-

ers but not at the expense of composing practice.
The emerging theories of writing of the late 1970s, drawing on
cognitive psychology, sociolinguistics, educational ethnography,
whole-language education, and applied linguistics, began to
spread so that there were powerful competing theories of writing
by the 1980s. These theories have finally begun to change writing
instruction. At the same time, it must be said that most writing
instruction is still oriented around ‘current traditional’
approaches. Many ESL classrooms still rely on writing textbooks
which stress paragraph models, grammar and usage rules, and
vocabulary development as their entire curriculum. The large
majority of these textbooks, while advertised as process oriented,
do not typically emphasize purposeful writing activities. ’Current
traditional’ approaches are also alive and well in LI writing
instruction, particularly in secondary and university level classes.
In many elementary classes, students still do little writing of any
kind beyond occasional paragraphs, short-answer seat work, and
taking objective question examinations (Applebee 1981).
Despite this state of affairs - in which many teachers appear to
be immune to, or unaware of, current writing research - writing

instruction is improving in a num ber of contexts. In elementary
schooling, whole-language approaches to literacy are having an
increasing effect on the amount of writing done and the methods
by which writing is taught. While the evangelical overtones of
whole-language advocates can be disturbing, more teachers are
adapting certain whole-language teaching practices in their own
classroom routines. Many teachers now stress early composing
practice for children, encouraging students to express themselves
in writing that is both meaningful and purposeful. In these con-
texts, the translation from theory to practice is just beginning to
take hold in a large num ber of classrooms. Its applications to the
teaching of language minority children appears to be as successful
as it is in the teaching of LI elementary school children (Dyson
1993, Faltis 1992, Peregoy and Boyle 1993).
Research on the composing process is now having a strong
influence on the teaching of writing in US tertiary institutions and
pre-tertiary programmes. More composition curricula are adopt-
ing instructional approaches which emphasize relevant topics,
class discussion, brainstorming (invention), multiple drafting/
revision, peer feedback, and realistic tasks. Many writing textbooks
have been rewritten to stress a ‘process approach’, though how
successful such textbooks are in developing students’ writing prac-
tices is still an open question. Writing process research has also
often been combined with Freirean notions of liberatory literacy
practices and with the philosophy of stressing meaningful commu-
nication to focus more attention on the content of writing
courses. Gradually, composition programmes have been moving
towards content-centredness, maintaining a consistent and coher-
ent theme or set of themes about which students have personal
knowledge and in which they can invest their time meaningfully
(Bartholomae and Petrosky 1986). In L2 contexts, similar efforts
are underway in some places with content-centred courses and
sheltered classes (Brinton et al. 1989, Crandall 1993, Krueger and
Ryan 1993, Short 1994). Writing across the curriculum move-
ments - a major issue in public education in the UK - constitute
another manifestation of this trend; more programmes are seek-
ing to add more writing to the content courses rather than more
content to the writing courses.
Research on the nature of texts and on text construction has
also had an impact on writing practices, though many of these
changes are only now being disseminated to classrooms and

teachers. Research on the various aspects of textual organization -
such as topic progression, information structuring, discourse
mode variation, logical relations among discourse units, the role
of inferencing, and patterns of lexico-syntactic usage - has led to
increasingly productive instructional practices.
Perhaps the major obstacle to the translation from theory to
practice in the case of text-linguistic research is that there is no
one overarching linguistic perspective on the application of the
research; unlike whole-language and composing-process research,
many researchers in text linguistics come from disciplines as sep-
arate as psychology, linguistics, applied linguistics, artificial
intelligence, educational psychology, communication, and rheto-
ric/criticism. A fair amount of the research is considered basic
rather than applied, and little consistent attention has been given
to the implications for instruction (cf. Christie 1992, Halliday
1993b, Martin 1993, Miller 1995). Applied linguists, however, have
been seeking to combine this research into an overarching
research perspective which is relevant to instructional practices.
Work to create an underlying synthesis of textual coherence is
leading to effective instructional practices particularly with L2
students (Cerniglia et al 1990, Grabe and Kaplan 1989, Johns
1986, Reid 1993).
The overall picture emerging from the various research
approaches and their different translations into writing practice is
that there is some disagreement as to what will be the most
appropriate curriculum for writing instruction. When the
research-influenced approaches to writing instruction are mixed
with the practices of many teachers who still espouse some form of
‘current traditional’ instruction, the resulting confusion is under-
standable. A large problem lies in the earlier (and persistent)
assumption that one standard m ethod can be developed which will
be appropriate to all students in all settings.
It is, however, possible to sort out the different instructional
practices and, in the case of the recent trends, formulate coherent
rationales that recognize and exploit the strength of each perspec-
tive, identify the different student populations most likely to
benefit from each perspective, and, most importantly, combine
those perspectives into a coherent theory of writing instruction
which bypasses many of the apparent tensions among the perspec-
tives (Leki 1992, Raimes 1991, Reid 1993). The most notable of
these tensions is the overstated distinction designated ‘process ver-

sus product’, a strawman which has been created by some
composition researchers (cf. Kaplan 1988, Raimes 1991).
On a practical level, the goal of writing research is to move
beyond simple dichotomies and gain a better understanding of at
least the following issues comprising a comprehensive theory of
writing and writing development:
■ the nature of coherent written discourse which is appropriate to
the task, topic, genre, and audience;
■ the set of processes which are called into play to generate the
discourse;
■ the social contexts and attitudes which influence the writing
(and the writer); and
■ the nature of the various learning processes that coalesce to
support the development of student writers.
From this point of view, writing is a matter of combining comple-
mentary perspectives rather than oppositions, and it is the goal of
any theory of writing to account for these various factors in one
m anner or another (see, for example, Flower 1994, Witte 1992).
In the following chapters the many issues associated with theo-
ries of writing as well as instructional practices, will be explored in
greater depth. The combination of theoretical discussions are syn-
thesized into a coherent framework for writing theory. The latter
chapters then focus on the transition from theory to practice, vari-
ous instructional approaches for different groups of students, and
issues in feedback on writing and writing assessment. The
extended discussion in the following chapters build upon the
many insights noted in this chapter and explore issues and prac-
tices that should promote more informed discussion and lead to
improved writing instruction.
Notes
1. This pre-m odern history is adm ittedly E urocentric. Literacy in Arabic,
Chinese, Japanese, and o th er languages was also lim ited to an élite or
used for special purposes (e.g. Koranic literacy). Dealing with literacy
developm ent in non-European languages is an undertaking beyond
the scope of this book, and we apologize for the adm ittedly E urocen-
tric lim itations to this history o f literacy.
2. In the USA, the term ‘im m igrant stu d en t’ is being used to include n o t
only recent im m igrants b u t also refugees, p erm a n en t resident aliens,
political asylum holders, an d o th er categories o f individuals p art o r all
of whose education has o ccurred in an o th er culture and who do n o t
speak English as a first language.
3. A rgum ents have been p u t forw ard th at teaching socially an d culturally
diverse students to ‘buy in to ’ the literacy practices o f the do m in an t
group is assimilationist an d prevents alternative voices from being
heard. T he same post-structural an d social-constructivist supports for
this argum ent, however, would also argue th at the contributions of
new individuals necessarily alter the existing social structure at the
same tim e that the social structure influences these new individuals
(e.g. Wells 1994).
2
Textlinguistic research
2.1 Introduction
Informed instruction presupposes theories of the arrays of knowl-

edge required (what one needs to know), the sets of skills which
may be brought to bear upon successful learning (what one needs
to be able to do), and an account of the social contexts influenc-
ing instruction (the factors influencing what one needs to be able
to do). Writing instruction, as of yet, lacks clear consensus in all
three of these contexts. What is more, much current research in
writing presupposes these issues without actually coming to terms
with them.
A more successful counterpart, reading instruction, has, in the
past 15 years, become the focal point of much theory construction,
and, more recently, the locus of attempts to translate theory into
practice. Reasons for this development are not hard to find. In
large part, reading has always been strongly related to the cognitive
activities of comprehension and understanding, a central concern
of cognitive psychology and artificial intelligence. As such, research
on comprehension, understanding, mental representations, mem-
ory, information retrieval, and learning have all tended to devise
experimental research with reading as the language task involved
(Adams 1989, Oakhill and Garnham 1988, Rayner and Pollatsek
1989, Rieben and Perfetti 1991, Stanovich 1992).
The result of such activity in reading research has provided
both theorists and practitioners in reading instruction with mod-
els of the fluent reading process, with factors which influence the
process, and with extensive studies of individual differences in
reading. Following the wider acceptance of more recent interac-
tive models of reading, the implications for instructional design,
36
Textlinguistic research 37
and the motivation for practical instructional concerns (i.e. how

to teach reading) have increased substantially. Reading instruc-
tion has lost some of its mystery and such slogans as ‘reading is
learned by reading’ have been recognized as oversimplifications,
even if partly correct, as the result of insightful theory construc-
tion and research in theoretically grounded instruction.
2.2 Writing research needs
The analogous situation for writing demonstrates much less evi-

dence of progress, despite the equally well-articulated concern
about the difficulties students have in learning to write. There is,
in fact, a fair amount of research on the writing process, as well as
on successful writing instruction. However, writing has yet to enjoy
the same input from cognitive psychologists generally, who, per-
haps justifiably, feel that research on production models first
requires a strong comprehension model. While many rhetoricians
and composition researchers are developing theories of writing,
and simultaneously exploring the implications of such theories for
instruction, little consensus has emerged which would allow for
significant advances in writing instruction parallel to the achieve-
ments in reading instruction.
The range of research on writing theory and practice is diverse,
reflecting a wide range of theoretical interests as well as student
constituencies. What emerges from this diversity is a field of
research and instruction based on individual preference, subjec-
tive study, tradition, and, for practical purposes, on notions of
pragmatic issues - i.e. of what seems to work. North (1987) comes
to a similar conclusion from the perspective of composition
instruction.
More complex models of writing are needed - ones that not
only appeal to both process- and product-oriented perspectives on
writing, but also account for audience and social context con-
cerns. In the past, too much discussion has been given to debates
over process versus product. It should be clear from analogous
interactive reading models that a study of the text product alone
will not lead to the kinds of models of fluent writing necessary to
support improvements in writing instruction. At the same time,
theories of the writing process do not, of themselves, form a com-
prehensive interpretation of the written text.
As a way to break free from past attempts to formulate theories

of writing, it is necessary to return to basic questions. A first stage
in moving towards a theory of writing abilities, then, lies in
addressing anew questions such as the following:
1. What is a text?
2. How may texts be examined as product?
3. How may texts be analysed as process?
4. How may product and process be viewed as reflecting an inter-
active model of writing?
5. How do process and product issues embed themselves within a
coherent interpretation of wider social contexts for writing?
6. What types of texts do we want learners to produce?
7. What do these issues suggest for writing instruction in the class-
room?
What is being proposed here is a set of essential criteria for estab-
lishing a theory of writing (supported by textlinguistics, cognitive
psychology, rhetoric, and composition research). In this and the
following chapters, essential elements of such a theory will be
sketched. These issues may be seen as definitional in the sense
that any adequate model must begin with a definition of the phe-
nom enon supported by research findings. It is not suggested that
the present work constitutes a comprehensive model of the writ-
ing skill; however, responses to the questions posed above will lead
at least to a preliminary statement.
2.3 On the nature o f written text
The analysis of written text is in many respects a relatively new

research area. Text analysis has grown rapidly over the past twenty
years in quantity and variety with contributions from sociolinguis-
tics, psycholinguistics, applied linguistics, cognitive psychology,
and artificial intelligence research. It is also im portant to recog-
nize a num ber of ongoing traditions for their contributions to text
analysis.
One of the earliest of these traditions is hermeneutics
(Eagleton 1983, Ong 1982). Originally intended to serve biblical
interpretation, the discipline has grown in the last century to
involve the interpretation of all literary forms, though its focus is
on historically validated text. While of central importance to liter-
ary criticism, this tradition also has implications for text analysis
(Britton and Black 1985a, Rabin 1986). Britton and Black, for
example, note that biblical interpretation was related to Bible
translation, and translation requires the analysis of discourse,
rather than grammar, if the translation is to be comprehensible
and not merely trivial. A second tradition, the rhetorical, again
easily pre-dating m odern text research, goes back to Aristotle and
other classical Greek scholars. The rhetorical tradition is having a
greater impact on recent discourse analysis and text analysis
(Berlin 1984, 1987, Neel 1988, Phelps 1988, Young 1987). A third
tradition, possibly also pre-dating m odern linguistics, is that of
m odern literary criticism and stylistics (Comprone 1987, Crowley
1989, Eagleton 1983, Fowler 1986). While this kind of research sel-
dom generates specific methodologies for m odern discourse
analysis, it raises fundamental questions about the nature of texts
- questions which must regularly be reconsidered.
These three traditions, combined with recent developments in
linguistics, form the historical base for the rise of text analysis as a
field of study; de Beaugrande and Dressier (1981) and Tate (1987)
provide good summaries of these developments. Although text
analysis has been perceived as an im portant discipline in Europe,
the same cannot be said for text analysis in North America. The
developments in linguistic theory in the USA over the past 30
years have, for the most part, hindered the development of text
analysis. Functional linguists, some working with sociolinguists
and some from a non-generativist background, have focused con-
siderable attention on discourse analysis, both spoken and written
(see, e.g., Brown and Yule 1983, Grabe 1992). Work by Coulthard
(1994), Givon (1983), Halliday (1994), Halliday and Hasan (1976,
1989), Mann and Thompson (1988, 1992), van Dijk (1985) and
others has explored aspects of discourse structure, and from this
research has come both a better understanding of text and a set of
techniques for examining the nature of text.
The range of research influencing text analysis requires the
view that text is a multidimensional construct; that is, no unidi-
mensional analysis of text can offer an adequate interpretation of
the nature of text. As Hudson (1980: 131) notes:
T he m ost obvious fact about discourse structure is th at many differ-

en t kinds of structure ru n through discourse, and any attem pt to
reduce them to a single type is b o u n d to fail.
Biber has demonstrated this assertion convincingly in his research

on textual variation in spoken and written genres. As Biber (1988:
22) notes:
Linguistic variation in any language is too com plex to be analyzed in
term s of any single dim ension. T he simple fact th at such a large
num ber of distinctions have been proposed by researchers indicates
that no single dim ension is adequate in itself. In addition to the dis-
tinctions ... such as restricted versus elaborated and form al versus
inform al, linguistic features vary across age, sex, social class, occupa-
tion, social role, politeness, purpose, topic, etc. From a theoretical
point of view, we thus have every expectation that the description of
linguistic variation in a given language will be m ultidim ensional.
Many previous efforts have assumed, however, that textual varia-

tion could be discussed along single overarching continua which
would capture the essence of the variation. Only recently, how-
ever, have researchers been willing to entertain the notion of
multidimensional constructs and what such constructs might
mean (Biber 1988, Schiffrin 1987). Once it is recognized that text
is a multidimensional construct, it becomes clear that no simple
definition can possibly suffice; however, it is possible to provide a
working definition that will serve as a reasonable heuristic for
determining what is not a text.
A text is a structural equivalent of language in real use which
conveys meaning in all four senses of Hymes’s (1972) communica-
tive competence (whether a text is: possible, feasible, appropriate,
and perform ed), and which suggests a topic of discourse (how-
ever minimal). Further, As Halliday (1978) argues, a text should
not be solely defined as having formal opening or closure, as text
is seamless, with predictable structure. The beginning and end of
texts are more likely to be determined socially and semantically
from the context, rather than from a set of formal structural pat-
terns of organization. Taking into consideration the above
constraints, text may be defined simply as follows:
A text occurs when the discourse segm ent is identified as possible,
feasible, appropriate, and perform ed, and has a topic.
This definition amounts to a textual version of the previously dis-

cussed notion of writing as composing. (See also Brown and Yule
1983.)
2.4 The Text as communication
Language, it is assumed, is primarily for communication, as com-

munication is broadly interpreted (see Widdowson 1980). For this
reason, writing may be said to represent an attempt to communi-
cate with the reader (Widdowson 1980; see also Raimes 1983b).
The writer has certain intentions and purposes, as well as certain
information to convey. Writing will, therefore, be structured to
communicate information within certain accepted linguistic, psy-
chological, and sociological principles; e.g.:
1. Gricean maxims - the need to be informative, factually correct,
relevant, and clear; and the systematically interpretable viola-
tions of these maxims (cf. Empson 1961, Steiner 1978).
2. Conventions for conveying status, situation, intent and attitude.
3. Mechanisms for indicating newness of information, rate of
information flow, and probability of information.
4. Predictability of cognitive structures which anticipate larger
patterns of organization: schemata, scripts, frames, goals,
etc.
Any theory of how a text is constructed will have to consider how
these principles influence the organization and structure of writ-
ing. However written texts are examined, it is necessary to
recognize that theoretical assumptions must include influences
beyond low-level linguistic descriptions. Text analysis will have to
be relational and depend on the context within which the text was
written. The following extracts should illustrate this point.
Nearly eight decades ago, for instance, a South African zoologist
n am ed J.E. D uerden published his observations on the d ecorator
crab, a species widely distributed across the Indian and Pacific
O ceans. Its tools, living sea anem ones. W hile conducting his study,
D uerden discovered th at the crab removes two anem ones from the
ocean floor and grasps one in each o f its fro n t claws. T hen, holding
its living ‘tools’ ou t in fro n t o f itself, the crab moves along the
ocean bottom in search o f food. As the an e m o n e’s fingerlike tenta-
cles, or polyps, reach o u t into the currents, they g ath er tiny food
particles, which the crab th en removes from the anem ones with its
second set of claws. Moreover, the anem ones serve an o th er p u r-
pose. If any p art of the crab is tou ch ed in fro n t it autom atically
extends its stinging tools in that direction. ‘It would be im possible,’
D uerden wrote, ‘for any m arine anim al o f m oderate size to m olest

the crab in fro n t w ithout touching the polyps.’ T he d eco rato r crab
has an effective tool of defense. P erhaps because m ore scientists are
spending m ore tim e these days studying invertebrates, they are
finding m ore and m ore such exam ples of tool use am ong the prim i-
tive creatures.
(Abraham son 1985: 25-6)
In this first text the author is addressing a general audience; he

does not presuppose specialist knowledge. The attitude of the
author is one of equal status with the readers, though the author
is definitely the expert. The topic is carefully elaborated with clear
examples.
Jobs that you wish to ru n u n d er batch processing should be subm it-
ted at the In p u t window. Each jo b th at you subm it m ust have a valid
pre-punched UCC Jo b Card as the first card o f the deck. After your
cards have been read through the card reader, the right side of the
jo b card is torn off and kept by the o p erato r and your deck is
retu rn ed in the jo b deck retu rn bins (immediately to the right o f the
In p u t window) according to the first letter o f the jo b n am e (printed
in the u p p er left han d co rn er of your jo b card).
(USC UCC USM A A 02 1985: 8)
This text, by contrast, is terse; the author presupposes a great deal

of information and assumes a specific audience in a specific situa-
tion. The rate of information flow is more rapid because of the
presupposed shared knowledge. The writer takes a position of
authority with respect to the reader, as indicated by the numerous
second person pronouns.
These two texts clearly illustrate how larger organizing princi-
ples play a role in text construction. These examples show that
texts convey information to the reader on many different levels.
Ajiy theory of writing will, therefore, have to provide explana-
tions about such multilevel communication; that is, understanding
and producing texts will not only require knowledge of the sur-
face structuring of texts but also of the underlying textual
structuring.
The final section of this chapter will review in detail a num ber
of research approaches which attempt to explore these various lev-
els of structuring at the surface syntactic level. The following
chapter will extend the analysis to considerations of the underly-
ing textual level.
2.5 Research on the surface/sentence level
2.5.1 Syntactic analyses

The Chomskyan enterprise of the 1950s and 1960s generated an
interest in examining the writing of students in order to deter-
mine syntactic features which would indicate writing
development. Basing their analyses on Chomsky’s notions of ker-
nel sentences and transformations, H unt (1983) and others
(Loban 1976, O ’Hare 1973) developed the concept of the T-unit
as a central structure of student writing. Discourse, for them, was
viewed as a series of structural T-units (defined as a main clause
and all dependent modifying clauses). The large majority of this
work was aimed at efforts to explore the output of student compo-
sitions. H unt’s well-known study (1965), ‘Grammatical structures
at three grade levels’, demonstrated a consistent increase in the
syntactic complexity of student writing at grades 4, 8, and 12.
From the basic T-unit, H unt created a num ber of measures
claimed to demonstrate writing development. Most popular
among these have been:
Number of words per T-unit
Number of T-units per sentence
Number of clauses per T-unit
N u m b e r o f w o rd s p e r clau se.
Despite numerous criticisms of the T-unit approach to writing

research, H unt’s findings have since been replicated sufficiently to
warrant some reliability for his measures (at least for LI contexts;
Hillocks 1986; cf. de Beaugrande 1984, 1985 for LI criticism;
Gaies 1980 and Larsen-Freeman 1978 and for L2 discussion and
criticism). The best results using this methodology have been with
measuring writing differences across clearly distinct age/grade
cohorts. Its effectiveness is more debatable when the compared
groups are less obviously distinct. Overall, his methodology has
been the source for innumerable studies and dissertations as well
as the instructional methodology known as sentence combining
(see Daiker et al. 1985 for a review of LI sentence combining
research).
While the theoretical grounds supporting sentence combining
have been strongly criticized, there is evidence that sentence com-
bining does lead to writing improvement to some extent; the instruc-

tional approach has achieved some measure of success and cannot
be labelled a failure. Both de Beaugrande (1984) and Hillocks
(1986) provide a number of suggestions why this would be the case.
Included among these factors are that sentence combining:
■ brings to conscious attention syntactic formats;

■ provides an orderly tactic for revision;
■ promotes student confidence;
■ increases familiarity with syntactic patterns;
■ improves fluency and frequency of sentence writing;
■ promotes sentence variety;
■ allows for denser informational load in sentences;
■ aids the controlling of chunking in discourse; and
■ integrates syntax with other levels of discourse processing.
Certainly sentence combining offers a way to raise student aware-

ness of various construction types and their combinatorial
possibilities. In the larger sense, however, a basic difficulty
remains; that is, there is no real evidence that either ‘syntactic
complexity’ or ‘syntactic maturity’ is a major reason for improve-
m ent in writing ability. Every L2 teacher has encountered students
who can write perfectly correct syntactic structures (or can parse
an isolated sentence accurately) but cannot generate written text.
Some degree of syntactic maturity seems to be a prerequisite to
discourse ability, but the precise relationship is not well under-
stood.
In addition to T-unit research and the development of sentence
combining, other syntactically based findings from composition
researchers and developmental psychologists provide results
similar to T-unit analysis. In particular, the following gross devel-
opmental changes can be seen in ‘more successful’ written text
(see Hillocks 1986, Perera 1984, Witte and Cherry 1986):
■ increased use of adjectives

■ increased nominal complexity
■ increased use of free modifiers
■ increased use of sentence adverbials
■ increased use of relative clauses
■ increased use of finite adverbial clauses
■ increased use of stylistic word-order variation
■ increased use of passives

■ increased use of complex NP subjects
■ increased range of tense and modal usage, and
■ decreased use of unmodified NPs.
Based on the research to date on syntactic features (including T-
unit analyses), there are a num ber of measures which differentiate
among a range of grade levels. It is uncontroversial to say that a
mature writing style would contain more of these features in an
equivalent task comparison. The more interesting and more con-
troversial question is the extent to which any of these features is
genuinely a ‘m arker’ of a mature writing style (as opposed to
being, for example, a marker of register variation). For many of
these features, this question remains open; further, it is not clear
that a similar general developmental progression occurs in the
writing of L2 students. Perhaps, more importantly, it is not clear
that one can assume the reverse - that promoting or teaching
these features in instruction will lead to better writing quality. The
finding from LI sentence combining research is suggestive for the
teaching of complex syntax to L2 students. This use of sentence
combining, however, should not be confused with the develop-
m ent of writing.
Perhaps most serious as a criticism of surface-feature research is
the inability to demonstrate a clear relation between syntactic
complexity measures and judgem ents of improved writing quality.
While T-units and other similar measures distinguish broadly
defined grade differences, there has been no clear connection
between these measures of syntactic complexity or fluency, and
writing quality as judged by measures of coherence a n d /o r by
rater evaluation (Hillocks 1986). Thus, one of the most serious
problems of the T-unit approach is the ambiguous relationship
between control of sentence level complexity and overall writing
quality. And while some significant gains have been made in
understanding a writer’s growing syntactic development, the rela-
tion to overall writing development is not well established.
Research by Faigley (1979), Nold and Freedman (1977), and
Stewart and Grobe (1979) has provided little correlation between
complexity and quality. Hillocks (1986) suggests that sentence
combining promotes ‘syntactic facility’ rather than syntactic com-
plexity, and the effect of facility on overall writing quality remains,
at this point, an open question. Hillocks does note, however, that
on the basis of experimental research, sentence combining

appears to be more effective than a num ber of other instructional
strategies such as free writing and model presentation.
Although syntactic analysis of compositions continues today,
and provides im portant insights into the students’ text products, it
omits much information on the nature of text construction; if
writing were simply the stringing together of complex clauses, the
problems currently evident in student writing would be relatively
easy to diagnose, predict, and cure. A second syntactic-based
approach to text analysis, current textlinguistic corpora research,
promises to provide additional insights by employing a more
sophisticated methodology and by stressing the discourse-func-
tional basis of groups of surface structural features. While many of
these studies began as stylistic analyses, recent developments in
corpora research suggest that new insights into text construction,
as presented in the next section, are emerging from this line of
research.
2.5.2 Corpora research

The general dilemma facing most projects on corpus research is
the lack of a theoretical foundation for the interpretation of the
results prior to the analysis. Thus, most corpus research has been
of a post-hoc nature, looking at the frequency counts and decid-
ing what can be said about these results (cf. Sinclair 1994, who
argues for minimizing the effect of theory in corpus analysis).
This is true not only for stylistic studies of literary works, but also
for more recent text analysis studies which seek to find patterns in
the results rather than employ a theory to interpret the results
(Sinclair 1991, 1994). It is widely recognized that texts are multidi-
mensional constructs; however, this claim seldom appears to be
accounted for in corpora research in ways that will allow for the
discovery of these dimensions.
Perhaps the best effort to overcome these limitations in corpora
research lies in the work of Biber (1988, 1992, 1995). As described
in Chapter 1, Biber’s research examines textual variation among
spoken and written texts. In his early major study (1988), he
argued for the existence of at least six identifiable textual dimen-
sions of variation. His research procedures were similar to those in
a num ber of other corpora analyses using multivariate statistics,
and some might argue that his approach is little different from
other similar efforts. Indeed, his findings are the result of inter-
preting a factor analysis of many lexico-syntactic features, and he
had no prior indication of the num ber of textual factors he would
create. The major differences lie in the fact that:
■ he used many more texts than most corpora studies have previ-
ously examined;
■ he measured more lexico-syn tactic features, each of which was
included because it had been discussed in the linguistic litera-
ture as having certain functional characteristics in discourse;
and
■ because he was able to define functional interpretations for
each feature, he was able to establish plausible overall interpre-
tations for clusters of features as they co-occurred in the factor
analysis.
Biber has since established the validity of his textual dimensions
by using a num ber of confirmatory statistical procedures, and by
replicating his results with a num ber of independent corpora.
A major finding of Biber’s research is that text genres may be
identified by the co-occurrence patterns of groups of surface lin-
guistic features. Surface structure reflects discourse variation
and, by inference, discourse structure, though the way in which
surface structure reflects discourse function does not rely on the
real or notional relationship between individual features and spe-
cific textual genres. As an illustration, Biber’s fourth dimension,
‘Overt Expression of Persuasion’, comprises necessity modals
(e.g. must, should), prediction modals (e.g. will, shall), suasive
verbs (e.g. agree, arrange, ask, beg, pledge, propose, request, suggest,
urge), infinitives (e.g. to go, to change the rule), and markers of con-
ditional subordination (e.g. if ... , unless ... ). These features
individually may do little to define a textual pattern. Taken as an
aggregate, however, they appear to coalesce into a textual dimen-
sion which is only definable in the aggregate. These textual
dimensions can then be used as barometers to see which textual
genres, defined functionally, appear high or low on these dimen-
sions.
Grabe (1987) applied Biber’s approach to varieties of exposi-
tory prose as a way to define textual variation within this larger
category. One hundred and fifty texts of 15 functionally defined
types were analysed in terms of 31 syntactic, lexical, and cohesion
variables. Using Biber’s programme for counting the linguistic
features, Grabe employed factor analysis and also derived six inter-
pretable factors. The first four factors proved to be highly reliable
and were used to examine different patterns among expository
prose text types. The textual/functional interpretations were
based on the im portant co-occurrence patterns of linguistic fea-
tures for each factor. Following Biber’s situational and
communicative interpretations for the linguistic features in the
study, the four im portant dimensions were labelled as follows:
1. Non-narrative versus Narrative Context (Immediacy of Context)
2. Interactional versus Informational Orientation
3. Abstract/Logical versus Situation Information
4. Objective versus Expressive Style.
These four dimensions served to define three types of expository
prose among the 15 text genres used in the study. The importance
of this research is that it demonstrated textual structure in the sur-
face linguistic features of a text, and it suggested a way to explore
variation in student-written texts which would have greater
explanatory power than simple counts of individual surface fea-
tures and their correlations to writing development. One such
study explored the general nature of freshman student writing at
an American university.
Grabe and Biber (1987) perform ed a pilot study using 40 final
essay exams from the University of Southern California Freshman
Writing Program. The goal was to see how freshman student writ-
ing compared with a variety of edited prose types from Biber’s
major corpus. The essays were entered into the larger corpus of
spoken and written texts used by Biber. The 40 texts showed no
significant differences among themselves though they comprised
ten high non-native English speakers, ten high native English
speakers, ten low non-native speaker essays, and ten low native-
speaker essays. The 40 texts were therefore treated as a group and
compared along five textual dimensions. Results of this pilot study
showed that freshman compositions were most similar to
Humanities Academic Prose on three of the five dimensions
(‘Narrative versus Non-narrative Concerns’, ‘Elaborated versus
Situation Dependent Reference’, and ‘Abstract versus Non-
abstract Style’). It would seem, therefore, that for three of the five
dimensions, student writing is following genre expectations.
On two other dimensions, however, student writing differs from
the pattern. On dimension one (‘Involved versus Information
Production’), student essays were unlike any professional exposi-

tory type but similar to general fiction, having many more markers
of involved style than do other categories of expository prose
texts. On dimension four (’Overt Expression of Persuasion’), the
student essays are unlike any of the professional genres, being
extremely marked in their use of overtly persuasive features.
Overall, the student essays did not match any of the professional
genres along all five dimensions. Rather, freshman essays, at least
as demonstrated in the final examinations in a writing course,
appear to constitute a hybrid form of writing which combines
salient features of a num ber of text genres. Composition writing
may be a somewhat unique genre form, raising certain questions
about its usefulness as a learning experience, at least as it is cur-
rently taught.
2.5.3 Functional sentence perspective: informational

structure
A third area where the study of syntactic level phenom ena leads to
insights about the structuring of texts and writing development
has emerged from the early work of the Prague School of
Linguistics (Firbas 1986). Vande Kopple (1986: 72-3) describes
their orientation as follows:
As their nam e indicates, Functional Sentence Perspectivists take a
functional approach to language. ... They ... investigate w hat lan-
guage does, how people use it in various ways to achieve various
purposes. Thus, their focus is prim arily on connected texts, n o t on
isolated or random ly connected sentences, since people rarely use
the latter for com m unicative purposes. A nd in m uch o f their work
they proceed by exam ining the relationships between the structure
and the m eaning of a text, the extralinguistic situation the text exists
in and for, the com m unicative function the text apparently has, and
the w riter’s or speaker’s ap p aren t assum ptions about the state o f his
or h er addressee’s motivation, knowledge, an d consciousness.
2.5.3.1 S orting th e term in o lo g y
There are, in fact, a num ber of overlapping notions in the

research on information structuring. These are commonly
referred to as given-new relations, topic-com ment relations, and
them e-rhem e relations. While each describes the functional
arrangem ent of information in texts, they are at times defined

somewhat differently, depending upon the researcher. It is useful
to point out these distinctions briefly so that the reader may rec-
ognize possible incompatibilities when reviewing other research.
A few examples will highlight these potential differences.
In a given-new analysis, the researcher defines given as informa-
tion that has already been mentioned.
1. Most people realize that wolves have to kill deer, moose caribou,
elk and other large animals to survive. The predators live in fam-
ily groups called packs, usually containing 6-12 members, and it
takes a lot of meat to feed them. The pack is well organized. ...
The first sentence begins with an indefinite general noun phrase
(most people), indicating that new information is likely to follow.
This new information is not likely to be closely related to the
topic just discussed. In the second sentence, the term predators is
used to refer back to wolves (signalled by the definite article, the).
The first occurrence of pack is introduced as new information
and then defined. In the next main clause, the term becomes
part of the given information signalled by the use of the definite
article, the.
Many researchers conflate a theme-rheme analysis with a topic-
comment analysis. When them e-rhem e structure is treated differ-
ently, the theme is typically assumed to be the first-mentioned
phrase in the main clause unit; usually this coincides with the
agent/subject/topic of a sentence. In the passage above the
themes would be, respectively, ‘most people’, ‘the predators’, ‘it’,
and ‘the pack’. The change of perspective that this type of analysis
offers may be found in the following examples:
2. In the late 1960s and early 1970s, I witnessed fluctuations caused
by adverse conditions. ...
3. Curtailed in the area, he stated that, ‘everyone knows the wolves
will wipe out the deer’.
In example 2, the theme of the clause is the temporal setting
rather than the agent subject. In example 3, the theme is a
fronted participle which sets the background (ungrammatically)
for the quotation to follow.
Them e-rhem e is now most commonly associated with the work
of Halliday (1985, 1994) and systemic text analysis rather than
with the Prague School directly (e.g. the work of Danes 1974,
Firbas 1986, and Vachek 1966). For Halliday, them e-rhem e struc-
ture in texts is treated as an independent concept. In his systemic
analyses, the notion of theme represents the point of departure in a
structure; in contrast, the rheme represents the move away from
the speaker’s starting point. This concept works together with
given-new relations, though they are not the same concept. The
distinction is that the given-new relationship is based on the per-
spective of the hearer/reader (and based on the intonation unit
of information), while the them e-rhem e relationship is based on
the perspectives of the speaker/writer (and based on constituent
sequence). Fries (1994) and Martin (1992) suggest, further, that
thematic structure represents the text’s method of development;
that is, the sequences of clause themes across a text point to the
development of the major ideas (or macro-themes) in the text.
Systemic linguistics, following Halliday, has contributed to this
line of work from a variety of perspectives (Benson and Greaves
1985, Coulthard 1994, Couture 1986, Halliday 1985, Halliday and
Hasan 1989).
Topic-comment structures are, when treated differently from the
above categorizations, seen as defining ‘what the sentence is
about’. This designation requires somewhat more interpretation
and intuition, though the work by Lauttimatti, discussed below,
uses this criterion in her analyses. In examples 1 and 3 above,
wolves are the topic of all the sentences, the ‘comment’ is what is
said about the wolves in each case. In example 2, the topic is T .
Finally, there is another set of terms which are sometimes dis-
cussed, and are sometimes confused with the previous sets:
focus-presupposition. This pair of terms refers to the information
that is highlighted or focused (and usually contrasted in some
unexpected way), and to the information which is backgrounded
(and is often treated as presupposed, or assumed, knowledge).
Two examples illustrate this relationship:
4. It has been my good fortune to have spent most of my career
researching such details of the wolfs life.
5. It is this type of stability that is often referred to as the ‘balance of
nature’.
In example 4, the it subject stands in the place of the understood
infinitive clause beginning with ‘to have spent ...’. This construc-
tion highlights the author’s ‘good fortune’. In example 5 the use of

the cleft construction similarly highlights ‘this type of stability’ as
the focus of the sentence. Focus is often treated as synonymous with
new information and presupposed as synonymous with given informa-
tion. However, it is possible to have information in focus which is
given, or is the topic of a sentence; it is also possible to have presup-
posed information actually introduced in the text for the first time.
Despite all of these potential confusions, there is much important
research on information structuring and, for the most part, the
blurring of distinctions does not seem to invalidate the findings.
2 .5 .3 .2 T op ical se n ten ce structure
A second major line of research on information structuring is rep-

resented by topical sentence structure, as developed by Lautamatti
(1987); in this case, using topic-com ment analysis to examine
written text and discourse simplification. Her theory of topical
development in discourse examines the relations between the
topic of discourse, the topical subject of a sentence, the syntactic
subject, and the initial sentence element. Noting that the latter
three notions do not always overlap, she explores the various pos-
sible patterns in written texts. Her goal is to isolate the topical
subject of a sentence and then to examine the patterns of progres-
sion which the topical subjects form in a text. Her approach is
im portant for a num ber of reasons.
1. She provides a functionally based taxonomy of topical and non-
topical linguistic material in a sentence; the categories explain
what each segment is expected to do in the sentence.
2. The specification of topical subject is presented in a way which
indicates that topic for Lauttimatti is not simply given, as
opposed to new, information.
3. She applies her analyses to written discourse to show that cer-
tain patterns of topical progression may be more readable than
others (i.e. those texts which have fewer competing subtopics,
fewer complex sequential progression (A-B, B-C, C-D), and
more series of parallel topic progressions (A-B, A-C, A-D)
appear to be more readable). This suggestion seems to be well
supported by the work of vande Kopple (1986).
Lautamatti’s approach to topical development should be applica-
ble to the analysis of student writing. In particular, it may provide
specific teaching suggestions for students who appear to be writ-

ing texts with deviant, and probably less readable, topical
organization. This latter application is investigated by Cerniglia et
al (1990).
2 .5 .3 .3 T op ic con tinu ity
A third line of research, focusing on the function of topic devel-

opm ent in discourse, involves the work of Givon (1983, 1985) on
topic continuity. In his research, Givon treats topics as noun
phrases (NPs) which receive continuous m ention in the ongoing
discourse. He does not distinguish topic-com m ent from
given-new information, conflating the two concepts. In fact, his
work is more in line with the work on given-new chaining in dis-
course in that it requires specific prior mention to count
something as a topic under analysis.
Beginning from the functional perspective that topic structur-
ing is essential to the continuity of discourse, Givon has proposed
that a NP which becomes a topic is restated in the ongoing dis-
course in different ways depending on:
• how far back in the text the last previous mention occurs;
• the num ber of potential competing NP referents in the immedi-
ate discourse; and
• the strength of its persistence in the oncoming discourse.
These quantifiable measures provide a means to account for the
different ways topics are encoded in the structure of the grammar
and suggest the writer’s decisions in regulating/signalling the flow
of information. In a sense, this approach amounts to a type of
cohesion analysis. The difference is that his measures relate to the
degree of difficulty of maintaining or recovering the topic of
discourse with increasingly more marked grammatical forms.
Specifically, Givon (1983: 17) suggests that the following scale of
topic realization is generally applicable to English:
Most continuous/accessible topic:

1. Zero anaphora (I went to the store and 0 bought some cheese)
2. U nstressed/bound pronouns or grammatical agreement [Does
not apply to English]
3. Stressed/independent pronouns (He went to the store by him-
self)
4. R-dislocated DEF-NP’s (It’s on the desk, the book you want)

5. Neutral ordered DEF-NP’s - definite article initiates search for
antecedent match ( The man came back again)
6. L-dislocated DEF-NP’s ( The book you found, we left it on the bus)
7. Y-moved NP’s (The dictionary, I found yesterday)
8. Cleft/focus constructions (It’s the dictionary that I found yester-
day)
9. Referential indefinite NP’s (A dictionary was found yesterday)
Most discontinuous/inaccessible topic.
Givon (1985) has since revised his approach into a more complex
set of continua which are appropriate to cross-linguistic research.
The importance of this line of research is that it provides a specific
agenda for examining how information continues and how topics
are maintained. It becomes possible to explore the difficulty read-
ers should have in processing a particular text depending on the
deviation of topic marking based on the expected continuum. For
writers, this continuum indicates the importance of knowing when
it is not necessary to stress the topic and when it is im portant to
mark it strongly for easy recovery.
2 .5 .3 .4 T op ical structure analysis
Drawing on the work of Danes and Lautamatti, a fourth line of

research in sentence-based functional discourse analysis centers
again around a conflation of topic and given information. Witte
(1983a, 1983b - see also Connor 1987, vande Kopple 1986) devel-
oped a topical structure analysis to study differences in high- and
low-quality writing and differences in revision strategies. Looking
for topical and sequential chaining patterns in stude nt essays,
Witte (1983b) found that low-rated essays did not provide enough
appropriate given information and forced the reader to make too
many inferences. The texts were not reader-friendly. Witte further
found that low-rated essays used fewer sequential chaining pat-
terns, making it harder for the reader to perceive main topics in
the essay. Overall, differing patterns of topical structure analysis
appeared to provide good predictors of student writing quality.
In a series of related studies, Connor (1987) has applied topical
structure analysis to writing instruction to see if students could be
brought to recognize the information structuring patterns in their
writing. She reports positive results from teaching students to

determine the main topics and subtopics, as well as the progres-
sion of supporting information. As Connor (1987: 685) states:
S tudent responses have b een positive, an d we have seen improve-
m en t in student writing, specifically in regard to clearer focus and
b etter developm ent of subtopics. We feel th at topical structure
analysis is a useful check o f coherence in writing.
2 .5 .3 .5 G iven an d n ew in fo rm a tio n
A fifth research approach in information structuring is the spe-

cific examination of given and new information in texts. This
approach differentiates given information from topic on the basis
that given information must appear in the prior discourse
(whereas topic of discourse does not have to appear in prior dis-
course), and given information does not have to be limited only to
the discourse topic. Research by vande Kopple (1982, 1983, 1986)
using a variety of patterns of information structure has shown
that, generally speaking, the organization of given information
before new information makes texts more readable and memor-
able. His research on information structures and their cognitive
correlates suggests that further complex patterns of informational
progression probably exist, that students can be taught to identify
these patterns of organization in their texts, and that researchers
may be able to examine differences among text genres, authors,
or writing purposes on the basis of information structuring.
Vande Kopple (1986) also discusses other recent advances in
the theory of given-new relations. In particular, the work by
Prince (1981, 1992) suggests that given and new relations may not
be a two-part division. Instead, she proposes seven categories of
new, inferred, and given information for a more refined analysis
of information structuring in texts (see also Brown and Yule
1983). Under ‘new’ information, Prince distinguishes information
which is unused (readily recognized by the reader when intro-
duced), and brand new\ the brand-new entities may be either
anchored or unanchored depending on whether they contribute
to the ongoing discourse or are never repeated. An intermediate
category comprises two types of ‘inferences’: inferables and contain-
ing inferables. Inferable entities refer to information that can be
presumed from the information presented in prior discourse (e.g.
‘a camera ... the lens’). Containing inferables refer to entities which

are recoverable as parts of a collective reference (e.g. ‘one of the
eggs’). The ‘given’ entities can be evoked in two ways; either explic-
itly by the prior text, or by reference to the situation (e.g. ‘you
should understand what this means’; what Halliday would term
exophoric reference). While this taxonomy of relations in texts is
relatively new, it could be used to examine variations among many
different types of texts.
2.5.4 Cohesion in texts

Cohesion is the means available in the surface forms of the text to
signal relationships that exist between sentences or clausal units in
the text. A few of the signalling systems have been indicated
above, particularly in the work of Givon on topic continuity.
Cohesion research (Halliday and Hasan 1976, 1989) focuses on a
comprehensive examination of systematic devices used to connect
the surface form of texts. It is the surface manifestation of the
underlying relations that bind a text. While cohesion does not
provide a full account of the textual interpretation of a text, it is
an important indicator.
There are various means by which cohesion operates, princi-
pally including reference, substitution, ellipsis, conjunction, and
the lexical relationships of repetition, inclusion, synonymy/
antonymy, and collocation (Brown and Yule 1983, Halliday and
Hasan 1976, 1989, Hoey 1991). There is, as might be expected,
considerable controversy over the specific uses, specific domains,
and specific forms that cohesion encompasses. The seminal work,
Cohesion in English (Halliday and Hasan 1976), though most com-
monly cited, does not represent the totality of cohesion as it will
eventually be defined (e.g. Halliday and Hasan 1989, Hoey 1991).
Nevertheless, Halliday and Hasan’s earlier research is, by any
account, the place to begin. The basic categories they define are
illustrated briefly below in the following four text segments
(6)-(9).
(6) Barber (Samuel) managed to reverse the process; his last
movement, a driving moto perpetuo, fully realizes the tradi-
tional function of a concerto as a showpiece. It seethes and
stomps with unrelenting fury, and it ends in a satisfying climax.
(Goldberg 1985: V, 1)
This text includes a num ber of cohesive devices. A good example

of pronominal cohesion is the use of his and it to refer across
clauses. The text also makes use of the conjunction and. Lexical
cohesion occurs in this text in two forms: inclusion relations are
indicated by Barber, concerto, and movement', concertos are one form
of music composed by Barber, and movements are basic parts of con-
certos. Lexical collocation is also represented in this text. Process
collocates with a series of actions: driving, moto perpetuo, seethes,
stomps, unrelentingfury, and ends.
(7) Scientific doubt about using tool behavior as evidence of
advanced intelligence is indeed a fairly recent phenomenon.
One reason for this is the fact that for many years, the majority
of the research on the subject was conducted on more
evolved species. ... Perhaps the most dramatic example was first
reported by behaviorist Jane Goodall in the mid-1960’s, when she
discovered that chimpanzees sometimes use small twigs to
probe into termite mounds. Other scientists have found that
some fifteen species of birds exhibit tool use behavior.
(Abrahamson 1985: 25)
Text (7) contains instances of demonstrative and comparative refer-
ence, as well as a full range of lexical cohesion devices. A
demonstrative connection occurs with one reason for this, where this
refers back to the previous sentence. Comparative cohesion is indi-
cated by most dramatic example and by other scientists. Lexical cohesion
forms represented here include repetition, inclusion, and colloca-
tion. Repetition occurs with using tool behavior, tool-use behavior,
species: species, scientific, scientists, and behavior, behaviorists. Inclusion is
found in the forms evolved species: chimpanzees, and scientist behaviorist.
Collocation is extensive throughout this segment: Animal colloca-
tion is indicated by chimpanzees, termites, birds, and species. Science
collocation is indicated by scientific, evidence, phenomenon, research, sub-
ject, example, reported, behaviorist, discovered, species, and exhibit.
(8) Another way of putting it is to say that they treat their forests
much as we in America ravished our woodlands a hundred
years ago. This reference to owrown forest history is Myers’, and
it is not a casual one.
(Wild 1985: 134)
Text (8) offers a nice example of substitution, where one refers
back to reference. In addition, this text uses conjunction (and),
demonstrative (this), pronoun (our, it), repetition (forests), and

synonymy (woodlands) as markers of cohesion.
(9) ‘I hope I will not trouble you again,’ the former prisoner of
conscience said, 'but there is a strong possibility that I may.'
(Amnesty Action Newsletter, Dec. 1984: 1)
Text (9) provides an example of ellipsis, as well as uses of pronom -
inal and conjunctive cohesion. Ellipsis occurs when may is not
followed by a predicate; the ellipsis refers back to ‘trouble you
again’. But and I are the other cohesive devices in this text seg-
ment.
All of these operations provide means for linking the surface
text structure. As such, they reflect both the communicative inten-
tions and the choices made by the author in the structures used
and in the linear ordering of the texts. The concept of cohesion
has created a certain amount of controversy, particularly over
whether it provides appropriate measures of textual or compre-
hension development, and whether it represents a complete
description of textual relations. The next chapter argues that
cohesion is only part of the organizational structuring of texts. As
to the former controversy, Brown and Yule (1983) argue that
Halliday and Hasan’s theory of cohesion offers a taxonomic inven-
tory of textual resources. It is not clear, however, that Halliday and
Hasan (1989) would accept the taxonomic limitations which
Brown and Yule ascribe to their theory.
Hasan (Halliday and Hasan 1989) has recently revised her
description of lexical cohesion by limiting its function in research
to specific categories which are more easily replicable.
Recognizing that collocation (though critical to the texture of a
text) could not readily be verified reliably among researchers,
Hasan now employs the lexical categories of synonymy, antonymy,
hyponymy (superordinate-subordinate class), meronymy (part-
whole), and repetition as types of coextension (lexical cohesion)
in texts (cf. Hoey 1991).
Cohesion research on texts and on student composition has
been extensive. In particular, research by Cox et al. (1990, 1991),
Spiegel and Fitzgerald (1991), Tierney and Mosenthal (1983),
and Witte and Faigley (1981) presents a complex set of research
results. Cohesion research does not appear to be a complete
answer to understanding writing development; at the same time,
cohesion analyses do provide certain useful insights. Witte and
Faigley, for example, note that cohesion is a reasonable predictor

of writing quality across grade levels. They also contend that cohe-
sion is a useful indicator of differences in students’ invention
skills.
Halliday and Hasan (1989) have more recently extended cohe-
sion analysis into a theory of cohesive harmony. Recognizing
criticisms that cohesion quantification, in and of itself, did not dis-
tinguish coherent from non-coherent texts, they have proposed
an analytic approach which creates chains of identity or ‘similar-
ity’ cohesion ties. Further, the m anner in which the chain
elements interlink across chains represents the central elements
producing the coherence of a text. Thus, the cohesive elements
which form chains, and which interlink with other chains, are the
quantifiable features of cohesion which indicate differential
coherence in texts. (See also, Singer 1990 for supporting argu-
ments from cognitive psychology research.) At present, some
amount of research has been performed using cohesive harmony
(Cox et al. 1990, 1991, Spiegel and Fitzgerald 1991; cf. Hoey
1991).
The connection between cohesion and coherence is an issue
that has been raised numerous times: To what extent does the sur-
face form contribute to the perception of underlying coherence
of a text? For Halliday and Hasan, taking a functional perspective
on the constructive use of language, the surface structure must
make a considerable contribution (see also Christie 1992, Martin
1989, 1992, 1993). O ther researchers have argued that an analysis
of the logical coherence of a text for comprehension must be aug-
m ented by using procedures not specifically marked by surface
forms. These procedures are the subject of the next chapter.
3
Towards a model of text

construction
3.1 Introduction
The development of text analysis over the past 15 years has been,
at least in part, an attempt to develop a model of text construc-
tion, a description of how the text structure is assembled, taking
into account the message, the writer’s purpose, the topic, and the
expectations of the audience (cf. Bereiter and Scardamalia 1987,
Halliday and Hasan 1989, Martin 1992). As the review of research
on text linguistics indicates (Chapter 2), the effort to create a
comprehensive model of text construction is a complex task. A
model will have to account for the research of psychologists on
text structure, the research of engineers and linguists on artificial
intelligence, the research of linguists and applied linguists on dis-
course analysis and text genres in both synchronic and diachronic
modes, the findings from studies of writing development, and the
insights from rhetoric and critical studies.
It may well be that an explicit model of text construction is
beyond current research capacity. Nevertheless, an understanding
of how texts are constructed is an essential part of understanding
the nature of writing and writing development. Over the past 15
years, efforts to develop a description of text construction have
been undertaken by de Beaugrande (1980, 1984), de Beaugrande
and Dressier (1981), Brown and Yule (1983), Dillon (1981),
Halliday and Hasan (1989), and Martin (1992). In each case, the
description stresses certain aspects of text analysis in favour of
others. In this chapter, a descriptive model of text construction
which is appropriate for a wide range of research on writing and
writing development is proposed. This model, in turn, represents
one com ponent of a larger theory of writing that incorporates
knowledge of text construction.
60
Toward a model of text construction 61
In order to develop the model, it is first necessary to review the

im portant hypotheses and findings which such a model should
address. Among the most im portant hypotheses that appear to be
supported by research are:
1. Written language is distinct from oral language along a num-
ber of textual dimensions, and the construction of written
language must be studied according to its own structural and
rhetorical emphases.
2. Texts have hierarchical structure, most likely constituted as a
set of logical relations among assertions, or as elements in a
discourse matrix, or as cohesive harmony.
3. Different types of texts will have varying larger structuring
because of requirements of purpose, audience, status, author,
and information load.
4. Texts have a top-level structure which appears to vary with dif-
ferent text types, purposes, and audiences.
5. A discernable top-level of text structure is related to better
comprehension, recall, and coherence assessment.
6. Systems for analysing text structure can be used for research
even if each system in current use has particular strengths and
weaknesses.
7. A theory of text type variation is possible and is needed for
comprehension, production, and assessment research.
8. A theory of coherence is important to any model of text con-
struction.
9. Any theory of coherence must incorporate an analysis of
information structure - given-new, topic-comment, etc.
10. The surface form of texts plays a more im portant role in text
construction than previously predicted.
11. Learning to write requires the manipulation of many com-
plex structural and rhetorical dimensions, with greater
complexity occurring in expository/argumentative writing.
While the concept of coherence, which is central to many of the
points noted above, may be controversial for some researchers -
who assume all of coherence to be a construction of the reader -
there is now considerable evidence that the structure of the text
itself contributes to, or interferes with, the coherence that is per-
ceived by the reader (e.g. Beck et al. 1991, Britton and Gulgoz
1991, Martin 1992, Singer 1990). In addition, the notion that a
text is a multidimensional construct is now well accepted, as are
the many ancillary assumptions that are natural extensions of the

basic assertion.
From the various research areas present in the literature (and
previously reviewed above), it seems that a descriptive model of
text construction requires at least seven basic components which
must somehow coalesce as multiple interacting strands.
1. Syntactic structures.
2. Semantic senses and mappings.
3. Cohesion signalling.
4. Genre and organizational structuring to support coherence
interpretations.
5. Lexical forms and relations.
6. Stylistic and register dimensions of text structure.
7. Non-linguistic knowledge bases, including ‘world knowledge’.
Within each of the components are numerous subcomponents,
interacting among themselves as well as with other components
and subcomponents, rather like the atoms in chemical molecules.
In the following discussion, the components of a text construction
model are presented in four parts:
1. The elements of text structure
2. A theory of coherence
3. The functional-use dimensions of texts
4. The non-linguistic resources interacting with the elements and
functional use dimensions.
3.2 Elements o f text structure
In text structure, four potentially independent components exist

on two levels: two on a sentential level and two on a textual (or
intersentential) level (Figure 3.1). There is also a major division at
both levels between surface structure and underlying structure.
This division may be seen generally as constituting a form-mean-
ing distinction. A fifth component - the lexicon - is a diffuse
com ponent underlying the other four. Together, these five com-
ponents comprise the elements of text structure - the
fundamental building blocks from which all texts are constructed.
Surface j Deep
1
Li
Sentential Syntax El Semantics
XI
11
T I
Cl
Textual Cohesion Ol Coherence
N1
1
Figure 3.1 Elem ents of text structure
3.2.1 The sentential level

At the sentential level, syntax and semantics are conceived as they
are normally understood in linguistic theory. They operate within
the structural level of the written clause, representing systems of
surface structure and underlying interpretation (senses and map-
pings from word meanings to sentential meanings). The syntactic
component involves types of phrasings, types of clause construc-
tions and clausal combinations, and the ordering of the phrases
and words within the sentence.1 For example, one might want to
examine the num ber and types of passive structures, the types of
phrases that appear sentence initially, the num ber and types of
prepositional phrase groupings, or the num ber of rhetorical ques-
tions (cf. Biber 1988, Perera 1984). A researcher would want to
adopt or propose a syntax component that describes adequately
all the syntactic devices available in the language (e.g. Quirk et al.
1985).
Perhaps in the future, as linguistic research continues to
advance, the preferred syntactic description would be a matter of
empirical research (Sinclair 1991, 1994; see, e.g., the ‘grammar’
developed in the Collins COBUILD series, based on the analysis
of a massive corpus of written English). For the moment,
researchers must accept the syntactic approach most compatible
with their interests and training. Syntactic analyses, most com-
monly, will involve the counting of various constructions and
categories, and their co-occurrence in various combinations, typi-
cally as described in Chapter 2.
The semantic com ponent is also relatively open to workable

alternative frameworks. Since the study of linguistic semantics is a
recent development, a concise, complete theory is not yet well
established (Frawley 1993, Ladusaw 1988, Martin 1992). Semantic
theory attempts to relate the linguistic form to objects and events
in the phenomenological world. The primary function of this
com ponent is to assign meanings to words and phrases, and to
interpret how the meanings of phrases combine to form meaning
interpretations of entire clauses or clausal combinations. The
semantic component also interprets the meaning of certain lexical
classes of words within the clause. Included in such classes are pro-
nouns, reflexives, modal verbs, as well as verb groups which are
accompanied by particular semantic or syntactic constraints (i.e.
factive verbs, perceptual verbs, aspectual verbs, suasive verbs, pub-
lic verbs). The semantic component will also account for facts of
scope interpretation and other semantic issues which can operate
within the level of the sentence (Frawley 1993, Jackendoff 1972).
For example, in the two sentences below, the negative marker
changes its scope of interpretation from the lower clause in (1) to
the entire clause in (2).
(1) I regret that Tom doesn't like me.
(2) I don *t regret that Tom likes me.
These and other features of semantic interpretation are often
addressed under the term ‘sentential semantics’ (as opposed to
more microstructural ‘lexical semantics’ on one hand, and more
macrostructural ‘pragmatics’ on the other).
3.2.2 The lexicon

The lexicon (our mental word list) pervades all four of the other
components, in both surface form and underlying organization. It
both affects, and is affected by, each of the other four compo-
nents. The lexical entries used in text construction provide the
basic meaning and inference signalling from which syntactic struc-
tures, semantic senses, and pragmatic interpretations are
produced. The lexicon assists the syntactic component by provid-
ing sets of syntactically useful forms such as prepositions, articles,
existential ‘there’, etc. Further, the lexical entries of specific words
include syntactic information for the generation of the syntactic
structure of the sentence. The verb put requires that a direct
object and a locational prepositional phrase follow, while the verb

sit will not allow a direct object but may require a locational
prepositional phrase; e.g.
1. The boy put the book o n th e desk. [put what where?]
2. The boy sat o n th e chair, [sat where?]
The lexicon also provides the semantic forms which represent the
ideational content of a text. The lexical forms themselves are most
likely organized according to semantic criteria (as well as syntactic
criteria) such as objects related by schema structures or scripts, or
more abstractly as, for example, mental verbs, verbs of perception,
psychological verbs, public verbs, verbs of motion, etc., or adverbs
of manner, location, time, evaluation, etc.
At the level of text structure above the clause, lexical forms sig-
nal textual information in terms of cohesion. The lexicon
provides the units for this purpose, including, for example, pro-
nouns, demonstratives, ellipsis markers (e.g. x does too), and
substitution markers (e.g. one). These, and other forms, may all
signal aspects of cohesion. Finally, coherence structures may be
lexically motivated in that necessary inferences, rhetorical predi-
cates, and logical relations among assertions can be readily
interpreted from specific lexical forms. As a simple example, a
first clause using the word ‘problem ’ usually indicates the begin-
ning of a problem-solution discourse (cf. Hoey 1986, 1994,
Tadros 1994). Viewing the lexicon as central to a theory of text
construction is not a radical departure; a num ber of researchers
argue that lexical cohesion may be the most im portant aspect of
cohesion (Halliday and Hasan 1989, Hoey 1991, Witte and Faigley
1981). Stylists have also long been concerned with the importance
of diction in conveying exact meaning and attitude.
3.2.3 The textual level

The two components of the sentential level are rather straightfor-
ward, as is the lexicon, in so far as most researchers will accept
them as essential to language research. The components on the
textual level are, however, somewhat more controversial. Cohesion
and coherence parallel syntax and semantics on a ‘greater than
the clause’ level as surface and underlying textual structure. The
‘structure’ at this level must be different since it is not possible to
predict accurately that any sentence will determine the form or
the interpretation of later sentences, though it is likely to influ-

ence later sentences (unlike words in a sentence which may
directly predict the words or structures to follow). The structure at
this second, larger level is paradigmatic. This means that it is pos-
sible to choose certain sentence forms in lieu of other sentence
forms. The range of options available for a given clause position
represents the limits of its paradigmatic structure.
Paradigmatic structure can be demonstrated in the following
manner. It is possible to begin a paragraph with any of the three
sentences below. They all represent viable options as paragraph
initial constructions:
3. What led to the change in perspective was the different angles
of the sun’s rays throughout the day.
4. The different angles of the sun’s ray throughout the day led to
the change in perspective.
5. A change of perspective was created by the different angles of
the sun’s rays throughout the day.
The texts following sentence 3 would probably involve a discus-
sion of ‘what leads to the change in perspective’; the text
following 4 would probably involve a discussion of ‘different
angles,’ and that following 5 would probably involve a discussion
of ‘a change in perspective’. By the same token, the determination
of the opening structure and the ensuing environment creates a
probability for the immediately following structure; e.g.
6. It was Mary’s birthday._____. And her m other gave her some
perfume.
(a) I gave her a rose.
(b) A rose was given to her by me.
(c) A rose was my gift to her.
(d) She received a rose from me.
(e) Giving her a rose was my contribution to the day.
Technically, any one of the five options is possible in the blank
space in text 6, but there is a strong probability that native speak-
ers of English would choose option (a) because the text creates a
probability for a list of gifts Mary received for her birthday and
because option (a) is grammatically parallel to the final structure
in the text. For example, there seems to be no logical reason to
choose a passive construction, and the other options seem to
change the focus of the structure. In fact, in actual trials with
native and near native speakers, in excess of 80 per cent of sub-

jects chose option (a).
There is less than consistent agreement on the nature and rela-
tive contributions of cohesion or coherence to the overall
construction of texts. Moreover, there is little consensus on the
m atter of an overall definition of coherence (cf. Brown and Yule
1983, Johns 1986, Phelps 1988, Singer 1990). The features defin-
ing cohesion are generally accepted, though the recent changes
in lexical cohesion outlined by Hasan (Halliday and Hasan 1989)
make this im portant area of cohesion more amenable to serious
research uses. These issues concerning cohesion have been dis-
cussed in Chapter 2. The concern here is to explore the relation
between cohesion and coherence, the definitional nature of
coherence, and the role of coherence in a text construction
model.
3.3 A theory o f coherence
The concept of coherence in writing has traditionally been

accepted as a com ponent of writing research and writing instruc-
tion (e.g. Bamberg 1983, de Beaugrande and Dressier 1981). In
the last 20 years, however, researchers in psychology, linguistics,
and applied linguistics have begun to explore this notion in an
effort to understand how readers interpret a text as coherent and
how writers control language structure to convey a sense of coher-
ence.
At the same time, research in pragmatics and conversational
analysis has suggested that at least some part of coherence is con-
structed by the reader’s interpretive systems regardless of the text
structure itself. Garnham has pointed out that readers make a
num ber of bridging inferences which appear to be controlled
largely by the reader’s efforts to assemble a text-model representa-
tion of a text into long-term memory (Garnham 1985, Oakhill
and Garnham 1988). An understanding of the extent to which
these bridging-inferencing processes (as opposed to elaborative
inferencing) reflect processing of a specifically linguistic nature as
opposed to general cognitive processing, might determine
whether these processes are part of the linguistic system or have
their origin elsewhere in the cognitive processing model. If they
are largely linguistic in nature, then the argum ent can be made
that textual form is the trigger for these inferencing mechanisms

(cf. Barsalou 1992, Rayner and Pollatsek 1989, Singer 1990).
A second potentially non-linguistic source of coherence inter-
pretation is discussed by Brown and Yule (1983). They suggest,
following Widdowson (1978) and others, that coherence is the
result of conventionalized knowledge and sequences which a
hearer (reader) will be able to call upon to impose a coherent
frame onto a message. These sources are typically referred to as
scripts, frames, a n d /o r schemas for organizing knowledge of the
world. Assuming these external knowledge sources, Brown and
Yule argue that coherence is essentially the creation of the reader
rather than a product of the text. Drawing upon a general princi-
ple of analogy, the reader looks for similarities between the
message and the knowledge he or she already has and, as a conse-
quence, the reader is compelled to make connections. As they
state (1983: 65-6):
T he imperative ‘need to fin d regularities’ which P opper speaks of,
coupled with B artlett’s ‘effort after m eaning,’ constitute a powerful
expectation in hum an beings that w hat is said o r w ritten will make
sense in the context in which it appears. ... T he natural reaction of
m an appears to be to make sense o f any sign resem bling language,
resem bling an effort to com m unicate. ... T he natural effort o f hear-
ers and readers alike is to attribute relevance an d coherence to the
text they encounter until they are forced n o t to. ... It is n o t the
sequence o f sentences which represents ‘co h eren t discourse.’
R ather it is the reader, driven by the principles o f analogy an d local
interpretation, who assumes that [a sequence] describes a series of
connected events and interprets the linguistic cues u n d e r that
assum ption.
A third version of the non-linguistic basis of coherence may be

seen in the more recent theory of Relevance proposed by Sperber
and Wilson (1986). In this pragmatic interpretation of language
use, the principle of relevance determines how coherent a mes-
sage is to be taken. The basis of the theory assumes a general
cognitive processing principle that hum an beings are designed to
seek relevance in texts by comparing the text to other information
resulting in the creation of new information, the contradiction of
old information, a n d /o r the confirmation of one’s commitment
concerning something (Smith 1989). The theory would not
specifically state that relevance = coherence; indeed, a text may be
coherent and meet none of the three criteria for relevance.
However, the cognitive principle is very similar to that stated

above by Brown and Yule and is suggestive of a non-linguistic basis
for a theory of coherence.
There is likely to be a fair am ount of truth in the perspectives
on coherence presented above. At the same time, arguments of
this nature are a bit like saying that the printed page is unimpor-
tant in the reading process. While advances in discourse analysis
and pragmatics have opened new research perspectives on text
comprehension and interpretation, the basis for these interpreta-
tions nonetheless rests first with the text message itself. Ignoring
this fact leads to an unbalanced overemphasis on top-down pro-
cessing; ignoring the constraints on interpretation imposed by the
text leads to the logical conclusion that the text may be altogether
irrelevant. However, the vast majority of people, receiving a letter,
will decide to read it rather than merely assume that they know
the information in the letter and not open it (except perhaps in
the instance of unsolicited advertising). The very act of reading a
text implies that the text will influence the coherence interpreta-
tion constructed by the reader. For this reason, it is reasonable to
assume that the text itself has a considerable role to play in the
construction of textual coherence.
More recent research on coherence in texts, in fact, suggests
that much of the coherence-building information is not simply
imposed on the text by the reader, particularly in the cases of
more m undane everyday texts. It is currently popular to cite
reader response theories and constructivist views on comprehen-
sion and coherence in text. Unfortunately, their arguments rest
primarily on relatively abstract logic and the illustrative literary
examples to argue the matter. In contrast, cognitive psychology
research now provides strong evidence for the impact of text
structuring itself as a prime contributor to coherence in texts.
Certain aspect of coherence are directly traceable to the text
structure itself, and other aspects are best seen as an interaction
effect of the reader and the text information together. Examples
of research which support this view may be found in Anderson
(1990), Beck et al. (1991), Britton and Gulgoz (1991), Singer
(1990) and van Dijk and Kintsch (1983).
In addition, over the past two decades, in contrast to accepting
a non-linguistic perspective on coherence, many text linguists
have explored the structuring of text to determine how text may
be said to be coherent. One early effort to examine this potential
property of texts was the theory of cohesion of Halliday and

Hasan (1976). A num ber of responses in the 1980s argued that
cohesion was not equivalent to coherence (Brown and Yule 1983,
Carrell 1982, Mosenthal and Tierney 1984, Widdowson 1979).
However, Halliday and Hasan’s more recent (1989) theory of
cohesive harmony agrees much more closely with research on
information structuring and the influence of local clausal rela-
tions in building text coherence.
While it is fair to say that cohesion represents the formal sig-
nalling features of texts beyond the limits of the sentence, and
that coherence is probably more than this, it is also likely that writ-
ers, in using this surface signalling, are guiding readers to achieve
the preferred coherent interpretation intended by the writer. In
this respect, the many linguistic signals and markers of cohesion
that appear in a text provide a framework within which to estab-
lish the coherent logic of the textual information. To assert
otherwise would be to argue that the many signalling mechanisms
in texts are essentially arbitrary and meaningless bits of conven-
tion and stylistic options. The research of cognitive psychology
and educational psychology on text comprehension clearly does
not support the arbitrariness of structural signalling in text.
The controversies that have arisen over the nature of coherence
point to the fact that surface structure in texts does not provide a
perfect match with the organizational logic of the text. First, a cer-
tain amount of ambiguity is inherent in language and in texts.
Second, no set of linguistic signalling will provide all the informa-
tion needed to construct the text logic completely. Finally, writers
will employ the formal resources for signalling text organization
to differing degrees depending on their familiarity with intended
readers and their knowledge of the topic and genre, the goals of
the writer, and the proficiency of the writer in the written
medium. These factors together suggest that, while it is relatively
easy to explore the surface forms of cohesion, the nature of coher-
ence is much more difficult to discern.
Beyond the surface form, text is organized by the writer’s rela-
tion to it, to the reader’s assumed knowledge, and to the subject
matter. Coherence as a theoretical construct in text structure
refers to the underlying relations that hold between assertions (or
propositions) and how these assertions contribute to the overall
discourse theme (or macrostructure) (e.g. Beck et al. 1991,
Britton et al. 1993). This set of relations assumes that coherent
texts will be unified by one overarching theme, whether stated or

implicit (Britton and Gulgoz 1991, Kaplan et al. 1983, Singer
1990). It is the coherence in text structure which allows the reader
to build, at least in part, a mental model of comprehension
(Garnham 1985, Singer 1990).
Various approaches of coherence structure have been proposed
(e.g. Mann and Thompson 1988, 1992, Martin 1992, Meyer 1975,
1985, Sperber and Wilson 1986, van Dijk and Kintsch 1983). In
the majority of models, coherence is defined (implicitly or expli-
citly) as:
■ having a discourse theme (overall topic of discourse);
■ comprising a set of relevant assertions relating logically among
themselves by means of subordination (cause, condition, com-
parison, specification), coordination (addition, restatement),
a n d /o r superordination, from the level of the sentence to the
top-level structuring of a text, and
■ being organized by information structure imposed on asser-
tions most effectively to guide the reader in understanding the
theme or the intent of the author (topic-comment, them e-
rheme, given-new, focus-presupposition).
The crucial issue for this textual component seems to be how the
relationships between logical assertions and information structure
work within the framework of the topic of discourse.
Of the three subcomponents of coherence, discussions of the
topic of discourse and of logical relations among assertions are
relatively straightforward. Within the subcomponent of topic of
discourse there must be some attempt to account for the relations
and interactions among clausal units, larger organizing notions
and macrostructure (Hoey 1991, Kaplan 1972, Mann and
Thompson 1988, Meyer 1987, Singer 1990, van Dijk and Kintsch
1983). These notions can be illustrated in the following
problem -solution text:
How delicate is the balance of nature ?
1. D uring two decades o f wolf research, conducting studies in

n o rth e rn M innesota and on Isle Royale in M ichigan, I have learn ed
that, far from always being ‘balanced,’ ratios o f wolves an d prey ani-
mals can fluctuate wildly - an d sometim es catastrophically. Wolves
may actually starve after killing off alm ost all the moose an d deer in
an area. This explains why wolf-control program s may sometim es
ensure greater and m ore stable num bers of both wolves and the ani-
mals they hunt.
2. Most people realize th at wolves have to kill deer, moose, cari-
bou, elk and other large animals in o rd er to survive. T he predators
live in family groups called packs, usually containing 6 to 12 m em -
bers, and it takes a lot o f m eat to feed them . T he pack is well
organized, with each wolf occupying its own place in the social lad-
der. Each pack possesses a territory large enough to encom pass
hundreds of prey animals and delineates that territory with u rine
marks and howling.
3. It has been my good fortune to have spent m ost o f my career
researching such details o f the wolf’s life an d of the creatu re’s inter-
actions with its prey. O ne of my studies th at helped fix the
balance-of-nature idea in the public m ind was the Isle Royale
wolf-m oose research th at I conducted as a doctoral candidate from
1958 through 1962. Flying over the snow-covered, 210-square-mile
national park in Lake Superior each w inter in a small ski-plane, I
learned that there were 20 to 25 wolves on the island an d approxi-
mately 600 moose. T he wolves were harvesting the old an d sick
m oose and the surplus calves, and b oth p red ato r and prey num bers
seem ed stable.
4. It is this type of stability th at is often referred to as the ‘bal-
ance of n atu re .’ Prey animals are superbly adapted for escaping
wolves, and wolves are well suited for catching prey, and the result is
a rough balance between the two. U nfortunately, it d o esn ’t always
work so smoothly over the short ru n . H um an interference, unusual
weather, or o th er ‘outside’ factors can cause disruptions in the
predator-prey relationship. For this reason, scientists are increasingly
hesitant to use the word ‘balance.’ Many o f us now prefer the phrase
‘dynamic equilibrium ,’ which b etter describes the ph en o m en o n .
5. In the late 1960’s and early 1970’s I witnessed fluctuations
caused by adverse conditions: a series of severe winters that struck
across N orth America. My students and I were studying wolves and
d eer in n ortheastern M innesota at the time, using aerial radio-track-
ing. O ne of my bush pilots w ondered why we were doing the study.
N oting that wolf-control program s h ad generally been curtailed in
the area, he stated that, ‘Everyone knows the wolves will wipe o u t the
deer.’
6. It looked like he was right. Year after year, we watched the
wolves decimate an overwintering h erd of white-tailed deer through-
out a 1,500-square-mile region. Almost all the deer were inaccessible
during the hunting season, so wolves caused most o f their m ortality....
7. But the wolves had help from the severe w inter weather.
W inter severity played a dual role in the d eer decline. First, the deep
snow m ade adult deer easier to kill. T hus in w inter 1968-1969,
wolves even took m ore d eer than they were able to consum e. Each
time we found a wolf kill, we would land o ur plane on a nearby
frozen lake and snowshoe to the carcass. We saw several that had
been killed and left with little or n o th in g eaten. T h at m ean t fewer
deer the following sum m er to produce new fawns.
8. Secondly, the fawns th at were p roduced that sum m er were in
trouble from the start....
9. Any prey population can safely sustain interm itten t fawn or
calf crop loss. However, o u r h erd in n o rth e rn M innesota was h it with
a series of seven severe winters, from 1966 to 1972, while wolf n um -
bers were high....
10. Initially, during this d eer decline, wolf num bers actually
increased - a fact I co u ld n ’t account for until I realized th at they
were cashing in on the increased vulnerability of the d ee r....
11. Wolves, therefore, were suddenly faced with a severe food
shortage....
12. T he same severe winters battered Isle Royale, which only lies
20 miles from M innesota....
16. However, there is little disputing the results of a recen t well-
controlled experim ent in central Alaska. Some 38 to 60 p er cent o f
the wolves were rem oved each year from a test area while wolves
were n o t controlled in several adjoining areas. Moose an d caribou
calves and yearlings increased two- and four-fold where wolves had
been taken com pared with their num bers before wolf control and
were consistently higher than in the areas with no wolf removal.
Actual m oose and caribou h erd sizes followed the same trend.
17. W hat would have h ap p en ed if wolves had n o t b een con-
trolled? Because the herds had been declining before the
experim ent, I expect th at they eith er would have co n tinued to
decrease, would have rem ained stable b u t low, or m ight have
increased only slowly. Meanwhile, from what I saw in n o rth e rn
M innesota, wolf pups would have starved to death, wolf productivity
would have declined, and adult wolves would have killed each other.
C ontrol program s allowed recovery o f b oth prey an d wolves so th at
m ore of each could live over a longer period. It is som ething I am
rem inded of every time I fly over my M innesota study area and look
at lakeshores that were speckled with d eer and wolves in the late
1960s, and that now lie empty.
18. W hen prey herds are low for whatever reason, wolf control is
often proposed as a ready m eans of relieving pressure on them . T he
non-hunting public then usually responds with cries o f indignation.
It looks like wolves are being used as scapegoats. Because many o f
these people view wolf-prey systems as constantly in balance, they fail
to u nderstand how wolf control can aid prey recovery.
(Mech 1985: 57-8)
This text contains larger discourse patterns of organization which

can be examined in terms of subordinate, coordinate and superor-
dinate units. Paragraph 1 introduces the topic of discourse and the
intentions of the author (wolves, and the balance of nature). This
topic is presented as a problem (ratios of wolves and prey animals
can fluctuate wildly - and sometimes catastrophically). Paragraph 2
is coordinated with 1 in that it provides additive information -
appropriate background information on wolves. The third para-
graph is also additive information, being parallel to 2 and
coordinate with 1. It provides appropriate background informa-
tion on the author. As such, it legitimizes the authority of the writer
to speak on this issue. The final sentence of paragraph 3 refers
back to the ‘stability’ in paragraph 1 as a transition device to con-
tinue the hierarchical ‘problem’ organization of the text.
Paragraph 4 expands and quantifies the notion of ‘stability’ (the
balance of nature) of paragraph 3, and also of 1; as such, it is sub-
ordinate to 1. Paragraph 5 is an example of ‘imbalance’ or loss of
‘dynamic equilibrium’; it is subordinate to paragraph 4.
Paragraph 6 provides apparent support for the generalization at
the end of paragraph 5, but with signals that this view is simplistic.
This unit constitutes a transitional paragraph to a more complex
‘problem ’ re-analysis. Paragraphs 7, 8, and 9 develop a second line
of analysis where a second cause of imbalance (severe winter
weather) is examined. The discussion in these paragraphs refers
primarily back to paragraph 4, where the principle of ‘dynamic
equilibrium’ was first introduced. Paragraph 7 is parallel to 5 in
providing a second factor responsible for an imbalance, both fac-
tors having been initially m entioned in paragraph 4.
Paragraph 10 reaffirms the problem, and paragraph 11 reaf-
firms the immediate cause. Paragraphs 12 through 15 describe
parallel events in another location. Paragraph 16 suggests the solu-
tion to the problem. These paragraphs reaffirm the description in
paragraphs 7-9. The text continues until paragraph 18, where the
author once again states his intention to re-examine the notion of
‘balance of nature’. Here the author reaffirms the topic of dis-
course. This paragraph, then, is superordinate to all but the first
three paragraphs of the text.
This text exemplifies some of the logical patterns of textual
organization which form a part of the text’s coherence. The para-
graph relations could be graphed using a discourse bloc analysis, a
clause relational analysis, a rhetorical structure analysis, or a
propositional analysis to examine the different relations operating

in this text (cf. Britton and Gulgoz 1991, Hoey 1991, 1994, Kaplan
1972, Mann and Thompson 1988, 1992). Alternatively, the text
could be diagrammed in terms of the cohesive relations linking
the surface forms with those which support coherence relations.
The third major subcomponent, information structure (see
Chapter 2), is not as clearly recognized as a part of coherence. In
particular, research on information structure is confused by the
wide-ranging sets of definitions for the assorted terminology.
Information structure appears, however, to be centrally involved
in the creation of coherent texts. Information structure implicates
the issue of how linearized texts signal rate of information flow,
amount of information, and the relations involved in the ongoing
flow of information. At issue are questions concerning:
■ how more important thematic information is highlighted in
texts;
■ how given (old) and new information are signalled;
■ how particular aspects of information may be presupposed in
relation to what is expressed; and
■ how too little new information or too much new information
(which can only be defined with respect to some intended/
assumed audience) affects the coherence of texts.
All of these issues of information structuring are constrained by
the linear ordering of texts as well as by how rapidly the author
wants to present any given body of information, and from what
perspective. A careful discussion of the linearity of texts is found
in de Beaugrande (1984).
Finally, the role given to inferencing systems in the construction
of text coherence needs to be considered. While these are not
specifically aspects of the text itself, such systems interact with tex-
tual structure to constrain the w riter/reader in determining the
limits on interpretation of text meaning. These systems, in their
turn, must be constrained by the mechanisms of text structure if
appropriate inferences are to be made. For example, a common
distinction made in text inferences is between bridging and elabora-
tive inferences in comprehension of texts. Many researchers now
assume that most inferences are elaborative and are not part of
the basic comprehension cues directly signalled by texts. Some
inferences, however, are required to create connections between
the new information and the information already stored. This sort
of inference, a bridging inference, is assumed to be produced and

stored as part of the basic text analysis (Oakhill and Garnham
1988, Rayner and Pollatsek 1989). These findings would suggest
that inferencing, as a coherence-creating mechanism, is strongly
constrained by the structure of the text, and defining the concept
of coherence in some manageable way seems to be crucial to any
understanding of how texts are constructed.
3.4 Fimctional-use dimensions o f texts
There must be a component of text construction which identifies

the dimensions along which text elements are functionally orga-
nized to create the text. This dimension may be construed as the
interpersonal level of text construction, though more may be
involved (see, e.g., Halliday 1985). The functional/style level
explores the correlates of the writer’s attitudes to the reader, to
the subject matter, to the situation, to world knowledge, and per-
haps, reflexively, to himself. This level of analysis also implicates
the parameters of appropriateness within the text; e.g. which
structural elements match which textual contexts - after all, one
does not expect to find direct speech structures in a technical
manual. This level of analysis should not simply be considered a
matter of style, though it may be closely related to style. Style
reflects the personality of the writer. The components discussed
here represent the language parameters a writer uses to manipu-
late the text for various purposes; such manipulations, in and of
themselves, may not represent all of a writer’s personality, nor is
writer’s personality all that such manipulations represent.
Unlike the first five text-construction components, this compo-
nent is not concerned with elements of text structure, but rather
is concerned with how the elements are combined to form partic-
ular texts. A contribution to textual-dimensions from the field of
rhetorical studies has been the establishment of a traditional clas-
sification system for the logical organization of texts. More
recently, linguists have been exploring ways to examine texts for
textual dimensions which organize the text without presuming tra-
ditional top-down patterns of text classification. Dillon (1983)
suggested that the interpersonal dimension of texts could be
understood in terms of five dimensions of stance, representing
the relation of writer to the text and to the reader:
■ personal-impersonal
■ distance-solidarity
■ superior-equal
■ oblique-confronted
■ formal-informal.
While these five parameters represent simply one attempt to
define interpersonal dimensions in text construction, most have
an empirical reflection in the other research (see, e.g., Biber
1988, Scollon and Scollon 1983). It should be noted that the supe-
rior/equal dimension is more commonly discussed in terms of
parameters of equal/unequal or politeness/power.
Dillon does not intend this set of parameters to be interpreted
as style options, as in Joos (1967), but rather as a set of basic com-
ponents defining any text (see alignment in Tierney and Pearson
1983: 572-6). Dillon refers to the parameters as the ‘social sig-
nalling functions in a text’. The real issue is how to use these
parameters to establish measurable aspects of text structure.
Chafe (1982), in an early attempt to address this issue and
define dimensions of text structure, suggested that texts may be
classified according to how they vary along two dimensions of text
construction: text involvement/detachment and text integra-
tion/fragm entation (see Tannen 1987, 1989). More recent
research by Biber and his colleagues (Biber 1988, 1989, 1992,
1995, Biber and Finegan 1988, 1989, Grabe 1987) suggests that a
num ber of textual dimensions are constructed from the complex
co-occurrence patterns of lexico-syntactic elements. Biber (1988),
in his most comprehensive study, defines seven textual dimensions
underlying spoken and written texts. From this research, he has
proposed that a theory of text types must account for these
dimensions: types that are empirically identifiable, rather than
determ ined a priori. Following this line of research, the text con-
struction model proposed here includes a component of stylistic
use - the textual uses of linguistic structures to create communica-
tive dimensions of text structure. Nine such dimensions of text
structure, each of which receives support from text analysis
research, are proposed:
1. Rhetorical intention reflects that dimension of text construction
whereby texts are constrained by the top-level logical structur-
ing of texts (Meyer 1984, 1987).
2. Interactivity is a textual dimension which combines various
features to convey relative interaction/involvement between

writer and reader through the text (Biber 1988, Grabe 1987).
3. Referentiality describes the degree and type (logical, text-
internal reference or reference to situation) of referring which
takes place in texts and can be related to the ‘endophoric ver-
sus exophoric reference’ distinction of Halliday and Hasan
(1976, 1989).
4. Immediacy of context reflects the relative use of different tempo-
ral signalling to define text information - it typically delineates
primarily narrative texts from other text types (Biber 1988,
Grabe 1987).
5. Suasion refers to the use of features which signal a writer’s
attempts to persuade the reader, most typically indicated by
modals and certain subclasses of verbs (Biber 1988).
6. Abstractness reflects the relative degree of neutrality or objectivity
appearing in a text; texts are marked for technical and formal
styles versus other types of textual style (Biber 1988, Grabe 1987).
7. Elaboration refers to the degrees and types of elaboration which
occur in different types of texts; it can be equated with the
notion of constraints on planned versus unplanned discourse
(Biber 1988, Ochs 1979, Reppen and Grabe 1993).
8. Evidentiality refers to the need for writers, in careful profes-
sional prose, to indicate their degree of commitment to the
factuality of what is being stated (Biber 1988, Chafe and
Nichols 1986).
9. Text type (genre) refers to the variation to be found, more gener-
ally, among different text types and the ways that textual
features combine to define these types (Biber 1989, Dudley-
Evans 1989, Martin 1985, 1992, Swales 1990).
Each of these dimensions (and there may be others as well) repre-
sents ways that texts are shaped and constrained according to the
various purposes of the writer and the demands of the context. A
model of text construction needs to account for these dimensions
operating on the creation of texts.
3.5 Non-linguistic knowledge
The final com ponent of the text construction model provides the
w orld/background knowledge for appropriate interpretation and
production of text. Certain aspects of this knowledge will be

encoded in the lexicon though there is also much more informa-
tion of a non-linguistic nature that must be accessible as well.
There are a num ber of non-linguistic factors which must be
included in a model of text construction:
■ reference
■ world background knowledge (and intertextuality)
■ memory
■ emotion
■ perception
■ intention
■ logical arrangem ent (deduction, etc.)
■ situation.
All of the members of this set have powerful influences on aspects
of texts but are essentially independent of the linguistic domain of
the text model. Reference, for example, is the set of abilities func-
tioning to connect cognition to the real world - an activity that
can be carried on in the absence of verbal language. The fact that
one of its typical manifestations is linguistic does not make the
entire notion linguistic. Similar arguments can be made for emo-
tion, perception, and intention. World background knowledge and
memory, and situation (which perhaps cover the same territory),
are sets of schemata or frames for organizing content and context,
and are not dependent on language; it would appear reasonable
to imagine a set of knowledge in the absence of language, and
there is considerable psychological support for such a perception
(Cohen 1983, Johnson-Laird 1983, Paivio 1986).
3.6 The overall text model
The components for a text model outlined above should not be

seen as an accurate, fully formed model, nor even as a formal
attempt at psychologically valid model building; that would
require specific processing mechanisms and an explanation of
exactly how the pieces interact in the way they do. Such an explicit
model is beyond current understanding of text construction.
Rather, what has been provided is a descriptive model of the writ-
ten text which attempts (1) to include what must be considered in
a theory of text construction, (2) to explain the reasons why the
various components are important, and (3) to suggest ways in

which future research might explore component interaction in
texts.
With this goal in mind, theories of text construction would seek
to explain what is involved in the linguistic-production aspects of a
cognitive model of the writing process (e.g. de Beaugrande 1984,
Martin 1992). The model by Flower and Hayes of the writing
process does not provide a specification of what is involved in text
construction, but it would require such a specification if it were to
define specific text-construction research questions. By the same
token, any text-construction model would also need to be comple-
m ented by a processing model for a larger theory of writing. Such
a processing mechanism would describe how the various compo-
nents of text production are combined in the creation of text. For
the moment, however, the model described here does not need to
specify how text information is combined and created; its main
concerns are what is combined for which purposes. The text-
comprehension description by Garnham (1985, also Oakhill and
Garnham 1988) a n d /o r the processing models of Bereiter and
Scardamalia (1987), and Flower and Hayes (1981a, 1981b) pro-
vide appealing descriptions which are compatible with, though
distinct from, the text-construction model presented in this chap-
ter. This conception of text is represented as Figure 3.2.
The seven components represented in the figure (syntax,
semantics, lexicon, cohesion, coherence, functional dimensions,
and non-linguistic resources) form the basis of the text-
construction model. Each component should represent a signifi-
cant aspect of text structure or of text-structure constraints in its
own right. W hether this may be so can eventually be tested by
quantitative methods. This first model also strongly suggests that
earlier research, depending on only one or two components of
text, evokes the tale of the seven blind men and the elephant. At
the same time, it is im portant to recognize the basic nature of
texts as interactive, rather than simply componential in nature;
that is, the whole is greater than the sum of its parts in isolation.
That is why a componential approach (like that used in sentence-
level analyses) is insufficient. One cannot simply assemble the
pieces and thereby create the whole text.
There is an emergent quality to texts that does not allow for the
simple adding of the parts (Bateson 1979, Dillon 1981). An analo-
gous situation may well be represented by the chess board; that is,
Linguistic
boundary
SURFACE DEEP
Grammatical
features Ideational Reference
(a sentential level) Syntax Semantics
L World background
I knowledge
E
T Memory
Functional ,.:.:::::... :::..'................. ,.. ::<:'........... X
features Textu;t( ' I .r---t-t- Emotion
(a whole~text level) Cohesion : I Coherence ~
I Perception ~
.."""./ C a
l. Intension (pragmatics?) ~
0 ;!
.................... , .. I Logical arrangement c
Stylistic ./.::~:.:~~::::::: ... ...:::,
: (Rhetorical patterns)
features Interpersonal ,~ ~
(a writer·reader level) Posture Stance
Situation ~
1;t
t ~
Motivations 8
for i;;
form
~
~
§'
00
Figure 3.2 A model of text construction ......
the text can be conceived as a chess board on which the force field
of each piece is visible. Many pieces have multiple, though not
limitless, options; certain patterns of organization become recog-
nized generally as more appropriate in certain contexts, and each
piece interacts with every other piece in ways and combinations
that do tend to become limitless.
For texts, the results of the interacting elements are not only a
num ber of one-to-one effects but also of combinations of groups
in relation to any one, etc., forming unforeseen relations (and
creating real ‘covert’ categories in the language; see, e.g., W horf
1941). The text is a multifactored, multidimensional field, created
out of identifiable components, but not fully reducible to them.
Perhaps this is why writing is so much a practised art.
3.7 Conclusion
The purpose of this chapter has been to present a model of text

construction - one that would be compatible with im portant cur-
rent research on writing and would also conform to research on
prose comprehension and written discourse analysis. In all, seven
components of a model of text construction have been presented
in Figure 3.2. Each provides an essential element for any consider-
ation of how texts are created. The insights into prose
organization discussed in Chapters 2 and 3 provide much of the
support for the model presented here. These notions have been
elaborated and combined into a descriptive model of text con-
struction. Because the goal of such a model of text construction is
to understand the components of text, and at some point to
understand how they interact, this model fits within a more gen-
eral research objective which is to define properties of written text
in real use.
At this juncture, it is also important to recall that text construc-
tion represents a model of the text as product. In the next
chapter, the foundations for a theory of writing as process will be
explored. A mature theory of writing will have to account, mini-
mally, for both the text as product and writing as process, not as
independent concepts but as complementary perspectives on the
same phenomenon. As we shall see in Chapter 8, a viable theory
of writing will also have to incorporate an interpretation of social
contexts and their effects on writing.
Note
1. It m ust be understood th at the gram m atical m odel which is adopted

for research will determ in e the structures available to be analysed; that
is, gram m atical theories produce models which are, in K uhn’s sense,
self-defining. For exam ple, if one adopts an early transform ational
generative gram m ar - one utilizing kernel sentences an d a phrase-
structure gram m ar - the kernal structures will define what can be
examined, as they do in H u n t’s T-unit analyses (see C hapter 2). At the
same time, the m odel taken as the standard will also define the m eans
of analysis; that is, to the extent that certain gram m atical functions are
foregrounded in the m odel and others are m arginalized, the m odel
will define how various structures can be examined. T he p o in t is simply
that the choice of a gram m atical m odel is n o t a neutral choice.
4
Writing process approaches
4.1 Introduction
At the close of Chapter 1, the issue of the writing product versus

the writing process in composition research and instruction was
m entioned as perhaps the controversy of the 1980s. The develop-
m ent of the writing process approach is now widely recognized as
the major impetus for the emerging field of composition research
and also for composition studies as a graduate field of study.
Such a development - with its emphasis on tertiary-level com-
position - may be unique to the USA, as most research on
writing in other English-speaking countries has focused prim -
arily on elem entary and secondary students (e.g. Boomer 1985,
Britton et al 1975, Christie et al 1989, Wells 1986, Wells and
Chang-Wells 1992). It should also be noted at the outset that
much current research on writing in a L2 is based directly on
theoretical and instructional trends in writing-as-a-process
theory. In fact, current perspectives on writing-as-a-process in L2
contexts can only be understood properly when seen as an out-
growth of the process movement more generally (Hudelson
1989b, Johnson and Roen 1989, Kroll 1990, Leki 1992, Raimes
1991, Zamel 1987).
Up to just past mid-century, writing instruction in the USA was
approached in a rather uniform way. Students, who often did very
little writing (see, e.g., Applebee 1981), would learn to outline
and write ‘themes’ based on the four major ‘rhetorical’ distinc-
tions - description, narration, exposition, and argumentation -
and often framed in a three- or five-paragraph format. Through
expository assignments, students were taught such patterns of log-
ical arrangem ent as definition, classification, comparison and
84
Writing process approaches 85
contrast, etc., and told to write an essay imitating (or approximat-

ing) a given pattern. Essays were written in one draft, and errors
were corrected by keying to a writing handbook, these in turn
primarily providing answers and exercises on various aspects of
surface grammar. Ideas for writing topics were extrapolated from
literary source books and ‘readers’, containing models for analysis
and imitation or from lectures delivered by the writing instructor.
Even research writing was organized around published sets of
selected primary materials. Generally, the focus in writing classes
was on the form of the written product rather than on how the
learner should approach the process of writing.
In the 1960s, a num ber of factors coalesced to generate a strong
reaction to this instructional profile:
1. US tertiary institutions, during the 1960s, underwent a period
of extraordinary expansion, and as a result of a liberalizing
movement many schools adopted a policy of ‘open enrolm ent’;
that is, they admitted any student who applied, and, in many
cases, used ‘gatekeeping’ courses - such as composition - to
eliminate students from the institution.
2. As a result of open enrolm ent policies and equal-educational-
opportunity demands, many more minority student were
admitted into tertiary institutions; the minority students were
often eventually those who were screened out.
3. The concern among some tertiary institutions over the high
minority drop-out rate, as well as the growth in the num ber of
two-year community colleges throughout the USA, led to an
increasing focus on remedial/basic writing classes for high-risk
students. Many institutions began to test students’ writing abil-
ities on arrival and to place lower-ability students in remedial
writing classes before permitting them to enrol in the required
‘norm al’ first-year composition classes.
4. Teachers in remedial writing and freshman composition classes
became dissatisfied with traditional approaches to writing
instruction, recognizing that the conventional approaches may
have been acceptable in a system of education designed for the
culturally homogenous élite and middle class but were now
essentially irrelevant for students with vastly different life experi-
ences. Instructors in tertiary-level institutions began to move
away from a focus on purely expository essay writing and liter-
ary criticism as the content (or non-content) of writing courses.
5. Tertiary-level institutions moved to become open forums for

inquiry rather than bastions of authoritative knowledge.
6. At the same time, researchers noted a downward trend in
nationwide test scores which were believed to index literacy
levels; that slide in standardized test scores fuelled a general
concern in the 1980s over the perceived erosion of literacy
skills among students.
7. In the 1960s the psychological and philosophical foundations
of human skills research began a major shift in the direction of
what is now defined as cognitive psychology (Gardner 1985); its
major underlying assumptions - (i) that the inner workings of
the mind can be studied, (ii) that complex skills are made up
of interacting components working together in complex pro-
cesses, and (iii) that learners are creative hypothesis generators
- led researchers to view language skills, particularly reading
and writing abilities, as legitimate domains for theoretical
research.
All of these trends, together, have led teachers, researchers, and
evaluators to reassess the nature of writing and the written
medium, and the ways in which writing is learned and taught.
While the outcomes of this reassessment are many, the focus here
is specifically on the rise of what has been popularly designated
the xvriting-as-a-process movement.
As a liberating concept, writing-as-a-process did much to change
the general perceptions of writing instruction and of the ways
students learn to write. An obvious sign that this major change has
become well established is the instantiation of this innovation in
the textbook market.1
The writing-as-a-process approach, as a reaction to earlier
instruction, freed instruction from:
■ the three- or five-paragraph model;
■ simplistic assumptions about the organization and ordering of
information;
■ the typical one-draft writing assignment;
■ the assumption that each student should be working alone, or
only with the instructor on summative feedback;
■ reliance on gram m ar/usage handbooks and lectures;
■ the linear composing model based on outlining, writing, and
editing;
and it freed instructors from imposed, artificial topics for writing.
In the place of these previous practices, the process approach

encourages:
■ self-discovery and authorial ‘voice’;
■ meaningful writing on topics of importance (or at least of inter-
est) to the writer;
■ the need to plan out writing as a goal-oriented, contextualized
activity;
■ invention and pre-writing tasks, and multiple drafting with feed-
back between drafts;
■ a variety of feedback options from real audiences, whether from
peers, small groups, a n d /o r the teacher, through conferencing,
or through other formative evaluation;
■ free writing and journal writing as alternative means of generat-
ing writing and developing written expression, overcoming
writer’s block;
■ content information and personal expression as more import-
ant than final product grammar and usage;
■ the idea that writing is multiply recursive rather than linear as a
process - tasks are repeated alternatively as often as necessary;
and
■ students’ awareness of the writing process and of notions such
as audience, voice, plans, etc.
The process approach is frequently discussed as a wholly positive
innovation allowing teachers and students more meaningful inter-
action and more purposeful writing.
In fact, writing-as-a-process has become so popular that some
researchers (e.g. Hairston 1982) claimed a paradigm shift (as
explicated in Kuhn 1970) in writing theory and instruction (thus
requiring an independent theoretical foundation). The fact that
this movement has developed so strongly and remains so popular
testifies to the fact that the writing process perspective has cap-
tured certain im portant truths about language. Like any
movement in education which asserts itself as the new truth, the
writing process approach has been the object of careful scrutiny in
recent years, leading to serious criticism, much of it well justified.
By the 1980s, the first critiques of this new orthodoxy began to
emerge and many of the writing process claims have now been seri-
ously questioned or qualified. At the center of much criticism is the
assertion that the writing process approach is an instructional
notion without a strong theoretical foundation, a foundation which
it has claimed to have as it attempted to revolutionize writing (see,

e.g., Hairston 1982, North 1987). In retrospect, it is evident that the
writing-process approach fares about as well as most other new per-
spectives; it offers significant insights, but it claims to offer much
more (cf. Applebee et al 1994). An understanding of writing-as-a-
process is best achieved by reviewing the major developments in the
movement over the past 30 years. In this way, the excesses may be
separated from the genuine strengths and insights.
The history of writing process approaches can be divided into a
rough set of four stages from the 1960s to the present. First pro-
posed in the mid-1980s (Faigley 1986), this perspective remains an
effective analytic scheme: Each stage represents new insights and
responds to difficulties identified in the preceding stage. These
stages can be outlined as follows:
1. The expressive stage
2. The cognitive stage
3. The social stage
4. The discourse community stage.
It is im portant to note that these stages do not represent general
historical transitions for all practitioners; in fact, many researchers
and instructors still prefer to use an expressive approach to writ-
ing instruction. Others, for example, still prefer a cognitive
approach without serious consideration of social-context factors in
their instruction.2
4.2 The expressive approach to the writing process
The expressive stage of the writing process can be traced back, in

recent times (cf. Myer [1985] for an historical review of process
claims asserted more than 50 years ago), to the 1960s and the
arguments put forth by Elbow (1973, 1981), Macrorie (1970,
1980), Murray (1968, 1980, 1985) and others, urging that writers
look for their authentic voices and be able to express themselves
freely. The goal, a romantic one, as both Berlin (1987) and Faigley
(1986) note, was to produce writing that was fresh and spontane-
ous and had integrity. Writers should say what they really thought;
they should be creative and take chances. Writers should let their
natural voices speak out.
North (1987) refers to the leading figures of this expressive
movement as writing practitioners guided primarily by pragmatic

insights into the nature of good writing and writing instruction. In
essence, their advice amounts to recounting what worked for them
as good writers and what should, consequently, work for others in
the classroom. As North notes, this approach to writing, either as
theory or as practice, had no guiding theoretical foundation which
could differentiate lesser insights from better insights or determine
how, when, or why pragmatic advice on writing instruction would
transfer to other contexts and other teachers. Relatively little atten-
tion was given to why these insights were appropriate as opposed to
what to do in a classroom, and how to do it.
Elbow, in his influential books on writing, added a touch of biol-
ogy - writing was to be seen as an organic, if not mysterious,
process, springing from a creative source; it was the writer’s
vehicle for self-expression.
However, despite the mysterious, romantic underpinnings of
the expressive movement, it should also be noted that the practi-
cal lore on writing instruction devised during this period
proposed great changes from the more traditional approaches to
writing. Many of the writing-as-a-process concepts explored in
later research have their origins in practical advice given to writ-
ing instructors by these expert practitioners.
The major problem with such an approach to writing is that it
assumed that the writer already has all the intellectual resources
he or she would need and was merely looking for an appropriate
outlet for expressions. It essentially ignored the context of writing
and the social context in which writing is perform ed in the real
world. Perhaps, more importantly, it assumed that the cognitive
processing of immature writers and expert writers during the writ-
ing process operate in the same manner, though to different
degrees of refinem ent (see, e.g., Bereiter and Scardamalia 1987,
Moffett 1968). Nevertheless, the neo-romantic assumptions of this
expressive view, combined with its weak theoretical methodology,
served at least to open the door to more scientific approaches to
writing-as-a-process.
4,3 The cognitive approach to the writing process
In the early 1970s, a competing process approach to writing - psy-

chologically based - arose out of research in cognitive psychology.
Janet Emig’s (1971, 1983) pioneering efforts at case study

research and protocol analysis represented a breakthrough for
writing research - a more scientific way to study the writing pro-
cess, and to see what writers were actually doing, at least on the
surface, when they were writing. She legitimized the case study
approach, the think-aloud methodology, the study of pauses, the
role of re-reading in revision, and the amount and type of revision
among writers (Faigley 1986: 532). On a more practical level, her
research led to a view of writing as recursive rather than linear;
she called attention to the importance of pre-planning and edit-
ing as ongoing activities, and to the importance of writer’s errors
as a source of data (cf. critique by North 1987). As such, she was a
precursor to a range of studies classifiable as clinical research
(North 1987), particularly the influential case study research of
Perl (1979), Selfe (1981, 1984), and Sommer (1980), and the
study of writer errors by Shaughnessy (1977). At the same time,
new interest in the composing of elementary and secondary
students, led by Calkins (1986), Graves (1983, 1984), and others,
opened up research using case study approaches but also, more
generally, ethnographic research methods. From the work of
Emig and other earlier researchers using the case study approach,
a theoretical dilemma arose. What Emig had not provided was a
theoretical foundation for the sort of research she had developed
(Hillocks 1986, North 1987). To fill this theoretical lacuna, writing
researchers turned to cognitive development/Piagetian theory,
audience awareness theory, and real-writing situations.
Researchers began to see writers as developing from egocentric
stages of writing, initially composing writer-based, rather than
reader-based, text (Calkins 1983, Graves 1984, Kroll 1981, Moffett
1968). Given that writing is an exceedingly complex task, and
given the voluminous research findings that have appeared in the
literature, it became clear that some attempt should be made to
provide a coherent framework - a model that would account for
the data reported, and would explain conflicting findings and
interpretations. The goal was to draw together findings from many
types of research and from many kinds of evidence, using perspec-
tives from many disciplines. During this time, two teams of
researchers who had figured prominently in the 1980s, began to
publish results of experimental research on the writing process,
developing cognitive models of the writing process: Flower and
Hayes (from 1977 on), and Bereiter and Scardamalia (e.g. 1987).
Beginning in the late 1970s, Flower and Hayes (1977, 1980a,

1980b, 1981a, 1981b, 1984) developed a cognitive model of the
writing process which attempts to provide a synthesis of research,
and which has been dom inant for the past 15 years in composition
research. Strongly influencing the writing-as-a-process movement,
Flower and Hayes have asserted that:
■ composing processes are interactive, intermingling, and poten-
tially simultaneous;
■ composing is a goal-directed activity;
■ expert writers compose differently than novice writers.
These three hypotheses have become basic premises of a theory of
the writing process. The research methodology proposed by
Flower and Hayes has been based on protocol analysis, drawing
legitimacy from the protocol analysis methods of H erbert Simon
and colleagues (Ericsson and Simon 1984, Newell and Simon
1972), a major research approach in cognitive psychology for
examining goal-directed behaviour. By collecting and examining
protocols, transcripts, and videotapes of students talking aloud
while they are writing, Flower and Hayes presented data which
supported their perspectives on composing. (For explanation and
discussion of protocol analysis in research, see Hayes and Flower
1983, Lauer and Asher 1988, Smagorinsky 1994.) From their
earlier efforts, they subsequently developed a model of the com-
posing process (e.g. 1981a), a processing model based on their
protocol research and fashioned closely after the Artificial
Intelligence (Al) processing models created by Simon and others
to explain problem-solving routines.
Figure 4.1 displays their model as it is most commonly repre-
sented. The model divides the composing processes of a writer
into three major components: the composing processor, the task
environment, and the writer’s long-term memory. Within the composing
processor, three operational processes generate the written text:
planning\ translating, and reviewing The three processes are man-
aged by an executive control called a monitor. Finally, in the
planning process, there are three subcomponents - generating
ideas, organizing information, and setting goals. In the actual genera-
tion of text, the ideas in planning are translated into language on
the page, which is then reviewed and revised. This model of writ-
ing has remained essentially unchanged since it was first proposed
in the late 1970s (cf. Flower 1994).
Figure 4.1 T he Flower an d Hayes writing process m odel
This theory of the writing process has generated much excite-

m ent and research using protocol analysis. It has also generated
criticism on a num ber of levels. Some researchers have noted that
writers are not likely to be uniform with respect to their process-
ing preferences and cognitive abilities. Rather, writing involves
numerous processing-model options, and different writers will
approach the task employing different processing strategies.
Further criticism of their influential model has centred on the
elaboration of the model itself. North (1987) argues that the
model as proposed by Flower and Hayes is much too vague to sat-
isfy criteria for formal model building; specifically, he notes that
the translation from the model to people actually writing is barely
explained. For example, there is virtually no specification of how
the text material might be constructed and what linguistic con-
straints might be imposed on this construction (see Chapter 3 of
this volume).
A final critique of the Flower and Hayes model is methodologi-
cal, and therefore one that strikes at the heart of their theory.
Essentially, it is argued that a protocol analysis approach may not
be a valid primary methodology for the study of the writing pro-
cess to the extent that Flower and Hayes claim (Dobrin 1986; cf.
Smagorinsky 1994). A more moderate perspective is that protocol
analysis can reveal certain im portant things about what writers do,
but it cannot be the primary source of evidence for a theory of the
writing process. A balanced discussion of this issue is provided by
Bereiter and Scardamalia (1987: 42-4, 195-6) who point out that
evidence from any source must be controlled, reliable, and con-
vergent with findings from other experimental sources if it is to be
persuasive (Scardamalia and Bereiter 1986: 724).
Despite the many criticism the writing process model developed
by Flower and Hayes must be credited with im portant findings
resulting from their research program:
1. They have raised a new range of issues for public debate.
2. They have raised understanding of recursion in writing to a
new level, and provided a perspective which is not presently
controversial. Essentially, they have brought to writing the
notion of interaction among processes in parallel, echoing
recent developments in reading research (see, e.g., Grabe
1988a, Stanovich 1980, 1986, 1992).
3. They have attempted to model writing processes, and thereby
have opened writing research up to more explicit claims, more
explicit and testable hypotheses (even if not by protocol analy-
sis), and more carefully defined research methods.
Following the work of Flower and Hayes, other efforts to synthe-
size research findings into a coherent model have benefited from
their efforts, though not all have attempted the same sort of syn-
thesis. A model by Cooper and Matsuhashi, appearing in 1983,
drew extensively from research on the text product to infer the
processes required for expert text production. Their model repre-
sented an ideal, and rather informal, description of the
composing process - one which relied more heavily on findings
and research in rhetoric and text linguistics than it did on obser-
vational studies of writers composing.
A third processing model, which draws together a wide array of
information, is the model devised by de Beaugrande (1982, 1984).
He proposed a descriptive account of the features that must be
included in a model of the writing process. His analysis draws
heavily from research in cognitive psychology, particularly the
research on memory, attention, and reading. His model provides
a coherent global argum ent for what expert writers do, and
extrapolates from the actions of children and poor writers, much
as Flower and Hayes have done. Although de Beaugrande’s model
is highly detailed and well supported by research in cognitive psy-

chology, it is still essentially a descriptive global model. It is not an
operational model making specific falsifiable predictions.
While the various models of composing all appear to have limi-
tations, they have all moved the debate on writing research
forward and have led to more refined analyses (cf. discussion of
more recent research by Flower and Hayes in Chapter 5). The
gradual development of model construction for composing pro-
cesses has led to the more current powerful model proposed by
Bereiter and Scardamalia (1987); their theory represents a major
advance in the understanding of what writers do when they write
and why different writers write in different ways. They provide a
more explicit theory which accounts for many different research
perspectives and sources of evidence in a coherent explanation.
Their theory accounts for what children and beginning writers do,
for what poor writers do, and for what expert writers do. It also
makes explicit predictions as to why writing differs from person to
person - predictions which can be confirmed or refuted. In this
sense, then, their model is not merely descriptive, but rather
explanatory to the extent that it is more easily verifiable by experi-
mental evidence.
4.4 A social-context approach to the writing process
Beginning in the 1980s, socially oriented views of writing

developed from a num ber of different sources, among them
sociolinguistics, Hallidayan functional linguistics, elementary edu-
cation research, socially based rhetoric, and the sociology of
science. These perspectives on literacy, knowledge, and writing
found a receptive audience among writing researchers who were
troubled by the educational status quo, Piagetian developmental
psychology, and the limitations of protocol analysis. Rather, the
new groups of researchers have proposed that a writing-as-a-pro-
cess approach has little meaning outside of the social context
which defines the particular writing purpose, a notion as applic-
able in the classroom as it is in the real world (e.g. a Vygotskyan
perspective). The essential point, as argued by Cooper (1989),
Faigley (1986), and Witte (1992), is that writing can only be
understood from the perspective of a social context and not as a
product of a single individual.
This general trend in writing research is a product of recent

years and is, understandably, not yet well formed in terms of a con-
sistent theory of the writing process (cf. Flower 1994). In fact,
there are a num ber of distinct perspectives within a social view of
the writing process: one from the perspective of educational eth-
nography (being more educationally oriented), a second from the
perspective of sociological linguistics (more sociolinguistically
oriented), and a third from the perspective of discourse commu-
nities (more theoretically oriented). Yet another perspective is that
from the sociology of science, treating the production of writing as
also being creative of a social context. This latter perspective will be
treated in Chapter 6 on writing in specialized contexts.
4.4.1 Ethnography in educational contexts

Ethnographic research in education, and particularly in the study
of writing, has emerged in the past 20 years, largely from the eth-
nomethodology and participant-observation research in sociology,
sociolinguistics, and anthropology (e.g. Cazden 1988, Heath 1983,
1986a, Spindler and Spindler 1987b). A researcher assuming this
methodological orientation eschews theoretical preconditions
and observes what is actually occurring without imposing an a
priori framework on the observations. The observer participates in
the community over a period of time so that the community
becomes familiar to the researcher. The ethnographic approach
makes an effort not to be biased by self-fulfilling theoretical mod-
els which may distort the observed reality, and to collect naturally
occurring data in its social context rather than to create and con-
trol data artificially.
From an ethnographic perspective, writing, much like speaking,
loses its purpose outside of its naturally occurring context. To
know how and why people write, it is necessary to see how they
perform under normal, natural conditions; and to see how chil-
dren and basic writers develop writing skills, it is necessary to
observe them in the process of their development without inter-
fering (Graves 1984, North 1987). The importance of this
perspective cannot be overestimated; too often, theoretical mod-
els emerge before there is adequate data on the range of
phenom ena for which the theory can account.
Ethnography applied to educational contexts and literacy
instruction:
(1) recognizes and studies the social contexts in which language

occurs;
(2) assumes that different language uses occur in different con-
texts; and
(3) attempts to interpret/bridge the uses of different languages
in their contexts, including the usually rigid demands
imposed on all students by the formal educational system
(e.g. Boggs 1985, Tharp and Gallimore 1988), and it investi-
gates language in terms of these assumptions.
In educational settings, the first group of ethnographic studies
focuses specifically on the writing tasks of students in classes as
they occur during the course of the day. Researchers such as
Calkins (1986), Graves (1983, 1984), and Harste et al (1984) are
all representative of this line of investigation. Because these stud-
ies are more in line with case studies and task intervention
approaches, they should be seen as distinct from more main-
stream ethnographic approaches, represented by the second
group of studies discussed in section 4.3.
4.4.2 Writing research by Graves and his colleagues

Graves (1984) first developed the use of ethnographic case studies
to study writing in classrooms in his 1973 dissertation on the writ-
ing of 7-year-old children. Since then, he and his colleagues have
studied the writing of children in Europe, the USA, Canada, and
Australia. Based on observational research projects of students’
writing typically extending a year or more, he has argued that chil-
dren are able to, and want to, write from the day they enter the
classroom. It is the educational system which alters this natural
tendency by providing little positive assistance to take advantage
of students’ desires. Rather, in most schools, writing is given very
little time, and students are not encouraged to write. Teachers, for
their part, typically do not know how to teach writing and seldom
write anything themselves; even less frequently do teachers model
writing in the classroom, and then share the writing with students
for their comments and feedback. Educational funding in the
USA has also, until recently, ignored writing issues. Graves (1984),
in particular, notes that there is a striking disparity in funding for
research on writing as opposed to research on reading at the pri-
mary school level.
Finally, many of the traditional uses of writing in school have

negative associations for most children: writing tasks are used for
disciplinary purposes, for testing and evaluation, for busywork -
artificial assignments with little relevance for students; and when
writing is produced, it is returned with teacher feedback in the
form of extensive correction of surface features. This dishearten-
ing picture of writing, particularly in US elementary school
classes, has led Calkins (1983, 1986), Graves (1983, 1984), and
others to develop participatory ethnographic-research approaches
to elementary school writing. From the results of these research
studies they have proposed a num ber of directions for instruction:
1. Writing is a process of discovery; its development progresses as
a problem-solving activity.
2. Drawing and talking are means for pre-writing and rehearsal.
3. Students learn to internalize their conversations to the point
where they do not simply write down what they say - they learn
to rehearse better.
4. Students learn to move from writing for themselves to writing
for others.
5. Students learn that they need to spend time on planning, pre-
writing, and rehearsal activities to improve their writing.
6. Students learn that they are capable of serious revision by
themselves, even at lower levels of ability, particularly by confer-
encing with teachers about their writing.
7. Students learn that they need to have control over their writ-
ing.
8. Since schools do not create student motivation and involve-
ment, teachers need to generate a commitment to writing;
students need to be involved and motivated to write, and
teachers also need to practise writing in class and to share their
writing with children.
9. Teachers need to have control over their classrooms and
become active teacher/researchers with respect to their own
practice.
These findings have had a strong influence on the teaching of
early writing in Australia (Walshe 1981a, 1981b). They also par-
allel similar research by Clay (1975) for New Zealand and are
consistent with the Bullock Report (1975) in Great Britain leading
to the ‘Language across the Curriculum’ movement (Marland
1977). In the USA, these findings, as well as a range of other early
literacy research, have had a strong influence on the promotion

and development across the country of National Writing Centers,
the purpose of which is to promote the teaching of the writing
process. Teachers from different area schools in the USA come to
summer workshops to learn how to teach writing effectively. In
return, these teachers must hold workshops on teaching writing
for other teachers in their school. Thus, every year, more and
more teachers become involved in writing and devote more time
to writing in their classes.
It is inevitable that a research approach which has made num er-
ous claims would come under close scrutiny as others scholars
have time to reflect upon the work gradually appearing in publica-
tion. In particular, Bereiter and Scardamalia (1987), Hillocks
(1986), Lauer and Asher (1988), and North (1987) have com-
m ented to some extent on the work of Graves and his colleagues.
The five criticisms which follow represent the general arguments
against this particular groups’ use of ethnographic research.
1. Graves’ research perhaps should not be considered ethno-
graphic. There is very little description given of the overall
classroom contexts in which these observations have taken
place; rather, the studies only focus on the time spent writing
during the observations (cf. Heath 1983). This criticism is a
common one directed at many writing research studies which
claim to be ethnographic but which are, in fact, types of case
studies.
2. Naturalistic data, such as that which has been collected, may
not be a valid source for making larger generalizations. The
goal of ethnography is to tell a particular story, to provide a
detailed description of what goes on at a particular place and at
a particular time, and to provide such information as it occurs
with those particular students. Such a methodology does not
perm it ‘averaging’ across case studies (cf. Spindler and
Spindler 1987a, Wells 1986).
3. Graves and others have made a num ber of causal statements on
the basis of relations between two co-occurring behaviours or
objects of investigation. They give power to one variable as cau-
sal of the other. However, Graves does not appear to use any
systematic control across observation subjects, which is mini-
mally necessary for cause and effect assertions (Hillocks 1986,
North 1987).
4. Graves and his colleagues present quite minimal data in the

range of their published writings. It is difficult to verify their
claims and impossible to assess the depth of insight they derive
from their data; perhaps other researchers faced with a larger
segment of the data might arrive at different conclusions
(Hillocks 1986).
5. Nowhere in the research publications in question is the issue of
researcher intrusion given careful attention. However, it clearly
is an issue when researchers are intruding into the ‘natural’
classroom setting. Graves, for example, in his early research, sat
directly in front of particular students when they appeared to
be ready to write. He copied down the students’ words, coded
the students’ behaviours, and asked students questions about
what/why they were writing (Hillocks 1986, Lauer and Asher
1988, North 1987).
As damaging as these criticisms are - and they do raise serious
questions - it is also im portant not to underestimate this line of
research which constituted a pioneering effort in the study of
early student writing, and which has generated a num ber of
im portant hypotheses.
1. It has confirmed notions advanced by other researchers (e.g. by
Flower and Hayes).
2. It has generated many new research questions which remain to
be answered.
3. It has had a beneficial effect on the teaching of writing in US
elementary schools, though the extent of the benefit has yet to
be documented.
4. It has influenced and reinforced the more general language
arts approach known as the whole-language approach
(Goodman 1986).
5. It has stimulated discussion which can only lead to improved
research in the future.
Hillocks (1986: 60-1), in a carefully balanced commentary, has
this to say about the research of Graves and his colleagues:
Research on com posing has developed many im p o rtan t ideas, m ost
o f which m ust still be regarded as hypotheses - b ut hypotheses wor-
thy o f fu rth er exploration an d testing. Ideas developed by Graves,
Calkins, ... and their colleagues are fascinating examples. In their
work we see children beginning to write very early in the first year o f
school by using invented spellings, learning to use the space on the

paper, sharing work with o th er youngsters an d the teacher, shifting
from externalized to internalized behavior in revising, moving from
egocentric to sociocentric, an d so forth. Researchers should exam -
ine these ideas as rigorously as possible, presenting rules for the
analysis of the events observed, displaying data across cases, and
using controls to account for alternative explanations.
A considerable amount of space has been devoted to a review of

the work of Graves and his colleagues because of the widespread
influence this work has had both on other researchers and on
teachers.
4.4.3 Sociolinguistic research in the ethnography

of uniting
The second ethnographic line of research might be described as a
more ‘regular’ approach to the study of writing. Its impact on the
writing process can be traced to studies of social communities
designed to observe the match between (a) literacy at home and
literacy expected in schools, and (b) studies of classroom interac-
tion designed to observe teacher-student and student-student
language uses. Major studies of the match or, more often, the mis-
match, between home and school begin with Bernstein’s
sociological research exploring the influence of socialization pro-
cesses on the language registers used by different groups of
students, and how these register uses match the demands of edu-
cational language (Bernstein 1972a, 1972b, 1990). Based on his
research in the 1950s and 1960s, Bernstein argued that children
who are failing in the school system do not have adequate control
of the registers used and demanded by the educational system -
not because they are incapable of such uses but because they are
neither encouraged nor expected to use these registers in their
home socialization environment (Cope and Kalantzis 1993,
Freebody and Welch 1993, Hasan 1989, Lareau 1989, Wells 1986,
1994). His research opened the way for ethnographic research on
the educational mismatch between home comm unity/culture and
the demands of the formal educational system (see also Halliday
1978, Stubbs 1980, Wells 1986).
O ther research following in this direction includes at least work
by the following: Boggs (1985), studying Hawaiian children;
Cazden (1988), studying variation in classroom discourse; Cook-
Gumperz (1986), studying patterns of oral discourse interactions

on later literacy development; Heath (1983, 1986a, 1986b, 1993,
Heath and Mangiola 1991), studying variation in home and
school language uses in numerous contexts ; Philips (1983), study-
ing variation in home and school language use among Warm
Springs Indian students; Scollon and Scollon (1981), studying
variation in home and school language use among Athabaskan
Indian students; and Tharp and Gallimore (1988), studying the
applications of a Vygotskyan approach to literacy development
with Hawaiian and Navajo elementary school students.
A consistent axiom running through all this research is to find
out what students do, how they do it, and why, both in the formal
educational environm ent and in the community outside the
school. Only when these issues are understood can some deci-
sion be made about how writing develops differentially, how the
writing process works, and why and when it works appropriately
(e.g. Heath 1983). On the basis of this research foundation, a
writing-as-a-process approach can be proposed which takes into
account: (i) the context in which writing occurs, (ii) the possible
instructional contexts which have ecological validity, and (iii)
contexts which have the power to guide more effective writing
instruction.
Perhaps most influential in this line of research is the work of
Heath (1982, 1983, 1986b, 1993). In her major study (1983), she
examined the writing of three groups of students from different
communities in North Carolina enabling her to suggest, by the
study of different language uses in different home contexts, that
the language of the educational system would be accepted and
used in very different ways, and would lead to systematic differ-
ences in educational success and failure.
Similar discontinuities between home and school uses of lan-
guage have been studied in a variety of other contexts. In Boggs’
(1985) study of the socialization practices among Hawaiian chil-
dren, he found that the children were socialized to learn one set
of language practices at home and a very different, conflicting set
of practices at school. One early outcome of this research by
Boggs and his colleagues has been the development of literacy les-
sons which incorporate aspects of Hawaiian socialization practices
(Au and Jordan 1981, Au et al. 1986). Phillips (1983) and Scollon
and Scollon (1981) explored the mismatch in language socializa-
tion practices between Native American communities and the
Anglo educational system. Great differences in socialization prac-

tices between home and school are seen as a major contributing
factor to the high Native American student failure rate.
Heath (1986b), in a study of three ethnic groups in California
(Chinese-American, Mexican-American, and Southeast Asian-
Americans), compared their home socialization practices with the
literacy practices expected in schools. She found that Chinese-
American students were more successful in schools, in part
because the home socialization practices for Chinese-American
children most closely resemble the school literacy practices in the
US educational system, and in part because the Chinese-
American children had greater exposure to public language use
in the community outside the home. Wells (1986), in a ten-year
study of children in Bristol (UK), compared various student
groups in their use of written and oral language in both the home
and at school. He found that socialization practices in language
use by parents affected student success in early grades (see also
Christie 1992, Hasan 1989, Lareau 1989, Wells 1994).
The comparison of home-school literacy practices has also
been, in turn, a resource for writing process instruction, asking
students to be active ethnomethodologists (Heath 1985, Heath
and Branscombe 1985, Heath and Mangiola 1991). Students
become real field workers in collaboration with the teacher and
an outside researcher, collecting language data, and writing up
their results for real researchers to use, respond to, and question.
The result of such an ongoing activity is that students:
■ observe language uses
■ become aware of language metalinguistically in order to talk
about their ‘research’ findings
■ develop new ways to express their ideas about what they have
observed, and
■ write for a real and rewarding purpose to a real audience.
For students, the experience thus becomes a purposeful writing-
as-a-process learning activity.
A growing body of research is demonstrating that oral language
practices in schools have a profound effects on children’s literacy
development (Bloome and Green 1992). How a child interacts
verbally with the teacher and with other students will determine to
a significant extent how that student is evaluated by the teacher,
and more importantly, how the child will learn to read and write.
4.4.4 Halliday’s social semiotic approach to literacy

M.A.K. Halliday has probably been more influential in the current
changes in writing instruction in many countries than some
researchers and teachers realize. To the extent that his linguistic-
sociological approach (Halliday 1978: 38-9) to language provides a
theoretical linguistic foundation for whole language instructional the-
ories, his views of language and of the development of writing are
permeating US, Australian, New Zealand, and British educational
systems. Over the past 20 years, Halliday (1973, 1975, 1978, 1985,
1993b) has developed a socially contextualized approach to linguist
tics which stresses functional use of language. In the early 1970s, as
an integral part of his overall linguistic theory, he articulated a
theory of language development which has since become very
influential among early-literacy development researchers. In par-
ticular, his ideas - that language is primarily used to serve functional
purposes, and that patterns of language development depend on
the needs of the child to express meaningful communication - have
been a major source of whole language philosophy in the USA
(Edelsky et al. 1991, Goodman 1986, Jagger and Smith-Burke 1985).
Halliday proposes that all language use serves social-functional
purposes. Children learn language initially by recognizing that
language serves the purpose of activities of which the children are
in need. To quote Halliday (1978: 57):
... w hat is learning to read and to write? Fundam entally it is an
extension of the functional potential o f language. Those children
who d o n ’t learn to read an d write, by and large, are children to
w hom it do esn ’t make sense; to whom the functional extension that
these m edia provide has n o t been m ade clear, o r does n o t m atch up
with their own expectations o f what language is for. H ence if the
child has n o t been oriented towards the types o f m eaning which the
teacher sees as those which are p ro p er to the writing system, then
the learning of writing an d reading would be o u t o f context, because
fundam entally, as in the history o f the hum an race, reading and writ-
ing are an extension of the functions o f language. This is w hat they
m ust be for the child equally well.
This notion of writing as a functional extension of oral language,

one which can be mastered when the child sees the purpose for its
use, is basic to the whole-language instructional philosophy.
Elsewhere Halliday expands upon this perspective in relation to
the educational context.
[Education] starts from w hat children can do with language, which

is already a great deal by the time they com e to school; and from
w hat they cannot do yet b u t are learning to do with language, as well
as what they will n eed to learn to do with it in ord er to succeed in
school and in life.... T he im petus for reading and writing is a func-
tional one, ju st as was the im petus for learning to speak and listen in
the first place. We learn to speak because we want to do things th at
we cannot do otherwise; and we learn to read and write for the same
reason. (1978:205)
W hat are the functions for which [a student] is now beginning to

n eed the xvritten language, an d writing as a m edium? ... Will learning
to read and write make sense to him , m atching his experience of
w hat language is and w hat it is for, so th at he sees it as a m eans of
enlarging th at experience; o r will it seem to be a m eaningless exer-
cise which is unrelated to any o f his own uses o f language?
(1978: 213)
T he key question, perhaps, is this: to the ex tent th at the school is a

new culture into which the child has been socialized (and, as we
have seen, this makes greater dem ands on some children than it
does on o th e rs), is the actual p attern o f language use in the daily life
o f the school adequate to the socializing task? If it is n o t - w hat can
be done to rem edy this situation.
(1978: 215)
The three quotations above raise questions which have been taken
up as a challenge by the founders of the whole-language movement
- a movement which refers to Hallidayan functional linguistics for
theoretical support.
Leaders in the whole-language movement - e.g. Edelsky, et al.
(1991), K. Goodman (1986), Y. Goodman (1985), and Harste et al
(1984) - have taken Halliday’s questioning of the educational
system as the key to literacy learning. From their perspective, ‘the
actual pattern of daily language use in the schools’ is not adequate
to the socializing task. Children find that learning to read and
write is difficult because the school system makes it difficult. As K.
Goodman (1986: 20) notes:
Schools frequently isolate language from its m eaningful functional
use. T hen they change language into non-language. Only in the social
context of language usage does it have a m eaning potential for the
learner, and only in such a context is it language and easy to learn.
The response of these various scholars to the reality of the school

environment is to reassert the Hallidayan notion that children
learn language by ‘learning how to m ean’, and they learn to read

and write much more successfully when the reading and writing is,
similarly, a meaningful (or whole) activity.
Perhaps most im portant for the idea of learning to mean, all
language skills are practised together to convey meaningful com-
munication rather than in isolation so that students learn one
language skill, reading for example, separate from other language
skills (see also Enright and McCloskey 1988). As Goodman (1986:
30) notes:
Speaking, listening, writing, an d reading are all hap p en in g in con-
text of the exploration o f the world o f things, events, ideas, and
experiences. T he conten t curriculum draws on the interest and
experiences children have outside o f school, an d thus incorporates
the full range of oral and w ritten language functions. It becom es a
broad, rich curriculum th at starts w here learners are in language
and knowledge and builds outw ard from th e re .... Integration
becom es the central m otif in a whole language curriculum .
O ther researchers concerned with the writing development of

children have tended to confirm a num ber of the observational
claims of the whole-language perspective. As early as 1970,
Chomsky argued that learning to write by students did not have
to wait until they were well developed in their language arts (see
also Clay 1975, Feitelson 1988, Gaskins 1994, S. Stahl 1994).
Rather, the exploration of language through the process of early
writing and the creativity of invented spelling tended to have a
beneficial effect on both later writing and reading development
(see also Adams 1989). Since that pioneering research, other
scholars have found evidence to support those claims. In addi-
tion to the findings of Graves and his colleagues, Bissex (1980),
Clay (1975), and Hudelson (1984, 1989a) have all provided signi-
ficant supporting research perspectives on early writing
development.
While the whole-language approach to early literacy development
has a strong logical appeal, it must be recognized that much of the
evidence in support of the whole language approach is in the
form of observation, case study, and teacher testimonial. There
has, in fact, been little empirical research demonstrating the
superiority of this approach over other instructional approaches
(Gaskins 1994, McKenna et al 1993, 1994, S. Stahl 1994, Stahl and
Miller 1989).
4.4.5 Post-secondary uniting and discourse communities

Given the widespread appeal of ethnographic research in grade
school settings and the considerable changes taking place in ele-
mentary school language arts instruction, it is somewhat
surprising that there is very little ethnographic research on writ-
ing at post-secondary levels. Instead, the discussion at tertiary
levels focuses more broadly upon reasons why standard research
does not translate well into writing instruction and why students
do not seem to transfer what they learn in writing classes to writ-
ing in other classes or in the workplace.
In the 1980s, Cooper (1986), Miller (1984) and Reither (1985)
all argued for an approach which sees writing as occurring within
the social-rhetorical situation. Recognizing that writing is
produced, read, and interpreted in social contexts, Cooper
argued that a model of writing guided primarily by ‘writing as
thinking’ is an inadequate representation of the writing process.
Writing cannot be seen as isolated from the social world of inter-
action. Rather, ‘writing is a social activity, dependent on social
structures ... ’ (Cooper 1986: 366), and
writing and w hat writers do during writing can n o t be artificially sep-
arated from the social rhetorical situation in which writing gets
done, from the conditions th at enable writers to do w hat they do,
and from the motives writers have for doing w hat they do.
(Reither 1985: 621)
This perspective on writing research has been extended further in

recent work by Berkenkotter and Huckin (1995), Flower (1994),
Freedman and Medway (1994), and Witte (1992).
Writers, much like speakers, take on roles that are defined by
social structures. And by writing within socially recognized con-
ventions and expectations, writers can also alter the situation (or
reconstruct a new version of interaction); thus, writing reflects
ecology - the totality of relations between organisms and their
environment. For Cooper, an ecological model of writing ‘is an
activity through which a person is continually engaged with a vari-
ety of socially constituted systems’ (1986: 367).
Taking such a notion one step further opens a set of basic issues
now discussed regularly; these discussions centre on socio-cogni-
tive approaches to writing, on audience, and on genre and
discourse communities. The gradual merging of social and cogni-
tive perspectives has been discussed over the past decade in
various LI writing contexts. It has been raised by Langer (1987)

for elementary and secondary school contexts (see also Davis et al.
1994, Needels and Knapp 1994), and, more recently, Flower
(1994) has proposed a social-cognitive theory of writing that
addresses more advanced writers. In these approaches, the essen-
tial argum ent is that a comprehensive theory of writing needs to
recognize the various social factors which influence writing; but at
the same time, recognition must also be given to the idea that writ-
ing itself is produced through the cognitive activities of the writer.
In this way, a theory of writing which disregards either major com-
ponent - cognition or social context - will be necessarily
inadequate. (Cf. Berkenkotter and Huckin 1995, Freedman and
Medway 1994, Nystrand 1989 and Witte 1992.)
Over the past 20 years, considerable interest has been directed
at the issue of audience as a major social-context factor, owing its
inception to the increasing rhetorical influence over composition
in the 1970s (Ede and Lunsford 1984, Kroll 1984, Park 1982).
While many perspectives have been promoted in attempting to
define audience (the precursor to the social context of writing),
the concept has remained elusive (cf. the discussion of audience
in Chapter 8).
A more recent perspective on the social context defined by
readers has been the emergence of the notion of discourse com-
munity (Bazerman 1988, Swales 1990, 1993). The notion of
discourse community includes writers, readers, texts, and social
contexts in their natural interaction, rather than artificially high-
lighted and separated relations between writer and reader based
on other considerations (Rafoth 1988).
At post-secondary levels of writing instruction, these views have
led to a num ber of developments which may change the nature of
tertiary-level writing instruction in the future. One issue lies in the
move to introduce students to ‘discourse communities’ so that
their writing becomes purposeful in the academic setting.
Another related issue is the expanded role that disciplinary aware-
ness creates for writing across the curriculum; writing is not a skill
that can be taught in isolation, but is the entire faculty’s instruc-
tional responsibilities. Still another issue raised involves the
development of content-based writing courses which teach writing
as the means to interact with the material and with other students.
This approach is compatible with the whole language approach,
but at the tertiary level (cf. Bartholomae and Petrosky 1986,
Brinton et al 1989, Shih 1986).
The notion of discourse community is appealing, and must be

explored if it is to be useful in the expansion of writing theory. The
idea underlying the academic discourse community is that students
need to initiate themselves into the academic discourse community
they wish to join. The discourse community in this case is the com-
munity of scholars who both discuss and write about ideas and
information relevant to their professional interests.
Recently, a num ber of reservations and concerns have been
raised about the notion of discourse community. One is a defini-
tional concern: what exactly is a discourse community? This
question, raised by Cooper (1989), is not easily answered (see also
criticisms by Nystrand 1990). Citing a conference paper delivered
by John Swales in 1987 (cf. Swales 1990), Cooper suggests that his
explanation provides a satisfactory working definition from an
institutional perspective (1989: 212-16), though the criteria also
are open to criticism on exclusionary grounds. Swales (1990:
23-7), in his later book, Genre analysis, outlines six criteria for
identifying a discourse community:
1. A discourse community shares common public goals, even if
the private goals of individual members are distinct. Stating
public goals of groups serves a real function: ‘it creates the illu-
sion of agreement by allowing members to interpret the stated
goals in their own way’ (Cooper 1989: 212).
2. The discourse community is a forum for discussion and debate;
discourse communities must have interaction among their
members. These mechanisms may include meetings, corre-
spondence, newsletters, etc.
3. Discourse communities will provide feedback and information
to members through various participatory mechanisms. A
major function of newsletters, journals, conferences is to sup-
port members with useful resources which match their interests
and needs.
4. Discourse communities, as they interact, will develop discourse
expectations and discourse genres as norms by which inter-
change regularly occurs. These genres will become norms for
the community and may also become conservative in nature.
5. The discourse community develops along with a regularized set
of genre expectations a specific set of terminology and special-
ized vocabulary. Community specific abbreviations, acronyms,
and references will become commonly used.
6. The discourse community must have enough members to dis-

cuss matters of importance to a wider group and to generate
content expertise and the regularized exchange of this infor-
mation.
Recent work by Fairclough (1992a, 1992b) from the emerging
perspective of critical linguistics is also pertinent with respect to
defining the shape of the discourse community.
A central issue raised by Cooper is not simply whether or not a
discourse community can be defined, but whether it is a structure
into which student writers should be initiated (see also Bizzell
1993). As Cooper (1989: 216-19) argues:
Knowledge, power, success are seen to flow from the com m unity
automatically to those who before their apprenticeship lacked any
relevant cognitive or social abilities.... We may wish to see the writ-
ing classroom in a foundational way, as a stable discourse
community, one that validates o u r interests, o ur values, o u r status,
o u r way of life, automatically. But we are also concerned th at o ur
students - all of our students - becom e effective writers, learn how
to en ter into discourses, which may n o t be the same thing as the
foundational notion o f jo in in g a discourse com m unity as a neo-
phyte.... Claiming that a foundational notion of discourse
com m unities should guide o u r teaching o f writing can only be seen
as cynical and self-serving. A nd it is far from clear th at students want
to - or should w ant to - jo in any of these com m unities we are creat-
ing for ourselves.... If we insist that students ad o p t w hat we see as
the values of our com m unity (our values), we will effectively with-
hold power within academ ic discourse from all students who com e
from a different generation, a different ethnic background, a differ-
e n t race, a different sex, a different econom ic class.
Cooper is not merely defining discourse community, but is also

examining the possible weakness of the notion as theory and as
guide for writing instruction (see also Fairclough 1992a, 1992b).
From the beginning, the discourse community perspective has
raised such concerns. Both Bartholomae (1985) and Bizzell
(1986a, 1993) have asked how someone becomes initiated into a
discourse community and what it takes to become accepted in that
community. Bartholomae notes that apprenticeship in the aca-
demic community may not be easy to achieve. Bizzell goes one
step further and asks whether it is desirable for students to
become members of the academic discourse community. As
Cooper notes (above), discourse communities can quickly assume
a foundation of rightness (righteousness) that appears to exceed

the actual collective will of the group. The community decides
what not to accept and who not to accept as if guided by some sort
of absolute knowledge rather than by the political/subjective deci-
sions of authority figures in the community.
If the notion of discourse community becomes defined as a
community of élite members, then the discourse community does
not represent an exchange of knowledge or a discussion of differ-
ing views; instead, it reifies itself into a body of information to be
absorbed much as pieces of cultural literacy are to be learned
without question (see, e.g., Hirsh 1987; cf. Nystrand 1990).
One approach to writing which appears to have made use of this
notion of discourse community without silencing the voices of the
student-initiates has been the adaptation of a Freirean approach to
literacy to the tertiary-level writing class. A Freirean approach, per-
haps oversimplified, requires the group to become aware of its
condition and to learn how to change its condition through dia-
logue, confrontation, and literacy development (Freire 1985, 1994,
Freire and Macedo 1987, Giroux 1988, Shor 1987).
Recently, Bartholomae and Petrosky (1986) reported on a long
sequence of curricular projects to develop a tertiary-level first-year
writing programme at the University of Pittsburgh based on the
notion of developing a discourse community within the writing
class. Special composition classes consisting of 12 students, an
instructor, and a writing lab assistant were set up; each such class
met six hours a week throughout the semester (six credit hours).
Students, all high-risk, were asked to read 12 books and produce
24 writing assignments. The essential features of the course which
made it unique were:
(1) the content theme of the course (’growth and change in
adolescence’);
(2) the requirem ent that all students participate in extended dis-
cussions of the readings and the writing assignments -
dialogue and exchange being seen as a crucial part of com-
munity building;
(3) the requirem ent that all assignments be done on time and all
classes attended - with the penalty that delinquent students
would be dropped from the class;
(4) the expectation that each student had something to contrib-
ute to the group since all students have experienced growth
and change in adolescence and therefore have ‘expert’

knowledge;
(5) the sequence of assignments, which moved from a personal
awareness to an understanding of how academic commu-
nities undertake research on the same topic;
(6) a low student-to-teacher ratio.
The goal of the curriculum in this class was not simply to create a
content-based instructional setting, but to confront students with
issues that should force awareness of their condition in three ways:
(1) in relation to the content of the course, (2) in relation to their
status as students, and (3) in relation to the expectations of the
academic institution. At the same time, the small class size, six
hours per week with the same instructor, and the extensive addi-
tional use of a writing laboratory assistant were all m eant to
provide support for struggling students who were willing to com-
mit themselves to the course.
This use of the discourse community to create awareness of the
social-context of writing represents one of a num ber of
approaches to writing that have developed recently as a result of
the move in writing research from a cognitive emphasis to a social
emphasis, and finally to a socio-cognitive emphasis.
Drawing on post-structuralist literary criticism, Faigley (1986)
notes the impact of reader-response theory (e.g. Fish, Eco), to
the effect that any effort to write about reality, or self, or author
always demands reference to prior text (intertextuality). Further,
it is impossible to say in the abstract what constitutes expertise
(knowledge) because it is defined by a community of writers.
Finally, for some writers, continuing problems may well be due to
ignorance of the privileged language of the academic community.
Most importantly, Faigley discusses recent sociology of science
research. Texts, particularly science texts, are not autonomous
presentations of information; rather they are ‘social tools (used)
in the complex interaction of the research community’. How this
is so will be a major focus of the next chapter.
Notes
1. Since m ost textbooks trail research innovations by 5-10 years, only the
m ore pow erful changes m anifest themselves quickly in a textbook
industry.
2. This chapter has drawn u p o n four useful sources o f inform ation for
reviewing the historical antecedents o f cu rren t writing-as-a-process
theory and practice, as well as com peting perspectives: Berlin (1987),
Faigley (1986), Hillocks (1986), an d N orth (1987).
5
Writing process research and

recent extensions
5.1 Introduction
The previous chapter presented an overview of writing process
approaches, as commonly interpreted. The present chapter will
explore research into the cognitive processes involved in writing
in three ways. First, it addresses the more current research on the
well-known Flower and Hayes model of the writing process; it then
discusses alternative approaches suggested by Bereiter and
Scardamalia, and by researchers concerned with the nature and
development of expertise. Second, the discussion is extended to
recent research and instructional practices which explore a
Hallidayan approach to language, and particularly to the empha-
sis on genre form in writing research and instruction. This
extension reasserts the balance between linguistic form and cogni-
tive processing in writing; it also moves the chapter away from
cognitive research on the writing process and towards a considera-
tion of the varying social contexts in which writing is practised.
Finally, the influence of the process approach on L2 contexts for
writing research and instruction will be discussed. Increasingly,
writing instruction which is informed by process approaches is
being employed in contexts involving L2 students. However, it is
im portant to establish the extent to which the contexts for L2 writ-
ing differ from writing in the LI, and the extent to which such
differences may suggest different instructional practices.
5.2 Theories o f the writing process

A careful account of the more recent research on the writing
process in the past decade is im portant because that research has
113
greatly refined our understanding of writing; it has incorporated

into its concerns a greater awareness of the complexity of writing;
it makes strong connections with the developing theory of genre
for writing, and it has placed a num ber of very strong claims
about writing into a more balanced perspective. In particular, the
more recent research of Flower and Hayes and of Bereiter and
Scardamalia has provided im portant insights into the develop-
m ent of writing abilities. In order to place these recent
developments in proper perspective, we shall first summarize
recent research on the well-known Flower and Hayes model of
the writing process. Given this background, it will then be possi-
ble to situate the work of Bereiter and Scardamalia (1987;
Scardamalia and Bereiter 1987), highlighting the importance of
their theory as a synthesis for a wide range of writing research.
The theory they propose also provides a strong foundation for
designing a more principled and effective writing curriculum in
academic settings.
5.2.1 The uniting process: the Flower and Hayes model

of uniting
In Chapter 4, we reviewed carefully a num ber of ideas which have
evolved out of research on the writing process, particularly with
respect to the earlier work of Flower and Hayes. Much of the work
of Flower and Hayes in the past ten years has been directed
towards building more explicit definitions of the various subcom-
ponents of their model (Flower 1988, 1989, Flower et al. 1990,
Haas and Flower 1988, Hayes et al 1987). For example, beginning
in 1980, Flower and Hayes (1980a) developed the notion of the
rhetorical problem in the task-based component of their model
(Chapter 4) as a part of the process of discovery in writing. They
presented a model of the rhetorical problem, the major elements
involved, and showed how expert and novice writers differed in
their application of this model. For Flower and Hayes, the rhetori-
cal problem breaks into two major units: the rhetorical situation
(audience, topic, assignment), and the writer’s own goals (involv-
ing the reader, the writer’s persona, the construction of meaning,
the production of the formal text). In short, the model is
intended to show the range of potential writing problems which a
writer could face during the composing process. This notion of
‘problem ’ in writing is extended in Hayes et al (1987) through the
Writing process research and recent extensions 115
discussion of problem solving in a theory of revision. In this case,

problem solving centres on responding to the text and developing
processes for detecting and diagnosing problems, then developing
strategies for resolving the identified problems.
The theory of revision presented in Hayes et al (1987) also con-
stitutes an effort to explore the subcomponent of reviewing in their
general model. In their theory of revision, there are four basic
processes: (a) task definition, (b) evaluation, (c) strategy selec-
tion, and (d) modification of text in the writing plan. The goal of
this subcomponent model is to explain, from their protocol
research and from earlier research findings, how writers revise,
why it is often difficult to revise globally (e.g. change organization
or plans, restructure information), and how expert and novice
writers differ in their revision processes.
In another set of studies (Flower et al 1990), the relationship
between the major components task environment and writing
processes is examined. In particular, Flower analyses the academic
task of reading-to-write. These studies explore how students read
in order to perform a writing task, how different students repre-
sent the task (and its influence on reading) differently, and how
task representation and reading influence the writing processes.
The studies also examine students’ own views on writing in the
context of a particular task. The combined study of teacher per-
spective, student perspective, students in the process of writing,
and the written text together created a complex set of factors
which argue that students primarily need to develop strategic
knowledge - having strategies to determine appropriate writing
goals, having awareness of the writing-task goals, and having strate-
gies for carrying out the goals set. The overall theme in this
research is to establish the interaction of context and cognition in
the carrying out of a particular writing task: writing is to be seen as
both a cognitive activity and a contextually constrained activity. In
cases in which the context might impose limiting constraints (e.g.
a complex writing assignment), the goal of writing instruction
would be to have students develop strategic knowledge and be
able to transform their knowledge rather than rely on simpler
strategies such as knowledge telling.
For Flower, the major transition which writers have difficulty
learning is not how to acquire new writing skills but how to apply
already-practised writing skills in new ways for new purposes. As
she argues:
For the student, ... the classroom content, the teach er’s concern
with content, and the role o f the p ap er as a tool in the grading
process ... are likely to fit a familiar schem a for them e writing. But
w hat is im portant in college is n o t the ap p aren t genre o r conven-
tions, b u t the goals. T he goals of self-directed critical inquiry, of
using writing to think thro u g h genuine problem s and issues, and of
writing to an im agined com m unity o f peers with a personal rhetori-
cal purpose - these distinguish academ ic writing from a m ore
lim ited com prehension and response
(Flower et al. 1990: 251)
Thus, a major outcome of this particular study, as well as many of

their earlier studies, is to see the growth of writing ability as the
development of strategies for appropriate goal-formation - con-
strained by task environment and content knowledge (i.e.
constructing a ‘rhetorically-bounded purpose’, Flower 1989: 292),
and the concomitant expert strategies for carrying out the goals
through the writing process.
Results of their research on this model argue that good writers
consider many more aspects of the rhetorical problem, and to a
much greater depth. Good writers also can respond to a unique
rhetorical problem with a fully developed representation of the
problem. Finally, good writers are able to reassess their goals
(vis-à-vis audience, persona, meaning, text production) in the light
of rhetorical problems which arise in the course of writing; that is,
good writers are able to alter their goals appropriately. Overall,
Flower and Hayes derive two major implications from their model.
First, good writers have a richer sense of what they want to do when
they write, and have a fully developed image of the rhetorical prob-
lem. Good writers are, in effect, creative in their problem finding
and in their problem solving. Second, recognizing and exploring
the rhetorical problem is a teachable process.
Flower and Hayes’s emphasis on students’ strategic knowledge
and the ability of students to transform information (and difficul-
ties students have in doing this) to meet rhetorically constrained
purposes is echoed in a num ber of other studies (e.g. Bereiter
1990, Bereiter and Scardamalia 1987, Brown and Palincsar 1989,
M. Carter 1990, Collins et al 1989) and opens the way for writing
instruction which tries to teach students to be more strategically
aware of their goals and of ways to carry out their goals in writing.
At the same time that this influential theory has evolved into a
more elaborated and socially contexted approach to writing, it has
also been limited in its assumption that there is a single writing

process which is essentially the same for all writers; that is, expert
writers would appear to do the same thing as novice writers, but
do it much better. In their model, skilled and less-skilled writers
exist on a single continuum. While Flower and Hayes devote con-
siderable effort to comparing skilled and less-skilled writers, their
model seeks to describe features that are common to all writers. A
different perspective is proposed by Bereiter and Scardamalia
(1987), who have argued that expert writers appear to go through
a different sort of writing process, which allows these writers to
transform information through an additional set of strategic
processes.
5.2.2 The Bereiter and Scardamalia model of the

uniting process
Bereiter and Scardamalia’s theory proposes that ‘the writing
process’ cannot assume a single processing model, but should
consider different processing models at different developmental
stages of writing. They argue that the writing process of a young
student and that of a mature skilled writer cannot be the same:
the skilled writer does not perform the same processing, but many
times more efficiently. Instead, the skilled writer performs a differ-
ent kind of writing process (when necessary) - one which the
unskilled writer is not (yet) capable of performing. Rather than
debate the polemical question of how many such models might
exist, it is enough, for their purposes, to demonstrate that a theory
proposing two models of writing processes accounts for the range
of accumulated research findings better than a theory proposing a
single processing model.
Since Bereiter and Scardamalia focus more on describing why
and how skilled and less-skilled writers compose differently, rather
than on describing the common features of all writers, they con-
sider a wider range of research. Their theory accounts for these
research findings coherently and allows for the generation of a
num ber of testable hypotheses as a primary means for elaborating
and extending their theory. (Flower and Hayes, in contrast, often
argue that they are exploring data to develop their theories of
writing, leaving others to test the various aspects of their model.)
In particular, Bereiter and Scardamalia are concerned with
accounting for the following im portant questions:
1. How does a processing model distinguish skilled writing from

less-skilled writing?
2. How do audience and genre differences create distinct writ-
ing difficulties, and why do some genres appear more difficult
to master, and some audiences more difficult to address?
3. Why are some writing tasks easy and others more difficult (not
only genre and audience, but also purpose, topic, and lan-
guage variation)?
4. Why do writing skills in one writing task or genre not transfer
to other writing tasks or genres?
5. Why do some writers have more difficulty than others on
some writing tasks yet appear to be at the same general profi-
ciency level?
6. Why do some children find writing easy and natural, yet
skilled writers often find it difficult and painful?
7. Why is advanced writing instruction particularly difficult and
often ineffective?
8. Why do some writers never seem to develop mature compos-
ing skills in spite of much practice and long educational
experience?
9. Why do expert writers revise differently from less-skilled writ-
ers?
10. How can the writing process account for the notion of ‘shap-
ing at the point of utterance’ (cf. Britton 1983)?
The essential mechanism which Bereiter and Scardamalia use to

explore these issues is the notion that mature, skilled composing
is based on a sophisticated interplay of problem recognition and
solution - a process that must be different for skilled writers of
complex prose than it is for children and less-skilled older writers.
The basic difference is captured in their two models of the writing
process: the knowledge-telling model and the knowledge-transforming
model. They draw the basic rationale for such a two-model theory
from research findings which point to very different types of com-
posing behaviour among skilled and less-skilled writers. In
particular, they present evidence that less-skilled writers begin to
write on a given assignment much sooner (less time for initial
planning):
(i) less-skilled writers produce much less elaborated or

abstract sets of pre-writing notes;
(ii) less-skilled writers primarily concern themselves with gen-

erating content during composing, rather than also
considering goals, plans, and problems;
(iii) less-skilled writers are seemingly incapable of making
major revisions which would involve reorganization of the
content;
(iv) less-skilled writers use less complex routes (no references
to goals, organizational strategies, etc.) for recalling ideas
used in their writing; and
(v) less-skilled writers do not make use of main ideas in their
writing as guides for planning and integrating information.
(Scardamalia and Bereiter 1987)
These findings, argue Bereiter and Scardamalia, suggest not sim-
ply a more efficient use of a writing process for more skilled
writers, but a qualitatively different kind of writing process for
skilled writers.
The knowledge-telling model of writing used by less-skilled writers
provides a streamlined set of procedures which allow writers to
bypass the sorts of complex problem-solving activities often seen
in the composing of skilled writers. When children and less-expe-
rienced writers begin to compose texts, they need to keep the task
relatively uncomplicated if they are to be successful. In order to
do this, they must solve the most basic problem, converting oral
language experiences into written form, without having to worry
about issues that will confound the process. In particular, they
have to shift from engaging in dialogues and responding to topics
initiated by a partner, to a monologue in which they can only
imagine the dialogue as a way to retrieve and create information
that will be used for writing. Thus, their primary problem is that
of generating enough useful information from their own internal
resources. The primary goal becomes to tell what they have
retrieved.
In order to simplify the retrieval procedure it appears that
beginning writers rely on a few specific strategies:
■ They consider the topic of the assignment and ask themselves
what they know.
■ They consider the genre of the assignment and ask themselves
what they know.
■ They read what they have just written and use this to generate
additional information.
These strategies work well for writing about personal experiences

or feelings, writing in journals and diaries, and for telling narra-
tives - recounts of specific events and stories. These topics, as
writing assignments, are typical of early writing instruction (see
Applebee 1984, Graves 1983, Martin 1989). Teachers have learned
from their teaching experience that students can accomplish those
assignments and practise basic writing skills without great confu-
sion or a lack of ‘something to say’. A further advantage of these
assignment types is that ‘coherence’, however it might be defined
theoretically, tends to take care of itself through the basic process
of simple content generation on a topic and possibly through the
simple device of chronological ordering of details and events.
A description of the task facing a young inexperienced writer
will explain the operation of the knowledge-telling model.
Starting from an example writing assignment such as
‘What did you do during the holiday weekend V
the writer will look for things to write about by looking at the ques-
tion, seeing the words ‘holiday weekend’, and thinking about what
he or she did that weekend. He or she can also think about what
people normally do on weekends that is worth writing about as
another way to generate content. The writer also knows that he or
she has to recount events, since that is the genre being called for,
typically within a simple time order. Because of the assignment,
almost everything that is retrieved will be associated with the
topic, and thus be coherent by association with the topic and
probably will be unified by simple chronology. Abstract logical
organization is not a major concern. As the student writes, the
prior written text also becomes a stimulus to activate and retrieve
additional information from memory. The result of this process is
a written text:
■ for which it is easy to find sufficient information,
■ which does not overburden the writer with too many cognitive
tasks (and so is relatively easy), and
■ which tends to organize itself coherently by simple association
to the topic and to time.
This general writing process is presented in Figure 5.1.
In this model, information is generated from the assignment, the
topic, the genre, and any terms or lexical items in the assignment.
Idea identifiers are retrieved and memory is searched for relevant
Figure 5.1 Structure of the knowledge-telling process
information. If retrieved information seems appropriate to the

topic, the information should be written down and used (together
with material retrieved through the previously identified strategies)
to search memory for more things to say. The processing demands
are fairly simple in this model, as are the retrieval and evaluation
demands. The knowledge-telling model accounts for a manageable
solution to composing which then becomes an efficient means for
writing.1 However, the model does not account for writing tasks
which demand more complex processing, in which considerations
of information ordering, of relative salience of information, of audi-
ence expectations, and of logical patterns of argument organization
will determine the effectiveness of the writing. This sort of writing

task is typically expected in more advanced writing for academic
purposes, and the knowledge-telling model sheds little light on the
task complexity involved in such advanced writing demands. For
this reason, Bereiter and Scardamalia have proposed that a second
model of the writing process be invoked when such writing is
demanded, and performed, by the writers. They have named this
the knowledge-transforming model of the writing process.
In Figure 5.2, the knowledge-telling model becomes just one com-
ponent of a different process. The larger model represents the
reflective problem-solving nature of expert writing when task com-
plexity requires such processing. In particular, problems are
solved by conscious resolution in either the content problem space or
Figure 5.2 Structure of the knowledge-transform ing process

the rhetorical-problem space - the output of one becomes the input

of the other.
In this model, the writing task leads directly to problem analysis
and goal setting. The resulting goals and problem anticipation
lead to plans for the resolution of the perceived problems,
whether they be problems of content generation, content integra-
tion, audience expectation, writer intention, genre form,
linguistic style, organizational logic, etc. The resolution of one
problem may create another. Thus, generating additional content
(solving a content problem) may lead to a new rhetorical problem
of how to organize the new information best in light of the previ-
ously presented information. As problems become resolved they
feed to the knowledge-telling com ponent which generates the
writing. As writing is generated, it also contributes sets of prob-
lems which must then go through the problem-solving processes.
As an example of this writing process, suppose that in writing
about the development of students’ reading skills the writer has
been using the term decoding skills to explain how words are identi-
fied rapidly, but, on reflection, realizes (a problem) that some
readers in the audience (rhetorical problem space) would have a
different understanding of the term decoding For some readers,
the term means only the mental translation from letters to sounds;
for others it can mean word recognition. The writer goes to the
content problem space and searches for a better term in content
knowledge. The term word recognition skills fits the intended mean-
ing without the misinterpretation potentially caused by the other
term. This new term, along with its connotations and connections,
is taken back to the rhetorical problem space to see how it will fit
into the organization of the text. Rhetorical problems will include:
■ where in the prior text to introduce the term
■ how carefully to explain the term
■ whether or not the new term will create some other types of
confusion for the readers
■ whether or not additional information will be needed elsewhere
in the text because of the change of terms.
The latter two concerns may move the writer back to the content
problem space to find supporting content information. If there are
potential confusions arising from the new term or if additional
information is needed, the writer will have to consider also whether
or not the organizational framework of the text will need to be
adjusted (back to rhetorical problem space). Such an interplay may

continue for all the problems raised by goal setting, problem antici-
pation, or problem recognition from the generated text.
The two writing processes, knowledge telling and knowledge trans-
forming, represent two different ways that writers compose, and
many writers seem never to master the knowledge-transforming
process. The knowledge-telling process provides adequate support
for less-skilled writers to generate sufficient on-topic material
while keeping cognitive complexity at a manageable level. The
knowledge-transforming process represents writing with reflection on
the complexity of the task, and leads the writer to find an appro-
priate way to address the complexities. A good writer, in many
cases, may depend solely on a knowledge-telling process because the
information has been through a knowledge-transforming process
from an earlier similar writing task, or because the writer can sat-
isfy the task requirements from a sophisticated use of the
knowledge-telling process. Thus, it is possible that even genres which
may be viewed as expository in nature can be generated by a
skilled writer who relies primarily on a knowledge-telling process,
generating sophisticated content with minimal problem solving
required. For students, the five-paragraph model often fills the
role of making an essay topic an exercise in knowledge telling.
The evidence which Bereiter and Scardamalia draw upon to sup-
port this theory is in two forms. Their own research program over
the past 15 years has explored many of the issues argued for by
their theory and, in combination with a wide array of other
research findings, strongly supports this theory. In addition, their
two-model theory provides a powerful explanation for many of the
differences noted in the novice-expert writing research, thereby
accounting parsimoniously for many research findings. In particu-
lar, their theory accounts for the research questions raised earlier.
5.2.3 Accountingfor skilled and less-skilled uniting

performance
The two-process theory provides an explicit account of how chil-
dren can write well and not find writing to be overly difficult
(knowledge telling), yet skilled writers often find writing (knowl-
edge transforming) to be difficult and complex. Similarly, writing
tasks will vary in difficulty depending on how complex the infor-
mation management demand becomes, and this will vary from
person to person depending on background knowledge of the

topic and the extent to which the particular topic has been prac-
tised previously by a writer. The more information that has to be
addressed in the problem spaces of the knowledge-transforming
model, the more complex the composing demand and the more
difficult the writing task for a particular writer. In this way, the the-
ory accounts for different performance on particular writing tasks
by equally skilled writers.
The knowledge-transforming model allows for the elaboration of
explicit hypotheses relating audience and genre differences to
writing task difficulties. It suggests that writers who have practised
one genre or writing task will not necessarily transfer their skilful
performance to other genres or tasks until they have had practice
resolving the sorts of complex writing problems which are typi-
cally created by such new tasks and genres. The difficulties
encountered by advanced-level instruction can also be explained.
If students seldom practise the sorts of writing tasks which develop
knowledge-transforming skills, they are not likely to be able to per-
form those skills easily. Bereiter and Scardamalia (Scardamalia
and Bereiter 1987) have argued that the combined effort to
resolve both content and rhetorical problems through the move-
m ent between these problem spaces invokes a dialectical process
for reflection, a process which is excluded in simpler writing tasks.
In this way, simpler writing tasks (typically narrative recounts, and
personal expressive writing) which do not require such a dialecti-
cal process, do not prepare writers for situations in which the
dialectical process is essential.
Evidence in Australia, Canada, England, and the USA indicates
that many English-speaking students do not receive serious
instruction or practice in cognitively demanding writing tasks
(Applebee 1981, R. Carter 1990, Christie 1989, Martin 1989). The
problem of writing process transfer across tasks and genres is one
which has received relatively little attention (except in Australia;
see Section 5.3 below), but which constitutes a serious issue for
writing development. For example, Peyton et al. (1990) argued
that dialogue journal writing is a better approach to writing
instruction because students write more in dialogue journals than
in an essay assignment. Following Bereiter and Scardamalia (and
the Australian genre-based theories), such a result ensues because
dialogue journal writing is knowledge telling and is easier; it is also
practised daily.
Both the work of Bereiter and Scardamalia and that of Hayes et

al. (1987) demonstrate that, also in revision, skilled writers are
capable of performing in ways that less-skilled writers do not seem
to be able to master, even with training and assistance. This differ-
ence is seen even at the tertiary academic level. The clearly
different behaviour is a strong argum ent for a two-process rather
than a single-process theory.
All skilled writers have experienced the phenom enon in which
they discover what they really want to say only after they put their
initial thoughts on paper, or as they reflect on what they have just
written. This discovery or transforming of what one has to say is
the result of recognizing problems explicitly which then need to
be resolved in the writing as it evolves. The act of writing becomes
a heuristic process.
At present, it would seem that Bereiter and Scardamalia’s two-
process theory of writing provides a good account of major
research findings on the writing process and on differences in
composing between expert and novice writers. At the same time, it
should be recognized that recent research by Flower and Hayes
has greatly refined their process model. Given the greater elabora-
tion now available for the subcomponents of the Flower and
Hayes model (exploring in greater details the components of
planning, problem solving, revising, and interactions with task
context), it is tempting to argue that their model is sufficient for
the many research questions raised earlier. However, the two-
process model has certain advantages which cannot be ignored.
First, the non-occurrence of certain writing process abilities
among less-skilled writers would not support a model in which a
continuous development would be predicted. If less-skilled writers
had the same general processing model but were able to employ a
less-refined version of it, one would expect them to perform like
skilled writers, but not as well. However, the less-skilled writers
actually perform in ways which suggest that they are doing some-
thing different.
Second, the two-process model focuses on differences rather
than similarities. Consequently, it provides a more direct account
of the many results deriving from comparisons of skilled and less-
skilled writers.
Third, the two-process model provides an account of how more
complex writing tasks create problems that are beyond the abili-
ties of less-skilled writers but can be handled by skilled writers; it
more directly addresses the problem-solving nature of more com-

plex writing tasks.
5.2.4 Limitations of the Bereiter and Scardamalia theory

It must be recognized, however, that the Bereiter and Scardamalia
theory of writing processes also has a num ber of difficulties which
need to be addressed if it is to become the strong hypothesis-gen-
erating theory that the creators assume it to be. It will be
necessary to account for the influence of context on the writing
process, as has been argued recently by Flower (1989, 1994,
Flower et al 1990). This becomes possible by elaborating on spe-
cific model components: the problem spaces, the organization of
content knowledge, and the organization of rhetorical knowledge;
and the ways the elaborated sources of information and problem
representations are connected.
A second problem concerns the ways in which a writer develops
a knowledge-transforming model of the writing process. It is not
clear, from the evidence Bereiter and Scardamalia have presented
to date, how - or when - a writer makes this cognitive transition
(cf. Bereiter and Scardamalia 1993). A related question is whether
or not it is possible to speak of a stage in which a writer has a par-
tially developed knowledge-transforming ability. If it is possible, how
can it be recognized, and how generalized is its applicability (e.g.
to what extent will it be context dependent)? Yet another related
question is whether or not the more complex writing process is
limited only to certain individuals and not equally accessible to all
A\iio would want to become expert writers. For example, many
studies of LI students in grades 4-8 indicate that, even with
prompting and training, students have difficulty carrying out
more expert approaches to writing and revision.
Bereiter and Scardamalia have argued that a major impediment
to learning this more complex strategy for composing is the non-
challenging nature of writing demands in school curricula.
According to Bereiter and Scardamalia (1985), among the ways
inert knowledge is promoted in schools include:
■ pre-writing activities when they are not needed for content gen-
eration
■ the extensive use of expressive ‘turn-on’ topics
■ acceptance of ‘tell what you know’ strategies - even when not
called for by the task
■ a willingness by teachers to put up with recognized weaknesses in

structure and content - even when students admit the existence
of these weaknesses but make little effort towards improvement
■ the use of conversational ploys to avoid rhetorical problems in
the writing (see also Scardamalia and Bereiter 1986).
More recently, Bereiter and Scardamalia have argued that the
development of school-based generalized learning skills, and
(knowledge-transforming) writing in particular, is possible by pro-
moting students’ capacities for ‘intending to learn’ (Bereiter and
Scardamalia 1989). This perspective on school-based learning is
that students create a ‘contextual m odule’ (Bereiter 1990) in the
course of their dealing with school activities and requirements.
A contextual module constitutes an acquired complex of knowl-
edge which combines and integrates procedural knowledge,
declarative knowledge, goal structures, problem models, affect,
persona, and a code of conduct. As Bereiter (1990: 613) states:
A contextual m odule consists of this entire com plex of knowledge,
skills, goals, and feelings. So long as all these com ponents are separ-
ately represented in the m ind, however, there is little reason to talk
about them as a unit, even though they are interrelated. A cquired
m odularity implies that over time these separate com ponents come
to form a m ore organic whole.
In the ‘schoolwork’ module that Bereiter and Scardamalia

describe, students perform school tasks as though these tasks were
a job, and a student’s goal is to find the most efficient way to get
the job done. The ‘intending-to-learn’ module, in contrast, refers
to breaking out of the simpler efficiency-driven ‘schoolwork’ con-
textual module and learning a second, self-driven learning
module - the intentional learning-module. In the second learning
module, students set learning goals which go beyond the school
task requirements and reflect a willingness both to understand
knowledge and processes in terms of problem-solving activities,
and to confront learning problems, all of which points to a learn-
ing system as an active goal-seeking process.
5.2.5 Expertise, cognitive-strategy training, and the

problem of ill-defined knowledge
Approaches related to the notion of intentional learning are
suggested by cognitive research on expertise. Applied to the
educational context, such approaches would explore how students

make the transition to a goal-setting and problem-solving orienta-
tion in writing. Bereiter and Scardamalia (1993) have recently
explored the nature of expertise, drawing on their work in writing
and synthesizing it with much of the recent work on developing
expertise. A num ber of other researchers have explored the grow-
ing field of expertise and applied this research to writing
development. Michael Carter (1990), for example, has pointed
out that the research on ‘expertise’ has debated the importance of
general learning strategies versus applications of local knowledge.
From the studies he reviewed, it appears that experts turn to gen-
eral problem-solving strategies only when they are faced with
unusual problems. In such a situation, experts can transfer skills,
but usually only with assistance, practice, and explanations.
Results suggest that the best way to transfer skills and strategies is
through consistent practice in a variety of similar contexts to the
point of proceduralization or automaticity. At the same time,
Carter notes that the influence of local knowledge on expert per-
formance has been similarly overestimated. The importance of
local, context-specific knowledge tends to become greater for
learners as they are asked to perform in specific contexts such as
those in graduate school or in professional work environments. As
Carter notes, ‘experts become experts because they possess very
highly organized schemata, developed over a long time, that are
related to a specific skill’ (1990: 272). Novices, by contrast, lack
the specific knowledge which comes from practise in the organiza-
tional abilities, content information, and genres required by a
context.
Many less-skilled student writers would appear to approach writ-
ing tasks with few resources in either local knowledge or general
strategies. Among the research studies developed to create such
resources, Bereiter and Scardamalia (1987) explored a general
strategy-training approach involving the use of procedural facilita-
tion and goal concretization. The steps that they recommend are:
■ Define a self-regulatory system which students should learn to
control.
■ Design a way of cuing the type of expert processing in which
student should be trained.
■ Set up a teachable routine which externalizes the process and
allows students to learn it.
One example involved giving sixth-grade students a set of cards

with strategy prompts for writing. For example, when students
arrived at a block, they would draw from a pile a card suggesting a
possible further direction; e.g.
‘an even better idea is
‘an example of this
‘I’m getting off topic so ...’
‘my purpose...’.
There were approximately 30 cards with instructions (for each
genre) and they were randomly ordered. The point was to provide
students with a cue that would foster their reflective thinking
about their writing, and would interrupt the knowledge-telling
flow. A similar training strategy was used to develop skills in revi-
sion. Students would examine a set of guidelines for comparing
their text with their composing goals. If a problem was seen in the
comparison stage, for example, a set of guidelines was available to
the students to suggest a revision strategy for fixing the problem.
The goal was to make students aware of revision processes and to
encourage them to practise appropriate strategies. In general,
results showed that students could learn to use these strategies in
their essay writing; the limitation was that these strategies did not
seem to transfer readily to other writing tasks which were not sup-
ported by the cues.
This limitation is typical of many strategy-training approaches
for writing. (See also Pressley et al. 1994 for procedural training
involving reading strategies.) Students must:
■ first become aware of the limits of their performance;
■ then become aware of specific ways in which their performance
is limited (some means for comparison);
■ then learn ways to resolve the problems in their performance
(guided by some technique);
■ then learn to transfer their improved performance to tasks not
directly supported by assisting cues.
Many strategy-training studies have demonstrated success in teach-
ing students specific techniques and strategies, but few have
demonstrated that the techniques can be transferred to contexts
beyond the specific training environment, or that the training has
an effect on performance in post-training tasks (Pressley et al
1992).
The general goal of having students become aware of different

performance criteria, having them learn to use these different per-
formance criteria, and then having them transfer these criteria
independently to other contexts has been the purpose of much
work under the name of metacognitive strategy research. Following
Bruner’s (1983) notion of scaffolded instruction, and Vygotsky’s
(1978) notion of the ‘zone of proximal development’, researchers
such as Brown (1994), Collins et a l (1989), Newman et a l (1989),
Palincsar and Brown (1984), and Pressley et a l (1991, 1992, 1994),
and have sought to move students through to independent use of
new strategies by having them become co-constructors of knowl-
edge and co-users of more expert strategies.
A major problem with the application of problem-solving train-
ing procedures to writing instruction is that writing itself tends to
be what some researchers call ‘ill-defined problem solving’; that is,
situations in which there are no single well-defined solutions.
Spiro et a l (1987) argued that notions such as ‘cognitive flexibil-
ity’ and ‘cognitive transfer’ are not well understood for ill-defined
problem-oriented tasks such as writing. In ill-defined contexts of
cognitive strategy use, it is not clear just what skills are transferable
to new contexts, or how this transfer can be accomplished.
Based on a review of research, Spiro et a l (1987) argued that,
for ill-defined knowledge to be transformed, students need to be
exposed to a variety of related situations in which they must assem-
ble a new schema structure rather than rely on established schema
structures (see also Bereiter and Scardamalia 1993, McGinley and
Tierney 1989). The best approach for developing this ability lies
in exposure to many variations of a content domain; that is, stu-
dents are presented with many related cases and contexts, all
bearing some relationship to each other. Spiro et a l suggest that
students will learn to handle ill-structured tasks best when first
introduced to a content via the presentation of a task with an
intermediate degree of ill-structuredness. (For example, a situa-
tion may suggest more than one good solution, yet each solution
can be demonstrated as reasonably effective.)
5.2.6 Balancing process and form

As suggested by Spiro et a l (1987), one of the major issues which
appears repeatedly in strategy research lies in the need for stu-
dents to be aware of the constraints of the type of writing they are
being asked to perform. Bereiter and Scardamalia (1987) point

out the different requirements of narrative and expository writing
and the need for students to write for different purposes to
develop problem-solving skills. Carter (1990), in his discussion of
writing expertise, points out the importance of extensive practice
in a wide variety of writing tasks which will require different strat-
egy use. Similarly, Swales (1990) has pointed out the need for
genre knowledge as an important part of students’ knowledge
about writing if they are to be academically successful. In the next
section, the notion of genre as it applies to writing development is
explored. Genres represent ways that groups of writers have man-
aged to solve problems in writing which conform to general
expectations, which facilitate communication for both the writer
and the reader, and which provide students with frames suited to
communication about different sorts of knowledge and different
ways of addressing the reader.
The renewed emphasis on genre also represents a corrective
reaction to an earlier emphasis on process as a simple set of proce-
dures for all writing tasks. Applebee (1986), Bereiter and
Scardamalia (1985), Christie (1985), and Martin (1989) have all
addressed the difficulties involved in the typically oversimple
application of a process approach - an approach which has led to
sometimes disappointing results despite strongly affirmative state-
ments (cf. Applebee et al. 1994). All argue that the disregard for
formal aspects of writing and the problem-solving required by dif-
ferent genres of writing, seriously limit students’ writing abilities.
They argue, rather, that it is necessary to recognize the wide array
of tasks which will extend students’ cognitive abilities.
The work of applied linguists and educational researchers in
Australia has extended this argum ent to a recognition that lan-
guage form, as it reflects genre and writing purpose, is also a
critical com ponent in writing development - one which has been
ignored in the false dichotomy between skills and content instruc-
tion (Cope and Kalantzis 1993; see also Coe 1994). These scholars
do not downplay the important role that writing processes play in
writing research and instruction, but they do see the notion of the
writing process, as currently interpreted in instruction, as a
severely limiting view (Christie 1985, Maclean 1989). Instead, they
propose that the incorporation of form should lead to a balance
between form and process in a functional approach to writing
development. Through a focus on the functions of language use,
the goal is to account for the cognitive and social influences on

writing as realized through the genre form. In the next section,
this perspective will be examined, and it will be argued that the
development of genre knowledge, in Halliday’s functional sense,
plays a particularly im portant role in a student’s writing develop-
m ent beyond the knowledge-telling stage. The Australian
perspective on writing development is particularly im portant as an
argum ent for student empowerment through writing - a notion
which has yet to be given careful consideration outside Australia .
5.3 Genre-based approaches to writing development
5.3.1 A Hallidayan view of language as function

A view of language which is important to any understanding of
writing development is that derived from Halliday’s functional the-
ory of language. In this approach to language, grammar develops
out of the need for speakers and writers to interact for functional
purposes; that is, language development evolves out of a child’s
‘learning to m ean’ (Christie 1989). In this view, language is not
separable from content or context, but varies systematically with
content and context, and is the medium through which meaning is
realized. Halliday’s emphasis on meaningful use of language in
educational contexts is a perspective adopted by whole-language
theorists in the USA. However, a Hallidayan orientation to mean-
ingful language use in context extends further to include the ways
in which language form gives structure to meaningful communica-
tion. As Christie states, ‘Success in mastering a content area is
actually a matter of mastering the necessary linguistic resources
with which to deal with that content - this implies knowing how
one’s discourse is to be structured’ (1989: 167).
Unlike the assumptions of American generative grammarians - in
which language form is seen as independent of language use -
Halliday sees language form and meaning-making as an integrated
system. For Halliday, language form is functional in nature; it is
meant to serve functional purposes. As Painter (1989: 21) notes,
Language is a functional resource in th at the language system as a
whole can be viewed as having the form it does because o f w hat it is
called up o n to do - the needs o f language learners have shaped the
linguistic system itself.
Children’s first language abilities develop out of the growing need

to communicate, first with parents, then with the expanding world
of hum an contacts. As children learn to interact and manipulate
their environment through language, language forms expand to
m eet new functional needs. In this way, children are engaged in
making meaning, and they do so through language forms which
serve children’s abilities to make meaning.
Students learning to interact orally and in writing need to come
to understand how language form and generic text structure pro-
vide resources for presenting information and interacting with
others. They learn to choose linguistic patterns which are appro-
priate to the meanings they are trying to make. One of the goals
of school, then, is to help students to recognize and use those lin-
guistic patterns. As Christie (1989: 198) states:
Knowledge is created in patterns of discourse, and successful mastery
of a ‘body of know ledge’ like the associated developm ent of various
m ental skills of reasoning, speculation, and inquiry, is entirely d ep en -
d en t upon mastery o f linguistic resources necessary for these to come
into being. It is for this reason, contrary to prevailing custom, that
good teaching practices should always have an overt and explicit
interest in the nature of the language students m ust learn to use.
Making meaning is the production of discourse and of texts. The

extension of this perspective is that students in school situations
also make meaning - and learn - through their use of discourse
and their construction of texts. Such a view argues that writing
and attention to form, as part of content exploration, is critical for
learning. This perspective is argued for forcefully by Halliday
(1993b) and Martin (1989), an argum ent which will be the subject
of the next section.
5.3.2 Extending language use to genre knowledge

Martin (1989), examining the causes for success and failure in ele-
mentary school literacy development, posits an argum ent for
extensive instruction in factual writing, and, in particular, exposi-
tory writing - writing which explores how the world works. He
contrasts this view directly against writing instruction which
focuses almost entirely on expressive and narrative writing - writing
which describes how the world looks (see also Bereiter and
Scardamalia 1987). This argument does not discount the impor-
tance of narrative and expressive writing as a way to provide
scaffolding for students as they make the transition from spoken

to written language use. Indeed, the utility of familiar genres for
learning to write is well recognized. Rather, the issue is whether or
not early school curricula recognize that students must practise a
range of genre forms in writing.
Assuming a functional view of language following Halliday,
Martin argues that school and teachers tend to reinforce the sta-
tus quo of success and failure by ignoring the connection between
language use and social purpose, and by disregarding any specific
analysis of language form which would result from children’s
efforts to work with information, whether through speech or
through writing. Learning how to use writing for understanding
the world offers, instead, the possibility (but not the guarantee)
for change and for student success.
Martin (1989) sees the learning of factual writing skills through
content studies as a way of giving students power over their own
learning. Students who can infer the requirements of controlling
information, in part because they have experienced a larger
socialization process which will allow them to do so, will succeed;
but other students who are less familiar with the expected genres
and their functional purposes will continue to fail in schools. By
this reasoning, the goals of school should include explicit instruc-
tion in those forms of language which many students will not be
able to infer indirectly. Students learn to control the information
presented in school by solving problems in integrating language,
content, and context, and by practising the types of writing which
demand this integration, not by writing practice in other genres
which do not.
In order to explain the notion of factual writing and the impor-
tance of genre for students’ writing development, Martin has
developed a framework for factual writing. In this framework, writ-
ing genres are identified which relate to functional goals of
knowledge presentation (see D’Angelo 1987, de Beaugrande
1984: 100). For Martin (1989), types of factual writing in which
elementary students should engage include:
Recounts (a specific event-related presentation);
Procedures (a general event-related presentation);
Descriptions (a specific object-related presentation);
Reports (a general object-related presentation);
Explanations (a specific argument on an issue, event, or object);
Exposition (a complex sequence of multiple explanations).
In a similar classification of factual genres which should be incor-

porated into a curriculum, Derewianka (1990) suggests a slightly
different set of options:
Recount (What we did?/W hat took place?)
Instructions (How something is done)
Narrative (Recounts with a twist)
Information report (What an entire class of things is like)
Explanation (Why is it so?/How does it work?)
Argument (Stating your case).
In both frameworks, the stress is on children learning how lan-
guage functions to present content in ways which are most
appropriate to a writer’s purpose. Neither framework refers to
matters of grammar practice or style exercises which might be
divorced from working with meaningful curriculum content
(Collerson 1990). Rather, in the process of exploring content, students
learn how to present the content through effective language use
and genre form. The critical point is that some notion of student
support from the teacher - either through such notions as lan-
guage apprenticeship (Collins et al 1989), scaffolding (Bruner
1983), or the practice of reciprocal teaching (Brown and Palincsar
1989) - should be extended to include the functional role of lan-
guage form as text (Christie 1989).
5.3.3 The importance of genre

To understand how a focus on language is not a matter of atten-
tion to arbitrary form, but is central to meaning-making, it is
necessary to pick up the argument presented in subsection 5.3.1
above - the argum ent that language and content are integrated in
the sense that language serves the needs of the writer to make
meaning. Kress (1989) develops the im portant role of genre as an
essential factor in language use by noting that all language use is a
m atter of making discourse. The making of discourse depends on
differences between speaker and listener, or writer and reader. As
certain discourses become more deeply embedded in the social
functioning of groups, these discourses become conventionalized;
they become recognized as genres which serve functional pur-
poses in communication.
Students in school contexts are expected to make use of genres
for learning information to the extent that they see how such
genres serve functional purposes, and to the extent that genre-

structures are made apparent to students. Assuming a close
relationship between language use and learning purposes - as a
Hallidayan approach would - students need to be taught to work
effectively with the language medium appropriate to the learning
task and informational content. When this does not happen, as
posited by Kress (1989) and Martin (1989), learning purposes are
obscured for those students who do not have the advantage of
socialization which reflects school-based expectations (see also
Heath 1986b, Painter 1989).
An argum ent has been raised at times that teaching students
genres will degenerate into teaching arbitrary models and textual
organization with little connection to a student’s learning pur-
poses (e.g. Freedman 1993, Silva 1990). While this can in fact
occur - e.g. in such pedagogical procedures as the teaching of the
three- or five-paragraph model - it is not the argument being pre-
sented here. Genre is not being presented as an end for
instruction, but rather as a means to understanding meaningful
content. Genres must be taught, understood, and critiqued in
terms of the potential they provide for working with informational
content and learning context. Since genres also reflect a cultural
ideology, the study of genre additionally opens for students an
awareness of the assumptions of groups who use specific genres
for specific ends, allowing students to critique not only the types
of knowledge they learn but also the ways in which knowledge is
valued and in which it reflects covert assumptions.
To be successful, then, students must learn how language works
to convey content through school-valued genres. In a sense, they
must learn the schemas for organizing different types of knowl-
edge, as well as those for presenting different types of information
(Cope and Kalantzis 1993, Mohan 1986). This is only likely to hap-
pen in contexts in which students get consistent practice with
different types of writing tasks, and in which teachers can point to
the language structuring in different genres for highlighting dif-
ferent ways of making meaning. Spiro et al (1987) suggest that
practice of this type would provide sufficient training in dealing
with relatively ill-defined problem-solving tasks. A slightly different
perspective is suggested by the various Australian researchers who
would agree with the need for extensive practice, but would argue
that explicit instruction is also needed to show how language
serves meaningful communication.
5.3.4 Extending genre use to academic contexts

Looking at the opposite end of the writing development spec-
trum, Swales (1990) also argues for the importance of genre as
having functional purposes in advanced writing development. He
similarly points to a knowledge of genre as an im portant way to
give students power over their learning. Taking as a starting point
the notion that genres serve meaningful purposes and provide
frames for various academic schema-based sets of knowledge, he
contends that the ability to use genre structure effectively will
transform students’ abilities to learn and function successfully in
academic contexts. Swales, referring to Martin and Rothery (1986;
see also Martin 1985), points out that genres provide ways to con-
strain information in socially recognized forms (e.g. a letter of
regret is not a good genre for inviting a person to a party). These
genre forms are intimately tied to cultures and discourse commu-
nities. Thus, as Swales states, ‘genres comprise a system for
accomplishing social purposes by verbal means’ (1990: 41). For
students to gain power over particular knowledge, and to gain
recognition in particular discourse communities, the ability to use
genre forms effectively is critical (see also Coe 1994).
Throughout this discussion of genre, it is im portant to recog-
nize the connections to all levels of writing development and
instruction. It is also im portant to see the relationships to research
on: the cognitive psychology of writing, cognitive strategy develop-
m ent and student difficulties in literacy developments. If genre is
the locus of local cultural assumptions, of language structure, of
writer’s purpose, of informational content, and of register con-
texts, then it is evident that genre is a key notion in writing
development, and in learning through writing. As students learn
to control genre, they concomitantly learn to control language,
writing purpose, content, and context.
5.3.5 Knowledge of generalized genresfrom rhetoric

A natural tendency is to look for superordinate categories for vari-
ous possible genres; this is a tendency which apparently has deep
historical roots as attested by the field of rhetoric. In the twentieth
century, such tendencies have led to a general educational recog-
nition of categories such as narrative, exposition, argumentation,
description, etc. This rhetorical perspective assumes that there are
different types of discourse which have systematic formal features

and reflect ways of organizing thought.2 The use of organizing
principles to guide discourse and teach patterns of arrangement
has been in existence since Aristotle. Much more recently, in the
eighteenth and nineteenth centuries, these principles were pre-
sented as representative of basic aims of rhetorical organization,
leading to Bain’s widely recognized classification: description, nar-
ration, exposition, and argumentation (Berlin 1984, Crowley
1990, D’Angelo 1987).
Among the currently discussed theories of discourse arrange-
ment, those of Britton et a l (1975), D’Angelo (1975, 1987),
Kinneavy (1971, 1983), and Moffett (1968) are the most promi-
nent. All propose conceptual theories for kinds of discourse and
functions of discourse arrangements, though there are enough
distinctions among the theories to make a synthesis difficult (see,
e.g., D’Angelo 1987, Kinneavy 1983). In all cases, however, the the-
ories conceive of writing as influenced by the functional roles of
the writer, the reader, the text, and the world (as opposed to the
traditional genres of Bain which were not functionally motivated).
Kinneavy, for example, translates these roles into four basic types
of written discourse: expressive, persuasive, literary, and referen-
tial. These basic types are supported by modes of written
discourse: description, narration, classification, and evaluation.
These theories of discourse types have had a powerful impact
not only on research issues relating to formal features of discourse
but also on writing pedagogy. In the USA, most English writing
textbooks (ranging from elementary language-arts texts to ter-
tiary-level writing textbooks), in both LI and L2 contexts, base
their design on some set of discourse types (the resulting peda-
gogy often referred to disparagingly as ‘current-traditional’
m ethod [Berlin 1987, Silva 1990]). The specific relevance of these
rhetorical theories is that they provide general classification
heuristics for formal organizational patterns which are intuitively
appealing even if they oversimplify or ignore the importance of
purpose, audience, attitudes, and topic in shaping discourse.
There is some evidence for the larger and more abstract classifi-
cation of genres (or rather, meta-genres) - e.g. expository,
argumentative, and narrative, depending on the context in which
these terms are applied. A considerable amount of research, con-
ducted in educational contexts, has demonstrated repeatedly that
students work with texts, defined as distinct by this classification
system, in quite different ways (e.g. Crowhurst 1987; 1990, Meyer

1987, Richgels et al. 1987, Spiro and Taylor 1987). It has also been
demonstrated that careful linguistic analyses of such meta-genres
reveal significant and consistent differences in the ways that lin-
guistic features co-occur and define these genres. W hether or not
such general classification is as useful as the strictly functional
genre schemes of Martin or Derewianka, or the advanced acade-
mic genres addressed by Bhatia (1993), Swales (1990), and others,
is an open question.
5.4 Research on second language writing

In the English-speaking countries of the world, increasing num -
bers of students find themselves in an educational system that
does not provide them with extensive support as L2 learners. In
the USA, for example, by the year 2000, there will be over 5 mil-
lion English L2 speakers in the school systems (Bernhardt 1994).
Yet relatively little effort is made to distinguish these students
from English LI students for purposes of instruction, or to recog-
nize different student needs. In the area of writing research in
particular, L2 writers have only recently been discussed (Leki
1992, Silva 1993). However, with so many students writers of
English being L2 students, it is essential that their situations and
their needs be given specific consideration. This section considers
the research issues that have been investigated with L2 students.
5.4.1 Supportfrom, and for, LI research

Throughout much of this chapter (and elsewhere), the discussion
has not focused specifically on learning to write in a second lan-
guage. The primary goal has been to assemble the various strands
which define the nature of writing in general. At various points in
this discussion, L2 research which supports and corroborates the
findings of LI research has been cited. It is important, however, to
address:
■ the ways in which L2 contexts create unique concerns;
■ the possible forms of these concerns; and
■ the research that has been done to address these concerns.
Until quite recently, L2 research on writing has been closely
dependent on LI research. Perhaps only in three areas - con-
trastive rhetoric, English for Specific Purposes, and writing assess-

m ent - has L2 research followed independent paths. (The first
two are addressed in Chapters 6 and 7; and writing assessment will
be discussed in Chapter 13 [see also Hamp-Lyons 1990, 1991d].)
One of the major catalysts for the rapid emergence of research on
L2 writing has been the growing numbers of foreign and lan-
guage-minority students entering English-speaking tertiary-level
institutions. Perhaps because of the ubiquitous freshman composi-
tion requirements at US colleges and universities, L2 writing
research has been quite prom inent in that country. Also, in the
USA and in Australia, a secondary emphasis on the difficulties fac-
ing language-minority students has become more evident. These
two contexts for research have generated a large proportion of L2
writing research; not surprisingly, most of this research stresses the
similarities rather than the potential differences between LI and
L2 learners.
Relying predominantly on case studies with small groups of writ-
ers, research on composing processes and on revision strategies
has shown that L2 students often behave rather like less-skilled LI
writers, as described in the LI research. In research on composing
processes, it appears that L2 writers make use of the same sets of
composing processes but, for various reasons, many L2 writers
apply these composing processes with less ability than is shown by
LI writers (Raimes 1985, Zamel 1983, 1985). Raimes (1985)
argues that writing difficulties for ESL students typically are
largely the result of difficulties with composing skills than with lin-
guistic skills. A second similarity, pointed out by Raimes (1985),
lies in the fact that L2 students may be driven by the need to gen-
erate content and write it as they initiate it (see also Bereiter and
Scardamalia 1987). Edelsky (1986), Gaskill (1986), and Jones and
Tetroe (1987) have also argued that LI writing processes transfer
to L2 writing. In addition, it has been argued that process-
oriented instruction has led to similar student progress, whether
learners are working in first or second language, particularly at
the elementary grades (Hudelson 1989b, Urzua 1987). In large
measure, the strong claims made for close similarities between LI
and L2 writers centre on research on composing processes (see
Krapels 1990 for an overview). This research orientation may
reveal im portant similarities; and this line of inquiry may also
mask evidence that would show L2 writing to differ more from LI
writing than has been suggested in writing process research.
5.4.2 Writing differences: LI vs L2

Apart from the clear differences in L2 writing illustrated by
research in contrastive rhetoric and ESP, and by assessment
research, a num ber of specific findings from recent research sug-
gest that L2 writing has a great many differences from LI writing -
more, perhaps, than it has similarities.
First, it is not clear that L2 writers who write at greater length or
with greater frequency typically write better; unlike the generaliza-
tion about LI writers, there is evidence which argues that weaker
L2 writers sometimes write more (Reppen 1995, Reppen and
Grabe 1993). This is a topic that needs further investigation since
a connection between writing quantity and writing quality is often
assumed in certain research studies (e.g. Carlisle 1989, Ferris
1994).
Second, it is widely recognized that LI students have some
implicit knowledge of rhetorical plans, organizational logic, and
genre form in their nadve language; it is not at all clear that stu-
dents have the same implicit knowledge with respect to L2, as is
indicated in the contrastive rhetoric research. This distinction may
actually be greater for research which assumes that writing natu-
rally emerges from students and cannot be taught directly (e.g.
Jones 1985, Krashen 1984). A Hallidayan perspective, by contrast,
would argue that many LI students may not be so different from
L2 students in the sense that many LI students have an inade-
quate knowledge of textual structuring and need the same sorts of
direct instruction on written discourse form that are required by
L2 students.
A third difference between LI and L2 students emerges in stud-
ies of revision. While LI research suggests that most revision has
little effect on writing improvement (Hillocks 1986; cf. Freedman
1987), feedback of various kinds (on form, on content, on both)
appears to be useful for improving the essays of certain L2 stu-
dents (Fathman and Whalley 1990). It has also been found that
teachers’ general comments referring to content and organization
seem to be useful to L2 learners. Similarly, Raimes (1985) found
that L2 students were less inhibited by teacher-editing and feed-
back. In effect, L2 learners appear to expect and accept greater
intervention, and to make greater improvements when they get
such feedback (Radecki and Swales 1988). Because there is not
sufficient L2 research to explore potential differences between LI
and L2 learners in revision strategies and the influence of feed-

back, this is an area that needs additional careful study (e.g.
Cohen and Cavalcanti 1990, Fathman and Whalley 1990, Ferris
1995, Leki 1990).
A fourth area of difference between LI and L2 students lies in
the impact and influence of the LI on writing in the L2. Clearly,
this is an added factor for L2 writing which is not present in LI
writing. Apart from the potentially positive influence of strategy
and process transfer from the LI, the L2 learner also faces con-
straints deriving from potentially limited linguistic abilities in the
L2. It is clear that limited knowledge of vocabulary, language
structure, and content constrains a L2 writer’s performance. The
entire question of the impact of level of language proficiency on
writing in the L2 has yet to be examined in careful detail (see,
e.g., Carson et al. 1990, Silva 1993). While Charney and Carlson
(1995), Hillocks (1986), and Smagorinsky (1992) all make a case
for the use of models and modelling in LI writing instruction, as
do the Australian proponents of genre-based curricula, the lim-
ited linguistic abilities of the L2 writer may also account for the
finding that writing models can have a beneficial effect for L2 stu-
dents in particular.
5.4.3 Towards a model of L2 writing

The results to date in L2 writing research suggest that:
■ there are many contradictory findings;
■ many results and interpretations of research reflect expected
findings and conform to LI research results;
■ L2 research must be developed more fully through carefully
developed studies and research designs, and in more carefully
controlled research situations.
Silva (1992, 1993) forcefully makes the argum ent for more con-
trolled research specifically addressed to L2 writing contexts. In
particular, he argues that theories of LI writing will not be ade-
quate for advances in L2 writing. Moreover, current L2 research is
primarily exploratory, limited in numbers of participants, uneven
in quality, and not accounting for the full range of factors which
influence the L2 context. These limitations on research are
reflected by a limited view of L2 writing theory. It is necessary to
develop a theory of L2 writing, making use of LI research when
such research is relevant, but incorporating the unique features of

L2 writing.
For Silva (1990: 19), such a model of L2 writing,
is one th at ..., at a m inim um , regards writing as an interactive activ-
ity; is reasonably com prehensive an d internally consistent; reflects
an understanding of historical developm ents in the field; is
inform ed by c u rren t work in relevant disciplines; and is sensitive to
the cultural, linguistic, and experiential differences o f individuals
and societies.
Silva (1990: 19) suggests that a model of L2 writing must incorpo-

rate the following five components:
(i) L2 writing theory,
(ii) research on the nature of L2 writing,
(iii) research on L2 writing instruction,
(iv) L2 writing instruction theory (approach), and
(v) L2 writing practice.
The extent of interaction among these components constitutes an
open questions, as does the sort of research which is capable of
supporting each of the five components. A second way to develop
a theory of L2 writing may lie in a reconsideration of the ethnog-
raphy of writing to be described in Chapter 8, and in the way in
which ethnographic information can be used to explore differ-
ences between LI and L2 writing, as well as the extent to which
such ethnographic data can generate testable hypotheses.
A model of L2 writing based on an ‘ethnography of writing’
approach would ask what is known about the various major com-
ponents of the writing situation:
the writer; the writing;
the reader; the content and form of the text;
the purpose; the writer’s underlying intention;
the location; the processes of writing.
One possible outcome of such an undertaking would be an argu-
m ent that L2 writing is so varied and distinct from situation to
situation that a general theory of L2 writing may not be possible
(e.g. Cohen and Cavalcanti 1990), at least not with the current
state of knowledge. Exactly how a descriptive theory of L2 writing
would differ from a descriptive theory of LI writing is an issue
going beyond the scope of this work, though we believe that an
effort of this type would be a step towards the generation of a the-

ory of L2 writing.
5.5 Moving from the cognitive to the contextual
This chapter has sought to bring together more current perspec-

tives on cognitive approaches to writing processes and, at the same
time, extend the current research in two directions that are not
typically discussed in much writing research. The notion that the
text form itself is re-emerging as an im portant component in the
study of writing follows from cognitive research on expertise as
well as from Bereiter and Scardamalia’s discussions of the
resources required to transform knowledge. It is also im portant to
recognize that many students learning to write in different con-
texts around the world are L2 students.3 Current LI writing
research gives minimal attention to these students, but there are
im portant research issues to consider, primary among them being
whether students learning to write in a second language have dif-
ferent skills, resources, needs, and expectations (see, in particular,
Kachru 1992, Sridhar 1994). Research on L2 students is needed to
assess the generality of claims made in the current LI writing liter-
ature. Moreover, to the extent that L2 students reveal differences
from LI students, research must explore these differences and
their implications for both first and second language instruction.
In the domain of cognitive research on writing in the LI,
researchers are increasingly recognizing the importance of differ-
ent contextual variables in understanding writing development. In
particular, the recent work of Flower (1989, 1994) demonstrates a
strong movement to consider the social contexts for writing that
operate in conjunction with the cognitive efforts of the writer.
This trend is an im portant corrective to the earlier discussions of
the Flower and Hayes model. Any appropriate descriptive theory
of writing will have to account for the many social contexts that
influence writing in various ways. This trend is also seen in recent
discussions of writing theory by Cope and Kalantzis (1993), Dyson
(1993), Kress (1994), and Witte (1992).
In the field of L2 writing and applied linguistics, there are two
additional contexts for writing which do not commonly receive
treatm ent in LI writing discussions. These are the areas of English
for Specific Purposes (ESP) and contrastive rhetoric. In both
cases, the interests of applied linguists are obvious and, perhaps,

the relative disinterest of LI writing research is understandable.
The clearest parallel for LI research is the growing interest in writ-
ing in advanced academic settings and in non-academic settings.
Current research is exploring how these contexts alter the needs
for writing, change patterns of use, and suggest different options
for writing instruction. In the next two chapters, these areas - writ-
ing in special-purpose contexts and contrastive rhetoric - will be
explored in detail and their contexts examined for the additional
dimensions they bring to our understanding of writing.
Notes
1. T he knowledge-telling m odel of writing is also used by m ature skilled

writers w hen it is sufficient to p erfo rm the task required (e.g. writing a
com m on m em o many tim es).
2. It should also be n oted th at many contem porary rhetoricians object to
this assumption.
3. This is, they are learning to write in any (national/official) language
o th e r than their first language; e.g. native speakers o f C ebuano and
Illocano in the Philippines learning to write in Filipino (P ilip in o /
T agalog), or native speakers of Tamil or Gugarati in India learning to
write in H indi. T he L2 is n o t necessarily English, and the issue is th at
little is known about the ways in which such teaching is actually con-
ducted or the ways in which learning is accom plished.
6
Writing for professional purposes
6.1 Introduction
From the discussion of the previous chapters, it would be easy to

get the impression that writing only occurs in school, or is only
used for schooling purposes. In the past 15 years, however,
researchers concerned with writing and literacy have now recog-
nized that most uses of writing are tied to situations which extend
well beyond the educational context. Writing is now studied in
many professional and technical contexts, as well as in the work-
place. The general goal of such research is to understand the
sorts of writing that are actually used in contexts outside of
school, to examine the connection between writing in the school
context and in other contexts, and to determ ine whether the
relation between school writing and writing in other contexts is
productive - that is, are students being trained in writing so that
they will be able to write effectively in other contexts once they
leave school?
This chapter will explore these questions by examining in some
detail four areas of writing specialization:
■ technical writing as taught in post-secondary courses;
■ writing actually perform ed in non-school contexts and its com-
parison with writing in school contexts;
■ writing approaches in English for Special Purposes (ESP) ;
■ a social constructionist view of writing in professional contexts.
These issues also raise such im portant concerns as the role of
English in professional communication world-wide and the need
to incorporate significant audience, topic, and purpose variation
in instructional approaches.
147
6.2 The teaching o f technical writing
In the USA, at the tertiary level, technical writing is taught to stu-

dents studying certain technical or pre-professional fields. These
courses are sometimes taught in English departments by technical
writing specialists; sometimes the courses are taught by the respec-
tive departments wanting their students to become familiar with
format and genre norms of that technical/professional/business
discipline. In each case, the basic assumption is that students need
to learn special ways to write (e.g. letters, memoranda, reports)
which are not taught in more general writing courses. It is also
often suggested that students are taught to break the bad writing
habits of secondary and tertiary instruction, inculcated by teach-
ers who do not understand how to be clear, objective, and concise.
In order to understand this seeming prejudice in writing instruc-
tion, it is necessary to review certain guiding assumptions in
technical writing instruction.
It is first im portant to introduce a num ber of distinctions. The
first of these is the distinction between writing as a technology and
writing for scientific and technological purposes. Often one finds
the assumption being made that writing can be reduced to a tem-
plate - to a set of techniques, or a set of simple step-by-step
procedures which will guarantee ‘correct’ and objective language
(see, e.g., Dobrin 1989, Winsor 1990). This notion of writing as
technology is not the same as the notion advanced in Chapter 1,
where writing was described as a technology in the sense that it is a
learned skill rather than a biologically conditioned one transmit-
ted through the hum an genetic system. Language-as-technology,
as it is introduced and discussed in this chapter, refers to an
assumed set of guidelines, a general ‘discovery procedure’ for
writing ‘objective’ prose.
A second distinction that must be made concerns the use of writ-
ing in scientific and technical fields versus the preparation of
documents to be used in instructions, letters, manuals, and other
explanatory material. In the former case, there is a lower percent-
age of routine writing tasks and more opportunities for planning,
collaborative revision, oral interaction, and presentation. In the lat-
ter cases, technical writing involves the use of certain formats and
rules of usage which may or may not be as effective as the technical
writing textbooks would lead one to believe. This latter approach to
technical writing would cover such conventional uses of writing as:
Writingfor professional purposes 149
■ memos for business purposes

■ business reports
■ guidelines for the use of heading, tables, and other illustrative
material
■ professional letter-writing conventions: salutations, formats,
closings
■ effective grammatical usage rules.
In fact, much of technical writing, as taught in the USA, involves
the sort of instruction that provides students with templates for
various non-academic genres. In essence, these courses constitute
training in document design and bureaucratic usage, and less
emphasis is placed on the actual uses of writing that scientists and
technical researchers employ as they work in their professional
environments. While it is important to fashion texts into standard
formats perceived as reader friendly in particular environments, it
is still necessary to solve rhetorical problems central to the con-
tent rather than the forms of various types of technical texts
(MacKinnon 1993). The more formulary approach of much disci-
pline-related writing instruction assumes relatively simple
form-function equations which have not been established and in
fact are not likely to be possible.
While there is a certain amount of pro forma discussion of scien-
tific and technical writing - typically the laboratory report, the
scientific m ethod paper, and the grant proposal - relatively little
time is given to the study of what professional scientists and tech-
nical researchers do when they write. Nor is much consideration
given to research on writing in the commercial workplace (cf.
Reither 1993, Reynolds et al 1995, Spilka 1993a). This chapter, in
contrast, will be much more concerned with the writing that is
actually done in professional contexts outside of school, and will
essentially ignore document design issues and the constraints of
bureaucratic text production.
A third distinction lies in writing for research and publication
purposes versus writing in other professional contexts. For the
purposes of this discussion, the creative use of writing for the dis-
semination of research information will be treated as more
im portant than the uses of writing which appear to be more pro-
grammatic and which reflect the day-to-day invocations of a
similar set format such as the business memo, the rejection letter,
the insurance claim, the police arrest report, etc. (cf. Matalene
1995).
Questions of information access and information transfer -

im portant issues in applied linguistics - centre to a greater extent
on research reports, grants, and professional publications than
they do on the day-to-day programmatic types of writing in non-
academic contexts. In particular, the focus on the presentation of
research findings for wider dissemination has the status of a differ-
ent type of writing than do other kinds of writing that typically
occur in work or school contexts. Although professional research
writing also is assumed to conform to certain conventions for
objective reporting, it will be argued that scientific writing does
not simply follow a set of mechanical writing procedures for
reporting results, though the articles which represent such
research typically appear to conform to a set of writing proce-
dures. Rather, scientific writing is as subjective and as rhetorically
conscious as other forms of writing, and may be even more rhetor-
ically sensitive in many situations (Geisler 1994, Kaplan and Grabe
1991). At the same time, research on the full range of writing
done in non-academic contexts provides im portant insights for
instruction; this matter will be considered in the next section.
Each of the three distinctions noted above influences the cri-
tique of writing instruction in tertiary courses. The teaching of
technical writing in tertiary-level courses, in general, seeks to
address certain uses of writing for technical and scientific pur-
poses. Such courses teach basic formats and writing conventions
which are widely recognized as marking writing in technical sub-
ject areas and genres. The courses stress the careful definition of
terms, the avoidance of ambiguous terminology, and an imper-
sonal/objective style of writing (‘neutral’ with respect to the
attitudes of the writer). On a relatively abstract level, such criteria
seem to be appropriate goals, though recent research would argue
that technically oriented writing may be much more complex than
these generalizations suggest (Spilka 1993b). Much technical writ-
ing instruction assumes that:
■ information may be ‘transferred’ from the writer to the reader
(in the sense that fresh produce is ‘transferred’ from the grower
to the consumer - a situation in which ‘treatm ent’ is minimal
and packaging is critical);
■ the reader can understand information better if it is broken
down into small, step-by-step, incremental explanations - if it is,
in a naive way, ‘program med’;
■ information exists ‘out there’, apart from the writer or the

reader, independent of interpretive background knowledge;
■ writing should follow an outline;
■ paragraphs have a structural shape, invariably introduced by a
topic sentence.
In a critique of this perspective, Dobrin (1989) examined the fun-
damental assumptions guiding much technical writing instruction
and concluded that many of these assumptions do not explain
how technical writing in the workplace is actually done; or, when
technical writing does follow this line of instructional training,
why it is done badly (cf. Reynolds et al 1995). Dobrin sees the
technical writing approaches practised in schools as making a
num ber of assumptions about writing which can also be traced to
cognitive writing models such as those following Flower and
Hayes, particularly the assumption that writing is a straightforward
problem-solving activity. As noted in Chapter 4, such an assump-
tion may ignore much of the social context in which writing - any
writing - occurs. For this reason, Dobrin argues that the notion of
technical writing as technology or technique is misguided. Further,
the notions of purely objective writing, unambiguous terminology,
and clear definition are impossible without bringing in the situa-
tional intentions of the writer and the interpretive assumptions of
who the reader may be. Dobrin also argues that the notion of infor-
mation transfer, when applied to the technical writing process, is
inappropriate because it suggests that writing can be reduced to a
mechanical translating procedure; research on writing in the work-
place, however, demonstrates that there is considerable ‘noise’
from (1) the writer’s background, (2) the intention of the message,
and (3) the lexico-syntactic choices made.
From his perspective, Dobrin suggests that technical writing is
distinct from other sorts of writing, but that certain features of
technical writing cannot be mechanically isolated as critical defin-
ing signals of the genre or subject matter. Rather, technical writing
is as difficult to understand and teach as any other type of writing;
its understanding and productive skill derive from knowledge and
experience, from steady practice, and from frequent failure. One
should not look to technical writing instruction as some sort of
panacea that has somehow escaped other writing teachers and
researchers (see also Anderson 1985, Doheny-Farina 1986,
MacKinnon 1993, Odell 1985).
A better understanding of writing in non-educational contexts

involves examining the actual writing and, where possible, study-
ing the writing processes being used by writers in these contexts.
6.3 Writing in non-academic contexts
In the 1980s, a num ber of writing researchers recognized the

need to understand how well, or how badly, US tertiary students
were being prepared for the writing they would have to do in pro-
fessional/business careers after graduation. It was widely
perceived that general tertiary-level composition courses were not,
by themselves, an adequate preparation, nor were technical writ-
ing classes a sufficient response. One approach introduced at the
tertiary level of instruction consisted of proposing ‘writing across
the curriculum ’, a notion which required that all tertiary-level
courses and departments - regardless of their technical content -
jointly take responsibility for students’ writing abilities (Ackerman
1993, Kaufer and Young 1993, Kinneavy 1987). Such an approach
to writing, it was argued, would guarantee a greater quantity of
writing practice, more specialized assignments related to a stu-
dent’s primary field of study, and an awareness that writing
development is the responsibility of all tertiary-level faculty (see
Chapter 4). A second approach was to study the writing tasks, atti-
tudes, and expectations of people on the job - to determine what
were in fact the writing demands placed on tertiary-level gradu-
ates in non-academic settings. This latter research approach has
demonstrated that writing in the professional and business con-
texts is both highly valued and complex (Couture and Rymer
1993, Odell and Goswami 1985, Reynolds et al. 1995).
In a comprehensive study and review of other survey research
on writing in the workplace, Anderson (1985) presented a picture
of writing needs in non-academic contexts which could not be
addressed simply by the kinds of technical writing courses cri-
tiqued by Dobrin (1989; see also MacKinnon 1993). He found
that writing in professional contexts involved many factors which
could not be reduced to simple pedagogical solutions. Anderson
(1985) carried out a survey of university graduates (7 different
fields of study, 841 respondents) to answer the following ques-
tions:
1. How im portant is writing in the workplace?

2. What sort of writing is actually done?
3. How much writing is typically done, by whom?
4. To whom do people write most?
5. Why do they write - that is, what is their purpose?
6. Does length of employment change the amount of writing?
Writing turned out to be very im portant in the workplace - 93 per
cent of respondents in his survey said that writing is of some
importance and 57 per cent stated that writing is of great/critical
importance. In general, writing is seen as an integral part of most
professional employment. WTiile a few surveys have indicated that
certain job categories require less writing (e.g. retail and whole-
sale trades, and blue-collar manufacturing jobs), most job
categories have been found to be consistent with the finding that
writing is basic to most types of employment.
Writing on the job varies to a considerable degree across a num -
ber of variables. Questions of audience, purpose, and topic all
shape writing. It is true that a num ber of writing formats predomi-
nate: letters, memos, forms, and instructions are frequently
recurring genres in workplaces. Less common, though important,
genres include: internal reports, progress reports, and project
proposals. Only a small num ber of employees are requested to
publish articles (though for this minority, such activity is
extremely important). One consequence of this finding is that
technical writing in schools is perceived as useful; it does assist stu-
dents with practice in a num ber of standard genres such as letters,
memos, and instructions. The standard forms of these genres are
seen as im portant in the workplace because they provide a quick
and clear cognitive frame for information transfer, and thus are
functionally useful for effective and efficient communication. The
importance of audience is stressed in all surveys. Writers in the
workplace always emphasize that audience selection strongly influ-
ences their writing - depending on whether the writing is
intended to be read by a subordinate, a colleague, a superior, a
customer, or a company-external official.
Most respondents to Anderson’s (1985) survey said that they
spent a substantial amount of time writing in the workplace:
69 per cent claim to write more than 10 per cent of the time,
38 per cent wrote more than 20 per cent of the time,
15 per cent wrote more than 40 per cent of the time.
Thus, almost 40 per cent stated that they spent at least the equiva-
lent of one full day of work during the work-week doing some sort
of writing. These findings have been amply supported by other
subsequent research.
It was also found that writers in the workplace write to persons
within the organization much more frequently than they do to
persons external to the place of work. Paradis et al. (1985) found
that internal written communication was clearly the most impor-
tant use of writing in the workplace, particularly in the
professional research context, and they found that writing served
many different functions. Other surveys were divided on the per-
centage of writing sent to superiors, colleagues, or subordinates.
The general conclusion to be drawn is that there is broad varia-
tion in the intended audience.
The purposes for writing are:
■ to distribute information
■ to give instructions or orders
■ to respond to memos and letters
■ to complete forms for record keeping
■ to propose new alternatives and options.
Most of this writing is done in the major genre formats: letters,
memos, and forms. Ethnographic research on this question has
brought out a num ber of other purposes for writing in the profes-
sional context which are not normally presented as functions for
writing, though their roles in the workplace are crucial (Faigley
1985, Spilka 1993b). Paradis et al. (1985) state that these roles -
which define, and are defined by, the social context of the work-
place - include:
■ establishing cooperative research networks
■ verifying work progress
■ allowing self-promotion
■ stimulating ideas
■ educating colleagues.
In general, writers on the job felt that their own writing skills were
good, though writing skills in the workplace overall were per-
ceived to be weak. Newer employees tended to overstate their
writing skills more than older workers. Most workers thought that
the writing courses in college were important, particularly those in
technical writing and business communication.
Odell (1985), in ethnographic research on writing in the

bureaucratic context, pointed out two additional factors to con-
sider. Much writing involves collaboration, checking with
colleagues about facts, relevant background information, identifi-
cation of possible additional readers of the document, etc. In
addition, audience concepts are not necessarily determ ined by
knowing specifically who will be reading the document; rather,
much of the sense of audience develops from past experiences
with writing and feedback received from supervisors. Thus, collab-
oration and prior experience with very specific contexts play
im portant roles in writing in the workplace (see also Reynolds et
al 1995).
Research on writing in the workplace points to a num ber of
issues for writing instruction. Anderson (1985) includes the fol-
lowing four general suggestions:
■ explaining the importance of writing in these contexts;
■ attending to general writing skills for clear, concise and well-
organized prose;
■ addressing a variety of audiences;
■ practising a variety of writing tasks and types.
Odell (1985) suggests, in addition, that general frameworks for
organizing writing may be less useful than tasks involving very spe-
cific background and social context information typical of the
workplace environment (see also Reynolds et al 1995).
6.4 English for Special Purposes and advanced

writing instruction
A first concern for ESL/EFL contexts is whether writing is suffi-

ciently im portant for students that it needs to be taught explicitly,
to a high level of sophistication, and on a large scale. It would
seem that in many EFL contexts, the great majority of students
(particularly those learning basic English skills to complete sec-
ondary school requirements, to sit national examinations, or to
seek improved employment opportunities), do not need extensive
writing instruction. The extent to which they might be called
upon to practise sophisticated writing in English is extremely lim-
ited. Instead, the focus of writing instruction in EFL contexts may
more profitably be placed on the training of skilled professionals
who are more likely to interact with English-speaking profession-

als, whether for job-related purposes, for scientific research, or for
professional/academic exchanges. Until recently, it was rare for
an EFL curriculum to require such high levels of writing sophisti-
cation.
It is now well established that English is the primary world lan-
guage of science and technology. A num ber of researchers have
provided consistent evidence of this fact. Large (1983) was among
the first to present the arguments by examining the language dis-
tribution of scientific publications. After looking at language of
publication and language of article citation from a num ber of
indexes and abstracting sources, he concluded that the use of
English for scientific and technical publications is enormous and
growing (e.g. for biological abstracts, 77 per cent [1965], 87 per
cent [1977], 88 per cent [1980]; for medicine abstracts, 51 per
cent [1965], 76 per cent [1978]; for physics, 73 per cent [1965],
87 per cent [1977]). Similar results from a num ber of other
sources confirm this overall trend (Baldauf 1986, Baldauf and
Jernudd 1983, 1987, Grabe 1988b, Kaplan 1993, Laponce 1987,
Maher 1987, Medgyes and Kaplan 1992). The dom inant role of
English in these fields creates serious potential problems for infor-
mation access and retrieval in places where English is not a
commonly used/learned language. This situation arose as a result
of three historical phenom ena that occurred within a relatively
short timespan.
First, the nature of scientific research changed in certain direc-
tions. The present century is often described as a technological
age; but all ages before the present century have also been tech-
nological ages as the human species has always been involved with
technology. Indeed, the presence of technology is a defining char-
acteristic of the species. What is different about the current time is
the speed of technological change rather than the character of
such change, making the management of, and access to, informa-
tion a major enterprise.
Second, at the conclusion of the Second World War, the United
States was the only major industrialized nation with its educational
and technological infrastructure fully in place. Following the war,
the scientific influence of the USA advanced, and the USA
became both the largest contributor to the pool of scientific infor-
mation and the greatest user of such information. Consequently,
the USA succeeded to the control of this information (Grabe and
Kaplan 1986, Kaplan 1992, van Naersson and Kaplan 1987), and
control of the system also m eant control of the definition of the
categories through which the system wasx organized. Thus, the
underlying infrastructure now relies upon an English-based socio-
logy of knowledge.
Third, the great automated information storage and retrieval
networks which have gradually evolved over the past half-century
are based on the capacity of the computer to deal with huge quan-
tities of information. Because the inception of information
network development coincided with the end of the Second
World War, the Allies played a significant role in determining how
those systems would be structured and, by political agreement, vir-
tually all information entered into the great networks must be
either written in or abstracted in English, French, German, or
Russian. Furthermore, there has been little requirem ent for
Western scientists to worry about learning other languages, and
there has been little pressure on system managers to hurry the
development of the capacity to deal with non-Roman scripts
(Large 1983).
The coincidence of these three factors implies that nations
which do not have access to either English or the information sys-
tems are limited by their inability to cope with information or to
modify the information to make it locally useful. The transfer and
adaptation of information is not only a problem of technical con-
tent but also a problem of language, of language variety, of
cultural bias, and of intercultural understanding (Shuchman
1981). The developing nation is, in this sense, a dependent on the
advanced country’s information storage and access sophistication.
As a consequence, English has become the language of science
and technology, and the teaching of this language has become a
world-wide activity (cf. Phillipson 1992).
Given this scenario, most developing countries have found it
necessary to train enough translators and researchers to a suffi-
cient level of English language proficiency to be able to access
scientific and technical information. In many cases, the advanced
language skills have been combined with professional training - a
strategy accomplished by sending students to the USA or the UK
for university and professional studies/research.
There have been two primary responses to the ESL/EFL sci-
ence and technology focus. In ESL contexts, the primary response
has been to prepare students for all the academic language skills
required for tertiary study (English for Academic Purposes

[EAP]). Defined in many cases as pre-academic study at the ter-
tiary level, this training then is intended to prepare students for
the ubiquitous composition course requirement, at least for most
US tertiary level institutions. The EAP curricular sequence leads
international students through the same writing instruction and
writing problems faced by English LI students. Thus, at advanced
levels of study, many of the issues raised for writing in LI contexts
apply equally to ESL students. In EFL contexts, a major response
on the part of applied linguists has been to provide various types
of English for Specific Purposes (ESP) programmes, and, in par-
ticular, English for Science and Technology (EST) courses.
English for Specific Purposes syllabi have been the source of con-
siderable controversy when applied to situations for training
advanced level students in sophisticated English language skills
(as opposed to short-term occupational training).
The history of ESP is primarily UK-oriented, ESP being an
instructional approach for students who needed some English lan-
guage skills, usually a reading ability, and who only had limited
time in their training to learn English amid all their other acade-
mic requirements. A brief review of this history is essential to
understanding the research directions which grew out of ESP as
well as the criticisms and ultimate limitations of this approach for
advanced level language learning (including writing).
In two mzyor sources which trace the development of ESP up to
the 1980s, Robinson (1992) and Swales (1985) examine:
■ its pragmatic origins,
■ its early emphasis on register analysis and morphosyntactic fre-
quency studies,
■ its use of discourse analysis for the study of rhetorical structur-
ing in scientific texts, and
■ the stress on communicative language use for teaching.
(See also Johns and Dudley-Evans 1991, van Naerssen and Kaplan
1987.)
These trends have continued through the 1980s and the early
1990s with additional questions being raised about the effective-
ness of ESP as a separate entity, distinct from more general
academic-oriented instruction (i.e. EAP).
In the 1960s, pragmatic questions about the need to train stu-
dents in English in technical contexts led applied linguists to look
for those formal features which distinguish technical and scien-

tific writing from other types of writing. This approach provided a
num ber of insights into register variation, discourse analysis, and,
more recently, genre analysis. Barber (1962, in Swales 1985), in
one of the first studies of scientific writing, identified a num ber of
formal features of this register, many of which are now generally
taken as received knowledge:
long sentence length (words/sentence); many complex noun
and adjective phrases; many non-finite subordinate phrases
and clauses; many infinitive clauses; many occurrences of the
verb BE (in all conjugations); many conditional structures; a
relatively greater num ber of passive structures; a relatively
greater num ber of relative clauses; a relatively greater propor-
tion of simple present tense verbs; a relatively greater num ber
of past participles; a smaller num ber of progressive tense
forms; a smaller num ber of past tense verbs; few questions;
and few contractions.
Later research on the formal features of register analysis has
added other features (e.g. a greater num ber of lexical repetitions,
few pronouns, a greater num ber of prepositional phrases), and
combined these features with functional interpretations explain-
ing why certain features of text would appear and, in the overall
structure of the text, when they are most likely to appear (e.g.
Lackstrom et al. 1973 for tense shifts; Tarone et al. 1981 for passive
voice uses).
In the 1970s, the study of scientific texts shifted from specific
formal features to the organization of the overall discourse.
Rhetorical analyses of texts and the application of patterns of
arrangem ent in the discourse became a major research focus (e.g.
Trimble 1985), and this line of study continues to the present in
research on genre analysis and text type variation, particularly
among functional/systemic/Hallidayan linguists (Bhatia 1993,
Dudley-Evans 1989, Halliday and Martin 1993, Hoey 1983, 1994,
Martin 1985, 1992, Martin and Rothery 1986, Swales 1981, 1991).
The debate over the importance of ESP is focused on the ques-
tion of whether scientific and technical language is sufficiently
distinguishable from more general academic language to merit
separate instruction. Hutchinson and Waters (1987) and
Widdowson (1983) argue for an approach to ESP which does not
require a strong separation for more general language training of
an EAP orientation. While Widdowson makes the case that scien-

tific and technical prose is identifiable as a distinct genre type, he
argues that many of the skills and strategies for more general lan-
guage learning transfer directly to the study of technical
information; therefore, a larger EAP orientation to language
learning provides a good foundation for students in technical
fields of study. This is a position which is now popular, particularly
among applied linguists who have become disillusioned by the
promises of earlier ESP training curricula. Hutchinson and Waters
take the somewhat more extreme position that scientific and tech-
nical prose is really not different from other prose text types (a
somewhat dubious view, given the large volume of current
research on text types and genre variation). For them, then,
whether one employs a general curriculum, or a curriculum
including technical material makes little difference in language
teaching.
Whichever position is more accurate matters less than the more
general observation that advanced language instruction for acade-
mic/professional purposes may now be viewed as more closely
related to discussions of academic - and writing - instruction
which centres on English LI contexts. Both are concerned with
the general preparation of students to write successfully in acade-
mic contexts, principally at tertiary level institutions in the USA
and the UK.
Swales (1990) suggests that the above debate, in some respects,
misses the primary issue for ESP instruction. The debate is per-
haps not whether students should receive training in general
academic English for special purposes; rather, the issue concerns
when students should receive general academic English training
and when they should specialize in discipline-specific English-
language training. For Swales, the answer is that ESP instruction,
at least at academic institutions, should represent one aspect of
advanced training, and in particular, involve specialized writing
instruction. The argument is that learning to write in sophisti-
cated discipline-specific genres is time-consuming and is only
essential as students move into advanced training. In advanced
learning contexts, then, students need to be trained in the genre
structures that are valued by a given discipline; in such cases, the
considerable time invested in this training is worth the effort. Of
course, it is also at this point that many native-English students
need specialized training in specific genres such as proposal writ-
ing, grant writing, letters of enquiry, letters of application, and the

natural sciences research articles, etc.
Thus, the history of ESP has come full circle for advanced lan-
guage learners. The primary locus of training for these students is
in advanced tertiary - undergraduate and post-graduate - institu-
tions, where the ability to write well becomes a critical skill. This
writing skill, in advanced professional and academic contexts, is
shown to be more than a matter of knowing various language
structures and discourse strategies. The problem is that scientists
and technicians communicate within their own networks based in
various professional communities (see the discussion of discourse
communities in Chapter 4).
A physical phenom enon may be studied from a variety of back-
grounds by many scientists who may arrive at the same basic
result, but individual scientists - and individual disciplines - may
approach the problem quite differently. These differences may be
the result of various ways of thinking, which are in turn shaped by
cultural phenom ena including educational, rhetorical, political,
and philosophical systems. While there is no question that physics
is physics, regardless of where in the world it is perform ed - and it
is quite clear that there is no ‘Chinese physics’ as distinct from
‘English physics’ - the fact remains that the application of physics
varies enormously from place to place and from time to time. The
most obvious cases of this variation occur across cultures, but sig-
nificant variation also occurs within cultures across time (see
Atkinson 1993) and even within professional discourse communi-
ties. This variation is often the product of groups of individuals
who decide, consciously or unconsciously, what is legitimate
knowledge, and how to construct their disciplines. The basis for
such a view of science and scientific knowledge is known as social
construction, and that view now plays a significant role in both
knowledge theory and writing theory.
6.5 Writing in professional/research contexts
The social construction of scientific information is, then, a final

area of research on specialized writing for scientific or technical
publication. This research direction has become especially im-
portant for a num ber of reasons: first, it provides a very different
view of the text construction features of scientific texts; second, it
confirms the importance for writing research on the social and

rhetorical contexts of writing; and, third, it creates a locus where
applied linguists and rhetoricians converge in the study of writing
in a real-world context. The central issue concerns the nature of
scientific writing as it occurs in specific contexts - in particular, in
the context of the research laboratory.
Science writing, as seen through the finished text, appears to
conform to certain well-documented generalizations said to repre-
sent the scientific ‘style’ of writing. The underlying assumption
concerning this style is that it faithfully and accurately reflects the
experimental process that occurred in the laboratory. In fact, how-
ever, research in the sociology of science indicates that the
relationship between the experimental m ethod of the laboratory
and the text in which it is reported is anything but straightfor-
ward. The shaping of a written text by a writer reflects deeply
embedded cultural and rhetorical assumptions about what materi-
als may be presented, how those materials are to be organized,
and how they may be represented in a maximally acceptable way -
not necessarily in a way that is objectively most transparent. Such
writing may, then, be seen not only as constituting a social act, but
as constituting an act falling within a total social construct
(Bazerman 1988, 1993, Geisler 1994). To put it a slightly different
way, writing is both context constrained and context creating; not
only is writing a response to some need to express, but it is also
simultaneously contributing to, reinforcing, and redefining
notions of what may be construed as part of the knowledge-canon
and of what may be construed as ‘writable’.
It is frequently claimed, with considerable validity, that science is
cumulative - that is, that the discoveries of a given scientist depend
in no small degree upon the work of previous scientists, and that
the new discoveries as a matter of course will, then, become part of
the scientific canon available to ensuing generations of scientists. In
fact, the situation is probably much more complex (see Bazerman
1983, 1988). Scientists seeking to publish, or to have grant requests
favourably reviewed, are constrained because they are impacted not
only by the passive weight of previous science defined as a knowl-
edge canon, but also by the ascendant ‘knowledge matrix’,
including the dominant paradigm and the view of the most
respected researchers currently being published a n d /o r funded.
The ascendant knowledge matrix influences strongly what can
be defined as science:
■ what may be observed

■ how it can be reported
■ what is considered evidence
■ how is evidence arranged
■ how is scholarly argum ent conducted so that it becomes accept-
able to the scientific community.
The notion of scientific writing as pure objective reporting is a
myth promulgated by the science community itself. Objective fact
is only what the dom inant group says it is (except for what
Bazerman (1983, 1991) calls passive resistance - a set of incontro-
vertible facts derived from the past) ; and the reporting of
objective knowledge becomes the means by which the myth is
maintained.
Scientific writing is also substantiated through the concept of
replicability; in principle, any experiment reported in the scien-
tific literature can be replicated by other scientists and, through
such replication, verified. In reality, few experimental descrip-
tions, as they appear in the published literature, provide adequate
information to perm it independent replication. Thus, the crite-
rion of replication from description becomes, similarly, a myth
(see Collins 1985, Gilbert and Mulkay 1984, Mulkay 1979).
The examination of science writing - indeed, of any writing -
normally should occur at three levels:
■ the surface structure level,
■ the rhetorical level, and
■ the level of assumption - in this case, scientific assumptions,
and their potential instability.
In the past, much research has occurred at the surface structure
level; some (though not a great deal) has occurred at the rhetori-
cal (or functional) level (see Bhatia 1993, Biber 1988, Grabe 1987,
Kaplan 1987, Swales 1990, Swales and Najjar 1987, Trimble 1985);
and perhaps less has occurred at the level of the sociology of sci-
ence (which may be considered the genuine rhetorical level),
though a considerable output has occurred at this level over the
past ten years.
In order to understand how the genres of science writing are
produced, it is necessary to take into account all three levels (or
layers, or courses) of their construction (or organization). Should
any of the levels be disregarded, the understanding of science
genres will necessarily be distorted. Differences among various

genres, then, can be investigated through the surface structural
form, through the functional and organizational logic of the text,
and through the rhetorical features (including aims) which cir-
cumscribe the text as a whole.
The third level of research into science text - the level of
assumptions and aims - requires some preliminary discussion of
the sociology of science as it applies to writing (specifically to text
organization). However, before discussing science genres in this
context, a brief digression dealing with rhetorical and composi-
tion theory may be appropriate. It is through such theory that the
role of the sociology of science in science writing can be under-
stood and interpreted. Faigley (1986), in a review of process
theories of composition, argues that contemporary theories of
writing are turning increasingly towards considerations of the writ-
ing context and are drawing on the recent research in writing in
scientific communities to illustrate how the assumptions of a par-
ticular discipline impact both the reading and the writing of text.
At the heart of his review is the im portant notion that writing
occurs within a structure of power, and, consequently, that the
structure of power must be taken into consideration in any analy-
sis of writing.
Similarly, research on social construction and on the sociology
of knowledge are introduced by Bruffee (1986) into the context
of studies on rhetoric and composing. He shows how such consid-
erations determine the subject matter, the audience expectation,
and the rhetorical organization of text. Drawing on the important
work of Bakhtin (1981), Geertz (1973, 1983), Kuhn (1970), and
Rorty (1979), Bruffee defines social construction as the creation
of ‘reality, knowledge, thought, facts, texts, selves, and so on ... as
generated by communities of like-minded peers’ (1986: 774). This
conceptualization is in contrast to the cognitive view of knowledge
and language - a view that is widely held among scientists and
laypersons. The social constructionist view, instead, perceives real-
ity as constructed out of social relations within the research
community. This co-constructed reality maintains the coherence
and the power structure of the research community (through,
perhaps less so, sets of well-established physical facts). According
to Bruffee, there is, in this theoretical view, no tension between
objective and subjective; everything is subjective. In sum, individu-
als use writing (and language more generally) (a) to help them
become members of communities, (b) to cement relations with a

community, and (c) to determine and define who they are and
what they believe within a discourse community of peers.
Bizzell (1982, 1986a, 1986b, 1993) concurs with this view that
discourse communities are central as the locus of power and as a
source of acceptance for the writer functioning as a scholar con-
tributing to a given discipline. Similarly, Atkinson (1993), in a
study of the rhetorical development of the Philosophical
Transactions of the Royal Society from 1695 to 1975, has shown the
ways in which the preferred structure of scientific articles has
changed under the pressure of changing views of science and of
the collective personae of scientists (see also Bazerman 1993,
Berkenkotter and Huckin 1993, Olsen 1993, Rafoth 1990).
The importance of social construction as an ongoing force in
composition is reinforced by the research of rhetoricians and soci-
ologists who have analysed scientific discourse from the
perspective of social construction a n d /o r from the perspective of
the sociology of science. Beginning with Kuhn (1970 - a bench-
mark analysis), researchers have shown through careful
participant observation how scientific knowledge is constructed.
As Bruffee notes, scientific knowledge is a social construct, not a
discovery of ‘what is out there’ (1986: 779). The clearest indica-
tion of the validity of this conceptualization occurs when
knowledge ceases to be knowledge: this phenom enon occurs
when a scientific community disbands, or when its members sim-
ply die off - as, for example, in the cases of American structural
linguistics, and later Generative Semantic - two theoretical lin-
guistics movements which faded in the 1960s and late 1970s,
respectively (cf. Langacker 1995).
Bazerman (1988, 1993), Collins (1985), Gilbert and Mulkay
(1984), Latour (1987), Latour and Woolgar (1979), Lynch (1985),
and Myers (1985, 1990) have all observed and analysed scientists at
work, tracing the progression from experiment to description to
publication in ajournai (or grant submission). All of this research
shows quite clearly that writing for publication (and/or grant sub-
mission) is not a matter of reporting findings objectively; rather, it
involves the interpretation and modification of information, the
citation of appropriate ‘authoritative’ literature, and the imitation
of appropriate models in order to be accepted, and, in turn, to
become part of the authoritative establishment and its literature.
As Gilbert and Mulkay point out, the seriously difficult part of
scientific journal writing is that part of the process in which key

questions are asked is refined rhetorically so that it will be accept-
able to the peer community. These issues are discussed extensively
by Bazerman, perhaps the most productive rhetorician/researcher
of writing in scientific communities from a sociology-of-science per-
spective (e.g. 1988, 1993). His arguments, because they are
germane to the discussion throughout this chapter, are outlined
here in detail. In an early (1983) study, Bazerman comments on
the difficulty involved in the study of scientific writing:
1. Published articles in science do not reflect what actually hap-
pens in a laboratory; they omit false starts, digressions, and
failed procedures. They also omit those factors initially underly-
ing the choice of problem as well as factors motivating the
research design used to investigate the problem.
2. Published reports seldom supply adequate information to
perm it other independent researchers to replicate the experi-
ment.
3. Reader research among the readership of scientific papers sug-
gests that the acceptance of claims made in scientific papers is
dependent on factors other than the proof actually offered in
the paper itself.
On the basis of these observations, Bazerman (1983: 158) raises
an im portant question:
If a scientific p aper is n o t a com plete account o f a scientist’s observa-
tions and doings, n o r a tightly reasoned deductive p ro o f o f claims,
n o r an unproblem atic conveyor o f claims to be evaluated fairly and
promptly, w hat indeed is the scientific p ap er com m unicating, and to
whom?
Bazerman answers the question by proposing that scientific writ-

ing is the result of a writer’s effort to be placed into an accepted
research programme deriving from (and defined by) currently
accepted theory.
Each scientific contribution is to be understo o d against the back-
ground of the existing research program , the problem s the program
proposes, and the evolution o f the program in response to new find-
ings. Consequently, in o rd er to com m unicate the p o in t and value of
new work, the scientific writer would be well advised to un d erstan d
how his or h er new contribution fits within the continuity o f the
problem s of the relevant research p ro g ra m ....
More concretely from the w riter’s p o in t o f view, the w riter m ust

know the problem s of the field, the ideals an d ethos o f the field, the
accepted justificatory argum ents, the institutional structure in which
the knowledge is to be com m unicated, and the criteria o f adequacy
by which the innovative work will be ju d g ed .
(Bazerman 1983: 160-1)
It is certainly true that scientific writing builds upon the work of

prior researchers - that scientific research is a cumulative process;
but this trend implicates more than the objective impartial devel-
opm ent of science. A researcher ignoring the major accepted
views runs the risk of being ignored once in print, attacked on
irrelevant grounds, or simply not published at all. Escape from
this pressure is difficult and is rarely achieved. Major shifts in para-
digm - in theory, in possible research questions, or in permissible
research designs - are not often the result of singular reported
research; rather, such shifts come about through the gradual
accretion of dissatisfaction or difficulty with the accepted
approaches. When two distinctly different research matrices com-
pete for acceptance, the tenor of scientific writing changes to
become more persuasive, as appeals and arguments are mar-
shalled to reassert the value of individual pieces of the matrix.
All of this is not to claim that the social nature of scientific
research is capricious; even in the most turbulent periods, as
Bazerman points out, competing approaches must take into
account what has been called ‘passive constraint’ (Fleck 1935) -
observations about nature that have been collected and measured
and are so widely accepted as basic that they must be accounted
for in any theory. Scientific research is constrained by facts too
widely accepted to be ignored (cf. Gilbert and Mulkay 1984). This,
then, is the background for any discussion of scientific writing as a
social act within a social context - the social construction of scien-
tific reality.
Bazerman (1983, 1988) suggests that there are four main areas
which collectively illustrate the social co-construction of science:
■ the writing process
■ the textual form
■ the dissemination process
■ the audience response.
It is im portant to note that the writing process and the research
activity appear to be co-extensive. The pre-writing stage is
coincident with the beginnings of the research activity and usually

extends over a considerable time period; by contrast, the actual
writing of the results often occurs over a very brief period. During
the extended pre-writing phase, the scientific paper acquires its
shape. Laboratory research regularly involves tinkering - m oment
by m oment rethinking and adjusting as the research progresses -
to the point at which the specific problem and its possible analysis
crystallizes; and this process is recursive, continuing throughout
the research phase. In laboratory research, the choice of problem
or question to be addressed is selective; that is, the researcher is
led towards certain questions and away from others by the con-
straints imposed within the dom inant research matrix. To some
extent, research questions are defined by available instrumenta-
tion; that is, a paradigm defines the instruments needed to
evaluate it. To some extent research questions are economically
defined; the available economic resources, the potential for
achieving recognition, and the state of experimental resources all
constrain funding support. In sum, the development of scientific
research is dependent upon a range of decisions which are social
and strategic.
With respect to textual form, the rhetorical level plays a greater
shaping role than the surface structural level. Scientific writing, if
it is to achieve acceptability, must appear to be more universally
applicable than the confines of any particular research laboratory.
As a consequence, a significant amount of persuasion is required
to establish new criteria for evaluating research and to lend credi-
bility to the results being reported (including the judicious use of
appropriate citations); the published report is a persuasive docu-
m ent whose purpose is to increase the value of the researcher’s
ideas (and the persona of the researcher) within a particular mar-
ket. As Bazerman (1983: 169) notes:
T he paper m ust first reconstruct the m arket, define the needs o f the
m arket, and identify the research being rep o rted as the p ro p er vehi-
cle for the satisfaction of those needs. T he p ap er m ust th en fulfill
the m andate it has constructed by dem onstrating that its solution to
the m arket needs was in fact achieved in the laboratory.
In short, the textual form of a scientific paper is less objective and

more persuasive than would normally be assumed.
With respect to research papers, two primary modes of dissemi-
nation are available - the informal and the formal. The informal
mode represents what might be term ed ‘word-of-mouth’ commu-

nication, through conference presentations and the circulation of
draft manuscripts. This mode may provide the best means for the
maintenance of stable research areas and for the replication of
results (Gilbert and Mulkay 1984).
The formal mode occurs through the publication of papers in
journals, invoking the referee system. Journals reflect the con-
cerns of referees as representatives of the larger audience which a
paper is intended to reach. As research fields achieve greater con-
sensus with reference to what constitutes important research
issues, the likelihood of acceptance of a paper which conforms to
the expectations of referees is great; by the same token, the
acceptability of a paper which does not clearly conform to those
expectations diminishes. Thus, scientific writing arises within a
social context wherein the writing is responsive to the accumu-
lated writing of others, and where the author is able to manipulate
what actually appears in text, within the established limits of ‘pas-
sive constraints’ (Bazerman 1983). The long-term accumulation
of evidence and argument provides a continuum in research act-
ing as a corrective to the excesses of any particular research
programme or theory; thus, mistaken findings and misunder-
standings are corrected.
The social co-construction of science requires salesmanship as
an essential characteristic of scientific writing (Czubaroff 1989,
Prelly 1989); if other scientists can be persuaded and intrigued,
the greater will be the num ber of scientists working on similar
problems and reporting findings of a similar nature (Bazerman
1985). Myers (1985), examining the work of two particular biolo-
gists, found, as Bazerman had, that the biologists were required to
account for the interests of their peer audience, for the domain of
knowledge which constituted the accepted canon, for the need to
sell their product on the open market, and for the rhetorical sig-
nificance of their own writing. Looking specifically at grant
proposal writing, Myers (1985: 220) notes:
Most researchers would g ran t th at proposal writing is a rhetorical
activity in which the writer seeks a strategy for persuading a reluctant
audience. T here is a paradox here because the proposal form at,
with its standard questions about background goals and budget, and
the scientific rep o rt style, with its passives and impersonality, do n o t
allow for m ost types of rhetorical appeals; one m ust persuade with-
out seem ing to persuade. Yet almost every sentence is charged with
rhetorical significance. In classical rhetorical term s, the form s o f the

appeal in the proposal are ethical an d pathetic as well as logical; one
shows that one is able to do the work, that the work is potentially
interesting to o n e ’s audience o f o th er researchers, as well as showing
that one is right.
Myers takes the view that this demand on scientific writing is

essential to the health of scientific inquiry because it reinforces
the consensus-building process which is necessary to the ongoing
discovery of scientific knowledge, even if the formal writing act
itself contributes only a small part of that necessary consensus-
building. Thus, scientific knowledge is a socially constructed belief
system, explainable in relation to historical and social forces. The
objective façade of science masks the intense rhetorical manipula-
tion within the restrictions imposed by the surface form of the
research report. In the short run, empirical fact may not always
carry the day; rather, the most successful persuader - the writer
with the most appealing argument - often triumphs (Myers 1986).
Thus, there is the need for the kind of consensus-building which
both Myers and Bazerman refer to as the corrective for temporary
misunderstanding or error.
Of particular interest in Myer’s review is his effort to relate the
sociology of science to writing research and writing instruction.
He notes that the view of writing available through social construc-
tion makes possible a top-down analysis, beginning with the social
forces which shape the underlying assumptions of the writing act.
This contrasts sharply with the more linguistically oriented analy-
ses of science writing, which are bottom-up structured, and which
examine local surface forms to determine the constraints of par-
ticular text genres. Myers suggests that neither approach is, in
itself, sufficient; he believes it to be essential for researchers to
understand the discourse community in order to unravel the
structure and purpose of text.
T he m ajor lesson for writing researchers is th at we cannot limit our
studies to written texts, or to the controlled observations o f individu-
als developed by cognitive researchers, b u t m ust get o u t into the
field and en ter into the flow of language and work. Again and again
I realized w hat (perhaps necessary) lim itations I was making in con-
sidering ju st texts, even draft after draft o f revision.
(Myers 1986: 606)
The understanding of scientific text genres, the constraints on

their creation, the forces shaping the (published) product, and
the persuasive intent masked within the ostensibly objective for-

mat collectively require a perspective on scientific writing which
goes beyond, though it does not exclude, traditional applied lin-
guistic approaches to research on scientific text. In precisely the
same way that it is difficult to penetrate the differences between
texts in two different linguistic systems, it is difficult to penetrate
the differences between scientific texts and other types of text
within the same linguistic system (cf. Hutchinson and Waters
1987). The latter difficulty may result from the fact that scientific
text belongs less to a particular linguistic system than it does to an
international interdisciplinary community unique to science writ-
ing (e.g. Merton 1973).
Whatever the cause of this difficulty, there are central questions
to be asked about the construction of science genre and the avail-
able resources to teach such writing skills to anyone; it is
particularly difficult to teach scientific writing skills to individuals
who are primary participants in language systems other than
English and who may simultaneously be participants in communi-
ties in which literacy is recent and non-in tensive. When such
students enter traditional language programmes to learn to write
science text, or when they enter special-purpose language instruc-
tion, they are, in fact, being asked to understand not only the
lexical, syntactic, and rhetorical systems of the target language,
but also the assumptions underlying science writing and the
requirements implicit in science publication.
In sum, science writing constitutes a value-laden rhetorical activ-
ity. Science writers - particularly if they are writing in English
(generally accepted as the international language of science and
technology), and particularly if English is not their first language
- need to possess a complex set of skills which are only partially
definable in linguistic terms. They need to:
■ have a highly sophisticated sense of appropriate areas of
research, of proper research questions, and of acceptable
research designs;
■ be cognizant of intertextuality - that is, have a clear sense of
what others have had to say on the topic and know how to incor-
porate that material into their text;
■ be able to align their own work with the recognized leadership
in the field so that the leadership will (seem to) legitimize their
work;
■ have a highly sophisticated sense of audience; that is, be aware

of what will influence peer reviewers and the general readership
in the discipline;
■ be able to convey the necessary (and necessarily) rhetorically
charged information through appropriate mechanisms of per-
suasion within the constraints of a formal objective format.
This listing of requisite skills illustrates the problem of most lan-
guage instruction:
First, the continuing insistence on the substitution of grammati-
cal instruction for discourse instruction, or the equation of the
two, simply prevents attention from being directed to the problem
site.
Second, the persistence of the view that external reality can be
objectively represented means that productive rhetorical forms in
the creation of science text will be undervalued; this trend is
powerful even in those instructional programmes that have recog-
nized the importance of rhetorical instruction,
Third, the absence of any attention in the teaching of discourse
to epistemological issues forces the focus in such courses to fall on
secondary issues; that is, courses ignore the fact that science writ-
ing can only be understood in terms of the contemporary
sociology of science and in terms of the whole pre-existing body of
science texts. Further, when the learner is a member of a dis-
course community alien to all of these considerations, the
problem is vastly more complicated. It is in part complicated by a
universalist tendency to claim that there is no significant differ-
ence between texts in various languages, or to claim that, if any
differences exist, they only occur at the surface level.
That these difficulties persist in instructional programmes
should be no surprise since the same issues beset research in dis-
course analysis. Researchers in the discourse of science do not yet
know what it is necessary to know in order to understand the
rhetoric of science text, whether in the form of the published
report or the grant proposal, the introductory science text, or the
popular science article (cf. Bazerman 1993, Bhatia 1993, Lemke
1995, Swales 1990).
Contrastive rhetoric may offer one informative approach to sci-
entific text from a discourse perspective; that is, it provides both
motivation and mechanisms for text analysis, though it needs to
adopt the sociology-of-science orientation to be useful for text
analysis within particular languages. In the final analysis, however,

it is far more im portant to broaden pedagogical bases so that lan-
guage learners can in fact become aware of the complex
reticulated systems of constraints that exist in text.
It remains highly unlikely that the unmodified ‘process
approach’ to teaching composition will prove to be sufficient. The
kinds of knowledge that a writer is expected to have in science
writing - in addition to a thorough grasp of the lexical, syntactic,
semantic, and rhetorical systems of the language - is not likely to
be acquired through exercises in keeping a journal or in compos-
ing narrative or descriptive texts of a far more general type. It will
be necessary to rethink writing instruction, and in that rethinking
to recognize that there is more to the composing of texts than is
normally assumed in the linguistic system; traditional linguistic
knowledge is necessary but not sufficient. To the extent that sci-
ence texts are about science, epistemological issues must be
addressed; indeed, to the extent that any text has a content
carried through its linguistic and rhetorical structure, epistemo-
logical questions need to be addressed.
It is possible that the constraints on science writing are special;
but it is also possible that similar types of constraints apply to all
writing intended for any but the most general audiences. For
example, though the preceding discussion has focused on scien-
tific and technical text, similar problems are involved in the
teaching of writing in law. Legal text has not been submitted to
quite as much scrutiny as scientific text. On the other hand, law
schools often provide special courses in legal writing. Such courses
tend to suffer from the same kinds of concerns for convention
that occur in virtually all technical/specialist writing instruction.
However, because it is clear that the law is a linguistic construct -
in that sense rather different from science - much more attention
is given to the kind of formal logic which operates in writing about
the law (Conley and O ’Barr 1990, O ’Barr 1982, Paradis 1991,
Shuy 1987, 1991, Stygall 1991).
6.6 Specialized writing contexts and the teaching

o f writing
In the preceding sections of this chapter, a num ber of specialized

contexts for the study of writing - each providing a somewhat
different perspective - have been reviewed. The technical writing

perspective, while criticized, nevertheless represents a general
movement to respond to student needs as they leave tertiary edu-
cation and enter the workplace and the professional environment.
Technical writing courses provide students with extensive practice
in the use of common formats such as letters, memoranda, and
reports; they also stress the importance of form and the need for
clarity, concision, and logical organization. These basic skills were
perceived to be very important by writers in non-academic con-
texts when researchers actually explored writing in non-academic
organizations.
Specialized writing skills are im portant for students whose
native language is not English, whether in the UK, the USA,
Canada, Australia, or in their own countries. As argued above, the
need for sophisticated English language skills, including writing,
constitutes a critical skill for some percentage of the professional
and academic population of any country. This need explains the
rise of English for Specific Purposes and, in particular, of English
for Science and Technology courses.
The final extension of research on writing in specialized con-
texts has followed from research by rhetoricians and sociologists
of science. In this approach, researchers have examined the uses
of writing in advanced research contexts, primarily in academic
laboratory settings, and a num ber of rather startling insights have
emerged:
1. This research suggests that much science writing, assumed to
be objective, disinterested, and factual, is in fact highly rhetori-
cal and value-laden in nature.
2. This research supports social construction theories of knowl-
edge wherein all knowledge is understood to be the creation of
discourse communities rather than to exist ‘out there in the
world’.
3. This research demonstrates the importance of recognizing
audience, establishing the intertextuality of ideas, and being
persuasive without appearing to be so.
It is im portant to stress that the elimination of the ‘objective’
ground from beneath knowledge does not mean that knowledge
is less important, only that knowledge is something that is negoti-
ated; students need to understand the nature of such negotiation,
the rules for engaging in this negotiation, and the translation of
the rules deriving from negotiation of meaning into written dis-

course. Thus, rhetoric returns to writing instruction as a
foundational discipline which informs all types of writing, scien-
tific writing being just one of many domains in which the impact
of rhetorical principles may be explored. One of these extensions
of rhetorical principles involves the study of contrastive rhetoric.
Research in this field has been evolving over the past 30 years and
is now providing useful insights into both the writing product and
the writing process. This, however, is the subject of the next
chapter.
7
Writing across cultures:

contrastive rhetoric
7.1 Introduction
There have, historically, been two quite different ways of looking

at the nature of language, and each has led to a different theoreti-
cal posture. On the one hand, in what has been called ‘general’
linguistics (also ‘formal’ or ‘autonom ous’ linguistics), the object
of inquiry has traditionally been seen as an independent language
system composed of unique and invariant structural and semantic
rules. In contemporary linguistic thinking, this system is seen as
innate to hum an beings - a species-specific phenom enon
encoded into the genetic structure; given this biological explana-
tion of its ontogeny, it is perfectly logical to investigate language as
a separate entity because it has an independent existence unre-
lated to human production or use. The relationship between that
system and the investigator is straightforward and unproblematic
- subject-object. The object of formal inquiry is a system description
in ‘neutral’ scientific language quite separate from the value-laden
characteristics of everyday language. Such description is thought
to give rise to rational predictions about the internal operations of
the system and about its future directions and development.
This view derives from the traditions of logical positivism and
scientific realism, and holds the promise of a parsimonious and
invariant description - a notion that has captured the imagination
of some linguists and has now extended into work in artificial
intelligence. Without doubt, this enterprise has produced useful
information and has discovered certain cognitive-linguistic struc-
tures that appear invariant, but it has fallen increasingly under
attack from scholars holding alternative views.
These alternative views are not to be conceived of as a single
176
Writing across cultures: contrastive rhetoric 177
powerful theoretical thrust, though they do hold in common cer-

tain basic assumptions. Here, language is perceived not as an
independent system, but rather as a hum an product and a social
tool. The perception in this view is that, while the physical sci-
ences deal with inanimate objects outside the hum an sphere,
language is the product of the hum an mind and is therefore
inseparable from that mind and all its attendant subjectivity, value-
orientation, and emotion. The investigator is simultaneously both
the subject and the object of inquiry; the study of language is the
study of hum an beings; and the relationship can be described as
subject-subject. Such a perception challenges, on logical grounds,
the notion of the existence and objectification of language as well
as the possibility of devising an invariant abstract model.
Given the complexity of language, the fact that language
changes over time, and the fact that language exists within various
cultural systems, it would be impossible to discover invariant laws
as in physical sciences; thus, the study of language, at certain lev-
els, must be descriptive rather than predictive and explanatory. In
addition, it would be impossible to describe language in a context-
free, neutral scientific language because there is constant
movement between the parts and the whole with no absolute
beginning and ending points. Contrastive rhetoric derives its ori-
gins from such a view.
Discourse study in general constitutes one member of this group
of views. Discourse analysis recognizes the significance of social
(and perhaps physical) context and, most importantly, of prior text in
constituting meaning. And it has insisted on ‘grounded’ research.
The recognition of context and prior text, however, increases the ten-
sion between an interpretive stance and an empirical one - a
tension not yet satisfactorily addressed in discourse analysis and
one which, in turn, exacerbates the rift between quantitative and
qualitative research. Quantitative researchers preserve and often
reify the language system’s independent existence in at least some
of its aspects, emphasizing certain structural continuities and regu-
larities (on the basis of which empirical research can be
undertaken), and showing that the system does have predictable
structural properties which, appropriately defined, can be quanti-
fied and employed to build an abstract model (cf. Biber 1988,
1995, Biber and Finegan 1989, Grabe 1987).
At the same time, there is an increasing awareness that quantita-
tive research needs to be augmented by some theory of the
linguistic object that can account for its dynamism - by some inte-
gration between the characteristics of a particular corpus and
both contexts and prior texts. Thus, there is a need to reach an
understanding of the linguistic system before attempting struc-
tural and quantitative descriptions of it, and there is a need to
acknowledge that such an understanding can probably only be
achieved through interdisciplinary approaches.
One particular problem associated with the quantitative
approach has been what is known as the ‘disappearance of the
phenom enon’. In scientific realism, the objective of empirical
research is to capture an invariant objective reality through
repeated testing of hypothetical correspondences that occur
between models and observed phenom ena (but, as has been
observed elsewhere, such repeated, consistent testing is very diffi-
cult because published methods tends to limit consistent
repetition). In the alternative view, deriving from the notions of
Husserl, that sort of empiricism was conceived as an error trace-
able back to Galilean systematization because the notion
hypothesis-test-verification is based on an assumption of the con-
stancy of any given phenomenon. Such an assumption ignores the
practical problems inherent in setting up a consistent measure-
m ent system, or an experiment which always does what it is
supposed to do, or a survey employing a team of assistants; it is not
a question of sound methodology, but rather of practical organiza-
tion, and the result is the potential disappearance of the
phenomenon. The need for objectivity, consistency across investi-
gators, and so on, led to the replacement of the phenom enon by
the artefacts of the methodology (e.g. data runs, variables, statisti-
cally derived factors, and so on).
To avoid this problem and to perm it analysis of the phenom e-
non, a different, interpretive, empiricism developed, leading to
the extraordinarily detailed analyses of texts, but the problem of
replacing the text with artefacts of the method has not been com-
pletely solved. The concern with this problem has led to two
alternative strategies: one deriving from the work of Sachs and his
colleagues and gradually enlarging into the whole ethnomethod-
ological enterprise (Schiffrin 1991, 1994), the other giving rise to
the various techniques being employed in written discourse analy-
sis (cf. Cooper and Greenbaum 1986, Coulthard 1994, Purves
1988, van D ijkl985).
Another problem lies in the fact that both subject-object and
subject-subject notions of the relationship between the investiga-

tor and the object of analysis are perceived as ahistorical. The
interest in history in the context of text analysis is a very recent
phenom enon. It became possible only when the notion of lan-
guage as an invariant structure independent of hum an activity was
abandoned; as long as language is perceived as genetically condi-
tioned and independent of hum an agency, history is irrelevant. As
soon as language is seen as a product of the hum an mind and as a
tool, its continuing existence over time constitutes a theoretical
problem which can only be approached from the historical per-
spective. But these opposing views also bring into focus an older
dualism between structure and action (function) and the question
of the impact of larger social structures on individual behaviour
(cf. Kress 1991). The question of individual behaviour raises, in
turn, the question of learning (acculturation, socialization) and
the difficulty of modelling learning behaviour.
Contrastive rhetoric has its origins in notions of language struc-
ture, learning, and use which are not strongly autonomous, and its
goal is to describe ways in which written texts operate in larger cul-
tural contexts. It has sought to arrive at some understanding of
the ways in which written language operates and the ways in which
written language diverges from spoken language. It is interested
in questions relating to the description of various genres - the
question of whether or not these genres occur in various lan-
guages, and the questions of what constitutes evidence and what is
the best arrangem ent for such evidence in various genres as they
occur in different languages. It frankly derives some, but not all,
of its orientation from the weak version of the W horf/Sapir
Hypothesis - a notion reintroduced into both linguistic and liter-
ary study by the French structuralist and post-structuralist writers
(see also Berman and Slobin 1994, H unt and Agnoli 1991, Lucy
1992, Slobin 1990).
7.2 The question o f genres
Elsewhere in this volume, it has been noted that there are not only
differences between genres of the same language but important
differences also exist between similar genres in different lan-
guages (Berkenkotter and Huckin 1993, Biber 1988, 1995, Kachru
1987, Purves 1988, Swales 1991). While certain scientific and
technological genres have taken on global, translanguage fea-

tures, im portant differences still remain in the nature of evidence,
the means of organizing evidence, and even the matter of what
can be discussed in writing as opposed to speech (cf. Taylor and
Chen 1991). It has been clear for some time that such differences
are not merely manifestations of varying surface features across
languages but are caused by important underlying cultural and
historical differences, which contrastive rhetoric has consistently
sought to understand.
7.3 Spoken and written language
In Chapter 1, the notion that writing is a late-added technology

different from spoken language, was discussed in the context of
literacy. To recapitulate very briefly, while spoken language seems
to have been a part of the human genetic baggage for at least
100,000 years, written language represents rather distinct post-bio-
logical evolutionary changes that have occurred, in terms of
species time, over quite a short period.
At the present time, many of the functions served by written lan-
guage are not duplicative of the functions of oral language though
a fuzzy overlap does exist. That overlap is based on the existence
of genres which straddle oral and written forms; e.g. sermons writ-
ten out but delivered orally, or academic lectures, or the nightly
television news.
Evidence has shown that written and oral language developed
separately. Written language developed differentially among those
populations that have it, but is neither universally distributed across
the species nor even universally distributed in societies that possess
it. Written language is not merely transcribed oral language; on the
contrary, over time, written language has acquired a number of
unique functions not shared with oral language (Olson 1994).
7.4 Written language across cultures
Because written language is not universally distributed but has

developed differentially, it has come to serve somewhat different
purposes in different languages. To a minor degree, these differ-
ences may reflect surface variations between discrete languages.
In English, for example, knowledge is not countable as it is in

Japanese; this difference suggests that the concept would be dif-
ferently managed in discourse structure. But such obvious factors
as the placement of modifiers pre- and post-nominally will result
in differing discourse manifestations. The way in which verb sys-
tems are actualized (e.g. the presence or absence of tense, aspect,
mode, etc., in the verbal system; or the relative richness of distinc-
tions which they engender) will produce different discourse
treatments. By the same token, other surface structure features
will likely result in somewhat different discourse treatments (e.g.
forms and uses of passives in English and Chinese). But the more
essential questions deal with semantic and logical issues as those
issues are encoded in a language system. The questions raised ear-
lier:
■ What can be encoded in written language (as opposed to spo-
ken language and as opposed to not being discussable at all) ?
■ What is evidence?
■ How is evidence arranged to produce the most effective, per-
suasive, attractive text?
are the key questions that must be addressed if contrastive
rhetoric is to serve its purpose (see also Connor 1995).
It is also im portant to point out that contrastive rhetoric is
responsive to cultural-use preferences, and not simply to differ-
ences in language structures and their frequencies of use. While it
is true that culture and language influences may appear to be dis-
associated in certain situations, such as when a Chinese scientist
writing in English chooses a rhetorically Chinese approach, the
explanations for such choices may nevertheless reflect both cul-
tural and linguistic preferences (cf. Taylor and Chen 1991).
7.5 The historical development o f

The notion of contrastive rhetoric1 emerged in the middle 1960s

from an essentially pedagogical impetus. As the population of
international students grew at US tertiary institutions over the
decades of the 1950s and 1960s - a period when the audio-lingual
method was widely used for teaching English to speakers of other
languages - and as it became evident that the approaches being
used did not offer a means for teaching connected writing to

those students, it became clear that a need existed for a better
approach to the teaching of writing. In early research reported in
Kaplan (1966, 1972, 1988), a large num ber of international stu-
dent compositions were examined by hand and a num ber of
patterns emerged from those examinations. It seemed clear that
the writing in English of students whose native languages were
Arabic, Chinese, French, Japanese, Russian, etc., was systematically
different from the writing of comparable students who were native
speakers of English. The purpose of the research was intended to
understand those differences and to suggest pedagogical strate-
gies to assist students to bridge those differences.
Contrastive rhetoric does not primarily concern itself with such
matters as the basic placement of modifiers or with basic word-order
questions or with superficial differences in tense, aspect, or mode.
Rather, it is concerned with matters relating to topicalization, to the
various ways of achieving cohesion (as, for example discussed in
Halliday and Hasan 1989), to the combination of surface linguistic
features which reflect identifiable discourse functions, and to the
mechanisms through which coherence is achieved.
For example, in English, the following structure appears non-
sensical:
Last week, my brother and I went fishing. We didn’t catch a thing; the largest
was only a foot long.
The implication is, on the one hand, that no fish were caught, and,
on the other, that fish were caught but none over one-foot in lin-
ear length. On the face of it, this is a contradiction; it is impossible
to catch no fish and yet to catch some fish (regardless of their size).
But it is necessary to place the utterance in a context. If the
speaker is a fisher of sailfish, the utterance is non-contradictory;
that is, the implication is that no sailfish were caught even though
some ‘j u n k ’ fish, all relatively small, were caught. Thus, an utter-
ance which seems to lack coherence in isolation may acquire
coherence in a specific context.
The interesting question, of course, is how such coherence-mak-
ing contexts operate in languages other than English. Would it be
possible, for example, to provide a comparable utterance in a
comparable context in Mandarin, or would the discourse con-
straints on Mandarin require a different way of encoding the
implications of the utterance? The interesting questions, then, do
not lie in the syntactic structure of the utterance itself (which, in

Chinese, would necessarily be different because past-time is differ-
ently signalled), but in the way in which implicature is encoded in
coherent text.
Consider the two following utterances:
1. The two men were talking on the platform.
2. On the platform, the two men were talking.
Utterance 1 observes normal English word order, with the locative
prepositional phrase at the end of the utterance; utterance 2 dislo-
cates the locative prepositional phrase to the front of the
utterance. In purely syntactic terms, there is no significant differ-
ence between the two utterances; the only difference is that in one
case an additional moving rule has been applied. But in discourse
terms the difference is important. Assuming that these two utter-
ances stand, respectively, at the beginning of a text, the text that
follows utterance 1 may be about the two men, perhaps about the
subject of their conversation, but not about events on the plat-
form; on the other hand, the text following utterance 2 is likely to
be about events on the platform, and may not be about the two
men or the subject of their conversation (e.g. a setting in fiction
narrative, a police recall of a crime scene).
Of course, these discourse rules are fairly flexible, and it is
impossible to make categorical statements about the context of
the continuing text. To illustrate, the text below suggests the sort
of coherence normally expected to follow utterance 1:
T he two m en were talking on the platform . They m et there daily in
the course of their com m uting routine. Each day, they reviewed the
state o f their lives, the condition o f their marriages, the health and
achievem ent of their children and their wives. They were friends,
and the m undanity of the conversation only served to strengthen
the tie o f friendship - o f com m onality - between them .
The second text suggests the sort of coherence normally expected

to follow utterance 2:
O n the platform , the two m en were talking. T he rain fell gently, and
the twilight was encroaching on the scene. T he last train from the
city had arrived, and the platform was quiet, deserted by the hordes
o f daily com m uters, settling into the silence norm ally associated with
isolation and decay In the m orning, with the retu rn o f the com-
m uters, the platform would spring into renew ed life, the cycle o f the
seasons, of death and resurrection, repeated in a d iurnal chain.
It appears that topicalization can be accomplished in English, at

least in some environments, by leftward dislocation of sentence
elements in the initial text sentence. A question of considerable
interest is the extent to which similar dislocation serves topicaliza-
tion in other languages.
The fact remains that there are observable differences among
the rhetorical structures of different languages though those dif-
ferences are not necessarily manifested at the surface level.
7.6 External research evidence for

Supporting evidence for contrastive rhetoric from other discipli-

nary studies includes cross-linguistic language development
research in the past decade (Berman and Slobin 1994, Choi and
Bowerman 1991, Demuth 1990, Slobin 1990, Slobin and Bocaz
1988). In this line of research, Slobin and others have noted how
children acquiring different languages will sometimes exhibit
preferences for different sorts of linguistic structures (e.g. early
Spanish speakers make greater use of tense and aspect distinctions
than do English speakers). Slobin and Bocaz (1988), noting differ-
ences in the use of prepositional phrases to present the setting for
the narrative, state (p. 21):
Spanish-speaking children ... ten d to devote m ore effort in their
narratives to establishing the static locations o f objects an d partici-
pants in scenes, as if sensing th at it is difficult to independently
specify trajectories. English-speaking children, by contrast, devote
less attention to static description, attending m ore to the elabora-
tion of trajectories, and leaving m uch o f the arran g em en t o f objects
and participants to be in ferred ....
All of these language-specific characteristics o f the encoding of
states and processes, scenes and trajectories, seem to have conse-
quences for the eventual construction o f narrative discourse in two
types of language.
The structural differences uncovered do not argue for a strong

form of the Whorfian Hypothesis, but for the recognition of cul-
tural preferences which make greater use of certain options
among the linguistic possibilities.
Sociolinguistics and the study of literacy have provided further
evidence of variation in discourse which can only be understood
in terms of the sociocultural contexts; that is, patterns of discourse

use are socially and culturally shaped. The literacy research of Gee
(1990), Heath (1983, 1986a, 1986b), Scollon and Scollon (1981),
and Street (1984, 1993) provide extensive evidence to support this
line of argument.
Outside of the specific context of literacy, the sociolinguistic
exploration of language socialization, and its influence on pat-
terns of language use, supports a contrastive rhetoric perspective
(Clancy 1986, Ochs 1988, Philips 1983, Schieffelin and Ochs 1986,
Spindler and Spindler 1987b). Different cultures have different
ways of doing things with language. These different uses are cul-
turally/socially shaped and they have reflexes in the preferential
organization of discourse. It would seem natural, therefore, to
suggest that the same socially contexted language use preferences
in oral discourse would influence written discourse.
The development of social construction theory and its influence
on rhetoric as discussed by Bazerman (1988, 1991), Bruffee
(1986), Kaplan and Grabe (1991), and Myers (1986, 1990) provide
yet another source of evidence for a contrastive rhetoric perspec-
tive. In particular, the study of scientific discourse in the last
decade has demonstrated how even the most ‘objective’ and uni-
versal of discourse genres is, in fact, a product of rhetorical
socialization and preferred conventionalizations accepted by the
scientific community (see Chapter 6). For example scientists do
not write objective value-neutral articles; rather, they respond to
the current trends in theory making and shape their reporting
accordingly (much as this chapter is doing). A particularly interest-
ing analysis of this rhetorical shaping in linguistics is Czubaroff s
study of Chomsky’s review of B.F. Skinner in Language (1989).
A final area supporting the notion of contrastive rhetoric can be
found in rhetoric with the emergence of post-structural approaches
to the critical study of text and their emphases on the socio-histori-
cal forces which shape our writing and our reading of any text
(Crowley 1989, Neel 1988, Shapiro 1987). Following this line of
argument, any text is a product of multiple forces, all of which are
in some way contextual. The notion that any text could be objective,
free from contextual shaping, does not appear logically valid.
Given this diverse array of supporting evidence for the general
notion of contrastive rhetoric, it would seem appropriate to
acknowledge contrastive rhetoric’s broad theoretical appeal (see
also Leki 1991). At the same time, it is important to consider the
range of evidence emerging specifically from the direct study of

contrastive rhetoric across many languages. There is considerable
evidence from this line of research, discussed in the next section,
which converges with the arguments from other disciplines.2
7.7 Text-based research evidence
A typical criticism of research in contrastive rhetoric is that it has

not been able to control sufficiently the many confounding vari-
ables that enter into any comparison between writing in two
languages or in the writing of L2 learners in the second language.
There is no doubt that the ability to control many confounding
variables has been a problem in contrastive rhetoric research, and
will continue to be a difficulty in future research. However,
research in all areas of L2 learning have equal difficulties to over-
come, and more recent research in contrastive rhetoric has taken
this issue into account much more effectively than it had in the
past. In fact, the difficulty in exploring contrastive rhetoric, as well
as the general growing maturity in research methods in all of
applied linguistics, has led to increasingly more sophisticated
research and more creative discourse approaches (Connor 1995).
This evolution is first indicated by the increasing stress on the
comparison of texts in two different languages (Bickner and
Peyasantiwong 1988, Clyne 1983, 1991, Eggington 1987, Hinds
1987, 1990, Indrasutra 1988, Y. Kachru 1983, 1988, Lux and Grabe
1991, Montano-Harmon 1991, and Ventola and Mauranen 1991).
In these more recent studies, the focus is on direct comparison of
discourse features in the texts of both languages. Second, the
growing body of discourse-based research has led to the study of
syntactic/textual features central to the structuring of discourse
(e.g. Biber 1988, 1995, Indrasutra 1988, Reid 1988, Reppen and
Grabe 1993). These discoveries have been incorporated into more
recent research on contrastive rhetoric.
Third, recent research in discourse analysis has extended the
study of text to include analyses which assume larger divisions in
the organization of text. Earlier efforts have included Kaplan’s
(1972), Discourse Bloc framework (see also Ostler 1987). More
recent approaches include units of persuasive discourse (Connor
and Lauer 1988), storygraph analysis (Soter 1988), topical struc-
ture analysis (Cerniglia et al. 1990, Connor and Farmer 1990,
Lautamatti 1987), and content structure analysis (Connor and

McCagg 1987). Finally, discourse research has begun to focus on
textual analysis which examines co-occurring linguistic features of
texts (Besnier 1988, Biber 1995, Grabe 1987, Lux and Grabe
1991). Some of the more im portant and striking research in con-
trastive rhetoric will be reviewed to illustrate these developments
and also to suggest the strong array of research evidence directly
supporting the notion of contrastive rhetoric (see Connor 1995
for additional research).
7.7.1 German
In a set of studies examining the differences between German and
English, Clyne (1983, 1985, 1987, 1991) has argued that there are
indeed clearly different preferences for the organization of writ-
ten discourse. In particular, he notes (1985) that the role of
syntactic structures in German ‘learned’ prose is the major distin-
guisher from informal prose, whereas, for English, the role of
diction in distinguishing ‘learned’ from informal prose is more
important. Clyne also notes that essay writing in the educational
system is much more prevalent in English contexts than in
German contexts. In German contexts, the role of content is
much more im portant than formal style and organization in writ-
ing. As he notes:
Digressions from a linear structure are tolerated m uch m ore in
G erm an-language countries, as are repetitions. T he less linear and
less form al structure o f G erm an (academ ic) discourse also is evi-
denced in books and articles in fields such as linguistics and
sociology (Clyne 1981). T h ere one finds digressions, and digressions
from digressions - which entail recapitulation an d repetitions to
stress the main line of argum ent.
(Clyne 1985: 116)
A main source of evidence for these claims is the difficulty English

readers have with close translations of German academic works.
Clyne (1981) notes that a key difference lies in the non-linear writ-
ing style typical in German academic prose. In particular, the
translation of Norbert Dittmar’s Soziolinguistik has been described
by Americans as haphazard in organization and even chaotic. No
German or other continental European reviewer has made such
critical statements about the book.
Clyne (1991) replicates many of his earlier findings and

provides further evidence for differences in written texts in terms of
textual symmetry, text hierarchy, argument continuity, data integra-
tion, definition giving, use of advance organizers, and hedging and
modal verb use. He also finds many of the same German writing
patterns in the English writing of German scholars. A major conse-
quence of these distinctions is that German scholars view English
academic writing as superficial; in contrast, English scholars view
German academic writing as pretentious and badly organized.
7.7.2 Japanese
One of the more fruitful demonstrations of contrastive rhetoric
has been the study of major Asian languages in contrast to
English. Recent research has argued that Japanese, Korean,
Chinese, Thai, and Vietnamese all provide useful contrasts with
the discourse practices of American English. Hinds (1983a, 1987,
1990) has devoted considerable time to the study of discourse
organization in Japanese texts in contrast with English texts. In
particular, Hinds has argued that a major alternative rhetorical
pattern in Japanese writing is the organization of texts according
to a Ki-Shoo-Ten-Ketsu framework, a framework with origins in
classical Chinese poetry. The major contrast with English writing
expectations is the third elem ent of the development, Ten, which
develops a subtheme in a m anner that would be considered off-
topic in English. As Hinds (1983b: 188) notes, ‘it is the intrusion
of the unexpected elem ent into an otherwise normal progres-
sion of ideas’. The final element, Ketsu, represents the
conclusion, but that label is misleading in terms of English writ-
ing expectations. In this Japanese writing format, the conclusions
may only ask a question, indicate a doubt, or reach an indecisive
endpoint. By English standards, such a conclusion appears
almost incoherent.
7.7.3 Korean
Eggington (1987) has made a similar argument for an alternative
rhetorical pattern in Korean writing, one which would seem to
derive from the same source as the Japanese framework. The
Korean rhetorical structure, Ki-Sung-Chon-Kyul, follows a pattern
of (1) introduction and loose development, (2) a statement of the
main idea, (3) concepts indirectly connected with the argument,

and (4) a conclusion of the main theme. Eggington argues
that Koreans prefer this pattern if they are neither bilingual in
English and Korean, nor have been extensively exposed to English
prose.
7.7.4 Chinese
A similar pattern has been noted in Chinese by Cheng (1985),
who argues that a Chinese rhetorical style consists of a four-part
pattern similar to the four-part patterns noted above for Korean
and Japanese. This pattern is, again, believed to have originated
historically in Chinese poetry. Such sources may also explain the
extensive use of allusions and historical references noted in
Chinese writing (Tsao 1983). The four-part pattern may also have
a historical relation to the Confucian eight-legged essay (Scollon
1991). While the concept of the eight-legged essay has been a con-
troversial issue in Chinese-English contrastive rhetoric, Scollon
(1991: 7-8) makes the following argum ent in support of its influ-
ence:
Kaplan (1966) ... introd u ced the n o tion th at the ‘eight-legged essay’
was of structural significance in understan d in g contem porary
Chinese writing in English. O thers have taken exception to the use
o f this essay for com parison ... [and] ... have argued th at m ost con-
tem porary Chinese have little or no knowledge o f the ‘eight4egged
essay’ and, therefore, are n o t likely to be influenced by its structure.
This argum ent is com parable to saying th at m ost contem porary
Am erican writers have never read Aristotle and, therefore, the study
of A ristotle’s rhetoric will shed no light on W estern com position
practices.
While these issues will continue to be debated, the study of

Chinese-English rhetorical contrasts remains an im portant area of
research, both for contrastive rhetoric and for second language
writing more generally.
7.7.5 A quasi-inductive style and reader responsibility

In a more recent article, Hinds (1990) suggested that the above
Asian rhetorical frameworks are really all variants of a general
organizational strategy in writing which he terms quasi-inductive.
The major premise of this argum ent is that English writers (and
readers) are familiar with strictly deductive and inductive frame-

works, but not with a quasi-inductive framework.
Looking at Japanese in this way, Hinds notes that the thesis
statement is often buried in the passage. Japanese readers do not
expect a thesis to be explained to them at the outset. Instead,
Japanese readers are better at contextualizing a text than are
English readers. He notes a similar preference in Korean, citing
Eggington. In Korean, often no thesis statement is given; rather,
there is a list of points revolving around an unstated central
theme. Chinese, Hinds argues, also makes use of quasi-inductive
organization. In general, Hinds argues that there is an Oriental
style (though not the only style) involving a delayed introduction
of purpose.
This argument for a quasi-inductive style, with the topic implied
but not stated, and with subtheme development, may be a mani-
festation of H ind’s assertion that Japanese is a reader-responsible
language; that is, readers are expected to work to fill information
and transitions, and a writer who does all the work for the reader
is not as highly valued. In contrast, a writer-responsible language
places the burden on the writer to make relationships, purposes,
and main messages as transparent as possible within the conven-
tions of the text type. Reid (1988) argues that reader-responsible
preferences in Chinese may explain consistent findings for
shorter sentence length in the English writing by Chinese stu-
dents, this feature being a reflex of brevity and ‘reading between
the lines’ in Chinese writing. Similar arguments are given by
Matalene (1985) and Ostler (1987).
Jenkins and Hinds (1987) also explored the issue of reader or
writer responsibility as a major dimension in comparing business
letter writing in French, English, and Japanese. They found that
French conventions for letter writing were strongly writer respon-
sible. The writer is strongly constrained to follow expected
conventions of politeness and formal letter writing, and the writer
does not take into account the reader’s point of view. English let-
ters were seen as much more reader responsible; that is, the writer
is obliged to view the rhetorical situation from the reader’s per-
spective. Japanese letter writing was seen as oriented to the space
between the writer and the reader in that the form of the letter
took precedence over either the writer’s or the reader’s perspec-
tive. In Japanese business letters, writers will often search guide
books for set expressions to use.
7.7.6 Thai
Analysing rhetorical contrasts between English and Thai from a
somewhat different perspective, Bickner and Peyasantiwong
(1988) examined sets of students writing on the same task. They
found that Thai writers used more repetition, made extensive use
of lists, and often did not use conclusions. They also noted that
Thai student writing tended to be more impersonal, perhaps
through a lack of speculation or a lack of future oriented conclu-
sions; this may be due in part to the absence of counterfactual
statements in Thai writing (cf. Kaplan 1987).
In a similar study, Indrasutra (1988) examined English and
Thai students’ narrative writing. American students wrote essays in
English, and Thai student wrote essays both in English and in
Thai. While structure and cohesive measures did not indicate
im portant differences, there were clear differences in the prefer-
ential structure of the narratives. According to Indrasutra, Thai
narratives preferred analogy for narrative descriptions, making
more use of figurative language such as metaphor, simile, and per-
sonification. She argues that this distinction may be due to the
different role narratives play in Thai culture; narratives are an
im portant means for exposition and instruction - a practice
unlike that in the American educational context. Thai students
did not create stories to entertain or generate interest; rather,
their stories were taken only from real life and were intended to
explain or instruct. With respect to the Thai students writing in
English, Indrasatra (1988: 221) notes:
‘transfer’ occurs when the T hai students convey a p attern o r m odel
of w riting in Thai into writing in English as in the following cases.
T he use of nouns instead o f pron o u n s according to the degree of
formality shows transfer o f Thai conventional style into writing in
English. In addition, transfer o f conventional discourse structure is
shown w hen the T hai students wrote in the same way as they wrote
in Thai; exam ples are frequencies o f m ental states an d descriptions
of m ental states. Overall, the Thai students’ w ritten com positions in
English are m ore similar to the Thai com positions than they are to
the A m erican com positions.
7.7.7 Vietnamese
O ther evidence for distinct rhetorical patterning by Asian students
is argued for in So ter (1988). Examining the narratives of
Vietnamese students writing in English in Australia, she noted that

Vietnamese students focused less effort on the plot development
and more effort on the attributional features of the characters. As
So ter (1988: 198-9) states:
Various V ietnam ese narratives drawn on for the study reveal ... that
the V ietnam ese stories appear to be less goal-oriented and hence
less focused on plot than the typical English story.... A greater
em phasis appears to be placed on relationships am ong the partici-
pants in the telling of the story situation and on the in n er states of
characters within the story. T he Vietnamese students in particular
also drew m ore heavily on dialogue in their stories and the dialogue
inform ation did n o t include inform ation that, in general, forw arded
the action o f the story b u t was reflective o r attributive in nature.
7.7.8 Hindi
Evidence for culturally distinct rhetorical preferences has also been
argued for in a series of studies by Y Kachru (1983, 1987, 1988)
with respect to Hindi. In analyses of Hindi and English texts, she
has pointed out that syntactic and cohesive features differ in the two
languages as a consequence of distinct rhetorical preferences in the
organization of discourse. In one study of expository prose (1983),
she has argued that Hindi expository prose organization is some-
times spiral rather than linear, reflecting circular patterns of
organization in traditional Hindi culture and religion. She further
argued that traditional organizational patterns are also found in
Indian-English writing. Similar arguments have been made for
Marathi, another Indo-Aryan language (Pandharipande 1983).
In a more recent study, Y. Kachru (1988) notes that while cer-
tain Hindi expository prose essays follow linear patterns of
organization and obey English conventions of paragraph unity,
topic statement, and support for an argument following the claim,
other expository prose writing disregards these conventions. It is
im portant to note that she does not claim that all English exposi-
tory prose necessarily follows the above conventions, but that they
represent conventional reader expectations in English.
7.7.9 Arabic
The study of Arabic-English differences has been an interesting
source of contention through much of the history of contrastive
rhetoric. In Kaplan’s (1966) early discussions of contrastive

rhetoric, Arab students who were writing in English were seen as
writing in a pattern characterized by repetition and elaborate par-
allelism rather than in a linear pattern.3 Ostler (1987), in an effort
to explore Kaplan’s generalizations more carefully, argued that
Arabic ESL writers are heavily influenced by classical Arabic, a lan-
guage which is more reflective of oral traditions in language use.
As a result, she argues, the writing patterns of the Arab ESL stu-
dents reveal greater uses of coordination and apposition than
subordination as well as greater uses of parallelism and symmetry
in and across clauses. While her corpus was not large and the com-
parison problematic, the study did demonstrate an empirical
approach to exploring Arabic-English contrasts.
This study has been criticized in particular because it did not
examine Arab students writing in Arabic and because the text
analysis ignored a num ber of other possible interpretations. While
not directly countering the results of Ostler, Sa’Adeddin (1989)
reinterprets the many textually oral features in Arabic texts as rep-
resenting one rhetorical option in written Arabic, though one
which is typically preferred - an ‘aural’ mode of writing. This
mode is preferred because it indicates solidarity and shared cul-
tural beliefs. Its salient features will include a loose pattern of
organization, overemphasis, repetition of specific syntactic struc-
tures, development by addition and accumulation, etc. In
contrast, a ‘visual’ mode of writing is available in Arabic, but is
seen as distant and non-interactive. Its salient features include lin-
earization, elaboration of sentences and paragraphs, complicated
thematic structure, and a clearly delimited ending. Although the
characterizations for an ‘aural’ style is similar to Ostler’s analyses,
Sa’Adeddin makes the im portant point that Arabic writers have
the option to write in a ‘visual’ style, much as English writers do,
but that it is culturally dispreferred (see also Zellermeyer 1988 for
a similar comparison between Hebrew and English).
Hatim (1991) reinforces the point that Arabic writers have his-
torically had the option to develop arguments in writing in terms
of balanced counter-arguments, presenting the opponent’s view,
but then countering it. In m odern Arabic, however, this option is
not preferred; instead, a preference is given to argumentation
which either makes no reference to an opposing view or presents
a lop-sided argum ent with an explicit concessive (e.g. although).
For both Hatim and Sa’Adeddin, the issues of audience and an
interactive stance with the reader appear to play prom inent roles
in discussing linguistically observable differences between Arabic
and English rhetorical preferences. It is worth noting, in passing,
that a common critique of Ostler’s study, that it only examined
ESL writing, is not the issue with Sa’Adeddin or Hatim. Both of
their studies support distinctions noted in Ostler. The primary dif-
ference of opinion is with the interpretations as to what these
differences represent.
7.7.10 Spanish
Perhaps the most common contrastive rhetoric comparison is that
between Spanish and English. In the early 1970s, a num ber of
dissertations examined aspects of Spanish-English rhetorical con-
trasts. In the 1980s, additional studies have examined both
linguistic and rhetorical differences. Four studies, in particular,
lead to the notion that Spanish writers prefer a more ‘elaborated’
style of writing; that is, Spanish writers, whether writing in Spanish
or in English, will typically make greater use of both coordination
and subordination in clause structuring.
1. Reid (1988) first pointed out that Spanish ESL writers wrote
longer sentences, used more coordinate clauses, and used as
many subordinate clauses as English LI writers. At the time, she
suggested that these results might indicate a pattern of ‘loose
coordination’ as discussed in Ostler (1987).
2. Montano-Harmon (1988, 1991) compared secondary school
Spanish writers in Mexico and English LI writers in the USA,
and observed that Spanish writers wrote longer sentences, used
fewer simple sentences, and used more coordinating clauses.
In this case, the comparison was made across two groups of stu-
dents writing in their own first languages.
3. A third study by Lux (1991; Lux and Grabe 1991) compared
university level Ecuadorean Spanish writers and University
English LI students in the USA. Once again, the results
revealed that Spanish writers wrote longer sentences. They also
made greater use of subordinate clauses, but there was no dif-
ference in the use of coordinate clauses.
4. Reppen and Grabe (1993) compared the writing of Spanish-
speaking elementary ESL students with low socio-economic
English LI students. Once again, Spanish writers wrote longer
sentences (T-units), used more coordinate clauses and used

more subordinate clauses. While one might argue that the
coordination is a reflection of low proficiency in English, such
an argum ent does not match the results of either Montano-
Harmon or Lux.
Taken together, the results of the four studies consistently argue
for a more elaborated style preference for Spanish writers. This
consistent pattern applies across ESL student writing and Spanish
writing, across a range of distinct cultural contexts, and across
adults, adolescents, and children.
7.7.11 Research using multiple textual parameters

Another im portant area of investigation in contrastive rhetoric
involves the use of multiple textual parameters derived from lin-
guistic feature analysis. Carlson (1988), working with samples
from the Test of Written English (TWE), found that predictors of
holistic scoring varied by language group. This was in part due to
different overall language abilities in English, but she also sug-
gested that language background itself may be a cause of the
differences in the predictors that explained holistic scoring.
Looking specifically at a comparison between Brazilian-
Portuguese and American-English editorial writing, Dantas-
Whitney and Grabe (1989) examined the texts for linguistic
features that would reflect functional dimensions of text structure
(cf. Biber 1988, Grabe 1987). For example, texts which make
extensive use of linguistic features such as first and second person
pronouns, subordinators, infinitives, and general hedges appear
to be more interactionally oriented than texts which make low use
of these features but high use of repetition in prose. In this study,
Dantas-Whitney and Grabe found that there were significant dif-
ferences between the two samples in terms of abstract/formal wersus
situational/informal presentation of information. This dimension
was represented by high use of nominalizations, prepositions, and
words per sentence for abstract/formal style versus low use of
these features and high use of third person pronouns and locative
adverbs for situational/informal style. There was also a tendency
for Portuguese editorials to appear more interactional than the
English editorials, which were more informationally oriented,
though results were not statistically significant.
In a second analysis examining multiple functional dimensions,

Lux (1991, Lux and Grabe 1991), studying the writing of tertiary
level students in Ecuador and the USA, argued that Spanish-
language writers, whether as LI or as L2 students, wrote more
‘elaborated’ prose than did English-language writers. On the basis
of a multivariate analysis of 22 linguistic features, Lux proposed
four factors which were useful in interpreting the comparisons
among writing groups. Most im portant for the purposes of this
discussion was the ‘elaborated style’ factor noted above.
A third multidimensional approach to contrastive rhetoric,
though somewhat different from the above methods, follows from
the work of Purves and his colleagues in the International
Educational Assessment (IEA) study of written composition
(Gorman et al 1988, Purves and Takala 1982, Vahapassi 1988). In
this study, students in 14 countries and at three age levels wrote on
eight different writing prompts (Purves and Hawisher 1990,
Purves and Purves 1986). As a result, the project was able to com-
pare writing across different age levels, across countries, and
across task and text type. Purves and Hawisher (1990) describe
one study in which 100 essays from each country were rated on a
num ber of functional dimensions:
(1) personal-impersonal
(2) ornam ented-plain
(3) abstract-concrete
(4) single-multiple
(5) propositional-appositional.
These dimensions involved, respectively, references to the writer’s
thoughts; to the use of metaphors, imagery, and figures of speech;
to the amount of specific information; to the perceived narrow-
ness and unity of topic; and to the types and numbers of
connectives used. Results of their study suggested that writers
from different countries could be placed in a multidimensional
profile. For example, the Italian writers were rated ‘high’ in all
five categories; that is, the writing was judged as highly personal,
ornam ented, abstract, single-focused, and propositional. In con-
trast, the Finnish writers were rated low on all five dimensions. In
this way, students from 14 different countries were rated. While
this approach is intriguing, it remains to be seen what specific
consequences of this research will be forthcoming, and what
implications will be suggested for writing instruction.
These multidimensional studies are only suggestive of the

potential research options that are available. In the future,
research on contrastive rhetoric will require studies of a wide vari-
ety which relate linguistic and discourse-structure counts to
functional/textual dimensions of text. In this way, the many indi-
vidual findings can be corroborated and explained in terms of the
functional structure of discourse. Unfortunately, such research,
done on a large scale, requires considerable time and resources.
To summarize the review of the research on contrastive
rhetoric, there is considerable evidence that different cultures
have different rhetorical preferences for the organization of writ-
ten text. While contrastive rhetoric research is a complex
undertaking, there is now sufficient reliable evidence that con-
trastive rhetoric preferences not only shape written text in distinct
languages and cultures, but tend to manifest themselves consis-
tently, if subtly, in the writing of students learning a second
language. When combined with the evidence on discourse varia-
tion from other disciplines and the logical arguments for
contrastive rhetoric, the issue should centre on ways to under-
stand differences revealed by contrastive rhetoric, rather than ask
whether contrastive rhetoric is a fruitful avenue for research
(Connor 1995, Leki 1991).
7.8 The uses o f contrastive rhetoric
One difficulty in early contrastive rhetoric research lay in the fact

that it concentrated on the final product without giving adequate
attention to the ways in which text was produced. It has become
clear in recent years that the basic questions repeatedly raised with
respect to contrastive rhetoric cannot be answered by looking
exclusively at final products. A second problem in early contrastive
rhetoric research lay in the failure to compare similar text genres,
and therefore led to repeated attempts to compare student writ-
ing with the writing of professional writers or, at least, the writing
of skilled writers. Those researchers working in this area have
learned to control experiments so that genres actually are compa-
rable.
A third difficulty that marked earlier contrastive rhetoric
research was the failure to determ ine whether the writer being
studied was able to produce effective prose in his or her native
language. It is likely that an individual with constrained ability in

his or her native language will not exceed those constraints in a
second language. A fourth difficulty in early contrastive rhetoric
research lay in the fact that deductions were made by examining
deviation from the norms of English only, rather than examining
the discourse of the LI. This last difficulty has been substantially
overcome through the research cited above referring to particu-
lar languages; much more is known now about the rhetorical
structures of at least some other languages, and research in this
area is continuing. But the most serious problem lies in the fact that
there is no universal theoretical model for contrast; it is regret-
tably the case that the findings of various scholars cannot easily
be compared because results were often derived from different
research paradigms and from different empirical bases. (This is
not just a problem for contrastive rhetoric, but for all of discourse
analysis as a field of research; see, e.g., Kaplan 1991; cf. Biber
1995.)
These problems constrained the usefulness of contrastive
rhetoric both as a research base and as a base from which to make
pedagogical decisions. What is clear is that there are rhetorical dif-
ferences in the written discourses of various languages, and that
those differences need to be brought to consciousness before a
writer can begin to understand what he or she must do in order to
write in a more native-like m anner (or in a m anner that is more
acceptable to native speakers of the target language). Precisely
because written language has relative durability (it can be held in
place and examined) and perm anence (it can be stored and
brought forth at some other time and place), readers expect a
precision and a clarity in written language that they are willing to
forgo in oral language; that is, they will easily tolerate breakdowns
in pronunciation, syntax, lexicon, and top-level structure in oral
language that they will not tolerate in written language (cf.
Sa’Adeddin 1989). In oral language, communicability counts, but
in written text accuracy becomes more important.
It is not only in the area of mechanics that contrastive rhetoric
can be of assistance in the pedagogical setting. Students need to
be made aware of the fact that the ability to compose in one lan-
guage does not assure a comparable ability to compose in
another. It is very difficult for students to break out of a ‘simplis-
tic’ frame of organization (such as the five-paragraph essay) in a
second language if they do not have access to the ‘simplistic’
framework to begin with. While the teaching of organizational

models is not currently a popular pedagogical strategy (cf.
Smagorinsky 1992, Stolarek 1994), it is a well-recognized
approach in other areas of instruction involving creativity. In art
and music, students are taught basic forms so that they then have
frameworks to manipulate and to deviate from in original ways.
(It should be noted that this argum ent is not an argum ent for a
current-traditional approach to writing instruction; it only points
out the need for L2 students to understand organizational
aspects of written text.)
Similarly, it is necessary to have practice in particular genres to
becomes skilled in those genres (Bereiter and Scardamalia 1987,
Martin 1989). There may be no absolute constraints on the orga-
nization of larger blocks of text structure, but there are, in many
genres, preferred expectations about the way information should
be organized (Atkinson 1991, Swales 1991). These preferential
expectations can be examined in English prose; non-native stu-
dents can be made aware of the preferences and they can work
with these patterns of preferential organization pedagogically
through a num ber of activities (e.g. metacognitive skills training,
summarizing, paraphrasing, putting in paragraph boundaries,
unscrambling blocks of text, placing details and examples into a
summary, examining and writing organizing sentences for texts,
and looking for markers of superstructure; Evenson 1990, Harris
1990; Pressley et al. 1989).
7.9 Contrastive rhetoric and the development o f a

theory o f writing
It is always hazardous to move from a discussion of a theoretical

notion to its implications for teaching. It is more commonly the
case that theory does not lead directly to instructional implica-
tions; rather, the theoretical insights suggest a need to have
students understand a notion or set of notions, but not a particu-
lar m ethod which necessarily represents the pedagogical answer.
This relation holds true with the notion of contrastive rhetoric. It
is evident that contrastive rhetoric offers some implications for
what students need to know; it does not clearly provide answers
for exactly how to facilitate this understanding for a student.
Leaving aside the question of how to teach ‘towards an important
notion’, it is im portant to consider, in a general sense, the implica-

tions for instruction raised by contrastive rhetoric.
It is reasonable to suggest that contrastive rhetoric focuses
attention on seven types of knowledge important in the teaching
of writing.
1. Knowledge of rhetorical patterns of arrangem ent and the rela-
tive frequency of various patterns (e.g. exposition/argument:
classification, definition, etc.).
2. Knowledge of composing conventions and strategies needed to
generate text (e.g. pre-writing, data-collection, revision, etc.).
3. Knowledge of the morphosyntax of the target language, partic-
ularly as it applies at the intersentential level.
4. Knowledge of the coherence-creating mechanisms of the target
language.
5. Knowledge of the writing conventions of the target language in
the sense of both frequency and distribution of types and text
appearance (e.g. letter, essay, report).
6. Knowledge of the audience characteristics and expectations in
the target culture.
7. Knowledge of the subject to be discussed, including both ‘what
everyone knows’ in the target culture and specialist knowledge.
Clearly, this list is primarily, though not entirely, concerned with
product; however, it is worth reiterating that it in no way endorses
a ‘current-traditional’ approach to writing instruction (Berlin
1987; cf. Silva 1990). While the emphasis is on the text rather than
on the mental processes through which the composition is gener-
ated, these latter processes have not been, and cannot be,
ignored. Extended discussion of these practical implications of
contrastive rhetoric for writing instruction is addressed in Grabe
and Kaplan (1989). Many of the specific recommendations for
instruction in Chapters 10 through 12 also suggest ways that
insights from contrastive rhetoric support instructional practices.
It should be clear from the discussion in this chapter that
insights from contrastive rhetoric are applicable to research on
composing, and that there are a num ber of pedagogical implica-
tions that can be drawn for writing instruction in L2 contexts.
How contrastive rhetoric is combined with other theoretical
notions to build a theory of composition is the goal of the next
chapter.
Notes
1. Kaplan prefers the term ‘n o tio n ’ over th at o f ‘th eo ry ’. Contrastive
rhetoric does n o t yet com prise a fully articulated set o f principals and
m ethods that define this area o f research.
2. Labov (1971) makes a strong argum ent for the power o f converging
evidence.
3. While the distinction between parallel patterns versus linear patterns
was an oversimplification, it has served to galvanize a useful discussion.
8
Towards a theory of writing
If we think of a theory as a machine to think with, a

device for organizing and interpreting events with the
aim of bringing other questions and other forms of
evidence into conjunction, then it is not at all un-
reasonable to aspire to a theory of how writing
contributes not only to our understanding of the world
but also of ourselves.
(Olson 1994: xvii)
8.1 Introduction
In the first seven chapters, major trends in writing research and

differing contexts for writing and writing instruction have been
discussed. Together, these seven chapters provide the background
for a reconsideration of the nature of writing, covering issues in
textual structure, cognitive processing, and social contexts. In this
chapter, we would like to step back and consider once again the
basic question, ‘What is writing?’ This will be explored first
through an ethnography of writing; then the various issues raised
by the ethnography will be reconsidered through a taxonomy of
writing skills and contexts; finally, a descriptive model of commu-
nicative language processing will be suggested as a means for
integrating the cognitive, social and textual domains of a theory
of writing.
One often finds discussions similar to the model proposed in
this chapter in instructional contexts. For example, ‘the rhetorical
triangle’, treats writing as a combination of writer, reader, subject
matter, and text. It is interesting to note, in contrast, how so much
of writing research is not contextualized within such a larger
framework. For example, Cooper and Matsuhashi (1983) empha-
sized textual issues, de Beaugrande (1984) and Bereiter and
Scardamalia (1987) emphasized processing issues, and, more
recently, Flower (1994) has emphasized the combination of cogni-
tive and social contexts.
202
Towards a theory of writing 203
Witte (1992), in a general theoretical discussion, is one of the

few researchers who has argued that a theory of writing needs to
synthesize the cognitive, social, and textual (linguistic) factors.
We, in our turn, would like to suggest a synthesis which incorpo-
rates these three major domains commonly discussed in writing
research, but seldom integrated. As these domains are integrated,
the goal is to arrive at a balanced interpretation of what it means
to be able to write.
8.2 Towards an ethnography o f writing
Perhaps the best initial way to consider the overall set of concerns
involved in writing is to apply an ethnographic approach to our
current understanding of writing. This approach, applied to the
study of spoken language, has led to the sociolinguistic field of
ethnography of speaking and conversation analysis (Poole 1991,
Schiffrin 1991). An equivalent effort has not been applied to writ-
ing, in part because many linguists, including many sociolinguists,
remain convinced that written language is derived from spoken
language (e.g. Basso 1974, Biber 1988, Olson 1994). However, as
we have argued in Chapter 1, there are sufficient reasons to recon-
sider the relationship between spoken and written language, and
to attend as carefully to written language as we have to spoken lan-
guage.
One of the best ways to attempt a first ethnography of writing is
to ask the basic question (e.g. Cooper 1979):
Who writes what to whom, for what purpose, why, when, where, and how?
Providing a taxonomic answer to this question will lead to an ini-

tial approximation for an ethnography of writing.
8.2.1 Who
A first requirem ent of the ethnography is a taxonomy of writers. Is
the individual a beginning writer or a mature experienced writer?
Is the individual experienced in a wide variety of writing or only in
a narrow range of writing? Is the writer a student who expects to be
evaluated academically or a journalist who earns his keep by writ-
ing? These and many other related questions form a complex
matrix that must be analysed if any classificatory system is to result.
Knowing who the writer is (the characteristics of the writer

rather than, perhaps, his or her persona) can have an important
bearing on the nature of the writing that is studied. For example,
the young adult who does not have extensive experience in writ-
ing, but must write a report in a second language, will write quite
differently than the skilled adult writer who has experience in the
writing task and is writing in the LI. This difference in the charac-
teristics of the writer will be quite independent of other influences
(e.g. audience) which impact the writing situation - and it is not
simply a general proficiency issue.
8.2.2 Writes
The term ‘writes’ might normally suggest an action or process.
The issue of process, however, is discussed below (see ‘how’, sub-
section 8.2.8) and the notion of ‘writes’ is used here to examine
the linguistic nature of texts, the writing.
The study of the writing situation requires a theory of the text
itself, a theory of text construction. What are the linguistic parts
and how do the parts work together? What are the linguistic
resources? To what extent do linguistic features reflect some func-
tional purpose in the writing? How do sentences link together to
form a larger text (if indeed sentences are involved)? How are we
to understand the notion of coherence? And what part of this
notion resides in the text?
These and many other issues were explored in some depth in
Chapters 2 and 3. The text itself is an im portant independent
com ponent of the overall writing situation. Only through analysis
of the text can researchers examine the uses of particular linguis-
tic structures, transition devices, and lexical choices, as well as the
functional roles these uses might play in the context of the entire
text. Study of the text reveals the (in)appropriate use of formal
conventions such as opening statements, external reference,
stages in the sequencing of information (Atkinson 1991), and the
rhetorical arrangem ent of information (Bruthiaux 1993). Study
of the text also reveals patterns of information structuring in
terms of ‘given-new’ information ordering, ‘topic-com m ent’
arrangem ent, and ‘them e-rhem e’ structuring. A theory of text
construction contributes independently to the writing situation
in that it provides a framework for the various linguistic tools
available to the writer as well as combinatorial choices which
Towards a theory of mriting 205
create the flow of information and the notion of coherence. The

linguistic elements in text construction also aid interpretation of
other factors in the writing situation such as audience considera-
tions, writer’s purpose, the writing context, and the gepre
required by the task.
8.2.3 What
The most basic definition of what is written is some message and
some type of content. For our purposes, the what of writing will
be discussed in terms of content, genre, and register. These con-
cepts suggest a num ber of questions for writing: What are the
types of writing the writer typically engages in creating? What sorts
of general background information does the writer need? To what
extent is knowledge of specialized registers necessary for writing?
How can we define a theory of genre? To address these and
related questions, a theory of writing must take into account the
phenomenological world (a theory of world knowledge), a theory
of genre, and some specification of register.
Typically, we can think of the content as background knowledge,
for example, as schema theory. In addition to general background
knowledge, schema theory suggests that specific sets of knowledge
stored as integrated units are accessible for retrieval (or recon-
structing) and are used in understanding and producing content
knowledge. Schemas also provide frames for our knowledge of
appropriate register in different contexts and our knowledge of
genres as ways to organize discourse for specific purposes (Swales
1990). The basic influence of schema theory (or similar theories
of the mental organization of knowledge) on writing is apparent
in research which shows that students write more when they are
writing about information with which they are familiar (e.g.
Freidlander 1990 for L2). Background knowledge provides con-
tent and genre-structure resources for writing.
It should also be noted that background knowledge is, at least
to some extent, culturally derived. Our culturally shaped back-
ground knowledge can lead to misunderstandings because this
knowledge is not typically recognized as varying by culture. As a
result, we are often unaware of the assumptions and presupposi-
tions that guide our logic and our conclusions; rather, we think of
our own ways of knowing as common sense. However, when two
cultures come into contact, what is common sense, what ‘goes
without saying’, is by definition not the same. Even within a given

culture, differences may exist across generations, across social
groupings, across genders (Atkinson 1993, Lemke 1995). Thus,
some framework which explicitly represents background knowl-
edge is critical for text construction research.
Genres might simply be described as discourse types that have
identifiable formal properties, identifiable purposes, and a com-
plete structure (i.e. a beginning, a middle, and an end). Genre is
a concept which applies to both oral and written language,
though the primary issue, for our purposes, is the occurrence of
genre in written discourse. While it may be possible to discuss
genres in broad macro-generic terms such as expository prose,
persuasion, and narration, genres are more typically defined by
narrower contexts and more specific formal features. The follow-
ing provides a few examples of discourse types which can be
considered written genres (e.g. Martin 1985, Rothery 1989, Swales
1990):
fiction novels grant applications progress reports
ransom notes course syllabi survey articles1
Until recently the notion of genre has been relegated to a sec-
ondary status in writing theory, though it is used widely in writing
instruction and in text-linguistic and psychological research. In the
past decade, however, the notion of genre has been receiving
increasing recognition as an integral factor in the writing situation,
one which makes an independent contribution (e.g. Berkenkotter
and Huckin 1993, 1995, Bhatia 1993, Freedman and Medway 1994,
Lemke 1995). For example, the text of a fictional narrative will be
consistent in certain ways with other fictional narrative texts and
will differ from texts that are not fictional narratives in ways which
cannot be predicted specifically and completely from information
about the writer, the audience, the writing process, the topic, the
writing context, or the purpose of the writer. Thus, a reader might
expect a descriptive setting in a fictional narrative (even if not at
the very beginning) but will not expect a similar setting sequence
in a letter or a newspaper editorial.
Register; as independent from genre considerations, is defined
by the topic of the writing, the medium (always writing in this
case), and interpersonal tenor (Halliday 1978, Halliday and Hasan
1989, Martin 1992). Seen in this way, register also appears to have
an im portant influence on writing. Writing about vacations and
Towards a theory of xvriting 207
travel will be different from writing about economics, anthropol-

ogy, physics, medicine, or law, and will lead writers to use different
linguistic resources. While certain topics will predispose a writer to
select particular genres, it is also true that research articles, popu-
lar articles, and introductory textbooks dealing with music and
photography will be different from the same genres dealing with
physics and biology (Grabe 1987). Register aspects related to
interpersonal relations are treated in detail below in the discus-
sion of audience. Taken together, content knowledge, genres and
registers comprise social, topical, and cultural resources which
strongly influence writing.
8.2.4 To whom
Another major issue for a framework of writing is the develop-
m ent of a theory of audience. Audience is essential to the creation
of text and the generation of meaning. In terms of audience, the
following and many other related questions can be raised.
Who is the intended reader of the writing? Is the reader an
abstraction? Is the reader invoked equivalent to the reader
addressed (intended)? Is the reader a known individual? If the
audience is known, how close or distant is the reader? How much
shared background knowledge exists between the reader and the
writer? How much shared specific knowledge of a particular topic
exists between the reader and the writer?2
Some definition of the person (s) expected to read the writing has
a major influence over the discourse of the written text. Within
the general concept of the reader, or audience, are a num ber of
factors which constrain the decisions of the writer (Kirsch and
Roen 1990). It is perhaps even preferable to consider ‘parameters
of audience influence’ rather than specific features in order to
provide a more thorough account; at least five such parameters
would appear to play im portant roles in textual variation.
1. One param eter of reader influence on the writing is the num-
ber of persons who are expected to read the text. A text intended
for oneself, a single person, a small group of people, a large group
of people, or a general audience will influence the text structure.
A related issue is the extent to which the audience is an invoked
audience for a rhetorical purpose rather than a ‘real’ audience
definable by the writer (cf. Long 1990, Willey 1990); that is to say,
the writer provides cues which indicate the persona (invoked)
which the reader should take on during the reading. Another

related issue is the extent to which the specific audience for a text
is the intended audience. A diary may be intended for the use of
the writer, but ever since Samuel Pepys’s diary was published,
some diary writers write not for themselves but for posterity.
2. A second param eter of audience variation is the extent to
which readers are known or unknown. Writing to a known person,
a less familiar colleague, or a stranger is likely to alter the text.
This is well documented for oral language interaction in Wolfson’s
(1989) ‘theory of the bulge’; that is, those who are recognized to
be in the wide middle ground between stranger and close col-
league typically require more elaborate responses, and usually
require more hedged language expressions. Thus, degree of close-
ness to the reader is likely to determine the extent of interactional
and involvement features which appear in the writing (Biber 1988,
Tannen 1987, 1989).
3. Along a third parameter, that of status, the writing will vary
according to whether the reader has a higher status, an equal sta-
tus, or a lower status than the writer. In oral language contexts,
Wolfson (1989) has demonstrated that status also creates dis-
course variation, with higher and lower status listeners receiving
much less interactional negotiating. In a somewhat different con-
text relating to status, the knowledge that a particular person (e.g.
a well-known scholar) will be a reader may influence the writing,
though there is no research yet which specifically addresses this
issue.
4. As a fourth parameter, the extent of shared background
knowledge will influence the writing to a considerable degree;
that is, readers with a high degree of shared background knowl-
edge are likely to influence the writing in particular ways. Writing
for readers who are familiar with current events in certain cultural
contexts will allow the writer to anticipate general knowledge on
the part of the reader, and to allude to types of knowledge which
separate those who know from those who do not (e.g. newspaper
editorials, professional journal articles).
5. As a final parameter, the extent of specific topical knowledge
shared by the reader and the writer will influence the writing. The
extent and choice of detail, the need for defining ideas and
assumptions, the use of common versus specialist terms, etc., will
all affect the writing (Lemke 1995). Difficulties created by this
param eter are evident in the writing of grant proposals in which
only one or two (20 or 30 per cent) of the reviewing committee

may have equivalent specialist knowledge.
8.2.5 For what purpose

This analysis will consider purpose as a functional categorization.
Purpose raises im portant questions for writing, such as:
■ To what extent is it possible to define purpose in a writing task?
■ Are there multiple purposes in every writing task?
■ How does purpose interact with genre and audience?
and many others.
The purpose for writing may, in a general sense, be said to rep-
resent an attempt to communicate with the reader. Most
discussions relating rhetoric to writing instruction assume this
relation.3 Apart from personal notes, most recognizable forms of
writing are intended for an audience other than the writer;
indeed, even such ‘personal’ items as diaries and research notes
may be intended for other audiences. In considering his or her
functional purpose (s), the writer recognizes that he or she has
specific intentions as well as informational content to convey.
Because the reader assumes these purposes on the writer’s part,
both the reader and the writer will understand and interpret writ-
ing purposes from certain accepted linguistic, psychological, and
sociolinguistic principles:
1. Gricean maxims - the need to be informative, factually correct,
relevant, and clear; and their systematically interpretable viola-
tions.
2. Speech acts - specific features in the writing which signal
speech acts by the writer, and the degree to which they are
negotiable.
3. Conventions for conveying status, power, situation, intent, and
attitude. (These indicators are not necessarily the same as fea-
tures which mark genres.)
4. Predictability of cognitive structures which anticipate and
implicate larger patterns of organization: schemata, scripts,
frames, citation patterns, goals, ethos and pathos in rhetorical
persuasion, aims of discourse, etc.
All of the above are implicated in the writer’s purpose and the
ability of the reader to discern the purpose of the writing.
The application of Gricean maxims to writing provides the

means for making inferences as to the writer’s purpose in much
the same way as a listener can infer a speaker’s purpose. It is worth
noting that the writer’s awareness and application of Gricean max-
ims in writing allows him or her to convey purpose beyond that
signalled by a genre form. Thus, writers of letters of recommenda-
tion who violate maxims of quantity or relevance send a message
quite apart from the message conveyed by the genre itself. That is,
writing a letter of recommendation implies a recommendation;
however, the specific inferences which may be derived from the
amount, accuracy, relevance, and clarity of the letter convey much
additional information.
Much the same argument can be made for each of the other
three principles above. For example, specific speech act verbs may
indicate purpose independently of genre. Markers of writer or
reader status, choice of salutation, num ber of directives used, etc.,
all indicate writer’s purpose apart from genre choice. The choice
of rhetorical patterns of organization, types of persuasive appeals,
and types of references cited all indicate an author’s purpose in
ways independent of genre. So, too, will presuppositions and cul-
tural stereotypes (Schroder 1991).
It would seem, then, that the purpose for writing must be
addressed on at least two levels. On a general level, the overt pur-
pose is related to the concept of genre; the ‘purpose’ influences
the text structure itself, selecting appropriate genre options.
However, the parameter of ‘purpose’, as used here, also addresses
functional issues which are independent of recognized written
genres. For example, people write in order to apologize, invite,
inform, praise, threaten, complain, order, explain, reject, etc. -
speech acts for which there are no specific genres which may be
consistently invoked. This parameter of ‘purpose’ is an indepen-
dent dimension of writing, potentially separable from genre and
audience, since a person could write two texts to the same audi-
ence and in the same genre but have each text serve different
functional purposes (e.g. apology, reprimand, invitation) and con-
vey different messages according to applications of Gricean
maxims. It would seem that this functional category accounts for
speech act theory in written discourse as well as for applications of
Gricean maxims to writing contexts.
It is also important to consider the problem of lacunae in cross-
cultural writing environments; that is, certain genre/purposes
may simply not exist in one cultural environment as compared

with another.4 Lacunae occur at the discourse level, omitting cer-
tain genres, audiences, purposes, etc.
8.2.6 Why
The concept of why people write refers to the underlying intentions
or motives that may or may not be revealed by functional purpose.
Under what conditions does a writer not want to communicate
fully? Are there attitudes and notions which are difficult to convey
in writing? In what situations will some group of readers not be
able to see the purpose of the writing?
It might be best to see writer’s intentions as indicated along a
cline of transparency. Genres (subsection 8.2.3) represent the most
overt indication of intention and serve to facilitate schema instan-
tiation. As such, intention constitutes a level which is, in one way,
strongly constrained by audience and topic. The purpose for writing
param eter (subsection 8.2.5) is independent of genre (or the
question what?) in the sense that there may be many purposes not
existing in a one-to-one relationship with a given genre - indepen-
dent of genres that are signalled in writing. These purposes are
assumed to be related to communicative intention and, therefore,
to be relatively transparent. Even when a person violates Gricean
maxims, as in the letter of recommendation, it is done in a way
that is understood by the knowledgeable reader. The reader can
readily infer the writer’s purpose. A third level of writer intention,
underlying intentions or motives (the focus of this section), may be
represented by the extent to which the writer wishes to manipu-
late the reader to attend to the content, and the writer may not
necessarily value transparency for the reader above other consid-
erations. It is also possible that the writer is struggling with the
content to the extent that transparency for the reader is not
attainable.
The extent to which our present param eter (why) influences
text depends on two constraints. One is the extent to which a
writer wants the reader to recognize a hidden message - the more
hidden the underlying message, the more likely it is that the
reader will only recognize the functional purpose of the text (e.g.
Lemke’s (1995) discussion of technocratic discourse). The second
situation is one in which a complex or exacting content takes
precedence over reader friendliness (e.g. legal documents,
technical specification, gang graffiti). Situations may also arise

when a writer does not want to present a set of arguments too sim-
ply and thereby insult informed colleagues who may be the
primary audience (Swales 1990: 64). In this way, the discourse
community of the specialist group acts to make the writing less
accessible to other readers.
One conceptualization of this notion is Steiner’s (1978) four
levels of difficulty. This system of variable difficulty derives from
the author, the text, and the environment as they influence the
ability of the reader to comprehend. The first type is contingent dif-
ficulty, which arises from arcane and technical reference, but
which can be corroborated at some time and in some way. The
second type is modal difficulty, which arises from inaccessible or
alien interpretations of the hum an condition. The third is tactical
difficulty, which arises from the relative desire of an author to be
understood only up to a point. The last is ontological difficulty,
which arises from the constraints imposed by the language itself
(see Figure 8.1 on page 215).
8.2.7 When and where

It is not clear to what extent the notions of when and where a per-
son writes are critical to the general taxonomy for an ethnography
of writing. In fact, the relative non-importance of these issue for
writing points out a major distinction between an ethnography of
writing and an ethnography of speaking. The immediate situation
of language use in speaking is of critical importance in determin-
ing the purpose of speaking and interpreting specific language
use. The context of ‘here and now’ appears to be much less impor-
tant to a theory of writing. This is not to say that the influence
deriving from ‘when’ and ‘where’ a person writes is negligible or
non-existent; rather, these parameters play a much smaller and less
consistent role as factors which contribute independently to writ-
ten discourse. This is so because a writer may work late into the
night, under the influence of some substance, or in difficult cir-
cumstances (e.g. E.A. Poe, S.T. Coleridge), or, by contrast, a writer
may work in a comfortable office with state-of-the-art equipment,
but the reader is not likely to know. Such information becomes
available usually only long after the fact, in biographical studies,
and then only if the writer has achieved a modicum of fame.
Knowing the approximate date of a handwritten message (e.g. a
grocery list) may well be im portant for the person reading it.
Similarly, it is im portant to know when a letter was written to inter-
pret any deictic references contained in the letter appropriately. It
is also im portant to know when a certain research article was writ-
ten in order to interpret the importance of the claims being
made; but this is more a concern for making an informed reader’s
interpretation rather than a matter which influences the writing
of the producer. The same issue may influence letters of many
sorts as well as the writing of graffiti. Typically, the dating of a
piece of writing is often all that is needed (or available) to allow
appropriate reader interpretation, neither the process of writing
nor the form of the writing being otherwise influenced.
8.2.8 How
As the final param eter in an ethnography of writing, understand-
ing how written discourse is produced centres around a theory of
on-line writing production, or, in simpler terms, a theory of the
writing process. It is noteworthy that channel (physical means of
communicating) may be less significant in written text than in oral
text. W hether a writer generates text with a pen, a typewriter, or a
word processor seems to have limited implications for the struc-
ture of text, though this perception may be derived from the fact
that little research exists. Some tentative studies suggest that the
editing capabilities of a word processor do influence both length
and rewriting at least in teaching situations (Bangert-Drowns
1993, Cochran-Smith 1991).
The cognitive mechanism for production rests at the core of a
theory of writing. It provides methods of empirical research which
complement research on the written text, and also provides the
means for exploring notions such as audience, content, and writer
intention from a processing perspective. Among the important
questions which arise are:
■ To what extent can research inform us about important issues

in the writing process?
■ How useful are models of the writing process?
■ Does the writing process vary from culture to culture?
■ Is there more than one writing process, and if there are many
processing models, what implications does this variability hold
for writing theory?
Two groups have dominated research on the writing process (es)

over the last 15 years, and have shaped current conceptual under-
standings of how writing is produced. The work of Flower and
Hayes (Chapters 4 and 5) has opened up research on the writing
process and has presented theories which, though sometimes criti-
cized as too vague, have greatly influenced the way writing
research is conducted. Bereiter and Scardamalia, in a series of
research studies (see Chapter 5), have argued that there must be
more than one writing process, that the writing process used for
skilled academic writing is a development out of a more general
writing process and is not mastered well by all writers, including
university students.
What emerges from the current research on writing processes is
that many basic findings are now widely accepted and influence
instruction (e.g. Johnson and Roen 1989, Krapels 1990, Raimes
1987, 1991, Silva 1990, Zamel 1987 for L2 influences). At the same
time, there are still shortcomings in the writing process research
which will need to be addressed in the future. For example,
although some researchers (e.g. Perfetti and McCutchen 1987)
view linguistic skills as a primary concern, they do not extend such
linguistic concerns to the influence of the formal features of
genres or to the requirements imposed by certain rhetorical inten-
tions of the writer. More generally, cognitive processing research
privileges the reader/w riter contract over influences on writing
deriving from the topic, the genre, or the writer’s intentions. As
Swales (1990: 63) points out:
W hile Flower (1979) and h er co-workers may be generally right in
their theory th at the im m ature writer produces ‘writer-based p ro se’
and the m ature writer ‘reader-based p rose,’ it would seem equally
clear that in certain genres m ature writers also p roduce ‘writer-based
prose’ [appropriately].
8.2.9 Summarizing the ethnography of uniting

This attempt to develop an ethnography is tentative. Regardless of
how the final form of such an ethnography evolves, it should be
clear that the purpose is to provide at least a descriptive account
for the question posed at the outset of this chapter, ‘What is writ-
ing?’ The ethnography also provides a foundation for a more
comprehensive theory of writing: the availability of such a theory
would then be useful for situating results from any one parameter
of writing within a larger interpretive framework.
Sociolinguists studying oral interaction would not interpret
their results outside of some ethnography of speaking framework,
yet research on written discourse seldom seems to be similarly
compelled to frame research in larger contexts (cf. Kress 1991,
Lemke 1995, van Dijk 1988, Wodak 1989, 1990). This is, in part,
due to the extensive research on written discourse by cognitive
psychologists, many of whom do not see the need to situate writ-
ing socially as part of their explanations, and often do not even
see a need to recognize different types of writing. While the
ethnography sketched out here may not be an equally suitable
interpretation of the ‘writing event’ for all researchers, some fram-
ing context for writing is needed - one which goes well beyond
the simple communication schemes which relate communicative
function to writing only in a general way.
One way to organize the various parameters of this ethno-
graphy of writing is to retain the general sequence of the
communicative orientation to writing, but to incorporate into the
structure a wider range of influencing considerations. One frame-
work attempting to do that (Kaplan 1991) is reproduced in Figure
8 . 1.
Within a socioculturally defined universe

(when and where)
Constrained by Constrained by the Constrained

author’s intent shared phenomeno- by logical
(why) cultural conventions world and by
and available genres reader’s
(what) intent
(how)
A U TH O R ^ ► TEXT ► RECEPTOR
(who) (Text construction) (to whom)
(what)
And performative And further constrained And performative

ability (process), by Steiner’s four kinds of ability and ability
and recognition of difficulty to perceive author’s
audience expectation (for what purpose) intent
(writes) (why)
Figure 8.1 Model of param eters involved in writing

In this model, the phenomenological world may be said to

include the constraints imposed by various discourse communities
and the degree to which those constraints are understood by the
participants in the instantiation, including such variables as the
function and distribution of different genres.
8.3 A taxonomy o f writing skills, knowledge bases,

and processes
A second way of organizing the full range of information involved

in the above ethnography and the simple model of parameters is
through a detailed taxonomy of writing skills and contexts. The
development of a taxonomy of writing should not be seen as an
alternative to an ethnographic description but rather a way to
build upon it. The taxonomy organizes the full range of informa-
tion in a form that is readily accessible and useful for finding gaps
and establishing new areas of inquiry.
8.3.1 Background assumptions to the taxonomy

The taxonomy offered here tends to reflect two theoretical bases.
One is the general approach towards communicative competence
in language first developed by Canale and Swain (1980, Canale
1983, Harley et al 1990; cf. Bachman 1990, Celce-Murcia et al 1994,
Chapelle et al 1993); the other is the ethnography of writing frame-
work outlined earlier in this chapter. The development of the
taxonomy began by asking about the sorts of writing situations that
exist for writers (and particularly students) in terms of settings,
tasks, texts, and topics. These situational parameters are then
reflected in some way by the writer’s goals, combined with the par-
ticular writer’s intentions, attributions to the task from past
successes and failure at similar tasks, and emotional states. In addi-
tion, writing obviously requires control over specific linguistic skills
as well as over discourse and sociolinguistic skills. Earlier chapters
have made the argument for considering linguistic knowledge as a
critical component of writing ability and the foundation for text
construction. Incorporating a socio-cognitive perspective, one must
also consider sociolinguistic skills which play a role in writing (e.g.
audience considerations). The writing process also suggests many
component skills and strategies essential to writing instruction.
The taxonomy presented below reflects this effort to structure

the writing situation and the skills, knowledge, and processes used
by the writer. The taxonomy is not in any way intended to be a
definitive representation of all aspect of the writing situation - the
current state of knowledge would not perm it the development of
an inclusive taxonomy, and any taxonomic effort necessarily
reflects the biases of its compilers. Indeed, it can be anticipated
that significant alterations would be a necessary part of using such
a taxonomy since knowledge of writing is continually changing
and improving. To the extent that gaps become evident, this
awareness, in itself, validates the purposes for generating a taxon-
omy.
8.3.2 A taxonomy of academic writing skills, knowledge

bases>and processes
I. Educational settings for writing
A. Classroom
B. Office
C. H om e
D. Library
E. W riting ce n tre /la b o ra to ry /c o rn e r
F. C om puter centre
G. D orm itory lounge
II. Educational writing tasks
A. Notes and m em oranda
B Lecture notes
C. Letters
1 . Personal
2. Invitation
3. A cceptance
4. Refusal
5. C om plaint
(a) strong
(b) m ild
6. R ecom m endation
(a) positive
(b) neutral
(c) negative
7. Business
8. Apology
9. Request
10. C onfirm ation
Theory and Practice of Writing
D. Jo u rn als/d iaries
E. Free writing
F. Recounts (forecounts)
G. Narratives
1. Fictional (novels/short stories)
2. Non-fictional
H. Recipes
I. R eports/(expository) essays
1. D escription
2. Definition
3. Exem plification
4. Classification
5. C om parison/contrast
6. C ause/effect
7. P roblem /so lu tio n
8. Analysis/synthesis
J. Poster boards/diaram as
K. Interview s/surveys/questionnaires
L. Argum entative essays
1. Logical stances
2. Ethical appeal
3. Em otional appeal
4. Empirical stance
5. Appeal to authority
6. Counter-argum ents
M. Tim ed essay tests
1. In-class
2. Take-home
3. P art of standardized test (com m ercial or academic)
N. Newspaper rep o rtin g /co lu m n s
1. H eadlines/w orld and national news
2. Local news
3. Sports news
4. Book/m ovie review
5. S ocial/political/cultural colum ns
6. Editorials
7. Advertisements
8. C om ics/cartoons
O .P o em s
P. Plays
Q. Laboratory reports
R. C h a rts/ta b les/g ra p h s/m ap s/fig u res
S. Abstracts
T. Research papers
U. G rant proposals/applications
V. T heses/dissertations
III. Educational texts used and produced (most items in section II also apply here)
A. Textbooks
B. Novels
C. Short stories
D. Poems
E. Plays
E Jo u rn als/d iaries
G. Newspapers
H. M agazines/trade jo u rn als
I. Essays (narrative, expository, argum entative)
J. C h a rts/g rap h s/tab le s/fig u res/m ap s
K. W orkbooks
L. D ictionaries/encyclopaedias/gram m ar and usage books
M. Research jo u rn a l articles
N. Professional te x ts/b o o k s/ch ap ters
IV. Topics for academic writing
A. Personal expressive
B. Im aginary narratives
C. Personal recounts
D. Biographies
E. Bibliographic works
F. Topics from family, community, regional, national life
G. Topics from social, cultural, econom ic, political issues
H. Topics from academ ic hum anities fields
I. Topics from academ ic social sciences fields
J. Topics from academ ic natural sciences fields
K. Topics from professional disciplines
V. The uniter’s intentions, goals, attributions, and attitudes
A. W riter’s reinterp retatio n o f the task
B. Awareness of complexity o f task
C. W illingness to be understo o d (perhaps only up to a point)
D. Awareness of previous success with task type and topic
E. A ttitude toward task type an d topic
F. Willingness to elaborate and experim ent with task and topic
G. Motivation to p erfo rm to capacity
1. Grades
2. H igher proficiency
3. L earn new inform ation
4. Future jo b /p ro m o tio n
5. Impress te a c h e r/o th e r students
H. D egree of creativity in ten d ed

I. A ttitude towards teacher, o th er students, institution
J. Willingness to learn
K. Awareness of metacognitive strategies
L. Awareness of differences across languages an d cultures
M. Awareness of w riter’s own intentional, attributive, an d attitu-
dinal differences in different languages
VI. Linguistic knowledge
A. Knowledge of the w ritten code
1. O rthography
2. Spelling
3. Punctuation
4. Form atting conventions (margins, paragraphing, spacing,
etc.)
B. Knowledge of phonology an d m orphology
1. S o u n d /le tte r correspondences
2. Syllables
(a) onset
(b) rhym e/rhythm
(c) coda
3. M orphem e structure (word-part knowledge)
C. Vocabulary
1. Interpersonal words an d phrases
2. Academic and pedagogical words and phrases
3. Form al and technical words and phrases
4. Topic-specific words an d phrases
5. Non-literal an d m etaphoric language
D. S yntactic/structural knowledge
1. Basic syntactic patterns
2. P referred form al writing structures (appropriate style)
3. Tropes and figures o f expression
4. M etaphors/sim iles
E. Awareness of differences across languages
F. Awareness of relative proficiency in different languages and
registers
VII. Discourse knowledge
A. Knowledge of intrasentential an d intersentential m arking
devices (cohesion, syntactic parallelism)
B. Knowledge on inform ational structuring (to p ic/co m m en t,
given/new , th e m e/rh e m e, adjacency pairs)
C. Knowledge of semantic relations across clauses
D. Knowledge to recognize m ain topics
E. Knowledge of genre structure an d genre constraints
F. Knowledge of organizing schemes (top-level discourse struc-

ture)
G. Knowledge of inferencing (bridging, elaborating)
H. Awareness of differences in features o f discourse structuring
across languages an d cultures
I. Awareness of different proficiency levels o f discourse skills in
different languages
VIII. Sodolinguistic knowledge
A. Functional uses o f w ritten language
1. Apologize
2. Deny
3. Com plain
4. T hreaten
5. Invite
6. Agree
7. Congratulate
8. Request
9. Direct
10. C om plim ent
B. A pplication and interpretable violation o f G ricean maxims
C. Register and situational param eters
1. Age of w riter
2. Language used by w riter (LI, L2, ... )
3. Proficiency in language used
4. A udience considerations
5. Relative status o f interactants (pow er/politeness)
6. D egree of formality (deference/solidarity)
7. D egree of distance (detachm ent/involvem ent)
8. Topic of interaction
9. Means o f writing (p en /p e n cil, com puter, dictation,
shorthand)
10. Means of transm ission (single p a g e /b o o k /re a d a lo u d /
printed)
D. Awareness of sociolinguistic differences across languages and
cultures
E. Self-awareness o f roles o f register an d situational param eters
IX. Further audience considerations
A. Reality o f audience (perceived real o r invoked audience)
B. N um ber in audience
C. D egree of familiarity with audience (specifically known or
not)
D. Status o f audience with respect to writer (peer/em p lo y er)
E. E xtent o f cultural, social, an d world knowledge o f audience
F. E xtent o f specialist/subject-m atter knowledge o f audience
X. Knowledge of the world

A. Declarative (semantic, topical)
B. Episodic (events, personal experiences, interactional)
C. Procedural (processes, routines, conventions)
XI. Writing process skills (on-line processing skills; not linear)
A. Goal planning routines
B. Lexical and structural activation (generating content)
C. Propositional integration
D. Text-model production
E. M ental m odel interpretation
1. Inferencing
2. M atch to processing goals
3. M atch to w riter’s intentions (rein terp reted task)
E Rapid production routines
G. Revising routines
XII. Writing process strategies (executive control or metacognitive strategies)
A. M onitoring text production
B. G enerating additional co n ten t
C. C onsidering task problem s
1. A udience considerations
2. Purpose considerations
3. Rhetorical considerations (discourse knowledge)
4. Problem s created by additional content
5. Language o f preference for problem consideration
6. C om pensating strategies for weaknesses in linguistic, dis-
course, sociolinguistic knowledge
D. Using invention strategies, ‘topics’, brainstorm ing, free writes
E. C onsidering alternative solutions
E Re-reading already p ro d u ced texts
G. Using reading resources (texts, dictionaries, data, etc.)
H. Rejecting co n ten t/rh e to rica l inform ation/alternatives
I. H olding in storage co n ten t/rh e to rica l inform ation
J. S um m arizing /p arap h rasin g /reo rd erin g inform ation
K. Predicting future outcom es to m atch goals
L. Using notes/ou tlin es/d raw in g s/o th er self-created materials
M. G etting assistance
N R eassessing/changing goals
O. Recognizing m ism atch with processing goals, au th o r’s in ten -
tions
P. Editing texts
Q. Reassessing co n ten t/rh e to rica l strategies
R. Reassessing co n ten t/rh e to rica l revisions
S. G etting feedback from others
T. C onsidering individual style concerns (voice)
The above taxonomy is one attempt to account for the many vari-
ables that may need to be considered when describing the nature
of academic/professional writing, conducting research, interpret-
ing from theory to practice, and planning a writing curriculum.
Too often, research is conducted without adequate attention to
many other variables which may impact the outcome of the study.
Researchers also are sometimes willing to overstate the results of
research, without giving full consideration to a num ber of factors
which might mitigate their claims. A taxonomy can serve as a use-
ful rem inder in these cases. The taxonomy also allows researchers
to plan alternative studies which incorporate additional issues.
Further, The taxonomy allows researchers to reconsider research
questions and to adjust research hypotheses. Finally, and perhaps
most importantly, the taxonomy, by its nature, forces some type of
order on the many variables that impact our understanding of
writing.
On a more practical level, the taxonomy provides a framework
for curriculum considerations. Certainly any serious effort to plan
a writing curriculum will lead to some type of taxonomy of writing
skills, even if outlined only informally. The purpose for the above
taxonomy is not to examine every possible feature for a writing
curriculum but rather to decide what to emphasize and how to
order goals in the light of the many other concerns and con-
straints operating on an educational curriculum.
8.4 Towards a model o f writing
In both of the two previous sections (ethnography and taxon-

omy) , the goal was to use generative mechanisms to establish the
range of issues im portant in a theory of writing. In each case, the
mechanism provided a means for ordering the range of informa-
tion within the domain of writing. What neither approach does is
to suggest how the various factors might work together in an
integrated fashion to generate whatever writing would be appro-
priate. In this section, we shall sketch a model of writing abilities
which primarily views writing as a communicative activity, and
attempts to account for the skills, knowledge bases, and processes
as they are used in the course of writing. The model is not based
on any specific set of supporting data; in fact, from our perspec-
tive, it is unlikely that any relatively comprehensive model will be
established on any specific subset of data. Rather, the goal is to

account for the results of research in various domains of writing
and to integrate this information within a framework that could
also, in its turn, generate further hypotheses about the nature of
writing.
8.4.1 A communicative approach to writing

In Chapters 1 and 2, it was argued that the primary purpose for
writing in almost all contexts is to communicate. Most writing is
usually undertaken to communicate with one or more readers for
a variety of informational purposes. Even when writing for oneself,
there is a likelihood that the writing will, at a later point, be used
to communicate with others. When there is no other anticipated
reader, and the writing is truly personal and private, one could
argue that the writer serves as a reader, and thus the writing
remains as a communicative act.
Starting from the position that writing is a communicative act,
theories of communicative language use, or communicative com-
petence, provide an im portant resource for developing a model of
writing. Recently, a num ber of alternative communicative concep-
tualizations have been proposed for developing models of written
language. Flower (1994) has developed a socio-cognitive model
which incorporates contextual influences into the cognitive pro-
cessing of the writer, but the limitation of the Flower model is seen
in the minimal integration of textual factors. Martin (1992), from
a Hallidayan functional perspective, proposes a model of dis-
course which is adaptable to writing in particular. M artin’s model
accounts for textual and social issues, but ignores cognitive-pro-
cessing factors. Witte (1992) has recently proposed a social
semiotic perspective, which discusses the needs for integrating
social, cognitive, and textual components.
From an applied linguistics perspective, the models developed
by Bachman (1990), Canale and Swain (1980), and Hymes (1972)
provide a foundation for the model of writing described here.
Drawing on Hymes’s initial discussion, Canale (1983) proposed
that communicative competence could be discussed in terms of
grammatical (linguistic), sociolinguistic, discourse, and strategic
competence. These four components of communication would
account for a person’s linguistic skills in the following ways: (1)
phonological/orthographical, morphological, syntactic, and
semantic knowledge; (2) sociolinguistic awareness and rules of

appropriate language use; (3) knowledge of the ways that dis-
course is sequenced and abilities to structure discourse effectively;
and (4) knowledge of skills and strategies that either enhance
communication or repair miscommunication. This perspective on
communicative language use has been central to applied linguistic
discussions for the past decade (see, e.g., Bachman 1990, Celce-
Murcia et al 1994, Harley et al. 1990). Recently, Chapelle et al.
(1993, Grabe and Chapelle 1995) have proposed a model of com-
municative language use for academic purposes which converges
with the synthesis presented in this chapter. We summarize a
revised version of this model briefly and then discuss its applica-
tion to writing situations.
8.4.2 A model of communicative competence applied

to uniting
Chapelle et al (1993) developed a model of communicative lan-
guage use which is intended to account for academic language
performance in all four skill areas (listening, speaking, reading,
and writing). That model is adapted here specifically to writing
performance and is modified somewhat to take into account the
discussion in the earlier sections of this chapter. Overall, the model
specifies communicative language use as comprising a context for
language use and a representation of the language user’s verbal
working memory (see Figure 8.2). Included in the ‘context’ are com-
ponents that integrate situation and language performance output.
The ‘situation’ itself comprises participants, setting; task, text, and
topic. The ‘perform ance’ accounts for the actual textual output pro-
duced as a result of the processing in ‘verbal working memory’.
The ‘textual output’ provides an additional influence on compo-
nents in verbal working memory as it becomes available for
inspection in the ‘context’. All of the context variables together
comprise the external social context of the writing situation.
The ‘text’ component in the ‘situation’ accounts for expected
register constraints, genre constraints, communication purposes
(speech acts), norms and conventions of language use, and con-
straints of the communication channel, thus incorporating textual
resources as part of the situation. The other subcomponents of
the situation (participants, setting, task, topic), along with ‘text’,
are specified in detail in the earlier taxonomy (headings I, II, III,
IV, VIII, IX). More detailed interpretations of these components

may be drawn from the taxonomy of writing and the ethnography
of written communication presented earlier.
The second major component to the model is the processing
activities of the language user in verbal working memory (as opposed
to visual memory, for example) while carrying out a language
task.5 Generally speaking, all verbal operations are assumed to
take place in a person’s working memory (see, e.g., Barsalou 1992,
Just and Carpenter 1992). The ‘verbal working memory’ compo-
nent has three subparts: internal goal setting, verbal processing, and
internal processing output
‘Internal goal setting’ allows the language user to set goals and
purposes for writing based on the contextual situation (partici-
pants, setting, task, text, topic), internal motivations, performance
attributions (beliefs about how well similar past efforts were evalu-
ated), interest, etc. Internal goal setting also provides an initial
task representation consistent with the goals created. This task
representation (e.g. a one-page summary, a filled-out form) will
activate a cycle of operations in ‘verbal processing’. The arrow line
between internal goal setting and verbal processing indicates that
‘context’ influences on verbal working memory will always be
mediated by the internal goal setting. Metacognitive awareness
and monitoring is an im portant aspect of this subcomponent. It is
likely that metacognitive awareness and control abilities are possi-
ble throughout all of working memory space except the on-line
processing. It should also be noted that ‘internal goal setting’ is
the primary locus for the various affective factors which colour
and alter an individual’s perception of the external context, and
especially the writing task to be carried out. Thus, ‘internal goal
setting’ generates lenses through which the writer attempts to
match the external ‘context’ with internal resources.
‘Verbal processing’ itself is composed of three parts: language
competence, knowledge of the world, and on-line processing assembly.
’Language competence’ and ‘knowledge of the world’ are parts of
both long-term memory and verbal working memory - that part
of long-term memory activated for processing.6 This interpreta-
tion follows current views of working memory and its relation to
long-term memory (Barsalou 1992, Just and Carpenter 1992). For
purposes of the model, however, Figure 8.2 only represents these
components as activated in working memory.
In the figure, each of the subcomponents of verbal processing
are depicted as partly within, and partly outside, the ‘verbal pro-
cessing’ circle. This arrangem ent is intended to indicate that each
subcomponent ‘activates’ a set of information and resources in
relation to the ‘context’ and the ‘internal goal setting’; however,
only those aspects of each subcomponent which are used for a
given processing cycle are within the circle. O ther resources in
each subcomponent may still be activated, but they are not used
for the immediate processing task.
The ‘language competence’ component is made up of three
competencies discussed in earlier models of communicative com-
petence: linguistic (grammatical), discourse, and sociolinguistic.7
These are given detailed specification in the taxonomy presented

earlier (VI-IX; see also Bachman 1990, Celce-Murcia et al. 1994).
These three parts in the language competence activate relevant
linguistic resources based on internal goal setting and the cues
from the contextual situation. In a num ber of respects, therefore,
information specified in the ‘context’ will also activate parallel
language resources in ‘verbal working memory’ - particularly in
the discourse and sociolinguistic subcomponents of competence.
For example, the ‘text’ (e.g. narrative genre) and ‘participant’
(e.g. equal status, distant reader) components in the context will
activate, in turn, similar information from the discourse and soci-
olinguistic units of the language competence.
The second major component of verbal processing is ‘knowl-
edge of the world’. W hether this com ponent is specified in terms
of schemas, frames (e.g. Barsalou 1993), dual-coding theory
(Paivio 1986), or some other representation of knowledge
resources, the im portant issue is that the knowledge of the world
is also initially activated by the topic in the context and by the
internal goal setting, and it interacts strongly with the language
competence com ponent (perhaps sharing the lexicon). The ‘lan-
guage competence’ and ‘knowledge of the world’ components
generate further information interactively, engaging in iterative
problem solving (knowledge-transforming potential). The two
components, together with the internal goal setting, activate, at
the same time, the ‘on-line processing assembly’.
The on-line processing assembly is the component that inte-
grates the informational resources generated by ‘language
competence’ and ‘knowledge of the world’. On-line processing
refers to the various processing skills that are relatively automa-
tized and procedural in nature. These skills might include lexical
access, initial parsing structures, non-problematic propositional
integration into a text model, predictable bridging inferences,
writing production factors, etc. The results of on-line assembly will
be available internally as the ‘internal processing output’ and pro-
vide a mentally ‘observable’ unit of textual representation.
One of the major functions of the verbal processing unit as a
whole is to carry out ‘metacognitive processing’ which will sup-
port the on-line assembly processing and, when that doesn’t work
efficiently, to alert the ‘internal goal setting’ and ‘internal output’
components. All three components, then, can engage interactively
in repair strategies. The term ‘metacognitive processing’ clearly
represents a simplification for all the processing activity that

requires directed attentional resources, but it serves here to distin-
guish itself from on-line processing and account for the
problem-solving nature of many writing tasks.
It should be noted that metacognitive processing in the verbal
processing unit is the locus (along with goal setting) for activating
the type of processing which Bereiter and Scardamalia call ‘knowl-
edge transforming’. For example, as verbal processing employs
the resources from the language competence and world knowl-
edge components, goal setting may need to be adjusted, and new
plans created, in order to coordinate the array of language and
world knowledge information. In some cases, goals might be sim-
plified so as not to overwhelm an inadequate initial goal setting
with too much information.
The final com ponent of the internal operations is internal pro-
cessing output As processing is carried out, the output is compared
to the internal goal setting component in order to match goal set-
ting and processing output. This matching, then, will lead to
further processing, to an acceptable match and an end to the pro-
cessing cycle, or to an unacceptably poor match and a
discontinuation of the processing cycle. The ‘internal processing
output’ unit and the ‘internal goal setting’ unit are strongly influ-
enced by metacognitive strategies.
The details of how such a model might be specified would take us
well beyond the scope of this chapter, though various models from
cognitive psychology research suggest directions for specification.
The purpose of this sketch is to argue that such a model, or some
similar model, provides a way to integrate the three major concerns
for a theory of writing: a writer’s cognitive processing, the linguistic
and textual resources that instantiate the writing task, and the con-
textual factors which strongly shape the nature of the writing. As a
processing sequence it is not very different from either the Flower
and Hayes model or the Bereiter and Scardamalia approach. What
distinguishes it from these other models is the incorporation of tex-
tual influences, the explicit specification of the context, and the
built-in comparison mechanism between the goal-setting compo-
nent and the three sources of processing/processing outcomes
(verbal processing, internal processing output, textual output). In
the next section, the model is summarized and a writing situation is
outlined to illustrate how such a model might account for a given
writing activity and its context.
8.4.3 Applying the model

The model outlined above was originally intended to encompass
all language performance (listening, speaking, reading, and writ-
ing) in academic settings. In the case of writing, the processing
sequence would begin with an initiating activity/motivation com-
ing either from the internal motivations of the individual writer in
‘internal goal setting’ or from the ‘context,’ though these two
sources of initiation are not so easily separable. Instances of writer-
initiated motivation might be a diary entry, a letter, or a summary
of an im portant academic text. Instances of situation-initiated
activity might be an essay assignment, a memo, or a revision of an
essay.
W hether initiated from the writer or from the ‘situation’, the
first processing step would be to activate goal setting for writing.
This goal setting would involve an assessment of the context, an
initial representation of the writing outcome, an assessment of the
potential difficulties in carrying out the task, an initial activation
of features of the genre and conventional forms to be part of the
task, and an organizational plan. These initial processes in the
goal setting component will, in turn, activate the three compo-
nents in the verbal processing unit. The language competence
component will generate the language information needed to
respond to the task-setting requirements, including considera-
tions for audience, topic, setting, register, organization, etc. The
world knowledge component will activate relevant information
which, in turn, will generate more language resources.
The verbal processing unit is, at this time, using metacognitive
processing as needed, and assembling the set of world and lan-
guage knowledge for on-line processing assembly. This on-line
assembly, then, sends textual representations to internal process-
ing output. The internal processing output, in its turn, is
continually matching verbal output to goal settings. When prob-
lems are recognized, metacognitive processing may need to
reassess goals, task interpretations, organizational plans, types of
world knowledge used, types of language knowledge used,
a n d /o r the interactions among these resources and plans.
Reassessing the goal setting, at a certain point in the process,
would also be dependent on the developing text, now available in
the context (as textual output in perform ance); that is, from the
m om ent that text is produced, the many textual variables in the
‘textual output’ will begin to interact with contextual factors and

cognitive processing factors. O f course, the text already written
also requires that reading processes interact with writing
processes. One benefit of the model discussed here is that it is
also possible to account for reading contexts, goals, processes,
and outcomes, and such processing must also be an essential part
of any model of writing.
When the writer is satisfied that the processing output is suffi-
ciently close to the internal goal setting, and also to the task
requirements established by the context, the internal verbal work-
ing memory operations will discontinue generating information
to satisfy the task requirements. The text at that point is both a
representation (of some type) in the verbal working memory and
also fully part of the performance in the context.
An extended example with a specific writing task will illustrate
how the model would account for many of the considerations
raised by the earlier ethnography of writing and the taxonomy of
writing skills. In this example, the writing task, set by the teacher,
is to produce a one-page summary of a twelve-page reading. The
participants are the students as writers and the teacher as a real,
evaluating audience. The status of the audience will be superior
and the degree of solidarity expressed through the writing will be
minimal. The setting for the writing may be at home, in the school
library, or in the classroom. The task is to produce a one-page
summary - genre and organizational conventions should restrict
the writing to a few central ideas and their main supports. The text
constraints will include the following: the writing will not be a cri-
tique of the reading text; details, examples, lists, and processes will
be collapsed to single superordinate descriptors; the language
used will be formal and academic. The text may also be typed.
The topic is rain-forest ecology.
The student will engage the internal goal setting and set a physi-
cal goal of one page of writing. The summary genre and format
will be assumed as familiar to the writer, so success is expected.
Moreover, the reading is interesting, so the writer is motivated by
the topic. The writer also wants to receive a good grade, so he or
she will spend more time on this task. The functional goal is to
summarize the text - find the main points and report them. The
organization of the summary will probably follow the organiza-
tion of the text. There should be little need to generate
additional world knowledge resources because the writer is
familiar with the information in the reading text. The topic (re-
inforced by the reading of the text) as well as the internal goal
setting will generate world knowledge and language resources
that are appropriate to describing main points. The student may
begin production by brainstorming or noting main ideas in order
to generate a stronger organization plan. This pre-writing stage
will then lead to a more developed planning/drafting sequence.
For both the pre-writing and the drafting, the generated informa-
tional and language resources will be integrated by the on-line
processing assembly. As processing output begins, the writer
begins to produce text on paper (or computer screen).
The summary writing that appears, as textual output in perfor-
mance, can be evaluated against internal goal setting, against
expectations for text construction (e.g. text coherence, language
proficiency), and against summaries typically produced by other
students for such tasks. The performance, as text, is open to eval-
uation by the writer (and by others in collaborative settings), and
the many considerations of coherence and text construction are
open for assessment. The ongoing production may also create a
change in the internal goal setting processes and may lead to
new requirem ents for language resources and world knowledge
from the verbal processing unit. The repeated sequences of text
production, as well as the verbal processing and reassessment of
goal setting in verbal working memory, will eventually come to an
end when the text product achieves a reasonable fit with the goal
setting and matches the constraints and expectations of the con-
text.
From this example, one set of dynamics for the model is high-
lighted. As the various parameters of the context, and the
resources in verbal working memory, change, so also will the
dynamics of the model and the text output. Since there are virtu-
ally limitless options for different writing scenarios, a model of
writing will need to be sufficiently flexible to respond to these dif-
fering scenarios. We believe that this model, though tentative and
incomplete, has the potential to account for almost any writing
activity that typically occurs.
8.4.4 Uses and limitations of a theory of writing

This model of writing, adapted from the more general model pro-
posed by Chapelle et al. (1993), will need to be elaborated further,
as would any model, if it is to account for a num ber of ongoing

issues in writing research and instruction. For example, it will
need to incorporate something like the Hayes et al. (1987) discus-
sion of revision processes (Chapter 5). It will need to explore in
some detail how the various subcomponents of language compe-
tence interact among themselves and with world knowledge and
processing mechanisms. It will need to specify the full range of
factors that should be attributed to the internal goal-setting com-
ponent; this specification will help research to move beyond
listing types of goals and plans, and to consider a num ber of addi-
tional factors which influence goals and plans (e.g. motivation,
interest, prior experiences, emotional attitude, etc.). The model
will need to account for the notion of writing as a social construc-
tion (rather than a social interaction) and describe how notions
such as ‘shaping at the point of utterance’ are to be handled.
Further, it will also need to explain how different levels of a
writer’s language proficiency (and in different subcomponents of
language competence) will interact with other components and
explain different types of writing output. The model will also need
to be suitable for hypothesis generating, and it will need to gener-
ate relevant implications for writing instruction in many different
contexts (and cultures) as well as suggest possibilities for writing
curricula.
Aside from the many issues for which the model would eventu-
ally need to account, it also must respond to less tractable
questions in interesting ways. Foremost among these is the general
issue of how the model of writing changes when the various com-
ponents of the model have different values and salient features.
For example, how does collaborative writing or writing under time
pressure alter the dynamics of the model? How does a low level of
proficiency in the language or low motivation influence the model
dynamics? How do different levels of topical knowledge affect the
dynamics? These questions, and many more, become interesting
issues as research on writing attempts to situate itself in a model of
writing of the type presented here.
The limitations of this model are similar to the limitations
noted for many other models: the exact specifications are some-
what vague; the full set of interactions and outcomes are not clear;
and the direct implications for instruction are not obvious. At the
same time, this model does attempt to address three issues which
are not commonly considered in other models of writing.
First, the model incorporates issues related to the social con-

text, to cognitive processing, and to the textual product. This
three-way integration is distinct from most other models of writ-
ing. Second, because the model builds upon a notion of
communicative competence, it is able to incorporate a stronger
language com ponent than is normally considered in other mod-
els of writing. This greater emphasis on the language com ponent
allows more options for capturing the patterns of variation exhib-
ited by many different groups of writers. Third, the model is able
to incorporate the information described in the earlier ethnogra-
phy of writing and taxonomy of writing skills - particularly the
specification of the ‘situation’; this flexibility suggests that the
model is well adapted to the much wider range of concerns dis-
cussed by applied linguists as compared to composition
researchers. For example, there are probably more L2 students in
the world learning to write in English than there are LI students;
yet, aside from applied linguistics discussions, relatively little con-
sideration given to the many different groups of students who
would fall within L2 writing contexts, or how a theory of writing
would account for these students.
The goal of producing a descriptive model is to understand
better the nature of writing, account for im portant research and
writing practice, and rethink what is involved in writing under
many different circumstances. However, it is unlikely that any one
model will tell us most of what we need to know. One goal of
proposing a model is to generate critiques which, in turn, allow
our knowledge of writing to progress: the model described in this
chapter is proposed in this light rather than as a definitive state-
m ent on writing. The model is both a culminating synthesis of
research to this point and also a starting point to examine writing
anew. It allows for a consideration of useful insights from research
and also of shortcomings that need to be addressed.
8.5 Moving from theory to practice
In this chapter, we have presented three variations on a theme. The

theme is the question asked at the beginning: ‘What is writing?’ As a
first approximation, the initial ethnography of writing provided the
means to situate the parameters that any theory of writing should
explain and refer to when considering implications for instruction.
The taxonomy of writing skills, knowledge bases, and processes pro-

vided a means for organizing and ordering the many factors which
both influence and contribute to writing variation. This reordering
as a taxonomy permits an easier reassessment of the many issues
involved in writing, the ways that these issues should be specified,
and the alternatives that need to be included or substituted. The
model, in its turn, suggests ways in which the accumulated informa-
tion on writing could be integrated so that writing can be
understood as a combination of the cognitive, textual, and social
factors that are commonly discussed in various research
approaches. If we have not answered the basic question as directly
as one might wish, we leave it for others to do so and improve on
our attempts at synthesizing the relevant writing research.
A second major question which needs to be posed centres on
the pedagogical implications that follow from a theory of writing:
‘How does one develop writing instruction from a theory of
writing?’ The most realistic answer is that the movement from the-
ory to practice is not necessarily direct, nor is it necessarily
straightforward. One does not simply ‘apply’ a theory and thereby
produce a means of instruction. At the same time, the discussion
of theory in this chapter makes a num ber of strong suggestions
for designing writing curricula. These implications will be
explored in the next chapter as the discussion moves from theory
to application.
Notes
1. O th er possible genres include talltales, research articles, editorials,
society columns, local news reports, jokes, festschrifts, letters o f refer-
ence, business letters, fables, research reports, memos, personal
letters, condolence letters, biographies, proverbs, travel narratives,
business reports, legal briefs, legends, service m anuals, short stories,
poem s, textbooks, research m onographs, recipes, diaries, advice
books, shopping lists, legal summ onses, college handbooks, almanacs,
rep rin t requests, conference handouts, myths, abstracts, research ‘let-
ters’, classified ads, pro d u ct ads, serm ons, religious tracts, short-answer
responses, graffiti, jo b application forms, etc.
2. W iddowson notes that writers sometim es assume sophisticated knowl-
edge of an im m ediate audience for the sake o f n o t insulting those in
the audience who do possess such sophistication (in Swales 1990:
63-4).
3. Well-known examples include the ‘com m unication triangle’, B urke’s

‘p e n ta d ’, Jakobson’s ‘functions o f com m unication’, M offett’s ‘universe
of discourse’, Kinneavy’s ‘aims of discourse’ (cf. L indem ann 1987).
4. This problem of lacunae is m ost easily dem onstrated at the syntactic
level; e.g. languages such as French, G erm an, and Spanish contain, in
their pronom inal systems, the possibility o f distinguishing between for-
mal and inform al second-person m odes o f address (tu/vous in French,
sie/du in G erm an), bu t English lacks this distinction. This particular
feature does n o t generally pose serious problem s; o th er aspects of
m odes of address may create m ore serious difficulties. For example,
the use of first nam es, last nam es, titles, etc., have significantly differ-
en t im plications in different cultures, as do eye-contact and o th er
paralinguistic features.
5. ‘Verbal’ processing is the com m on cognitive psychology term for lin-
guistic processing. It is in ten d ed to apply to all linguistic processing in
any of the four traditional skills.
6. It is n o t possible at this time to argue w hether or n o t the ‘on-line pro-
cessing assembly’ is also directly p art o f long-term memory. It would
seem reasonable to think th at processing routines of many types are
p art of long-term m em ory and only the subset n eed ed for a given pro-
cessing task would becom e activated in verbal working memory. T he
question o f distributed processing versus stable representations is par-
ticularly problem atic for this co m p o n en t and takes us well beyond the
scope of this discussion, an d beyond the purposes of the m odel p ro -
posed.
7. T he earlier notion o f strategic com petence is n o t parallel to the o th er
language com petencies and it is n o t directly represented in this
model. Instead, it is accounted for prim arily by the metacognitive pro-
cessing throu g h o u t ‘verbal working m em ory’.
9
From theory to practice
9.1 Introduction
The previous chapters surveyed research on writing which encom-

passed issues of learning to write in both a first or second
language. While no one book can hope to provide a comprehen-
sive overview of research activity in all areas involving the
development of writing abilities, the previous chapters do perm it
us to emphasize those aspects of research which should influence
planning for instruction. The purpose of the present chapter is to
explore those design criteria which guide and constrain instruc-
tional practices. Chapters 10, 11 and 12 will then address practical
techniques for writing instruction in various contexts.
The transition from theory to practice will be based upon major
research findings on writing as well as a thorough analysis of the
social contexts of writing instruction. The research results influ-
ence the ways in which a writing curriculum should be designed
and suggest effective instructional practices. The social context
analysis ensures an appropriate and systematic transition from
theory to practice, which must take into account at least the fol-
lowing issues:
■ Insights from writing theory and writing research
■ The student writer
■ The writing teacher
■ The educational institution
■ The role of instructional materials
■ Instructional approaches.
237
9.2 Insights from writing theory and writing research
This overview highlights those research findings which should be

incorporated into planning for a writing curriculum. The section
will, for the most part, highlight L2 research findings that allow
for direct comparisons with LI writing research. It appears that a
num ber of L2 research findings differ from those of LI writing
research, and require a consideration of the L2 student writer in a
somewhat different light. A recurring theme running throughout
this review is that L2 writers, however they might be defined, do
not appear to be identical with LI ‘basic writers’, though there are
unquestionably a num ber of similarities.
9.2.1 The text product

The first set of insights from research argues that the written prod-
uct, and formal aspects of writing, cannot be disregarded in
instruction (Leki 1992, Raimes 1991). For L2 writers, greater lan-
guage proficiency (however defined) will lead, at some point, to
better writing quality. Recent research from a socio-cognitive per-
spective points out the need for models of writing and the raising
of student awareness with respect to the ways in which words,
structures, and genre forms all contribute to purposeful commu-
nication (Berkenkotter and Huckin 1995, Cazden and Gray 1992).
In this regard, even the five-paragraph essay and paragraph writ-
ing may be useful for students in certain contexts, though not
necessarily in the ways in which these notions have been discussed
as part of ‘traditional instruction’ (Smagorinsky 1992). In addi-
tion, Fathman and Whalley (1990) have argued that focusing on
written form during feedback is no less effective than focusing on
content for L2 students (cf., however, Robb et al. 1986, Zamel
1985), though focusing solely on form or solely on content
appears not to contribute to great improvement in revisions.
Similarly, the use of sentence combining has been shown to
improve students’ abilities to write more complex and varied sen-
tences. However, sentence combining, in order to have a salutary
effect, must not be dealt with mechanistically.
Considerations of how discourse is structured have also led to
numerous findings which emphasize the importance of discourse
features of writing. Vande Kopple (1986, 1990) and Witte (1983b)
From theory to practice 239
have shown that the general principle of ‘given’ before ‘new’ is

adhered to more consistently in better writing and that main ideas
are typically marked by topical chaining. Work on cohesive har-
mony has also shown that higher quality writing has a higher
index of cohesive harmony. Recent work on ‘involvement’ by
Tannen (1989) points out features of text which tend to draw in
the reader and build coherence. On a somewhat larger scale,
research on genre structure has indicated that its role in convey-
ing purpose and assisting the reader’s interpretation is critical in
effective writing.
Work on contrastive rhetoric has demonstrated that students
from different LI backgrounds, and with a history of culture-
specific educational training and socialization, will prefer certain
forms of textual organization over other forms in particular writ-
ing contexts. As a num ber of L2 researchers have argued (Leki
1992, Raimes 1991), making students aware of these different
preferences can have an im portant impact on L2 students’ com-
posing processes.
Silva (1992, 1993), comparing research on LI and L2 student
writing, has argued that many text-based studies point to distinct
issues which should be addressed in L2 writing instruction.
Among the differences he noted for L2 students (as compared to
LI students) are:
■ different organizational preferences;

■ different approaches to argum ent structuring;
■ different approaches to incorporating material from text into
writing (e.g. paraphrasing, quoting, style);
■ different perspectives on reader orientation, on attention-get-
ting devices, and on estimates of reader knowledge;
■ different uses of cohesion markers, in particular markers which
areless facilitative and create weaker lexical ties;
■ differences in the ways overt linguistic features of the text are
used (such as less subordination, more conjunction, less passiv-
ization, fewer free modifiers, less noun modification, less
specific words, less lexical variety, predictable variation in the
purposes of syntactic structures, and a simpler style).
Thus, it cannot be assumed that LI research findings apply auto-

matically to L2 contexts (Leki 1992, Silva 1992, 1993).
9.2.2 The uniting process

The notion that writing is not strictly linear in its planning or pro-
duction is central for instruction, as is the idea that writing is a
goal-driven activity. Research on writing processes indicates that
good writers:
■ plan longer
■ have more elaborate plans
■ review and reassess plans on a regular basis
■ consider more kinds of solutions to rhetorical problems in writ-
ing
■ consider the reader’s point of view in planning and writing
■ incorporate multiple perspectives into the drafting
■ revise in line with global goals rather than merely editing local
segments
■ have a wide range of writing and revising strategies to call upon.
Research comparing composing processes of ESL and Native

English Speaking (NES) students has often argued that compos-
ing processes are similar across LI and L2 contexts for writing
(Jones 1985, Krapels 1990, Richards 1990, Zamel 1982, 1983).
More recently, however, research has suggested a num ber of dif-
ferences between ESL and NES students (for overview, see Leki
1992). Raimes (1985, 1987, 1991) found that L2 students often
spent much more time rehearsing what they wanted to write, were
not as bound to local contexts or a concern for making errors in
their writing, and were not inhibited by efforts of teachers to cor-
rect or edit their work. She suggests that L2 students should not
be viewed as LI writers, nor should they be seen as ‘basic writers’.
Campbell (1990) found that in using a reading text as a basis for
writing, L2 students were more bound to the text in generating
their writing. Silva (1992) noted that ESL students could not make
use of intuitive editing the way LI writers can by re-reading texts
to see if they ‘sound’ right.
L2 writers’ performance does not always reflect language profi-
ciency; L2 students often reveal more sophisticated composing
abilities which have previously been developed in the LI. LI
students, in contrast, have high linguistic proficiency but may not
have adequate composing skills. Thus, both language proficiency
and composing abilities must be considered in evaluating L2 writ-
ing performance.
Other writing process research has accepted the distinction

between proficiency and composing skill, and has established that
a student’s LI often appears to have a positive influence on L2
composing, as opposed to the generally accepted view that the LI
is only a source of interference (see Krapels 1990). The LI repre-
sents a resource, reflecting what the .learner already knows about
writing (Edelsky 1982); it assists in the development of content,
organization, and details (Friedlander 1990, Lay 1982); it allows
the writer to access appropriate lexical items and phrases and to
consolidate ideational relations across languages (Cumming
1990a); it permits more sophisticated thinking on the writing
topic (Cumming 1989, Leki 1992), and it does not constrain writ-
ing time or writing quality (Friedlander 1990). Reviewing this
evidence suggests that many composing processes tend to transfer
positively, though they do not necessarily represent the same pro-
cessing used in LI writing (if for no other reason than L2 writers
combine resources from two languages).
Research on the writing process provides many im portant
insights into the ways skilled writers perform, and the ways in
which learning to write can be improved.
9.2.3 The social context

The third general insight is that the social context - that of the
classroom, the other students, the interactions among the teacher
and the students, and the larger world outside the classroom - has
a profound impact on the development of writing ability (Heath
1993, Johns 1990, Raimes 1991).
It also appears from recent research that differences between
skilled and less-skilled writers are strongly affected by the socio-
cognitive aspects of writing. For example, Collins and Williamson
(1984) have shown that better writers vary their writing more in
line with specific audience and genre constraints. They also have
shown that weaker writers are less able to provide explicit structu-
ral information for reader interpretation; that is, less-skilled
writers are less considerate of the audience - perhaps because
they are less aware of the audience (Hillocks 1986). Some
researchers have argued that students should not be overbur-
dened with information about audience until they are able to
handle the added planning complexity resulting from the avail-
ability of such information. Other researchers have pointed out
that even weak writers can handle audience concerns if other

aspects of the task are made manageable (Bonk 1990, Collins and
Williamson 1984).
The related emphasis on writing purpose and task variation has
led to a reconsideration of the concept of genre in writing devel-
opment. While writing instruction in many traditional approaches
was guided by model texts which demonstrated specific generic
patterns of organization, the current reassessment of genre relates
to its formal role in writing purpose and task. It is now recognized
that genre form has evolved out of a consistent and conventional-
ized means of addressing specific purposes and tasks in writing.
Thus, genres are conventionalized ways to achieve meaning - to
solve rhetorical problems - and students vary their writing accord-
ing to genre structures which match their purposes for writing.
Research along this line argues that the various genres and tasks
which are useful for academic contexts need to be practised
extensively, not as arbitrary models, but as means for achieving
academic goals and purposes (Atkinson 1991, Christie 1992,
Crowhurst 1990, Swales 1990).
The influence of genre on writing has been examined from a
num ber of perspectives. Empirical research has indicated that
genre is a critical factor in shaping the written text. Students,
assigned to write on different tasks and for different purposes,
systematically vary their grammatical, lexical, and organizational
choices to conform to expected genre constraints. Moreover,
skilled writers appear to be much more sensitive in this regard
than less-skilled writers (Bereiter and Scardamalia 1987,
Crowhurst 1987, 1990). Australian approaches to writing develop-
ment, influenced by Halliday’s functional theory of language use,
have similarly investigated the role of genre in content-centred
instruction and concluded that attention to genre structure is a
critical component of which students need to be made aware, and
with which they need to practise (Christie 1992, Gray 1990, Martin
1989).
Another socio-cognitive approach to writing development has
evolved out of Vygotsky’s theories of language and literacy devel-
opment. From this point of view, the student learns to write by
working with a more knowledgeable person on the skills and
knowledge needed to perform specific purposeful actions
through a kind of apprenticeship. Seen in this light, writing devel-
opm ent involves an apprenticeship which requires considerable
practice, under expert guidance. Moreover, students gradually

learn by appropriating the teacher’s goals and purposes for writ-
ing, as well as the appropriate language forms, in the process of
the writing activity and through feedback on the writing. Such an
appropriation occurs in what Vygotsky terms the ‘zone of proxi-
mal development’ (ZPD) - that level of knowledge between
normal student performance and what a student is capable of
attaining with expert assistance. Empirical research by Bereiter
and Scardamalia points out that students often are capable of writ-
ing performance well beyond their normal writing output.
The commonly assumed notion among composition research-
ers of generalized transfer of learned writing skills across tasks,
purposes, genres, and topics is not an inference to be drawn from
a Vygotskean approach to writing. Rather, children learn to inter-
nalize and self-regulate those writing purposes and tasks in which
they have had extensive practice and guidance. This approach:
■ accommodates the interactive roles of experts and peers;

■ stresses purposeful writing tasks;
■ stresses the interaction of language skills in the accomplishment
of specific tasks; and
■ highlights the importance of practising those writing tasks and
goals which students need to learn rather than assuming gen-
eral transfer of writing skills across purposes, tasks, topics, and
genres (Newman et al. 1989, Rogoff 1990, Tharp and Gallimore
1988).
9.2.4 Research on instructional techniques

The fourth general insight into writing from research derives
from studies testing the effectiveness of various instructional tech-
niques. These techniques include direct instruction - especially
metacognitive strategy training, types of group interactions among
teacher and students, techniques for guiding the drafting and
revising stages of writing, and techniques for providing effective
feedback on writing. The research which has im portant implica-
tions for instruction includes the specific training experiments of
Bereiter and Scardamalia, and those of Flower and Hayes. In
earlier work, Flower and Hayes have pointed out the importance
of multiple drafting, time for planning, the role of recursion
among planning, drafting and revising, and the essential problem-
solving nature of writing. In more recent work, they have

extended research into revision processes, pointing out the impor-
tance of, and difficulty with, global revision of writing. Very recent
work by Flower examines the interactive influences of the writer’s
cognitive processes and the social contexts of the writing situation.
In Flower’s recent work (Flower 1994, Flower et al. 1990), she
demonstrates:
■ that skilled writers consider more perspectives in relation to the
rhetorical problem at hand and are good at representing the
problem;
■ that task representation has a major impact on writing;
■ that students need to develop a strategic awareness for meeting
writing goals appropriately; and
■ that students have difficulty transferring their writing knowl-
edge from one context to another.
In addition, Bereiter and Scardamalia (1987) have shown:
■ that greater topic familiarity does not always lead to better writ-
ing (cf. Richards 1990);
■ that greater essay length tends to correspond with indicators of
writing quality and maturity (cf. L2 reverse findings, Leki 1992);
■ that general knowledge about how to perform a writing activity
will not ensure that students will be able to carry out the task;
■ that many writing assignments produce poor results not
because student are overloaded by the process but because they
are bored by the process;
■ that teacher modelling and support for revising strategically can
improve students’ revising abilities;
■ that the goal of writing instruction is to have students see writ-
ing as a problem-solving activity involving self-regulation,
evaluation, diagnosis, and reflection (see also Cumming 1990a
for a similar L2 perspective).
A key proposal for developing critical skills is to have the teacher
modelling and thinking aloud while composing (Bryson and
Scardamalia 1991). Such strategies for writing development have
also been suggested for L2 contexts (Raimes 1985, Cohen 1990).
Bereiter and Scardamalia also provide a num ber of suggestions
in order to help students develop a knowledge-transforming abil-
ity in writing:
■ instruction should induce a simplified version of the complex

strategies needed by students and then that these strategies may
be gradually appropriated by students;
■ writing tasks should be developed which will gradually help
students to transfer advanced skills to new domains;
■ students need to have teacher-modelling of problem-solving
composing processes;
■ students need to be aware of the goals of writing and need to be
given challenging tasks; and
■ students should be shown how to provide support and assist-
ance to others.
In related research on expertise and the problem of ill-defined
task representation, the skilled writer is again shown to be able to
work through writing tasks as problem-solving activities. What typi-
cally leads to expertise in writing is not well understood, since
writing is complex, locally dependent, and relatively ill-defined: it
can be made more or less complex depending on the writer, the
perceived reader, the complexity of the topic and the depth of
explanation intended, and the demands of the text (e.g. genre
constraints, length, purpose). Scardamalia and Bereiter (1991)
define writing expertise as a gradually building dialectic process
between domain knowledge of the writer and the specific tasks.
For many beginning writers, a challenging writing task leads to the
discovery of additional knowledge of the domain, which, in turn,
leads to a more effective execution of the writing task. As
Scardamalia and Bereiter note, ‘Experts acquire their vast knowl-
edge resources not by doing what falls comfortably within their
competence but by working on real problems that force them to
extend their knowledge and competence’ (1991: 174).
This line of research suggests that instruction should promote
extensive practice within a set of related contexts, using challeng-
ing tasks which build on the solutions of earlier tasks (M. Carter
1990, Spiro et al 1987). Bereiter and Scardamalia argue that
expository assignments offer the best context for developing prob-
lem-solving tasks in writing and opportunities for students to work
with their own knowledge. Expository tasks force students to clar-
ify meanings, discover implications, establish connections,
formulate problems for intended goals, and evaluate decisions.
An approach which relies predominantly on expressive writing
and on narrative recounting does not typically make as great a set
of demands or provide as many opportunities for knowledge

transforming.
Research on writing tasks in classroom environments has shown
that certain pre-writing activities are particularly useful for students.
Also the more traditional heuristics for writing ‘invention’ appear to
be useful for better writing (e.g. Hughey et al. (1983), Richards
1990, Young et al 1970; cf. Bereiter and Scardamalia’s (1987) con-
cern that pre-writing activities sometimes make writing tasks too
simplistic). Such devices for planning and goal-directed idea
generation are similar to Bereiter and Scardamalia’s instructional
interventions referred to as ‘procedural facilitation,’ but students
make use of the cues in procedural facilitation while they are writing
rather than before they write (Bryson and Scardamalia 1991).
Other insights from classroom contexts involve research on
feedback and revision instruction. While there are a num ber of
questions about the effectiveness of various revision and feedback
procedures, Leki (1992) points out that students need to learn
how to revise more effectively, whether the learners are inter-
national students, or immigrant or L2 minority student in tertiary
institutions.
9.2.5 Research and curriculum design

The research reviewed in this section points out the range of
issues and concerns which must be considered in developing cur-
riculum design. No theoretical perspective nor set of research
findings should be reinterpreted as having a one-to-one relation-
ship with curriculum design. Rather, the goal is to recognize that
research occurs in contexts not typically equivalent to the lan-
guage classroom; it also creates constraints a n d /o r opportunities/
resources which are not typical of the language classroom. The
goal in making the transition from theory to practice is to recog-
nize the relevance of research insights and their potential for
informing curriculum and instruction.
9.3 The student writer
One of the first tasks of curriculum design is to establish who the

instruction will be for, what the learner needs, wants, and is
capable of, what factors will cause individual learner variation
within a given group, and how other curriculum constraints inter-

act with this particular set of issues. The potential diversity of
learners in writing instruction is enormous.
In addressing this diversity, an applied linguistics perspective is
assumed, and thus various L2 contexts as well as LI contexts need
to be considered; no attempt to privilege a particular subset of
students or writing contexts for the discussion of writing has been
taken (except that learning-disabled students are not considered
in any context). In some respects such an approach is obviously
problematic, as this section will highlight. At the same time, any
strict delimiting of the perceived student in this discussion of writ-
ing would remove from consideration im portant research findings
and instructional issues, as well as interesting writing practices.
The purpose of this section, then, will be to highlight the different
types of constraints that accompany different groups of students,
to the extent that differences can be systematically recognized.
Variations in student background due to language, culture and
preferred learning style, educational training, age, and individual
differences all must be considered in relating theory to practice.
9.3.1 LI versus L2 student uniters

Within the domain of LI literacy development lies variation
created by the ages of the learners, the social and cultural group
identities of the learners, the extent of access to educationally val-
ued literacy practices, the extent to which the learners’
dialects/idiolects depart from the educationally valued dialect,
and the range of individual differences in terms of learning styles
and strategies (see Chapter 1).
The range of L2 learners is even more complex,1 as these learn-
ers may come from literally thousands of different LI
backgrounds. They will not only fit into all the ranges of divisions
for LI learners but will also vary considerably in terms of the con-
texts of L2 use and the purposes for needing/learning a second
language.
The picture which emerges from attempting to describe student
variables in planning writing instruction is one of great complex-
ity. In ESL contexts, consideration needs to be given to:
■ the extent of LI literacy;
■ social practices and expectation in LI literacy;
■ the LI educational background;

■ th e e x te n t o f p rev io u s a n d c u r r e n t c o n ta c t w ith E n g lish sp eak ers;
■ the social and cultural distance of the student from the domin-
ant and educationally valued English-speaking group;
■ the intended length of stay in an English-speaking environment
(any such environment, not necessarily the one in which
instruction occurs).
EFL contexts may be even more complex. Consideration must be

given to:
■ country of origin;
■ length of prior English study;
■ extent of access to English;
■ linguistic typological distance of LI from English;
■ social and political attitudes towards English and English speak-
ers generally;
■ training and expertise of the English teacher;
■ extent of LI literacy training;
■ social practices and expectations in LI literacy;
■ major field of study or educational track in school;
■ p o te n tia l fo r e c o n o m ic o p p o r tu n itie s ; a n d
■ cultural expectations for learning.
While the total picture is rather overwhelming, the fact is that
most teachers, institutions, and materials developers do not have
to deal with the full array of student variables. The purpose in
sorting through the range of variation is to sensitize the reader
towards the concerns of writing teachers in many different con-
texts. An example should illustrate the importance of this context
sensitivity and also point out im portant differences between LI
and L2 writing students - in this case, the differences between
international ESL students and English-speaking students in a ter-
tiary-level writing course context.
9.3.2 The international student

The international ESL student studying in the tertiary-level writ-
ing course in an English-speaking country works within a very
different set of constraints than do other students working at the
same level. Leki (1992) carefully reviews the many differences
between international students and English-speaking basic writers,
two populations often placed in the same courses in US tertiary-

level institutions and often assumed to be similar in backgrounds
and needs.
According to Leki (1992), international students have linguis-
tic, cultural/attitudinal and academic experiences which
distinguish them from the English-speaking basic writer.
International students are often seen as having composing prob-
lems by the teacher when, in fact, they are dealing with linguistic
problems. International students are typically trained in more for-
mal registers of English (‘book language’) which is an advantage
in academic contexts; however, teacher efforts to simplify lan-
guage by using informal registers often makes comprehension
more difficult for these students. International students:
■ are likely to misunderstand class interactions, teacher talk, and
orally given assignments;
■ have difficulty participating orally in class;
■ often misread non-verbal cues which cause numerous attitudi-
nal conflicts;
■ are not able to rely on native speaker intuitions when using lan-
guage, so reading papers aloud may not be a suitable teaching
technique;
■ expect to return home after finishing their studies, and may not
see English as critical for their future careers.
Attitudinal and cultural differences create major distinctions
between international and native English-speaking students. In
most basic writing courses, international students recognize hos-
tility among some native English-speaking students who resent
being placed in the same class as international students (whose
linguistic limitations may be wrongly interpreted as signs of stu-
pidity). The implication in such a situation is that the native LI
speakers do not know their own language. Since international
students do not have a cultural identity with the forms of English
they use, they can more easily abandon their uses for more aca-
demically ‘correct’ language. Many international students also
stand apart because they come from economically privileged
backgrounds (foreign education is intrinsically expensive) and
have very sophisticated world outlooks (their wealth/position in
their home communities has exposed them to high-quality educa-
tion, etc.).
Academically, these students also present quite different
profiles from other students, and are often surprised by the behav-
iour of their hosts in the classroom and by the level of informality
exhibited by both teacher and students.
International students also experience many disadvantages that
are not always recognized by teachers (particularly in the USA).
■ They do not have a full range of host country cultural experi-
ences (e.g. TV, sports, holidays, political system, economic
system (i.e. a credit economy), etc.).
■ They may come from countries with ambiguous political rela-
tions with the host country, marked by simplistic stereotypes of
host country people and situations.
■ They are perhaps often painfully reminded that host country
people know, or want to know, very little about their cultures.
■ They experience some degree of culture shock - disorientation
- and something as simple as shopping for shoes or buying food
can become overwhelming.
■ They may experience unique academic problems (especially in
the USA), and are often surprised by course requirements in
host country tertiary-level institutions - such as regular atten-
dance, homework, and periodic tests.
■ They are unaccustomed to the idea that any theory is relative,
open to criticism by the teacher and to comparison with con-
flicting views.
■ They are often unfamiliar with the rigidity of assignment dead-
lines, and the legalistic concerns of plagiarism.
■ They often find it difficult to express an honest opinion that
conflicts with the view by the teacher.
■ They may find such techniques as conferencing to be proble-
matic as they may out of politeness feel compelled to agree
with, but not comply with, all teacher suggestions.
■ They may experience frustrations when they recognize their
inability to express complex academic ideas which they know
they can express adequately in their own language.
When one recognizes that many detailed group differences can be
similarly drawn between Native American and Black students and
their white middle-class peers in US secondary schools, between
Spanish-speaking minority students and English-speaking Anglo
students in elementary classes, between EFL students learning
English in tertiary institutions in their own countries and inter-
national students learning English in the USA, etc., it becomes
clear that student considerations must enter into any planning for
a writing curriculum.
9.3.3 The role of the student

The obverse side of understanding the experiences of the
students and their range of variation is to recognize those skills
and attitudes which are im portant for the good learner. Hughey et
al (1983) describe good learners as students who take an active
role in their learning; these students set goals and take respons-
ibility for their learning. To explain what is needed to become an
active learner, Hughey et al (1983: 50-4) and Leki (1992) address
a num ber of areas which help define the good learner.
They note that, first, learners need to have a positive approach.
Learners must have some empathy for the language that they will
use and try to write whenever appropriate to do so. They will need
to:
■ believe that they will be successful in their writing development;
■ have an open attitude to their academic environment and to
the sorts of writing tasks they will be asked to perform;
■ willingly pursue issues and not think they already have all the
knowledge they might need; and
■ be willing to practise and revise their writing, recognizing that
writing development is a gradual process which requires much
hard work.
Second, learners need appropriate skills in order to carry out writ-
ing assignments. They need a reasonable degree of control over
the language and some ability to manipulate the language in
response to varying needs. They need to have an appropriate level
of control over vocabulary and the rhetorical structures of the lan-
guage. Students need adequate reading skills as well in order to
read what they write and read other sources for information, for
contrasting views, and for alternative ways to view writing.
Third, learners need some set of workable learning strategies which
are applicable to the writing context. They need strategies for
planning and setting appropriate goals, for working with words
and language structures, for developing information and argu-
ments, for attending to the rhetorical constraints of the task, for
re-reading texts and revising effectively, and for evaluating their
writing and comparing it to other writing.
Fourth, learners need to be motivated to invest the time and

effort, whether the motivation comes from perceived economic
opportunities, from better grades, or from a desire to develop bet-
ter communication skills. While a certain amount of extrinsic
motivation is typically used to motivate students such as well-
designed assignments, relevant topics, and interesting materials,
intrinsic motivation must be developed as well if the student is to
become an active learner.
The role of the learner represents one aspect of the matrix of
issues which contribute to curriculum design. Another major com-
ponent in the planning process involves the nature of teaching:
Who is the teacher and what role does the effective teacher play in
writing instruction?
9.4 The writing teacher
9.4.1 Teacher training

A writing curriculum must give consideration to the strengths and
training which the teachers bring to instruction. Once again, a
num ber of basic distinctions tend to group teachers in important
ways. Among the basic considerations are whether or not teachers
are themselves native (or near-native) speakers of English. Another
consideration is whether or not they have had explicit training
a n d /o r experience in teaching writing. A third issue is how skilled
and well-trained teachers are as teachers. While any programme
would like to assume that all teachers are excellent, well trained in
writing instruction (and native speakers of English), such is not
often the case, and curricula should be planned in light of realistic
estimates of teacher strengths and weaknesses.
The situation with respect to the non-native English speaker as
writing teacher is particularly complex. In certain countries in
which English is a foreign language without special emphasis in the
educational system, the extent of writing ability and practise
needed to become a writing teacher may actually be minimal, and
often teachers’ own skills in writing are rather weak. In these situa-
tions, teachers quite naturally are not interested in certain
approaches to writing instruction (e.g. the process approach to
writing, an approach which might reveal the English language limi-
tations of the teacher). Non-native English teachers in ESL
situations in which English is a major language'in the educational

system (e.g. the Philippines) may also have distinct instructional
preferences. Such teachers may not like to cede control of the
classroom to students since doing so might be interpreted as a
weakness in English language knowledge or writing skills; similarly,
allowing students to choose topics for writing - or not correcting
errors diligently - may be seen as signs of linguistic weakness. The
essential point is that local contexts of instruction often determine
the effectiveness of instructional approaches. Sometimes an
approach which is appropriate in the context of an English-speak-
ing country may be less effective in other contexts if for no other
reasons than those deriving from misperceptions by students.
Another set of issues pertains to the knowledge of language and
writing that teachers possess. Among these issues are the extent of
creativity and innovation with language and writing, degree of
skill in writing itself, knowledge of the pragmatics of the language
being taught, interest in examining/studying writing, writing hab-
its and practices, and motivation to write. Teachers also need
experiences with specific genres and registers that are important
for the writing curriculum.
It is true that the above issues are not all-or-nothing propositions;
teachers, both native and non-native, will be represented fully
along each continuum: English language skills, teaching skills, writ-
ing skills. In some respects, while in-service training may serve to
improve teaching skills and, in particular, the skills implicit in teach-
ing writing, it is somewhat more difficult to assess or to address the
training needs of the non-native English writing teacher.
A large num ber of particular teacher attributes should also be
considered in planning writing instruction. These attributes may
have an important, though not always obvious, impact on the suc-
cess of a writing curriculum. The following list points out
individual variables which occur among any group of teachers.
Individual teacher characteristics will vary at least by:
■ age
■ general educational and social background
■ cultural expectations
■ time since completion of teacher training
■ professional affiliations
■ years of teaching experience
■ academic subject knowledge/interests
■ philosophical views on teaching and learning

■ willingness to cede formal control in the classroom
■ preferences for class structuring
■ extent of adherence to m aterials/curriculum
■ gender (and gender stereotypes)
■ linguistic competence in LI and L2.
Differing perspectives on many of the issues noted above will

influence how any teacher will approach writing instruction.
9.4.2 The role of the uniting teacher

Aside from examining the variation which typically occurs among
groups of teachers and within these groups, it is also important to
recognize the types of tasks the teacher must carry out in the
course of writing instruction, and the strengths that many teachers
bring to the classroom. In planning a writing curriculum, the
teacher must at various times be a motivator, an interpreter of the
task, a designer of meaningful tasks, an organizer, a resource, a
support person, an evaluator, and a reader for information.
Somewhat more specifically, teachers need to show a positive atti-
tude, believing that students are capable of doing the work in the
curriculum; and when students are found to be struggling with
specific assignments, teachers need flexibility to adapt these assign-
ments to ensure reasonable opportunities for success. Teachers
need to provide students with a wide range of opportunities for
writing, opportunities which are interesting for students and which
serve important developmental goals. Teachers need, also, to
develop among students a sense of community and sharing so that
writing can become a collaborative and cooperative endeavour.
A final requirem ent for good teachers is that they have produc-
tive insights about writing and the writing process. They need to
be aware of all of the following issues in writing and convey these
insights to students:
1. Writing takes time to develop, and instruction should be
planned accordingly.
2. Students need to be made aware of the role of language form
as the medium of meaningful communication, as well as the
types of language constraints which are reflected in different
genres and purposes for writing.
3. Writing development requires extensive practice.
From theory to pra ctice 255
4. Writing can sometimes be difficult and frustrating, and stu-

dents need positive feedback and enough success to maintain a
willingness to work.
5. Students will occasionally vary in their performance and
should be made aware that this variation will occur.
6. Students should be encouraged, at times, to take risks, to inno-
vate, and to rethink assignments in more complex ways.
The role of the teacher is very complex; the teacher must balance
various pedagogical insights which will have to be rethought some-
what differently for each student in the class.
9.5 The educational institution
To this point, the emphasis for instruction has been squarely on

the teacher. But responsibility for what happens in the classroom
must be shared with the institution within which the teacher func-
tions and with the system in which that institution is embedded.
While the purpose of this section is not to suggest that any given
teacher must be prepared for all potential instructional contexts,
it does point out that planning a writing curriculum occurs in
many different places with many diverse institutional goals.
Institutions typically decide the goals for learning:
■ They plan and operationalize the curriculum, usually in accord
with the institutional norms for a country, a region, or a set of
related bodies.
■ They react to the requirements of an inspectorate.
■ They decide how much time in the curriculum to allocate to
writing instruction and support systems.
■ They decide how much money to allocate for writing instruc-
tion and support equipment.
■ They decide the teacher-training courses, and how much train-
ing teachers need in order to teach writing courses.
■ They decide how much to pay teachers, how many students
should be taught in any class, and how many classes should be
taught by any given teacher during a given term.
■ They decide how much writing instruction different students
need and for what purposes.
■ They decide whether or not to support teachers for further
training.
■ They decide the relative importance of writing in relation to

other components of the curriculum.
■ They decide how to integrate writing with other components of
the curriculum (ESP courses, etc.).
The institution’s decisions play a major role in shaping a writing
curriculum. As one example of this impact, many elementary
schools in the USA are restructuring language arts curricula to
integrate writing with other language activities and are simultan-
eously moving towards the consideration of writing in content
areas; thus, content-centred approaches and integrated-skills
approaches are bringing writing together with subject-matter
instruction rather than maintaining writing as a separate skill with
its own discrete and isolated block of time in the curriculum.
9.6 The role o f instructional materials
Writing instruction makes use of a wide variety of information and

resources. Some of these resources are straightforward in terms of
a writing context; others are less obvious but equally effective.
Resources for writing can be classified into texts, libraries/media,
realia, student-generated resources, activities, and discussions.
The latter four types of resources are discussed in detail in
Chapters 10, 11 and 12.
Texts for working with writing instruction would include rheto-
rics/writing texts (for academic contexts) ; language arts texts (for
lower grades) ; anthologies of readings for writing; workbooks and
usage/gram m ar handbooks; dictionaries; encyclopaedias/bio-
graphical sources/atlases; novels and short stories; articles from
magazines and journals; newspapers; non-fiction books; and text-
books from subject-area fields. Some of these resources are
designed explicitly for writing instruction and tend to reflect
specific approaches to instruction; other materials provide
resources for reflection, information, and argumentation.
Writing textbooks, of course, are not objective, scientific works;
on the contrary, they exemplify the biases of the writer, of the pub-
lisher, and of the market. It is a regrettable reality that publishers
are required by their Boards of Directors to earn money - they
are, presumably, profit-centred organizations. Thus, they are
inclined to make decisions in marketing terms, not in instruc-

tional terms. In sum, textbooks that are likely to sell are likely to
get published, and for this reason, teachers should not accept a
textbook at face value but should examine it carefully to deter-
mine whether it coincides with the methodology adopted for
instruction, whether it is appropriate to the intended instructional
audience, and whether it is reasonable in terms of the cultural
parameters in which the instruction is being offered.
Another obvious resource for materials is the library, either in
school or in the community. Library visits allow students to browse
through book topics and examine a range of topical possibilities
which might be completely new to the individual or the group.
The issue of materials and resources is another im portant area.
Often, one hears that classes cannot afford the materials and
resources they would like to enhance the curriculum. Resources
will obviously vary enormously in different instructional situa-
tions, but any class can generate its own range of materials and
activities to stimulate topics and tasks for writing if the most up-
to-date resources are not available. In some respects, the creative
use of student-generated materials provides the best avenue for
initiating exploration of issues and topics, both personal and aca-
demic.
9.7 Instructional approaches
There are three potentially separable domains into which one

can sort a discussion of instructional approaches: the general
‘schooling’ domain, the tertiary-level domain, and EFL/ESP
domain. The first context involves approaches at the general
‘schooling’ level but most particularly at elementary grade level;
discussions typically centre on whole-language versus more tradi-
tional language arts approaches (but also on content-based
approaches and genre-centred approaches). The second general
set of themes converge from discussions of tertiary-level and pre-
tertiary-level settings (e.g. intensive EAP programmes), whether
the students concerned are English LI students or ESL students
in English-speaking academic contexts. The third domain might
best be represented as an EFL/ESP or LI work-based grouping
which typically does not enter into the frays which excite the first
two groups. In fact, most discussions in this third group are
decidedly atheoretical - writing instruction is seen as following

the dictates of the general EFL language curriculum (usually with
little writing emphasis), or the dictates of specialized stu-
dent/institutional needs such as writing for airline mechanics,
medical technicians, missile battery operators, etc. There are, of
course, other potential groups, though they are seldom
addressed in the general writing literature and the professional
journals (e.g. adult education writing, deaf students, ‘learning-
disabled’ students, etc.).
9.7.1 Elementary contexts

In elementary contexts, there are three potentially discernible
positions with respect to writing approaches:
1. Traditional language arts approaches (with some process writ-
ing included)
2. Whole-language approaches
3. Genre-centred approaches.
In fact, there is considerable potential overlap among these
approaches, and sometimes the debate centres on philosophical
issues rather than focusing on radically different curricula and
techniques for teaching/learning writing.
9.7.1.1 Traditional lan guage arts
Traditional language arts curricula are typically textbook-based

and often - though not necessarily - combine language arts
texts with basal readers. Such approaches often use simple assign-
ments from the text which involve student-centred expressive
writing, narratives, and occasional reports/descriptions. This
approach includes options for teaching grammar, spelling, and
vocabulary. Most texts of this type now stress a process writing
approach, though it is not clear to what extent teachers follow
through with such text recommendations (Applebee 1986, Ley et
al 1994).
9 .7 .1 .2 W hole-language a p p roach es
Most whole-language variations represent strong reactions to the

frustration with traditional language arts approaches to writing
(Edelsky et al. 1991). The movement also represents, in general,

an effort to place the teacher and the students at the centre of the
curriculum rather than having instruction mandated by a school
or district plan.
There are many im portant principles and instructional prac-
tices which have been advocated by whole-language practitioners.
These principles and practices are currently having a major
impact on writing instruction in the USA and Canada, and some
influence in New Zealand, Australia, and the UK. Since whole-lan-
guage practitioners make strong claims for their approach, it is
im portant also to examine some of the limitations of this
approach. First, the theoretical framework for whole language is
somewhat protean in nature. It includes references to M.A.K
Halliday and Vygotsky; but, in contrast:
■ Halliday stresses the crucial role of language structure as media-
tor of social interaction and purposeful communication
(Halliday 1993b);
■ Vygotsky stresses the distinctions between oral and written lan-
guage in the sense that the latter requires socio-cultural
mediation and support - moreover, a Vygotskean approach
stresses attention to the skills to be learned through a social
apprenticeship and a gradual appropriation of others’ goals
and language use through direct instruction.
The protean nature of whole language also prevents comparisons
between itself and other approaches since evaluative measures
and empirical orientations arise out of the other approaches and
are seen as inappropriate means for evaluating whole language.
There are practical criticisms that whole-language classes, being
learner-centred, do not provide support for many students who
need a more structured learning environment. Such students may
need teacher modelling and assistance, and they may not, by
nature, be creative and inquisitive on the kinds of tasks/cycles
which focus the whole-language classroom. The notion that the
curriculum should come from the interests and choices of the
students is perceived by some as an abdication of the role of the
teacher as a more knowledgeable other, an evaluator, and a teacher.
All of these teacher roles, in the right contexts, are necessary for
effective student learning (Genesee 1994). Teaching the writing
process, for example, does not mean a teacher should avoid
actively guiding students’ writing processes. The assumptions of
universal learning processes and skills also de-sensitize teachers,

preventing them from seeing that students conceivably come from
very different social, linguistic, and cultural backgrounds, with dif-
ferent expectations about language use in class contexts.
A somewhat less ideological perspective on elementary lan-
guage development and whole-language activities is seen in
Enright and McCloskey (1988). They present a view of language
development among elementary-grade students which incorpo-
rates many whole-language methods, integrates language skills,
emphasizes purposeful language use, and combines language
learning with content area learning. In this more flexible
approach, the content plays a stronger role and creates the oppor-
tunity to relate language activities to content within existing
curricular frameworks.
9 .7 .1 .3 G enre-centred a p p roach es
The genre-centred approach is currently being practised in

Australia (Christie 1992, Christie et al 1991, Cope and Kalantzis
1993, Hasan and Martin 1989, Martin 1989, 1993). This approach,
based on Halliday’s functional systemic views of language and his
social semiotic theory of language development (see Chapter 5,
section 5.3), stresses the social nature of language learning but
also stresses the mediating role that language form plays in pur-
poseful communication (Halliday 1993b). In essence, this
approach brings in the role of language form in learning lan-
guage and combines it with notions of writing process, integration
of language skills, meaningful communication, peer collabora-
tion, and, in many instances, content-based instruction (Collerson
1990, Derewianka 1990). An im portant theme in this approach is
that students develop a sense of the shaping role of genres in writ-
ing, and how language structure and genre form constrain the
ways in which language communicates information.
9.7.2 The influence of instructional approaches

The role of different instructional approaches in the teaching of
writing has been quite influential. Often, teachers are exposed to
a general philosophy of writing based on one approach or
another and are assured that there is strong evidence for the vari-
ous insights claimed for the approach. In many cases, rationales
given for instructional approaches are themselves taken as evid-

ence and cited in later descriptions as research support for the
approach. It is im portant to sort the discussions of supporting
empirical and observational evidence from those sources which
describe an approach and provide logical/intuitive rationales for
the approach as well as appeals to similar earlier articles and
appeals to authority. In this way, it should then be possible to con-
sider the insights from instructional approaches in light of
research findings. From this relationship, it should also be pos-
sible to find generalizable instructional ideas which can underlie
the principles of a writing curriculum.
9.8 A generalized approach to writing curricula
The preceding discussion has attempted to determine the range

of evidence which is available from research and to consider how
that research can be categorized and related to other knowledge
(see also Chapter 8). The influence and roles of the student, of
the teacher, and of instructional material have all been con-
sidered. The influences of various instructional approaches to
writing have also been discussed. It has been necessary to look at
instructional approaches independently because many views pro-
posed in instructional approaches are presented as though they
are supported by evidence, without making the evidence explicit;
other views depend on appeals to teaching experiences or logi-
cal/theoretical arguments from rhetoric, psychology, education,
linguistics, or language learning. It is important, then, to compare
these approaches and to determine to what extent their underly-
ing philosophies overlap, their instructional methods match, and
their theoretical bases receive support from research. The final
aim of this chapter is to distill useful generalized notions of what
should be involved in good writing instruction on the basis of
these several sources.
9.8.1 An idealized uniting curriculum

An idealized writing curriculum would combine emphases from
the writing product, the process, the social context, and the sub-
ject-matter content (Raimes 1991).
1. The course would be content-driven and present topical issues

and writing tasks which motivate and engage students, while
at the same time being challenging and providing opportun-
ities for learning. Students would work within a coherent
content-based conceptual framework and would explore the
complexity of thematic issues, leading to more challenging
writing tasks, the learning of new content, and knowledge spe-
cialization as opposed to writing what the teacher already
knows (Bartholomae and Petrosky 1986).
2. The course would see writing as an apprenticeship training in
which teachers:
(i) act as knowledgeable experts who can model effective
writing practices;
(ii) raise student awareness of relevant strategies;
(iii) engage students in problem-solving planning;
(iv) help students sort through formal linguistic choices;
(v) guide the development of the strategies and skills which
will let students perform similarly without teacher sup-
port;
(vi) develop activities which allow the gradual transfer of
these skills to other (related) contexts.
Such a view of instructional guidance draws heavily on meta-
cognitive strategy research, research on the development of
expertise, and Vygotskean views on literacy learning.
3. The course would engage students in the writing process, rec-
ognizing that writing is a recursive process in which writers
continually plan, write, revise, and refine; thus, the course
would use multiple drafting, student selection of tasks, many
opportunities for feedback, and reaction writing and free writ-
ing for the exploration of ideas and issues related to the
writing tasks. Students would learn how to generate plans and
call up content information, how to organize plans through
invention and ‘topics’, how to write fluently, how to draft with-
out expecting a final text, how to revise - develop the inner
reader, and how to elaborate and refine.
4. The course would incorporate cooperative learning activities
for peer feedback, for exploration of ideas, for interaction
which will enhance writing strategies and skills, and for oppor-
tunities to improve learning as the student recognizes
alternative interpretations of information and of the writing
tasks themselves. Students learn from each other in many

ways, including access to prior successful performance
(improved attribution), positive attitudes (motivation), a
wider range of accessible content, and alternative sets of
underlying assumptions brought to the tasks. Group work
would usually be partly directive so that specific tasks would be
guided with work sheets or oral directions up to the point at
which students can provide effective feedback and be good
audiences (R. Stahl 1994).
The course would require the integration of language skills.
Students would read extensively; would spend much time writ-
ing about their readings and relating the reading to the
writing tasks; would discuss the readings and their writing;
would write more based on their discussions and those discus-
sions would lead to additional readings, which would, in turn,
lead to more writing, etc. Such a perspective combines
Freirean and Vygotskean views with research on the interac-
tive benefits of reading and writing together - research which
suggests that this interaction leads to better learning and pro-
motes greater interest/involvement.
The course would attend to formal constraints of the lan-
guage which serve to signal writing purpose, the role of genre
for reader expectation, and the flow of discourse information.
This aspect of the course would provide a focus for group
analysis of texts, for teacher modelling of writing, for verbal-
ized goal planning and problem solving, and for explanations
of formal language constraints as they arise naturally - giving
students the appropriate metalanguage to discuss these issues
directly and explicitly. Such an emphasis would allow students
to address more complex discourse and organizational issues
with a specific vocabulary - giving students true control over
the language through the metalanguage. Revisions of texts,
whether by teacher or by peers, would include analyses of for-
mal aspects of text, and not only in the final stages of drafting.
Lower-level writers would be given writing tasks such as com-
position modelling, guided composition, sentence combining
and other activities, which is particularly important for L2
contexts, allowing students to attend to formal aspects without
being initially overwhelmed. It should be noted that these
activities should not be done in place of other writing activities
or as a preliminary requirem ent to other writing activities.
7. The course would provide careful consideration of the audi-

ence, as well as the influence of the social context, on the
writing task. Students would be made aware, through the
teacher composing aloud and through misunderstandings by
other readers, how audience considerations must be given
careful attention, whether they be seen in terms of Gricean
maxims, power and politeness relations, or register parame-
ters and expectations. The course would look to incorporate a
range of audience sources, whether in the form of letters to
others; competitions for writing; presentations of projects,
posters, diaramas, or reports sent to other readers; or publica-
tions of writing production.
8. The course would provide ways to introduce early writing
activities and encourage experimentation without harsh evalu-
ation, promoting a sense of language play. Varieties of English
would be respected, though standard forms would also be
developed and used as the means for becoming bidialectal
and gaining access to the school-valued educational genres.
Writing development would move from the personal to the
academic, from those tasks in which students have access to
information to those tasks which require a search for informa-
tion and the complex development of information presen-
tation.
9. Writers would practise a range of writing tasks and learn to
work with a variety of genres and rhetorical issues. Students
would also write extensively, carrying out the idea that writing
proficiency requires constant practice, even if writing quantity,
by itself, is not sufficient for writing improvement. Students
and teachers would negotiate some subset of the total set of
assignments in the class; most commonly, students would
choose specific topics of interest within broader curricular
themes.
10. The course would offer a variety of options for feedback,
some of which would be evaluative; other feedback would be
non-evaluative. Assessment of writing development would be
carried out in a variety of ways, through in-class writing, out-
of-class writing, journals, conferences, free writing and
reflective writing, portfolios, evaluation of final drafts of essays
and projects. Students would take an active role in assessment
by participating in feedback, in developing content resources,
and in interaction and discussion.
9.8.2 The realities of curriculum planning

The ideal curriculum described above is, perhaps, not one which
is fully applicable to any single real context. As the other sections
of this chapter have pointed out, each context for writing instruc-
tion will consist of a unique set of constraints: different students,
teachers, institutions; different resources, time allocations, and
access to information; and different individual abilities, expecta-
tions, social/cultural backgrounds, and linguistic backgrounds. At
the same time, the above idealization does synthesize research and
practices which appear to improve writing. In this respect, these
feature should be incorporated into a writing curriculum to the
extent that it is feasible to do so. At the very least, they represent
guiding notions for the planning, implementation, a n d /o r evalua-
tion of a writing curriculum (see also Hilgers and Marsella 1992).
Given this set of guiding principles, it seems reasonable to turn
to the set of issues for which many practising teachers have a pri-
mary concern:
■ What specific techniques can be used in the teaching of writing?
■ How can they be adapted to different contexts?
■ How might they fit within a rationale such as that noted above?
Chapters 10, 11 and 12 explore specific instructional techniques
for beginning, pre-academic, and academic writing instruction.
While these chapters will, in effect, present ‘pieces’, it must be
noted that writing techniques must be considered within a given
curriculum design and a given theoretical motivation. Only those
techniques which are motivated by the goals and assumptions of a
given curriculum should be adopted. The techniques which fol-
low, then, are not a set of activities to be tried out ‘next week’, but
rather a set of options which should be considered for use only to
the extent that they serve the general goals of the curriculum
adopted and the preferred instructional approach of the institu-
tion.
Note
1. For ease o f discussion, the following treatm en t o f second language

students will assume th at English is the second language. T he follow-
ing categorization could be replicated with o th er second languages in
non-English speaking countries.
10
Teaching writing at beginning

levels
What is learning to read and write? Fundamentally, it is

an extension of the functional potential of language.
Those children who d o n ’t learn to read and write, by
and large, are children to whom it doesn’t make sense,
to whom the functional extension that these media
provide has not been made clear or does not match up
with their own expectations of what language is for....
Fundamentally, as in the history of the human race,
reading and writing are an extension of the functions
of language.... This is what they must be for the child
equally.
(Halliday 1978: 57)
10.1 Introduction
As we have stressed throughout this volume, we believe writing to
be first and foremost a communicative activity. Second, writing is
commonly an internally motivated activity; that is, outside the
school world, people often write because they believe they have
something to say. The teaching of writing in schools, by definition,
violates these two constraints. That is, teachers assign writing tasks;
the tasks are not internally motivated. Moreover, some tasks that
teachers must necessarily assign are not essentially communica-
tive. While we do not condone these artificialities in the various
teaching methodologies intended to instruct writing, we recog-
nize that instructional needs may outweigh other considerations.
At the same time, we believe that teachers can do much to
improve the communicative orientation of even the most m un-
dane writing tasks.
In this and the following two chapters, we shall outline 75
instructional th em es (see pp. 427-9) which are relevant to writing
and which suggest teaching tech n iq u es and ideas. The themes and
techniques draw on the theoretical positions outlined in Chapters
266
Teaching writing at beginning levels 267
8 and 9. Thus, an emphasis on certain aspects of the writing

process is assumed, particularly the notions of brainstorming,
multiple drafting, peer support, multiple audiences, and atten-
tion to form at the later stages. The notion that students should
write for meaningful purposes as much as possible is also
assumed, though the concept of meaningful writing should not
be taken to an extreme. There are a num ber of enjoyable and
interesting writing activities which do not have a purpose outside
of the class but which are useful for the development of writing
skills.
In this chapter, the various instructional themes present many
sample activities for writing practice. It is part of the nature of
beginning writing that students will need many different oppor-
tunities for writing, some of which will be, for one reason or
another, less carefully sequenced with larger curriculum objec-
tives. Nevertheless, the many specific activities suggest an array of
options for getting students involved in writing. Because of the
large num ber of specific activities, certain themes and techniques
presented in this chapter could be viewed as isolated practices and
techniques. We would prefer to see the various ideas for writing
instruction as resources which can be used to support content-
based instruction or some larger project in which the whole class
or student subgroups are engaged.
The themes and techniques in this chapter emphasize writing
activities which are appropriate for beginning students; we also
recognize that such a label is a convenient fiction. There is no one
group of students that can be labelled as beginning writers.
However, there are a num ber of groups of students who do fit the
characteristics of beginners learning to write in English.
Certainly children in English-speaking elementary school class-
rooms fit the characterization of beginning writers; this is true
whether the students are learning English writing skills in their
first language, in their second language, or in an additional
language. For purposes of general classification, beginning ele-
mentary students learning to write would include children in early
education. Most secondary school students do not fit within the
classification of beginning writers, though English L2 immigrant
students who enter secondary schools will be beginners in
English, regardless of their literacy skills in their LI. There is cer-
tainly a difference, however, between students who have learned
how to write in their LI and those who have not. Finally, there are
adult immigrants from non-English-speaking countries. They,
again, can be divided between those who have writing fluency in
their LI and those who do not. Adult courses will also sometimes
involve or include LI illiterate adults.
The extent to which these distinctions are im portant will be
addressed in our discussions of various themes and techniques. It
is im portant to note that the ‘them e’ and ‘technique’ designations
are not intended as a strict hierarchical division, nor are there any
strong claims attached to the designation of some topic as being a
theme rather than a technique. In general, themes characterize
larger issues a n d /o r guidelines for instruction rather than specific
tasks or assignments.
10.2 Themes for beginning writers
The themes for beginning writing discussed in this chapter involve

various issues which can make writing a more productive experi-
ence. While some fluent adult writers think of writing as a difficult
process, children initially do not have the same attitude, and the
teacher should try to ensure that writing is not seen as a frustrat-
ing and unrewarding skill. Rather, students should be encouraged
to feel that writing is an im portant and effective means of self-
expression, communication, and information-gathering. Themes
discussed here (though certainly not all-inclusive) are grouped to
highlight major concerns in beginning writing instruction.
The 25 themes are organized into five overarching principles:
1. Preparing students for writing
2. Assisting and guiding writing
3. Working with writing
4. Writing for different purposes (Genres)
5. Extending the writing curriculum.
These five principles - and corresponding themes - provide a way
to explore various aspects of writing instruction.
In beginning to write, students need confidence to feel that
they can do what is being asked of them. The section on preparing
students for writing offers five themes which promote this goal.
Second, students need guidance as they carry out a given writing
task. The section on assisting and guiding writing highlights ways to
support the development of student writing. Third, students need

to engage in writing activities which cover a wide variety of topics
and tasks. The themes of working with writing present a range of
ideas which can be applied to writing instruction; all of these
themes can be integrated easily into content-based or project-
oriented instruction. Fourth, students need to work with different
types of writing, even at early levels. The themes that address writing
for different purposes provide options for working with a variety
of text types and genres. Finally, students need to expand their uses
of writing, to be given more independent opportunities to write
and to pursue topics of personal interest.
Many of the activities and techniques discussed in this and the
next two chapters are described in a num ber of resources on writ-
ing. In quite a few cases, good ideas appear in multiple sources.
Rather than cite sources for each technique or approach, it is sim-
pler to refer to a num ber of basic resources which we found
particularly useful, and which the interested reader should also
find relevant. Among the most practical and insightful resources
for writing instruction referred to in Chapters 10 to 12 are the fol-
lowing: Frank (1979), Hedge (1988), NCTE (1983), Newkirk
(1993), Peregoy and Boyle (1993), Raimes (1983c, 1992), Reid
(1993), Tompkins (1990), and White and Arndt (1991).
10.3 Preparing students for writing (themes 1-5)
Five themes centre around preparing students for writing and

making students aware that writing is a means of communication.
Integrating these themes into classroom instruction can help
students develop writing skills as well as positive attitudes to writ-
ing. These themes include: (1) using print in the classroom, (2)
writing daily, (3) promoting discussion prior to writing, (4) tap-
ping student interest, and (5) providing reasons for writing. These
themes are designed to raise student awareness, generate interest,
create topics for writing, and get students started.
10.3.1 Using print in the classroom (1)

The teacher and students should find as many uses for print in the
class environment as possible. For classes composed of the same
students all day - common in elementary grades - there are many
options for print display in the classroom. Classroom wall space

can be used for organizing weekly and monthly activities; for gen-
erating lists of ideas and topics; for posting the class roster; for
displaying stories, in-class notes, out-of-class letters, and reports,
etc.
Many of these uses of print in the classroom can evolve gradu-
ally as units are taught through the school year. Others can be
recycled; for example, captions, labels, stories, letters, and reports
can be changed on a monthly basis. Students can work individu-
ally or in groups to create print for display or to assist others. The
possibilities for displaying print in the classroom are endless.
The principle behind these practices is critical. Students need
to see that print is used in a wide variety of contexts, and they
need to explore the ways that print informs, entertains, displays
intentions, and enhances learning.
10.3.2 Writing daily (2)

If writing is to be seen as a basic means of learning information,
exploring the world, and interacting with others, it must be a com-
mon practice. Writing should be practised in the classroom every
day. Following from theme 1, above, it is possible to use writing for
a variety of purposes beyond essay and narrative writing. Some
writing tasks can be brief; others may be a springboard for more
complex writing purposes.
Recent research has shown that students should begin writing
from their first year at school. Early writing, including drawings and
creative spelling, appear to have a positive influence on later writing
development. Early writing activities establish the importance of
writing for communication and raises students’ awareness to the
uses of writing. Early writing also appears to have an important posi-
tive influence on reading development (Adams 1989, Zamel 1992)
and is most readily accommodated through Language Experience
Approaches (LEA) (see subsections 10.4.1 and 10.5.1).
10.3.3 Promoting discussion prior to uniting (3)

Students should be encouraged to recognize the resources that
they each bring to writing, the information that they have but do
not know how to access. By means of class discussion, students dis-
cover and create additional knowledge for writing. The time for
Teaching writing at begi n ning levels 271
discussion also provides teachers with opportunities to shape and

extend student ideas, giving students additional resources to use
in their writing. Students also have the opportunity to learn how
to work in discussion groups to support one another. Often a dis-
cussion begun as a class activity can lead to a simple group task to
generate additional ideas that are reported back to the class. In
this way, class discussion before writing can lead to the beginnings
of cooperative group learning.
The generation of various types of lists and labels for objects
from class discussions provides an im portant resource for begin-
ning students. The generation of lists also builds vocabulary
resources for each student. Further, the recording/drawing-up of
lists, tables, maps, outlines and object labels on the board permits
both teacher and students to explore various ways in which infor-
mation can be reorganized.
10.3.4 Tapping student interest (4)

In addition to discussion which generates lists and other sets of
information, discussion can focus on tasks and topics that interest
students and about which they would like to write. Students need
the opportunity to contribute to the selection of tasks and topics
as a motivation for writing. This does not mean abdication of topic
choice to students, as some might suggest, but a judicious balanc-
ing of teacher and curriculum needs with student initiative and
involvement.
Students often do not know the range of writing activities and
topics open to them. Nor do they know about the various ways in
which a specific topic or task can be adapted as a source for a wide
range of different activities. Brainstorming with the whole class
and in groups allows students to recognize the options that are
open to them, and also to explore the types of information that
they will need for a range of specific writing activities.
10.3.5 Providing reasonsfor writing (5)

Students should be encouraged to carry out writing tasks which
serve a purpose. Not all tasks and activities serve purposes or audi-
ences outside of the classroom; in fact, many useful activities
involve self-discovery and reinforce im portant information from
content-areas such as social studies. While these activities do not
invoke audiences and uses for writing beyond the classroom itself,
they represent real uses of writing which are a basic part of on-
going learning in school. As such, they need to be regarded as real
and authentic writing activities.
Many writing tasks can also be created to extend beyond the
usual writing assignment. For example, writing can involve other
classes at the same grade level or at a different grade level. These
activities can include letters to other students, school personnel,
parents, city officials, etc. Projects and reports can be displayed for
other classes or in a project fair. Essays, stories, and reports can be
displayed on a thematic bulletin board. Writing can be used for
plays, debates, presentations, poetry reading, etc. There are many
extended uses of writing which can be incorporated into the writ-
ing curriculum. In the various themes that follow, a num ber focus
on audiences other than the teacher and beyond the immediate
classroom.
10.4 Assisting and guiding writing (themes 6-10)
A second set of themes centres around supporting students as

they carry out writing tasks in the classroom. In addition to pre-
paring students for writing, it is im portant to provide support for
students while they are writing, to assist them with ideas for organ-
izing their writing and adding information, and to respond as a
reader as well as the teacher. These themes focus on (1) using
sheltered instruction, (2) literacy scaffolding, (3)· attending to
vocabulary development, (4) responding and feedback, and (5)
helping those who cannot get started. In these themes, the
teacher provides the support and assistance that will give students
confidence and motivate them to take risks.
10.4.1 The use of sheltered instruction (6)

With beginning students, it is im portant to provide continual sup-
port, to shelter students as they learn to develop some basic
competence in writing. The teacher:
■ should offer students both verbal and non-verbal cues to help
them recall information, organize their ideas, and use strategies
for planning;
Teaching uniting at beginning levels 273
■ should remind students of the brainstorming and discussion

activities that took place before writing, show pictures and draw-
ings, demonstrate some concept or notion, and provide
redundant explanations;
■ should review and reinforce the purposes for writing as well as
the potential uses of the writing when possible;
■ should m onitor comprehension and help students make writing
more appealing and accessible for potential peer readers;
■ should also take advantage of the opportunities that arise to
suggest new vocabulary or other ways of presenting information
without insisting on students changing the text.
Motivation and encouragement are also important aspects of
sheltered instruction. The teacher needs to keep students
motivated and on task. Ways to increase motivation can include
short writing periods, ‘competitions’ for the best beginning, shar-
ing a paper, writing with students, pointing out good starts and
good organization, and presenting rewards for the best final
papers. Sheltered instruction also provides opportunities for
group work of various types. Students can be asked to help each
other find a good example, remember some information from a
class-generated list, comment on illustrations, or write a joint story
or report. The key concept is that the teacher needs to assist
students actively as they write, without taking responsibility away
from them. The teacher should, at the same time, exercise
good sense in responding to student writing. Not everything
needs to be graded. The teacher should also avoid imposing
interpretations on student efforts; if a text is ambiguous, a per-
fectly valid strategy lies in asking the student what the text is trying
to say.
A particularly effective way to provide sheltered instruction is
through Learning Experience Approach (LEA) activities. These
activities build primarily on stories dictated by students to the
teacher (or to class volunteers or students from higher grades) ; or,
the students’ written stories are read back to them by the teacher.
In this way, they experience success with reading and writing
beyond their normal abilities. This approach increases interest
and allows for systematic follow-up of the language in the stories.
Students can underline words that are most meaningful to them,
write them on word cards, and form a word bank. Students can
cut stories into strips and rearrange them. In some ways, LEA is
like Freire’s approach to generative themes and words since the

themes and words come from the students who can then reflect
on any information they find important. Teachers can also ask
students to re-read stories, illustrate stories, read stories to others,
revise stories, underline difficult words, circle words that they
know, and take dictation from their stories (Dixon and Nessel
1983, Peregoy and Boyle 1993).
10.4.2 Literacy scaffolding (7)

In assisting students in their writing, it is im portant to use lan-
guage that is repetitive and easily understandable. Teachers
should recycle important vocabulary and elaborate on words with
additional information and examples. Discourse routines that sup-
port student writing should be repeated so that students become
familiar with the expectations of both teacher and task.
Teachers should also provide students with basic metalinguistic
vocabulary and routines so that they can talk about their writing.
Even at early ages, students can talk about difficulties in finding
the right words, about a problem with organization, about a less-
than-clear example, etc. Students cannot be expected to do this
without learning how to do so from the teacher. The teacher
needs to present many models and guided examples of how to
look at writing, assess the strong and weak points, look for better
alternatives, and recognize the language structures and vocabu-
lary that go with specific genres and tasks. These insights and
routines should be incorporated gradually into ongoing student
interactions while they are writing, and thus become part of their
own active writing over time.
10.4.3 Attention to vocabulary development (8)

Much as writing leads to improved reading abilities and a more
elaborated knowledge of certain important vocabulary, the devel-
opm ent of writing itself calls for an ever-increasing vocabulary. In
the various activities involved in writing preparation (e.g. pre-
writing discussion or brainstorm ing), or in providing support for
an ongoing writing task, it is important to take advantage of useful
vocabulary which arise naturally out of the situation. As students
generate lists and ideas for writing, certain words will be new and
can be a source of discussion, or even of a writing activity. For
Teaching nrriting at beginning levels 275
example, the introduction of new words can lead to a num ber of

the following writing activities.
Lists of important, useful, interesting, or personal vocabulary can be
created from student writing. Students can, for example, keep a
personal word bank - a list of meaningful words in the back of a
notebook - and can share certain words when new activities call
for their use. Attending to vocabulary is a basic part of expanding
students’ repertoire of resources for writing. It also provides rea-
sons for cooperative group-work; students can ask each other to
explain certain words in their writing, or to use certain words as
part of a group writing activity. Vocabulary development not only
supports reading and writing, it also promotes syntactic flexibility
and creates a foundation for further learning. Further ideas for
vocabulary development and use are presented in a num ber of
themes (see, e.g., subsections 10.5.1, 10.5.2, 10.5.3, 10.5.5).
10.4.4 Responding and feedback (9)

At beginning levels of writing development, the purpose for
responding and feedback is to support student efforts; teachers
need to encourage students to continue, to feel good about their
writing, and to carry the activity through to completion. In this
regard, Frank (1979) provides a useful set of guidelines for teach-
ers working with beginning writers:
1. Build a helpful spirit and give directions for appropriate criti-
cism (e.g. find the funniest sentence; find two good words;
find something good about the opening; find any sentence
that is not clear; think of something that might be added).
2. Start with anonymous pieces from outside the classroom for
class criticism and ease into the process of critiquing slowly.
3. Focus on the positive.
4. Separate revising from editing.
5. Do drafting together.
6. Work often with short pieces.
7. Give specific responses: point out sentences that do not make
sense, strong or weak openings, the need for more descriptive
words, the over-repetition of vocabulary.
8. Decide what techniques need to be refined.
9. Recycle editing experiences into the next writing activity.
10 Avoid false praise.
11. Do not persist in an activity if students are resistant.
10.4.5 Helping those who cannot get started (10)

Even at the beginning, some students will have difficulty getting
started. Sometimes the task creates the problem; sometimes a stu-
dent is not having a good day, and sometimes a student is just
overwhelmed by situations that extend beyond the writing task.
Frank (1979) also has useful advice in this situation:
1. Spend more time collecting/generating ideas.
2. Give some time off from the writing activity.
3. Do not announce writing ahead of time.
4. Encourage other forms of expression (e.g. art).
5. Go back to words and word lists.
6. Return to the oral - talk a rough draft.
7. Accentuate the short as good.
8. Choose other high-interest writing topics for a while.
9. Let students write under a pseudonym.
10. Stop re-writing for awhile.
11. Write in class with a lot of assistance.
12. Combine writers into pairs.
13. Be aware of rigid constraints imposed by specific tasks.
14. Provide more direction.
15. Try transcribing ideas/copying information.
These suggestions should help a teacher to step back and consider
how to make writing less threatening and more interesting. The
teacher’s role is to assess the situation in which students are strug-
gling, find the best way to make writing tasks useful to the
students, and give them a sense of accomplishment so that they
will be better motivated for the next writing activity.
10.5 Working with writing (themes 11-16)
A third set of themes proposes basic activities for working with

beginning writers. These activities provide ways to get students to
generate text. They also represent sets of activities which are easy
to accomplish, giving students a sense of satisfaction and motiva-
tion to continue. Activities should capitalize on students’ early
positive attitudes towards writing and should support those
students whose prior encounters with writing have been insuffi-
cient or unmotivating. Six themes that present these basic writing
activities include: (1) using big books and Language Experience

Activities (LEA), (2) generating word lists, (3) working with
phrases and sentences, (4) copying and dictation, (5) working
with poems, and (6) working with pictures.
10.5.1 Using big books and Language Experience

Activities (11)
An im portant early goal of a writing programme is to familiarize
students with books and to connect books with their own writing
practices. Students will then accept books as a resource and sup-
port for writing. The use of big books represents one effective way
to introduce beginning writing. The teacher provides big books
for the class or the class makes big books. Smaller groups can also
make their own with teacher assistance. Using big books, the
teacher reads to groups and to the class. These books can be re-
read and the teacher can point out difficult vocabulary, assist with
letter-sound correspondences, check for meaning, and select
words and phrases that interest students. These books have pre-
dictable patterns and assist students with many word-listing and
sentence-making activities. A further activity employing a shared
book is a mapping of the story line. This activity teaches text
organization as well as comprehension and vocabulary. Students
can respond to big books in a variety of ways. They can fill in
response sheets in groups, make dioramas of a scene, develop a
timeline, expand on a character, or make a collage that fits the
story.
A variation on big books is the use of picture books without
words. Picture books have sequences of pictures which students
can present orally as stories, either individually or in groups.
These stories can then be written down and combined with copies
of the pictures. This can be done by having students select their
own picture sequences. Later, working in pairs, students share
their own stories with one another. These picture books can thus
be the starting point for writing, or for a Language Experience
Approach (LEA) activity.
Another activity involves the creation of concept books.
Students begin by choosing a simple concept such as big, small,
good, happy, green, ABCs, etc. Students can then collect pictures
and make illustrations which centre around the concept. They can
add words, captions, and simple explanatory sentences. A more
advanced variation, for example, could involve a no-number book-

let describing what the world would be like without numbers. Start
with a list of ideas: no bus-line numbers, no mathematics, no zero,
no checks, no money, no addresses, no telephone numbers, no
computers, no calendars, no birthdays, etc. Students collect illus-
trations and pictures which involve the use of numbers or which
depict life without numbers. These are combined into a book
along with supporting descriptions. The writing can be limited to
phrases and sentences, or it can be expanded into descriptive and
explanatory paragraphs. (This idea can be used to teach systems
using various num ber bases; e.g. 24 hours, 60 minutes, 7 days, 52
weeks, etc.)
Students can also make a class handbook which describes, for
example, the rules of the class. After discussing class rules and
deciding on a final set for the class, students put together the
rules in a class handbook. A group of students, appointed as an
editorial team, could be in charge of changing rules and adding
student-generated rules. This makes a good LEA activity.
Subsequently, students can look at and discuss historically import-
ant sets of rules (e.g. in the USA, the Bill of Rights).
10.5.2 Generating word lists (12)

Many beginning writers need to have a way to begin to generate
ideas and bits of language. One way to do this is by creating word
lists which become resources for writing. Students are not intimi-
dated by the activity, and they can be led to realize that they have
many personal ideas from which to draw. This activity also rein-
forces im portant vocabulary and allows students to explore
relationships between new and known words. Finally, the genera-
tion of word groups and word lists, when practised consistently,
provides students with an independent strategy for starting to
write.
Working with words and lists, in their most basic forms, involves
the teacher and students in collecting words from the class
environment. Students can be asked to collect different types of
words which can then be sources for stories, ‘poems,’ essays, and a
variety of LEA activities. Students can collect busy words, happy
words, convincing words, etc. Among the simple activities which
involve word collecting are:
1. Word tasting. Students select a word - milk, for example - gen-

erate words that tell how it feels, tastes, smells, looks, sounds.
They can add a word that describes how it feels in the mouth,
how it feels when it is swallowed, and an aftertaste word. These
words can be organized into sets of phrases and sentences, or
can be arranged in a poem-like pattern. As a follow-up activity,
students can add words for the types of people who like milk.
2. Prepositions for silly stories. After students select a preposition that
they like (e.g. out, down, in, on, up, with, for, to), they look for
it in a simple reading book. They can then tell a silly story
using as many occurrences of this preposition as possible.
3. Five key words. Students are asked to select a reading passage
from a text. They then choose the five (or x) most important
words that must be preserved from the text if all others are
taken away. Students explain orally, or write, why they would
keep these words by sharing them with others in a group, and
can then generate sets of phrases and sentences with the cho-
sen words. They can tell a story with these words, or make a
‘poem ’ with someone else’s words. The most interesting words
can also be placed in a personal word bank.
4. Feather words. After a discussion of birds or an experiment with
falling objects, students can put together a list of words that are
light or ‘like a feather’ (e.g. floating, soft, light, airy, slender,
twirling, free). Each student can make a ‘poem ’ or a paragraph
from the words, or students can compare word lists and com-
pose a group text.
5. Colour association. Students may pick a colour and assemble a
list of words and phrases to associate with the colour. These
words can then be used to make a set of interesting phrases
and sentences, which can also be written in different colours.
Students can try to tell a story with these words or compare
their words with those of other students. They can then con-
sider whether their statements and words about any colour can
be generalized.
6. Thinking like an animal. Students can pick an animal and make
a list of words that fit the animal. For example, if a student
picks a lion, he or she can list words that interpret the animal;
e.g. words describing how a lion eats or sleeps, or walks, etc.
The student can search for some ‘lion-like’ words, then tell a
lion story or explain how a lion lives.
7. Come to your senses. Students are asked to think of an emotion;

e.g. happiness, anger, sadness, excitement. Students then
describe (orally or in writing) these words by using the five
senses: how it tastes, feels, smells, looks, sounds. Students then
meet in groups with other students who have chosen similar
emotions. They may combine their words and make a word
chart. These charts can constitute part of a unit on the five
senses, on emotions, or on metaphorical meanings.
8. Pick a food. Students write ten words that describe a food they
like or do not like. Students compare these lists and determine
which foods everyone likes and which foods are not popular.
The two groups of foods can then be described through the
combined word lists generated by students.
A variation on student-generated word lists is to have students
look for relevant words in the print environment around them.
Instead of writing words themselves, students cut out words from
magazines and newspapers. These words can be pasted together
to tell a story or to describe/explain a concept. The words can
also be combined to make a bigger story or report as a group
activity. Students can then add pictures to illustrate the work. The
final products can be displayed on a ‘found’ writing bulletin
board. The teacher can keep a pile of such words and phrases in a
box, alongside a box full of interesting pictures, as a possible
ongoing writing option in the writing corner.
More organized and advanced activities which reinforce vocabu-
lary development and ideas from writing include semantic feature
analysis and word clustering. Semantic feature analysis can be car-
ried out in a num ber of ways. For example, based on a current
topic in class, students can select words from their word banks and
from elsewhere that may fit the topic. The teacher can set up a
grid of words so that the words themselves run down the side.
Students then select characteristics that fit some subset of the
words and fill in the grid to match words to their characteristics.
They can also write sentences which explain word groupings. This
activity builds word meanings, but also builds up a concept matrix
for im portant ideas.
In clustering activities, students may also nominate words on a
topic for writing. After many words are nominated, student groups
can cluster words together, then clusters from different groups
can be compared and, where appropriate, connected. The
students decide which clusters are most im portant for the writing
that they will do, and why, and use these clusters to write sentences
or a paragraph on the topic.
10.5.3 Working with phrases and sentences (13)

Another effective set of activities for beginning students involves
the creation of interesting phrases and sentences, which then
serve as a springboard for other activities. One of the simplest
ways to get students to write more complex sentences or work with
a new sentence pattern is through patterned sentences. After
choosing a simple concept or pattern, students experiment with
the patterned organization. Patterning examples include the fol-
lowing (doubled here to show the pattern):
I used to ..., but now I ...
I used to ..., but now I ...
I used to be ..., but now I am ...
I used to be ..., but now I am ...
I can see ..., I can see ..., I can see ...
I seem to be ..., but I really am ...
I seem to be ..., but I really am ...
I’m good a t ...; I’m not so good a t ...
I’m good a t ...; I’m not so good a t ...
I like ..., it looks like..., it feels like..., it tastes like ...
I am the one who likes ...,
I am the one who takes ...,
I am the one who has ...,
I am the one who plays ...
If..., if..., if..., then subject and verb
W hen..., when..., when..., then subject and verb
These patterns, which students repeat five or six times, can be
used to create simple poem-type arrangements. The patterns,
which act as a scaffolding, allow students to exchange ideas and
vocabulary as well as experiment with descriptions, concepts, and
specific structures.
Students can work with new sentence structures in other ways,
particularly as part of making sentences to organize ideas and
information. For example, students can think of warnings, prob-
lems, events, or changes. They can make a sentence list starting

each line with I f They can also make a list of daily habits that
people have or activities that they do. They then write a sentence
list with one sentence for each idea. Other options include mak-
ing a list of events that do not normally happen, or of events that
regularly occur during the year, and begin each sentence with
When.
A second type of sentence-generating activity is to present good
news/bad news. Students may select a person or situation and
think of five (or x) good things and an equal num ber of bad
things; they may then write a good/bad sequence; e.g. fortunately
A, unfortunately B, fortunately C, unfortunately D. Students can
recall a personal experience on a good news/bad news day. They
can also make a good news/bad news sequence for a famous per-
son.
As a third activity, students can write a set of sentences that
begin with the same sound (or letter). After writing five (or x) sen-
tences, they can decide if one could be the source for a story.
Students with the same sound (or letter) can compare sentences
and students in groups can also combine the sentence sets in
interesting ways. Similarly, students can select five (or x) foods and
write a sentence for each, describing each food in such a way that
it sounds awful/wonderful.
As a fourth activity, students in groups may select a sentence
from a text or one that a student has written. Each student in a
group adds a word to the sentence. This may also be done in
groups with some sentences from each student’s latest essay. In
this way, students focus on the types of words that can improve the
descriptive or explanatory power of sentences. Students can also
record particularly interesting words in their personal word bank.
Another useful approach to working with phrases and sentences
is with information gap activities. All information gap activities
involve a pair or group of students; one group or individual has
information that the other needs, and vice versa. Many activities
can be set up as information gap tasks. For example, students can
introduce pictures, ads, or posters and split these visuals in pairs
so that each member in student pairs has a set of pictures, etc.
Individuals (or groups) can then describe the pictures to the
other partner(s). Partner(s), in turn, give written description
back. Both work together to create a final version of a written
description. Students can examine how the description was mis-
leading and determine what is needed to make a description

accurate. These activity can work well when students are learning
to describe settings or backgrounds in larger writing projects.
A further set of activities which focus on sentences involves the
use of sentence strips. In these activities, the teacher selects a para-
graph that is relevant to a larger theme-unit or project and cuts it
into strips, or into words and phrases. The students are then
expected to reconstruct the original text. The teacher can also dis-
cuss with students how they made their decisions for the
reconstruction. The same approach for a more complex set of
sentences can use an illustration to help the students. In addition,
the teacher can scramble words or phrases in a student story and
ask the students to reconstruct the story. The reconstructed ver-
sions can then be compared with the original. Students can also
add sentences to the reconstruction and compare the expanded
text with the original text.
Finally, activities with phrases and sentences can centre around
the creation of lists for various purposes. The general guideline is
to generate a list of interest to students and follow-up with a more
focused writing task. A num ber of examples of list activities
include the following:
1. Ten (or x) friendly things to do. Students make a list of ten (or x)
friendly things to do and discuss why they make their choices.
Students alone or in groups may rank the list items in order of
importance. (How many of these things have they done lately?)
Student groups compare their lists with other groups and com-
pile a top-ten list for the class. Students can then make a poster
of ‘best things to do’ and discuss how many of these activities
they have done lately.
2. Your worst day. Students can list ten (or x) things that happened
on their worst day. From this activity, they can tell a story about
their worst day. The class can also discuss what someone should
say to a person having a bad day and make a short list of things
to say. Students can also list ten (or x) things that might hap-
pen on a very good day.
3. Yes. Students make a list of ten (or x) things to which they would
say ‘yes’. Students can start by thinking of good words in groups.
The class can then organize these things into various categories
and students can then think of headings for each group.
Students can discuss the values and ideas in the headings and
write a short explanation or description for one of the groups of

ideas.
4. Ten (or x) things to do. Students begin by thinking of ten (or x)
things to do last, to tell their mother, to learn about, to do
quickly, to do first, to do slowly, to do in their town, to forget
about, to remember, etc.
5. Excuses. Students can make a list of ten (or x) excuses for not
taking out the garbage, for not doing their homework, for
sleeping late, etc.
6. Five (or x) reasons. Students can make a list of five (or x) reasons
for growing tomatoes, getting good grades, going to school,
watching TV, having a pet, having friends, eating, seeing a
movie, writing (or not writing), etc.
10.5.4 Copying and dictation (14)

For students who have difficulty getting started, or for students
who need to gain fluency with the mechanics of writing, copying
words, sentences, and texts form an excellent beginning. While
some students have no inhibitions with free writing and creative
spelling, other students, particularly older ones, may want more
security in their early attempts at writing; both copying and
dictation provide a higher level of support for accurate and appro-
priate writing.
At the simplest level, students can copy words from the board;
copy a daily sentence; copy information of immediate importance
(such as announcements, class information, semantic maps on the
board); copy part of a class magazine; copy new words into word
banks, etc. Copying can also be a lead-in activity for word-generat-
ing activities and work with phrases and sentences.
Dictations provide students with additional practice in writing
mechanics, word recognition, and sound-letter correspondences.
They also improve fluency as well as related skills such as memor-
izing and note taking. With true beginners, the teacher can use
cloze dictations so that they only need to fill in selected words as
they listen.
10.5.5 Working with poems (15)

There are many ways to introduce students to ‘poem ’ writing.
Students often enjoy finding words that rhyme and sentence
groups that exhibit a metrical rhythm. While it is im portant to dis-

tinguish simple poem-like writing activities from good poetry,
students can produce interesting and enjoyable arrangements of
words and sentences. The extent to which the teacher extends
poem writing to encompass more complex aspects of poetry writ-
ing is an individual choice. These activities provide an ideal way to
engage in word play and structure variation. Many poem activities
can be an extension of other writing activities (e.g. word lists and
sentence patterns). Some options for including poetry in begin-
ning writing are as follows:
1. Texts to poems. Students can start with a simple news article,
advertisement, picture caption, or story, and physically re-
arrange the sentences or move phrases around to form new
sentences. They can also choose interesting words and arrange
them into a poem. The words can be organized into groups
with similar meanings, similar shapes, or similar sounds; they
can be moved into lines; they can be arranged into rhyming
sets; or can be subtracted from lines.
2. Mathematics poem. Students can generate a list of ten (or x)
mathematics words. They can write a poem using these words
and any other math words that can be added. Students can
then form groups and compare and revise their poems.
3. Alphabet poem. Students can pick a letter of the alphabet and
generate a list of interesting words beginning with that letter.
(They can be encouraged to use a dictionary or an encyclopae-
dia for this activity.) They can then arrange words into a poem
that shows their one letter of the alphabet, e.g.:
K is king of the Kentucky kneepads,
K has kennels, keys, kerosene, kites,
K knocks, kneads, knows, keeps, kicks, kids,
K is always on kettles, kilts, kinfolk,
K is a kindly kinetic keeper with a kit of ketchup.
4. Garbage. Students often like to explore ‘messy’ topics. One
example is the word garbage. For each letter of the word GAR-
BAGE, students can write words and phrases that describe
trash. They can then arrange these words into a poem. The
most interesting descriptions can be displayed on the class wall,
pasted on a cut-out poster of a garbage can.
5. Junk. Students can generate lists of words that name kinds of
junk. Using these words, students can write sentences that
include one or more pieces of junk. These sentences can then

be rearranged to form poems.
6. Cans. Students can generate a list of things that come in cans,
and can then make phrases and sentences from these words to
make a poem about items that come in cans. A second focus
can be on pleasant or unpleasant things that come in cans.
7. Diamond poem. Students can combine words with visual displays.
In this activity, students place words in a visual display that
moves from one extreme to the other and produces a poem in
a diamond shape:
10.5.6 Working with pictures (16)

Virtually every book on writing instruction includes activities
involving pictures, and there are almost no limits to the possibili-
ties for using pictures to generate writing activities. One good first
step before using many picture-based activities is to collect a large
num ber of interesting pictures from magazines and keep them in
boxes for student use. Students should also be encouraged to col-
lect and bring pictures for their own writing files to be kept in the
writing corner. Below are some of the options that can be tried as
writing activities. Teachers can also keep in mind that any theme-
based curriculum unit or project may include many pictures as a
focus for writing activities.
1. Make a photo essay. The teacher and students can collect some
photos and pictures and arrange them to tell a story. Students
can create drawings to fill in missing parts of a story line. After
sequencing the photos and drawings, students can write the
story in prose under the pictures and create a booklet. These
booklets can be posted around the room for other students to
read. Picture sets can focus on hobbies, cars, sports, family,

neighbourhood, etc.
2. Working with picture sets. Students can tell a story based on a set
of pictures distributed by the teacher (see also subsection
10.5.1). Each student can then write the story and recount it
to their group. Alternatively, groups receive the picture set,
tell the story, and write up the story together. In simpler ver-
sions, beginning students can describe individual pictures in a
set, then make a group story together. Groups of students can
rearrange picture sets to tell a different story. Picture sets can
also be used in information gap activities. With picture sets,
half of each student group gets the first two or three pictures;
the other half of the group receives the remaining two or
three pictures. Each subgroup tells a story from the pictures;
then the subgroups meet and tell a complete story from the
pictures. After this activity, individual students can generate
their own stories from a similar set of pictures.
3. Select an animal. After selecting an animal from a picture set or
a book, students can look up information on the animal in
the class library and in books on animals collected by the
teacher. Each can then write a small report of one or two para-
graphs, and all students in the group can combine the set of
reports into a class book on animals of interest.
4. Food pictures. Students can locate a picture of some food that
they find interesting. They can then write ten (or x) descrip-
tive words about the picture. Using these words, they can
describe a person who would eat the food.
5. Picture fortunes. Students can first make a list of words and
phrases that can be used to tell a fortune. They can then
select a picture of an unusual man or woman and can tell, or
write out, this person’s fortune.
6. Visualizing what happens. Students can choose a picture that is
unusual or interesting. They can write about what happened
before the picture was taken; about what happened after the
picture was taken; and about what is outside the frame of the
picture that might be important. If it is a picture of a room,
students can describe the contents of the next room, or the
characteristics of the entire house or building. Students can
use pictures, advertisements, cartoons, comics, etc., as a start-
ing point. They can also discuss what might have caused the
situation in the picture.
7. Room descriptions. The teacher can provide pairs of students

with pictures of two different rooms for description. Each stu-
dent pair writes a description of one of the rooms; then they
trade the descriptions. Pairs compare and discuss organiza-
tion and perspective; they can consider different ways to
describe a room, the num ber and kind of details to be
included, and the pattern of description to be followed.
8. Animals and people. The teacher and students can find pic-
tures of animals and people of proportional size on the
printed page. Students can cut the animals and people in
half and recombine the people with the animals. They can
then describe these new creatures: What are they like? What
do they do? Where do they live? Students can follow this
activity with one in which they tell a story about their new
creatures.
9. Pictures and poems. Students can collect pictures and poems on
a topic of interest: feelings, seasons, holidays, sports, sounds,
movie stars, music, nature. Students can paste or copy these
pictures into a notebook and can add poems or write their own
poems for the pictures in the book. Students can then com-
pare books and solicit the help of other students to find more
pictures/poems. They can complete the book by writing a con-
cluding essay that explains the book.
10. Captioned! The teacher and students can collect comics and car-
toon strips. The students can cut off (or white-out) the writing
and add their own scripts. Students can trade cartoons or keep
them. The teacher can also develop a large set of cartoons in
the writing corner. Students can trade their completed car-
toons and strips with other students and give advice or add a
second version.
11. Advertisements. Students can bring in advertisements that they
like or think are effective. The teacher can also keep a set of
advertisements in a box in the writing corner. The class can
discuss why some advertisements are effective and others are
less so. Students can then write an advertisement for x. In the
advertisement, they can try to convince people of the value of
x. In addition, they can design a larger brochure for the adver-
tised subject. They can include an endorsem ent by a famous
person and write out the lines that the famous person has to
say.
10.6 Writing for different purposes (themes 17-20)
The fourth set of themes focuses on four well-recognized text

types. With beginning students, it is reasonable to expect that they
learn how to tell stories, write simple letters, provide a description
or a procedural sequence of events, keep records, and write a
simple report. When they begin to write, most students already
know the basics of telling a story or relating an event so this genre
is a good text type to use for building fluency and mechanics (in
addition to using free writes, word generating, copying, dictation,
and journals). There are also many opportunities for students to
engage in letter writing, descriptions, procedures, and reports in
early grades, particularly in a content-based and project-oriented
curriculum. Four themes address these basic text types: (1) telling
stories, (2) writing letters, (3) providing descriptions, directions
and procedures, and (4) keeping records and writing reports.
10.6.1 Telling stories (17)

There are many ways to encourage students to generate simple
stories. In fact, many of the word-generating and sentence-gener-
ating ideas described previously can be concluded with a brief
story-writing activity. When students are engaged in story-writing
tasks, the teacher can take advantage of the opportunity to point
out the im portant aspects of writing good stories: a strong descrip-
tive opening, a vivid setting, interesting character sketches, a point
of conflict, an exciting opening episode, a sequence of events to
entertain, a climactic event, and a suitable ending. These stories
also allow the teacher to highlight the specific types of language
features which contribute to a strong story: action verbs, descrip-
tive adjectives with strong images, the specific words which detail a
character’s personality, the time-oriented transition words that
keep the sequence of events in order, adverbs which convey the
emotions and attitudes of the characters, etc. In short, there are
many ways to help students work with the structuring of a good
story. While many stories can be designed to follow reading activi-
ties and specific content-based activities, the following are a few
additional ideas which can lead to story writing.
1. Cliff hangers. The teacher can hang paper from a coat hanger
and spread the paper across students’ desks. Students begin to
write a short story on the paper. When students are at an excit-

ing part they stop. Students then move to another hanger and
finish the story. The teacher hangs the stories up for others to
read. In another variation, the teacher or students hang a pic-
ture or a photo from a hanger. On pieces of paper, students
add words and phrases that describe the picture and put them
on threads to hang from the hangers. Students then look at the
hangers and write sentences from other students’ words.
Groups then get together and write stories from the hangers.
2. Telegrams. The teacher should first show students a few actual
telegrams to provide an opportunity to talk about the condi-
tions under which telegrams are sent and received. Students
generate 15 (or x) words that can be used in a telegram and
use these words to write an urgent message. In the telegrams,
students refer to something important that has happened (or
in a story they have just read). Students can then switch tele-
grams and tell a story of what happened after they received an
urgent message.
3. Story telling. After a brainstorming session on a topic, the
teacher can begin a story and stop it at a certain point.
Students can then continue the story individually. Alternatively,
students can begin a story from descriptive words that have
been generated by the brainstorming activity (e.g. dangerous,
quiet, raining, explosion, noisy, fire).
10.6.2 Writing letters (18)

Writing letters is an easy and natural writing activity which can be
expanded in a num ber of interesting ways. Students can write to
many different people and organizations for a variety of purposes.
These possibilities provide genuine alternative audiences for stu-
dents and allow for discussion of audience expectations that
change as students write different types of letters. The differences
in audience and writing purpose can also provide opportunities to
point out the features of language that vary consistently along
with readership and writer’s purpose. In this way, letter writing
allows the teacher to discuss conventions of writing: down-toning
words to qualify assertions, adding strong descriptive adverbs and
adjectives, highlighting points or reasons in making an argument
or taking a position, and using modality appropriately to convey
the right attitude and emotion. Below are a few of the many
possible uses of letter writing in a writing class. Many letter-writing

opportunities will also follow from content-based instruction and
curriculum projects.
1. Writing to others - Letters can be written to a variety of people.
Students can write letters to students in other classes or in
other schools; or to the principal suggesting a change that
should be made in the school; or to the local authority describ-
ing the biggest problem facing the city and what the city should
do about it. Students can write letters that congratulate, invite,
apologize, or direct, and can also set up a class mailbox for
which the teacher can arrange ‘pen pals’ (or ‘pen pals’ can be
contacted through electronic mail; see subsection 10.7.5).
2. Protest Students can write a letter of protest to some company
about its product. The teacher and students can first talk about
words and phrases that are im portant for a letter of protest.
They can then decide on a company to whom they can write.
Students can also write a protest letter about some school rule,
or they can write a protest letter to the city about some issue
that is affecting their school.
3. Greetings. The teacher and students can collect birthday
announcements for famous people from calendars, newspa-
pers, television, and radio. Each student can select a famous
living person who was born on or near his or her birthday.
Each student can then make a birthday card and send it to the
selected person. On the back, students can explain why they
are sending their cards.
10.6.3 Providing descriptions, directions, and

procedures (19)
In beginning writing, it is im portant to expose students to a variety
of writing purposes and tasks. In addition to stories, letters, and
expressive writing, students need to practise types of writing which
engage them with content-centred informational materials and
tasks; these types of writing should be practised regularly for a
num ber of reasons, even though they may be perceived by
teachers as more difficult.
First, if they are indeed more difficult, students would benefit
from more practice rather than less. Second, informational genres
of writing are natural extensions of content-based and project-
oriented curricula and are im portant means for integrating and

learning content-area information, even in the early grades.
Third, these types of writing are natural extensions of stories and
expressive writing activities. Often, students can develop descrip-
tions or procedures as part of a story telling, and certain projects
involving expressive writing can be followed by descriptive and
procedural assignments that oblige students to explore additional
resources. Fourth, some students do not have much exposure to
informational types of writing at home. These students need early
exposure in school to informational genres if they are not to fall
behind more advantaged students in their writing development.
Fifth, students are naturally curious about the world. Information
genres provide another means for exploring the world, and they
let students organize this information for other purposes. While
many activities in this chapter begin with generating information
from personal resources, students should be encouraged to
explore topics which will lead to writing about the world around
them rather than consistently writing about themselves.
The teacher needs to focus attention on the language features
and organizing principles which tend to be consistently part of
descriptions, directions, and procedures. Students also need to be
aware of the words and phrases that clearly signal the intended
organization of the writing.
Early activities involving word lists, phrase and sentence writing,
and writing poems can easily be expanded into description, direc-
tion giving, and procedural writing. The following are just a few of
the many additional possibilities.
1. Favourite food. The class identifies favourite foods and makes a
list of class favourites. The recipes can be collected for a class
recipe book.
2. A special place. The teacher can describe a special place and
explain why it is special. Students can then think of their own
special place, visualize it, and describe it in a few sentences.
3. Shoe personality. Students can begin this activity by collecting
pictures of shoes. They can then describe the shoes and the
personalities of the people who wear them.
4. Giving directions. Students should think of situations in which
people often give advice and specific directions. The class can
brainstorm a list of situations - as well as words and phrases -
used in giving directions, and can then give directions for a
selected situation. For example, they can tell someone how to

do something: eat with chopsticks, make papier mâché, brush
their teeth, cook a turkey, make a hamburger, make a sand-
wich, water a plant, feed a pet, write a recipe, etc.
5. Favourite sport. Students can form groups and nominate their
favourite sports. They can then describe this sport and its rules
to someone from another planet.
10.6.4 Keeping records and uniting reports (20)

Students need to become familiar with report writing. They
should practise collecting and recording information on a topic,
sorting the information, and generalizing from it. They should
also practice writing about the information in a clearly organized
and logical presentation. This may seem to be a daunting task, but
activities can be developed which systematically lead students into
such writing and which follow naturally from other classroom
activities. Particularly in project-oriented and content-area instruc-
tion, students need to practise working with sets of information
that can genuinely inform others, whether they be classmates,
students in other classes, parents, or others.
The teacher can also point out formal features of report writ-
ing. This will represent natural opportunities to focus on the
language used in reports and the organizational frameworks that
will be effective. For example, students can focus on the need to
generalize away from specific examples, to classify information
into larger groupings, and to use qualifying terms rather than
absolute statements. Students can also become more aware of ways
to signal the organization of their texts through transition devices.
There are many possibilities for report writing with beginning
students. Students can begin by brainstorming ideas; generating
words and phrases; focusing on specific topics of interest; making
observations of objects or events; writing up simple recounts,
descriptions, directions, and journal entries; and trying to find
more information from the teacher and the class library. This
sequence of activities, over time, should lead naturally to simple
report writing. Below are a few examples of the types of activities
which lead to report writing.
1. Touch and tell scrapbook. Students collect items that are interest-
ing to touch (e.g. foil, plastic, paper, cardboard, sandpaper,
cheesecloth, corduroy, glue, burlap, foam, leather, satin, nylon,

leaves, sawdust, velour, wood chips, cement pieces, etc.). They
put samples on each page of a scrapbook and write down some
words for each piece (appearance, feeling, source, use). The
items can be organized into groups, appropriately labelled,
and the resulting scrapbooks can form the basis for a simple
report on types of materials in our world, what they look and
feel like, where they are found, and what they are used for.
2. Pet feeding and cleaning. In this project, students write out a
schedule for the care/feeding of a pet. The pet can be either at
home or at school. A record is kept of eating habits, times for
eating, preferred foods, etc. This notebook can then be the
source for a simple report on the feeding routines and eating
habits of the pet.
3. Data sheets. Students can keep a daily record of an ongoing
experiment. Such an experiment can include plant germina-
tion, ant farms, fish behaviour, humidity recording, etc. After a
set period of time, students can write a final summary of their
observations and determine whether or not such observations
can be generalized.
4. Making a questionnaire. The students can plan a class or group
project to examine student interests in some topic. Students
can nominate a topic of interest from a set listed on the board
and create a simple set of questions to gather information from
classmates. Students can also survey family members or
students from another class, and can write up the results of the
survey as a group project. Students will need to determine the
generalizations, if any, that can be made from the information
gathered.
5. Taking notes. Students can practise note-taking through many
simple tasks and activities in order to develop a sense of the key
words and phrases that are important. One such activity involves
taking notes on a short reading, a simple lecture, a video pres-
entation, or the comments of a guest speaker. Students can
compare notes in a group, then compare their notes with the
notes of the teacher. They can use the various sets of notes to fill
in useful words and phrases and to improve the accuracy of
their own notes. Students can then briefly write a summary, or a
response, or fill in a chart. This type of activity should be simple
and should be repeated often: it should also be done with infor-
mation that students need for another purpose.
10.7 Extending the writing curriculum

(themes 21-25)
A final set of themes suggests ways to enhance the writing environ-

m ent and assist students in carrying out writing tasks. These
themes indicate extensions that are natural and necessary for writ-
ing development, lead to positive attitudes towards writing, and
introduce a wide range of opportunities to practise writing. Five
themes appropriate for these purposes include: (1) using a dia-
logue journal, (2) publishing completed work, (3) establishing a
writing corner, (4) including extensive reading, and (5) providing
early access to word processing.
10.7.1 Using a dialoguejournal (21)

A good way to have students extend their writing opportunities and
also use writing to communicate is through dialogue journals. In
highly structured approaches, the dialogue journal requires writing
every day: at a given time, students regularly open their dialogue
journals and write an entry for that day. The journals are collected
periodically by the teacher who responds to student entries. The
responses are not meant to be evaluations of writing or suggestions
for writing; rather, the dialogue journal is an opportunity for the
teacher and the student to communicate informally in the written
medium. The teacher responds to a student’s ideas and engages stu-
dents in a dialogue on whatever topic they have chosen. Of course,
the teacher is free to nominate other topics of interest, much as any
conversational partner would do (Peyton 1990).
The dialogue journal does not have to be a daily routine, but it
should be a consistent one if it is to be seen as a genuine means of
expression and communication. The writing should occur often
enough to be seen as a regular part of classroom activities.
Enough time should also be given to journal writing to ensure
that it is not a hurried session squeezed in between other activities
or before the end of the class day.
There are a num ber of benefits in using journals. They provide
students with additional reasons for writing and further practice,
and they allow students and teachers to nominate topics of inter-
est which can become sources of other writing activities and
projects. The journals also become an outlet through which
students can express their feelings and their attitudes towards
school and other assigned activities. Journals can occasionally be

shared with other students to let them see what their peers are
writing about. Similarly, journals can be a source of discussion in
class and a resource for brainstorming (Staton et al 1988).
10.7.2 Publishing completed work (22)

A second useful extension of writing for beginning students is
publishing certain writings. This process can be relatively infor-
mal, using a space on the bulletin board for the publication of the
week. It can involve the production of a cover and some artwork,
which are then combined with the writing and stapled together as
a book for other students in the class to read in a reading corner
or class library. It can also involve a more elaborate publishing of
the writing. Many classes use a word processor to finish a writing
task for publication. The writing itself goes through an editing
process and then is published.
The use of publishing is a strong motivator for students. They
feel a sense of pride in their accomplishment because the pub-
lished work often looks very good. Nevertheless, the use of
publishing requires a sense of moderation and proportion.
Publishing should not be used for every writing task; rather, it
should be saved for special projects, longer and more involved
assignments, and group projects that are of some special value. A
class book of favourite poems or stories, for example, is a project
which might deserve publication. Publication should also be
reserved for good work. Students and teacher both recognize writ-
ing tasks that reflect special efforts; and these are good candidates
for publishing. Because publishing is special, it should only be
used for special writing otherwise it loses its value and students
will not see books as major productions that require considerable
effort (and provide great satisfaction).
The creation of a small newspaper is also an im portant authen-
tic outlet for student writing. A monthly two-page newspaper can
be taken home to parents and can include class announcements,
student writing, and plans for upcoming projects.
10.7.3 Establishing a -writing corner (23)

Writing activities can be extended into a writing corner, an area of
the classroom designed to promote writing, whether as part of a
writing assignment or as a resource for individual activities. In

many cases the writing corner can be the place where students
and student groups go to explore options for interesting writing
tasks. In classes that use student portfolios (see Chapter 11, sub-
section 11.7.2; and Chapter 13), the writing corner is where they
are kept, and students can then review their previous writing and
decide whether to complete it or revise it. Various editing and ref-
erences materials can be stored in the writing corner, such as
student dictionaries, resource books, encyclopaedias, and editing
guidelines, etc., for checking writing before it is published.
The writing corner can also be a place where many resource
collections are kept to motivate students to write. The writing cor-
ner can include cans and bags which have story starters: these can
be titles, first lines, or first paragraphs for a story. Students can
select one from a bag and finish the story. Another bag might
include papers with sets of words on them from which students
must make a poem or short story. Another bag can contain news-
paper headlines for which students have to write the
accompanying news story. A box can contain pictures which
students can describe or use as a starting point for a story or
poem. A bag can include addresses of people to whom students
can write letters (e.g. addresses of schools, city and governmental
offices, corporations, informational assistance addresses, etc.). A
large metal can, containing group-writing activities, can be used to
get groups started on a writing task or project. These and other
writing starters can all be components of a writing corner, one
which should motivate students to explore writing on different
tasks and topics, and for different purposes.
10.7.4 Including extensive reading (24)

One im portant extension for writing is the incorporation of
extensive reading in class. Reading and writing are reciprocal
activities; the outcome of a reading activity can serve as input for
writing, and writing can lead a student to further reading
resources. A good way to incorporate extensive reading into the
classroom is through Sustained Silent Reading (SSR). In SSR, the
teacher and students stop other activities in order to read silently.
The teacher must also read to model extensive reading. This
should be done on a regular basis (3 to 5 times per week) and can
last 10-15 minutes per session. The time can later be expanded as
students become more comfortable with sustained reading.

Students can introduce their own reading material or select
material from a good class library.
Another way to promote extensive reading which can lead to
many writing activities is by using high-interest class sets of readers.
Reading time is set aside on a regular basis and everyone quietly
reads the same book. After 15-20 minutes, students stop and record
their progress. They might also write a line or two on their response
to the day’s reading. Class readers provide opportunities for group
projects, dramatizations and acting out scenes from the book, addi-
tional creative writing about a character in the book, pre-reading
activities to set the scene before the book begins, or continuing the
story to tell what happens after the book ends. These and many
other options become available with the use of a class reader.
10.7.5 Providing early access to word processing (25)

Increasing numbers of students are gaining access to computers.
As a consequence, educational norms expect students to be famil-
iar with writing in this way. Many schools have a computer lab or a
few computers in the class. These resources should be used to
develop students’ facility with the technology and, more impor-
tantly, to give students writing practice with powerful resources.
The computer gives the students opportunities to write longer
pieces, to revise easily, to work with different formatting, and to
keep files of past work. Computers provide a new dimension for
group work because students can work on a given assignment
either together or at different times, and the text remains clean.
Finally, graphics options and clean print-outs tend to motivate stu-
dents to write for longer periods of time.
A further advantage of computers is that writing tasks can be
transferred to other sites. Students can trade diskette files with
other classes, either to complete a project or as a mail delivery.
Further, students can write to other classes through e-mail; not
only can they set up ‘pen-pal’ exchanges but also engage in infor-
mative projects which exchange information on cultures, holidays,
important people, and classroom practices, among other activities.
Computers provide important options for writing which increas-
ingly cannot be overlooked. Much like writing itself, computers
need to be seen by students as a medium for communication and a
powerful resource for their own development.
Teaching -writing at beginning levels 299
10.8 Adapting themes for the adult beginning writer
Much of the above discussion is primarily designed for the young

student beginning to write, and it assumes students in the elemen-
tary grades. Many of the issues raised, however, apply equally to
the adult beginning writer. Adult beginning writers need extensive
encouragement and support; they need to be given many oppor-
tunities to succeed; they also need to feel that the progress they
make is not trivial or useless. In this regard, many themes outlined
above can be adapted easily to the adult context. Adult students
need to be able to discuss what they will write about, and they
need to brainstorm information on a topic.
■ They need to build a productive writing vocabulary for express-
ing their more complex ideas.
■ They need to write on a daily basis and for a variety of purposes;
the essay is not the preferred goal for beginning adults, it is
rather frequent short writings for different real purposes.
■ They need to focus consciously on the wide variety of print in
the environment and on its many formats and purposes.
Too often, adults focus only on the difficult forms of writing that
they feel unable to produce.
In terms of writing support and guidance, adults also need to be
supported by sheltered instruction and literacy scaffolding. They
need much modelling, tasks that they can accomplish, and stimu-
lating and meaningful activities. In addition, they need to learn a
metalanguage for talking about writing so that they can discuss
their own writing and the writing of others, and can engage in giv-
ing and accepting feedback in a productive way, which is a
particularly difficult skill to develop as adult beginners are gener-
ally vulnerable to criticism from others.
The need to extend writing practice in a variety of ways is per-
haps even more im portant for adult beginners. Since adults have
less time to develop the necessary skills than young children who
will be in school for many years, adults need to practise many dif-
ferent types of relevant tasks and gain control over many genres in
a relatively short period of time. Dialogue journals are a particu-
larly useful way to engage adult writers in serious exchanges of
written information in writing without critiquing their writing
form: they allow adults to feel as though they are communicating
about ideas and issues from their real world as they exchange dia-
logue journals with teachers. Teachers need, however, to empha-

size that they are not correcting ivriting errors in the dialogue
journals; otherwise, students may assume that anything a teacher
reads and does not mark is correct.
A class which includes a writing corner may also motivate adults
to engage in interesting activities either individually or in groups.
The same is true in terms of opportunities to publish, most likely
in the form of well-written letters or newsletters to a variety of
people and groups. Similarly, access to computers can motivate
students if word processing is well supported and modelled. The
need to integrate writing with reading is also a more im portant
issue for adults than for young learners. Adults need to see various
types of printed texts and learn larger sets of informational and
topical vocabulary. They also have a greater need to feel a sense of
success and accomplishment with reading and writing than do
young children because they have little time and few opportun-
ities to recover from the influence of unfortunate past
experiences.
One added challenge involves those adults who arrive in class as
true illiterates. In these cases, student activities and teacher roles
become quite different. The teacher is faced not simply with
students who cannot write, but also with students who do not
understand the relationships between sound and print, who have
no knowledge of alphabetic units and little sense of how writing
forms are shaped by writing purposes. Also, they cannot read as a
way to gain access to forms of writing, at least not initially. In these
cases, a basic literacy programme which stresses the alphabetic
principle, the need to build a simple sight vocabulary of basic
words, and control over the mechanics of writing become the pri-
mary goal. In such a context, students can begin to write by
copying, starting with word lists which they self-generate and
which include vocabulary im portant to their needs. Only after
adults can control the mechanics of writing and have successful
experiences in short assisted writing tasks with teachers (e.g. LEA
activities) should they be given a wider range of opportunities for
self-expression. Adults, more than children, are aware that they
are not writing appropriately, and false encouragement and frus-
trating writing activities should not be allowed to damage a
sensitive adult ego. The transition to a beginning writing stage in
which some familiarity with writing is assumed is a decision that
should be negotiated between the student and the teacher.
10.9 Conclusion
The themes for writing introduced in this chapter are meant to

raise awareness of the many options available for writing with
beginning students. The compendium of themes, however, is not
m eant to be seen simply as ideas for ‘something to do’ the next
day. Rather, the many specific techniques discussed throughout
the themes are intended as activities that can be used within larger
projects and content-units in a class. The themes provide some
simple guidelines to assist their integration into the development
of larger curricular units.
It should also be evident that there are certain instructional
notions that run through many themes. First, it is im portant to set
up an assignment at the students’ level. Even for beginning writ-
ers, this is possible if the task is simple at first (and simple does not
mean simple-minded). Often the starting point is the generation
of words and phrases and their association with some props. In
other cases, the use of lists, pictures, and sentences all provide the
sorts of sheltering that students sometimes need and that can be
the beginning of a larger, more complex activity. Second, students
need to work in groups on a regular basis to see how other
students are conducting the activity and to compare their own
work with others in a non-threatening environment. The teacher,
however, should also use groups to provide feedback that is timely
and that does not always single out students individually. Third,
students should be encouraged to explore ideas and information
in a variety of ways. And while there is a greater emphasis on
expressive and narrative ideas in this chapter, there are also many
activities which can be used to explore the students’ world.
Fourth, students should also explore many aspects of language
form and language use through these activities. These types of
activities provide teachers with opportunities to focus on language
aspects as they arise out of the assignment. Such opportunities
represent im portant ways to raise student awareness about lan-
guage, point out the effective uses of language, and give students
opportunities to explore a variety of writing situations and their
attendant linguistic constraints.
In the two chapters to follow, additional principles, with accom-
panying themes and activities, will be presented with more
advanced students in mind. Many of those activities build on the
simpler activities presented in this chapter. It is also important to
keep in mind that the themes and activities in this chapter are not
only applicable to beginning writers; many ideas can be adapted
to become more complex tasks, appropriate for intermediate and
even advanced writers.
11
Teaching writing at
intermediate levels
11.1 Introduction
As students have more experience with writing in English, and as

they move into educational contexts with more demanding learn-
ing tasks, writing instruction takes on new dimensions. Students
need to be able to organize more complex sets of information and
develop fluency with a wider range of genres and formal struc-
tures of written discourse. While basic fluencies and abilities, as
discussed in the previous chapter, remain important components
of writing instruction, there are many additional abilities which
must be developed and many resources for writing which must be
invoked.
The intermediate student is one who is able to write on a basic
level1 and now must use writing to learn a wide range of other aca-
demic information; someone who uses writing to learn about
history, science, literature, social studies, mathematics, art, and
foreign languages. Students at this level must learn how to read
from multiple sources and write from these sources. Often these
students will be engaged in longer projects which require specific
types of analyses, syntheses, and critical evaluation. At the same
time, intermediate students are continually gaining control over
additional vocabulary and more complex sentence structure while
also gaining a greater degree of stylistic maturity, a reflection of a
growing sense of purpose and audience in their writing. All of
these changes in writing expectations reflect a greater emphasis
on informational writing and its various genres and constraints.
There are two major groups of intermediate writers who need
to improve their writing for academic purposes. For the student in
the public school context, the intermediate writer is seen as the
303
secondary school student, roughly grades 6-11 (ages 12-17).

These students presumably have mastered the basic skills for com-
posing simple messages and the mechanics of writing. They have
also typically gained a fair am ount of experience in writing narra-
tive and expressive tasks. A smaller subset of secondary students
have already gained a range of experiences writing various exposi-
tory and argumentative tasks: these latter students have developed
strategies for organizing information in logical frames which are
appropriate for their purposes. Many of these students have also
worked with thematic projects and content-area information inte-
grated with an emphasis on literacy and writing.
Unfortunately, research has revealed that, for the majority of
students, emphases on expository and argumentative writing have
not been a consistent part of the educational curriculum. For the
majority of students, then, it is imperative that they gain access to
the more powerful forms of writing during their secondary school
experiences. It is well known that expository and argumentative
forms of writing require much practice in a variety of contexts;
thus, students need many opportunities to work with these types
of tasks. A writer cannot move to an advanced level of writing with-
out such extensive practice, even if these students do find
themselves in a tertiary environment and believe themselves to be
ready for tertiary-level work. This group includes both English LI
and English L2 students. While it is important, on general prin-
ciples, to differentiate the needs of LI and L2 students, for the
purposes of this chapter, their similar backgrounds in secondary
school settings provide strong rationales for treating them alike.
(For other purposes, they certainly would have to be distin-
guished; Leki 1992, Silva 1993.)
A second well-defined group of intermediate students is the
international ESL student who seeks to enter an English-speaking
tertiary-level academic institution but has only an intermediate
level of English language skills. These students have typically mas-
tered the mechanics of writing in English, and have developed
some basic fluency in writing through practice. They commonly
take English academic preparation courses either in their home
countries or on arrival in the English-speaking country before
they are ready for tertiary-level work.
A third possible group, adult literacy students, often do not
move beyond basic literacy abilities in their classwork. For this
reason, they do not represent a major group of students requiring
Teaching writing at intermediate levels 305
intermediate-level writing instruction. While it can be argued that

adult literacy students should be explicitly represented in a discus-
sion of intermediate writing development, the political reality is
that they do not, in fact, form a large group of students in compar-
ison with the two above groups.
11.2 Themes for intermediate writing instruction
This chapter will focus on a set of 25 themes which are im portant

for developing writing abilities beyond a basic level. It builds upon
the themes and activities addressed in the previous chapter,
extending and incorporating basic writing activities into the con-
text of intermediate instruction. Again, it must be pointed out
that the themes in this chapter are not intended to represent iso-
lated sets of issues or stand-alone techniques which do not fit into
a larger writing curriculum. Rather, these themes are intended to
highlight critical issues for curriculum design and options for
assignments, and they should be integrated within an overall
instructional framework, preferably a content-based instructional
framework.
The themes for writing instruction in this chapter are again
organized into five general principles:
1. Preparing students for writing
3. Working with writing
4. Writing for different purposes (genres)
The first set of themes addresses pre-writing concerns, presenting
general issues which support subsequent writing activities. A
second set of themes focuses on issues which are im portant during
the time of writing, particularly in the case of larger writing pro-
jects over a period of time. These themes deal with ways to
improve writing as it moves through the writing process of multi-
ple drafts and extension activities. A third set focuses specifically
on various types of writing activities. A fourth set addresses the
need to introduce and practice additional genres. The final five
themes address issues which often exist outside the framework of
specific writing tasks. These ‘extension’ themes consider addi-
tional types of knowledge or abilities (e.g. sentence combining)
which can be integrated into a curriculum as they are most useful

and appropriate.
11.3 Preparing for writing (themes 1-5)
The five themes in this section highlight major approaches to

building a writing curriculum and ways to prepare students for
writing. These themes include: (1) cooperative learning and
group work, (2) content-based instruction as a framework for writ-
ing development, (3) an awareness of audience, (4) free writing
and brainstorming, and (5) semantic mapping and graphic orga-
nizers. Together, the five themes offer students ways to prepare for
writing and opportunities for generating many relevant ideas.
113.1 Cooperative learning and group work (1)

Group work is commonly discussed as an important component of
writing instruction. However, there are few consistent research
results which support the general notion of group work as an
effective approach to writing instruction (cf. Spear 1988). Rather,
there are certain types of group work which appear to be effective
for many types of learning, including the development of writing.
Foremost among these more effective approaches is the general
notion of cooperative learning and its several variations, almost all
of which are supported by a strong set of research studies
(Fathman and Kessler 1993, Slavin 1990; see also Chapter 13).
Cooperative learning is a type of group work (though certainly
not all group work is cooperative learning). The goal of coopera-
tive learning is to engage students cooperatively in inquiry and
interaction for the purpose of working towards some joint goal.2 It
is a carefully structured approach in which all learners are
responsible for the group outcome. Students in each group are
assigned distinct roles: to direct the group, to promote interac-
tion, to record progress, and to report to the outside world.
Students are engaged in solving problems together and support-
ing each other. The goal of the teacher is to model strategic
behaviour, facilitate group interaction, monitor progress, and clar-
ify the problems and the means to solve them.
Cooperative learning approaches have a num ber of points in
common. First, they allow varying options for grouping students;
second, they teach social and interactive skills explicitly; third,

they foster a sense of cohesiveness among group members; fourth,
they promote shared leadership within groups; fifth, they empha-
size the perspectives of various members of the group; and sixth,
they use structured tasks to promote cooperation and learning.
Many approaches also stress group and individual responsibility
for accomplishing the task.3
Five main types of cooperative learning are commonly cited: the
Group Investigative Method, the Learning Together Method,
Jigsaw, the Structural Approach, and Student Team Learning.
Each of these approaches is readily adaptable to writing instruc-
tion, particularly in a content-based, or thematic, writing
curriculum (Fathman and Kessler 1993, R. Stahl 1994). The
Group Investigative Method (Sharan and Sharan 1992) proposes a
large project in which each subgroup investigates one part of the
larger theme. Each subgroup is responsible for planning, carrying
out the study, and reporting back to the class. The Learning
Together Method (Johnson and Johnson 1991) emphasizes the
importance of working together on a project through team-
building activities, discussions, and teacher support. Jigsaw
(Aronson 1978) promotes group cooperation and cross-group
interaction. Each member of a group receives a different set of
information and students must leave their group to work with
other students who received similar information. Groups then
reassemble and combine their information to accomplish a larger
task. The Structural Approach (Kagan 1992) teaches students a
variety of means for organizing and presenting information to
others. These structures are the building blocks for larger activ-
ities, encouraging different types of cooperation. Finally, Student
Team Learning (Slavin 1990) is an approach specifically designed
for content-area learning. This approach emphasizes individual
accountability, team rewards, and multiple opportunities for suc-
cess.
The various approaches to cooperative learning all fit well with
more complex writing tasks and projects and are especially effec-
tive when writing is also used to learn and work with specific sets
of information. Students using cooperative learning approaches
have demonstrated significant gains in reading and writing abil-
ities as well as superior content-material learning (Fathman and
Kessler 1993). While there is a general preference for ‘group
work’ among writing teachers, few methods can match the
research results and demonstrated improvements that have been

shown by true cooperative learning. For this reason, writing
teachers need to be aware of cooperative learning methods and to
consider whether they can be applied to their various writing
instruction contexts.
11.3.2 Content-based instruction (2)

The use of content-based instruction is no longer a new idea, but
it is one that is extremely effective for both content and language
learning, and, in particular, for writing development (Batholomae
and Petrosky 1986, Crandall 1993, Enright and McCloskey 1988,
Mohan 1986, Readance et al. 1989, Santa and Alvermann 1991,
Tchudi and Huerta 1983). The basic goal is to keep content at the
centre of instruction and to build various language learning activi-
ties and projects around major themes. Themes and projects
should be interpreted broadly so that a variety of topics and tasks
can be sequenced to meet learning objectives. Writing activities
should develop directly out of the themes and allow for natural
discussion of ways that different genres of writing serve different
purposes; they should also highlight the ways that appropriate lan-
guage use supports specific genres and purposes for writing.
These notions are applicable to writing instruction from elemen-
tary grades through tertiary-level courses (Bartholomae and
Petrosky 1986, Enright and McCloskey 1988, Manning et al. 1994,
Meinbach et al. 1995, Tchudi and Huerta 1983, Walmsley 1994).
Content-based instruction is effective because students are moti-
vated to learn both content and language. The various uses of
language are not artificially developed but arise as part of the
more general process of inquiry. This approach also allows for
more realistic complexity of the information with which students
are asked to work. Students will each develop somewhat unique
sets of knowledge from the resources they refer to and read.
When they work together in groups, they then have genuinely dis-
tinct sets of knowledge and insights to contribute (cf. many
information-gap writing activities in ESL texts). Further, writing
assignments themselves can be sequenced so that simpler tasks
build into more complex tasks, all of which build naturally to the
completion of a major project.
A good way for a teacher to develop content-based instruction is
by collecting information and nominating useful themes for the
class to explore. These could include explorers, dinosaurs, mon-

sters and myths, trees, the 1950s, volcanoes, underwater
exploration, heroes, space, etc. Each of these themes, in turn, can
be the source for a num ber of more specific topics; for example,
the explorer theme can focus on geographical explorers (on earth
or in space), but it can also include scientific discoveries, the use
of robots for research, the first airplanes, or exploration of the
hum an mind. Topics might include the geography of exploration,
the record keeping required in exploration, financial aspects of
exploration, the sociology of exploration, the politics of explora-
tion, etc. Topics should be organized in such a way as to build up
from easier tasks to more complex and difficult ones. Further,
content themes and topics should be organized in a coherent
m anner so that an entire curriculum can work from an ever-
increasing set of knowledge as a resource for various writing
activities. (See subsection 11.5.2 for additional discussion.)
Content-based instruction also provides a natural framework for
cooperative learning, strategy instruction, the use of graphic orga-
nizers (visual representations of text structure and discourse
organization), and a range of Language Experience Activities
(LEA). Many of the specific ways to develop tasks and activities for
cooperative learning, as discussed above, are especially well suited
to content-area instruction. Students can keep learning journals
and progress logs for sequenced tasks. Well-designed information-
gap activities are also appropriate since students who deal with an
extensive amount of content information remember and synthe-
size the information differently; thus, they generate natural
information gaps through discussions and writing activities.
Students, in addition, will be able to ask, and respond to, informa-
tional questions without needing artificially constructed tasks.
Content-based instruction also lends itself well to discussions
about writing assignments and generates extensive information on
specific topics.
11.3.3 Audience awareness (3)

Intermediate students need to focus greater attention on a consid-
eration of audience. While it is possible to raise such concerns
with beginning writers, they do not often give serious considera-
tion to who the reader is, how the reader may influence the writer,
and how the writer may influence the reader. Writers need to
recognize that readers also have expectations, attitudes, interests,

and objectives. Readers assume that texts will be organized in cer-
tain ways and that sufficient signals of the writer’s intentions will
be available in the writing. Readers also have specific attitudes
towards the content of a text - whether the content is useful or
not, entertaining or not, sympathetic or not, etc. In addition,
readers have their own interests, and texts need to be shaped to
consider reader interests. Finally, readers have objectives in the
sense that they want a text to provide a thorough explanation, a
complete story, an interesting point of view, an honest interpreta-
tion, or a challenging assessment.
Audience factors may also include consideration of a num ber of
variables; whether or not:
■ the reader is known;
■ the reader is an individual or a group;
■ the reader will evaluate the writing;
■ the reader has approximately the same level of general knowl-
edge;
■ the reader knows a lot about the specific topic of writing;
■ the reader will be empathetic;
■ the reader has a different power status.
An attempt by a student to present an argument, and provide rea-
sons for the position taken, should be shaped to some extent by
the various audience factors noted here.
Activities to raise awareness of audience may include rewriting
an argum ent text from the perspective of the opposing viewpoint,
telling the same story from the perspective of a different charac-
ter, rewriting a text for a younger audience, rewriting a text in the
form of a letter to the local newspaper, rewriting a letter of com-
plaint or protest as a letter to a friend explaining your complaint
or concern, or reworking an informational text into a poster for
public display. In each of these cases, it is possible to include
exploration of how a different audience will influence the way
that a student will write (see also Schriver 1992).
11.3.4 Free uniting and brainstorming (4)

Both free writing and brainstorming provide additional resources
for student writing. Free writing, allowing students to write for a
short period of time on any chosen topic, is a useful way to get
students to ‘find’ topics of interest, to recognize ideas of which

they were not aware, and to nominate possible topics for writing
activities. A more directed form of free writing occurs when a
teacher suggests a general idea and students write on anything rel-
evant to that general idea. Similarly, students can free write after a
class discussion to explore ideas of interest or worthy of a response
of some sort. The most controlled type of free writing occurs in
response to a specific question raised by the teacher. Often, spe-
cific aspects of a student’s free writing can be nominated as a new
source for more specific free writing. This looping activity is a com-
monly suggested option in textbooks and among writing teachers.
Free writing has a num ber of benefits for student writing:
1. It is not evaluated and students can feel free to say what they
want about a topic or an idea.
2. It allows students to overcome writer’s block and develop
greater fluency in writing.
3. It is a good source for ideas that can be used for later writing
assignments or future class activities.
4. It gives students a chance to explore, or become aware of their
individual voice - how they write (how they sound) without
conforming to some set of external constraints.
In other brainstorming activities, students can reflect on some
event, concept, or controversy and discuss this as a whole class.
They can write individual reactions to some issue and then use
these responses to discuss the topic. Another possible way to
gather and organize ideas occurs through note-taking during a
discussion and subsequently through reorganizing content by
working with groups. Generating lists, making word associations,
and creating semantic maps (see also subsection 11.3.5) provide
further brainstorming activities. Responding to a short reading or
to a visual display such as a picture, chart, table, or graph can also
introduce ideas for writing.
11.3.5 Semantic mapping and graphic organizers (5)

Working out the relationships among sets of information and rec-
ognizing the levels of discourse and information structuring is an
im portant type of brainstorming, but such an activity goes beyond
brainstorming to matters of overall text structure and the organi-
zation of information. Through semantic mapping and graphic
organizers, students and teachers explore ways to examine the

organizational framework of a text and structure information.
Semantic mapping (or webbing) is now a well-known technique
for generating information, connecting ideas, finding ways to
organize the information, and using the information to develop an
effective written text. Typically, semantic mapping begins with
students generating words that they associate with a topic or cen-
tral idea. These words are written on the board and then organized
into groups of words that appear to go together. This reorganizing
of the words can be done individually, in groups, or as a whole-class
activity. Once groups of words are agreed upon, students can nomi-
nate topic headings for each subgroup of items. Class discussion
can then determine which groups of ideas are more or less import-
ant. The class can also decide whether or not each group of ideas
contributes important information to the central idea; whether or
not the writing tasks should focus on the central idea in the map or
on one or two of the sub-ideas as a more focused activity; and
whether or not there are other ways to organize/add information.
The results of these activities then serve as a major resource for
vocabulary acquisition, reading comprehension, information
learning, and writing assignments. Semantic maps can also be
‘translated’ into more linear informal outlines to assist writing
activities. Overall, semantic mapping is recognized as an effective
technique which improves both content learning and language
development (Heimlich and Pittelman 1986, Nagy 1988).
Graphic organizers are primarily intended to represent concep-
tual relations between sets of ideas. The term refers to visual
structures which organize information and guide reading compre-
hension and planning for writing. There are many types of
graphic organizers and many ways to use them. Information can
be organized into flow charts, hierarchical tree structures, classifi-
cation arrays, charts, tables, figures, histograms, diagrams,
matrices, maps, or outlines (Armbruster 1991, Mohan 1986, 1990,
Tierney et al 1990). These displays can be created by the students
and teachers; they can be given to students as a guide; they can be
partially completed, leaving gaps to be filled in by students; or
they can be presented as empty frames to be filled in by students.
Working with graphic organizers can be done individually or in
groups, and they present good tasks for cooperative learning activ-
ities. Graphic organizers can be used as pre-writing and as
post-writing activities. Ideally, the use of graphic organizers on a
Teaching uniting at intermediate levels 313
consistent basis should lead students to use them independently

to examine conceptual relations and organize information. Many
studies show that this approach is effective for both content and
language instruction (Armbruster et al 1991, Mohan 1990,
Pearson and Fielding 1991, Pressley et al 1989, Tang 1992).
The five themes addressed under this principle reflect the more
challenging demands placed on intermediate writers as well as the
more complex set of factors influencing writing activities and writ-
ing development. These themes include: (1) strategy instruction,
(2) peer response groups, (3) the writing process, (4) movement
from controlled to free writing, and (5) awareness of language
and genre structure. These themes extend the notion that effec-
tive writing requires attention to both the process of working with
text and the output which conveys the intentions and ideas of the
writer. Together, these themes make students more aware of the
options they have while writing.
11.4.1 Strategy instruction (6)

In preparation for writing, and at various times during the writing
process, a focus on writing strategies (e.g. summarizing mentally,
revising plans, providing examples) is an effective way to prepare
students for independent strategy use. Strategy instruction can be
presented to the whole class or as part of small-group activities. In
either case, the most effective means for presenting writing strate-
gies is through explicit modelling and follow-up discussions. The
teacher can model a writing assignment by talking aloud while
composing at the blackboard or on an overhead projector.
Students note im portant or interesting choices made, and during
the discussion after the composing activity, students ask the
teacher about various aspects of composing. At the same time, the
teacher is able to point out a num ber of conscious planning and
composing strategies that he or she has used (e.g. audience aware-
ness, added information). In this way, a num ber of important
composing strategies are brought to a conscious level and consis-
tent attention to these strategies becomes an important
com ponent of writing strategy instruction.
An additional goal of strategy instruction is to have students

consciously attempt important strategies on their own. One
m ethod for developing independent strategy use is through recip-
rocal teaching and a community of learners approach (Brown
1994, Brown and Campione 1994, Brown and Palincsar 1989,
Palincsar and Klenk 1991). In this approach, the teacher works
with a group of students, taking turns composing aloud. As
students compose aloud, they can explain the sorts of strategies
they are using. Students can also focus on planning strategies,
revising strategies, elaborating strategies, or evaluating strategies.
Feedback from the teacher and other members of the group will
highlight additional strategies used or not used. Consistent prac-
tice, with students taking the role of the teacher, will eventually
lead to independent control over a range of writing strategies.
This approach allows the teacher to point to specific aspects of dis-
course structure which support the goals for writing such as
attention to cohesive relations, information structuring, text
organization, and rhetorical goals.
A second m ethod for developing writing strategies is through
procedural facilitation (Bereiter and Scardamalia 1987, Pressley
et al. 1992, 1994). In a series of studies, Bereiter and Scardamalia
trained students to use cue cards while writing. The cards each
had a strategy written on the reverse side to cue strategic writing
behaviour when the student was not sure what to do next. Cuing
was designed both for narrative topics and expository topics, and
cues were developed for planning (think of a new idea, improve
plan, recall major goal, etc.), for evaluating, and for diagnosing
problems. Evaluating cues, for example, would include statements
such as the following:
1. People won’t see why this is important.
2. People may not understand what I mean here.
3. This is a useful sentence.
4. I’m getting away from the main point.
5. I’d better give an example.
6. I’d better leave this part out.
(Bereiter and Scardamalia 1987: 270-1)
As students gradually became comfortable with these cues, they
consulted them more freely. Results of research on this seemingly
mechanistic approach revealed significant improvements in writ-
ing and in appropriate strategy use when used over time.
Teaching mriting at intermediate levels 315
11.4.2 Peer response groups (7)

The use of peer response groups is widespread in writing instruc-
tion, and it is promoted at all levels of instruction, from beginning
to advanced. Theoretically, peer revision is well supported in that
it is consistent with Vygotskean perspectives on learning; it can
draw on research and instructional insights from the cooperative
learning literature; it is supported by social-construction theory as
an im portant way to establish dialogue about writing, nominate
appropriate information, and get assistance from real readers;
and it is consistent with strategy-instruction approaches such as
reciprocal teaching (Nystrand 1986, Nystrand and Brandt 1989,
Spear 1988, Zhu 1994).
Peer response groups, however, do not implicate a single
m ethod or technique; rather, the approach can be carried out in
many different ways. The choices depend on the views of the
teacher, the types of students and teaching situations, and the
time and resources available for instruction. For example, inter-
mediate-level students in grade 8 may have difficulty in
understanding how to comment on peer writing effectively. These
students cannot simply be left to their own resources but need to
be guided if they are to give helpful critiques. Zhu (1994) classifies
five parameters of peer response approaches: small group vs. large
group, student-selected vs teacher-assigned groupings, teacher-
directed vs student-directed groups, role-specific vs non-role-
specific groups, and oral response vs written response groups.
Many of these distinctions centre around the degree of teacher
intervention.
To work effectively, peer response groups need to be modelled
for students. Teachers need to guide students through several ses-
sions so that students become effective readers and responders.
Moreover, students and teachers need to work together to estab-
lish positive interactive environments in which all students work
together to help each other. Spear (1988) argues that effective
peer response groups are task oriented - that is, they stay on task;
they focus on more global aspects of student writing; they give
accurate and specific feedback; they interact as peers rather than
as surrogate teachers and students; and they promote student
trust and support for their writing efforts.
There are many ways to help students become more effective,
sensitive, and insightful readers for their peers’ writings. Spear
(1988) discusses feedback in terms of three types: supporting

feedback, challenging feedback, and editing feedback. The first
builds trust and support; the second provides ways to rethink the
ideas and the structure of the text; the third attends to the formal
aspects of writing that typically are expected by readers. This tax-
onomy highlights different types of feedback that can be useful to
students; it does not suggest that these types of feedback are nec-
essarily exclusive of one another. Spear also suggests that, through
the use of reader free writing, reader note taking, and close struc-
tural analysis of reading, effective guidelines for responding can
be developed. At the same time, she argues against the use of
highly constrained task-oriented revision checklists which reflect
categories of a writing handbook (e.g. diction, paragraph struc-
ture, sentence structure, mechanics). These latter guides are
typically stated negatively and encourage group members to slip
into surrogate teacher-student roles rather than negotiating the
text under discussion: such checklists block real discussion of the
text. Useful guidelines for effective peer responses are also sug-
gested by Elbow (1981) and Reid (1993) (see also Chapter 13).
11.43 The uniting process (8)

The writing process has become a widespread instructional
approach over the last 15 years. It is seen as a major improvement
over traditional methods of writing instruction. Many case studies
and observational accounts support a writing process approach
(Calkins 1986, Graves 1984); however, there is relatively little
empirical evidence that the writing process itself leads to signifi-
cantly better writing or that it is being carried out effectively even
in classes labelled as writing process oriented (cf. Applebee et al.
1994). This is not a counter-argument to process instruction as
much as it is a caution that a m ethod for organizing writing
processes cannot be expected, by itself, to overcome many of the
difficulties encountered in teaching writing; the approach cannot
be seen as a mechanical sequencing of techniques that automati-
cally produces good writers.
The writing process, in general, is described as a five-stage pro-
cess: pre-writing, drafting, revising, editing, and publishing. The
first three steps will typically be recycled as many times as needed.
Activities which are usually associated with the writing process
include brainstorming, semantic mapping, free writing, journal
writing, reading, class and group discussion, peer response,

teacher conferences, mini-lessons on aspects of language, revision
and editing based on the student papers, and teacher feedback
for revising and editing. The teacher facilitates topic selection and
helps students to focus their writing activities; the teacher also
helps students find further information and encourages them to
develop their ideas before being concerned with formal editing.
For younger students, Frank (1979) suggests, in addition, that
students write on a daily basis, that the process be varied for differ-
ent assignments depending on student interest and energy level,
and that not all phases of the process need be involved in every
activity. For example, she notes that students sometimes do not
have the concentration to take a difficult task through all the
stages. Frank also suggests that teachers make the writing process
fun, make it easy, take small steps, avoid over-evaluating, save
unfinished pieces, make a non-writer into a class observer who
takes notes, and not worry about the progress of every student on
each assignment. With younger students, such a relaxed atmos-
phere may often be more motivating in the long run than one
which is consistently demanding.
11.4.4 Controlled, guided, parallel, and free uniting (9)

Students who are making the transition from beginning to inter-
mediate writers need to gain access to the means for more
effective writing for differing purposes, and they need to be
shown explicitly how to do this. They need not only the words and
ways to talk about writing, but they also need specific models
which point out how language and text form are tied to specific
purposes for writing. They need to see how certain discourse con-
ventions are used in clear and simple examples (Smagorinsky
1992). This use of examples does not mean that students are
taught to write in simple ways that they will not use later; rather,
the goal is to give students an initial framework for building more
complex routines and organizational patterns as they encounter
more complex information and more demanding writing tasks
(Stolarek 1994).
Controlled writing can be as simple as changing a paragraph from
first person to third person, filling in certain cloze gaps, adding a
final sentence, combining simple sentences, or filling out an out-
line from a matching text. These and many other activities can be
done not only with commercial materials (e.g. Kaplan and Shaw
1983) but also with texts generated by students or by the teacher
from previous joint or group writing. Thus, there is no need to
treat controlled writing activities as unrelated to larger curriculum
topics and projects. Even the simplest activities involving changes
in writing, or requiring analysis of the structure of writing, can
come from materials generated by students or for other purposes
by the teacher. These activities then provide a way to examine the
form of purposeful writing and to explore relations between lan-
guage form and textual function. These controlled activities also
allow students to gain fluency in writing, to see appropriate writ-
ing models, and to know that the changes they make are not
merely guesses but are likely to be appropriate because the task
has not overwhelmed the student.
The gradual movement from controlled to free writing is m eant
to provide support for student writing, not to require error-free
writing. As students practise various activities in controlled writing,
they can also engage in guided-xvriting assignments in which they
produce one of the paragraphs of an essay from information
given, fill in an outline, or develop an essay from sentence com-
bining. In parallel writing, students can read a simple model
paragraph or two, fill in an outline that allows students to write on
a parallel topic, then produce the one or two paragraphs. Free
writing follows from reading some model texts and writing either a
similar text or a response of some kind. In this final step, students
are not directly guided or supported through the text materials
themselves. It is worth reiterating, in describing this sequence,
that students are as free to be creative, or to proceed without
being completely accurate, as the teacher will allow. In many ways,
the movement from controlled- to free-writing activities can be
seen as an extension of Language Experience Approaches to writ-
ing into an intermediate writing context (e.g. Allen 1976, Cantoni
1987, Tierney et al. 1990). Students’ writing development is
limited or assisted not by the technique itself but by the teacher’s
choices in using the technique.
11.4.5 Awareness of language and genre structure (10)

An im portant component in writing instruction is the integration
of formal aspects of writing with the writing process. To say that
formal aspects of the language be considered resources for writing
does not imply an emphasis on grammar instruction in the tradi-

tional sense. Hillocks (1986), in a careful review of research on
formal grammar instruction, has argued convincingly that formal
grammar instruction provides few positive results when taught as a
school subject. However, many researchers and practitioners rec-
ognize the need for students to attend to formal aspects of their
writing when they work on revising and editing. Moreover, many
writing activities lend themselves naturally to examining formal
aspects of language structure as they apply to writing. These
opportunities may be used to stress how language works as a
resource for better writing.
One of the major shortcomings of the relation between the
structure of language in written genres and the teaching of writ-
ing has been the training that most writing teachers receive in
m odern grammar (if they receive any). There is little that North
American generative grammar can contribute to writing instruc-
tion; rather, the completely theoretical emphasis of m odern
grammar, as it is commonly taught in the USA, for example, offers
no explanation for the use of language structure as a resource for
making meaning in various registers and genres of writing. In con-
trast, various researchers in systemic linguistics, most prominently
in Australia, have pointed out the strong relation between varia-
tions in language form and the specific demands on language
resources made by particular genres of writing (Christie 1992,
Cope and Kalantzis 1993, Martin 1989).
This functional-theoretical perspective on language emphasizes
the relations among a theory of language structure, a social the-
ory of language use, and their implications for literacy
development. For example, systemic research points out ways that
informational writing makes greater use of complex noun groups
and nominalizes many verbs and adjectives. It also demonstrates
ways that argumentative writers rely on a modality system to pro-
vide the rhetorical power of persuasion. These and many other
aspects of written texts can be explored directly with students as
they become aware of the different language resources which can
serve various purposes (Christie 1990, Christie et al 1989,
Collerson 1990, Derewianka 1990, Hasan and Martin 1989). Of
course, teachers also need to be made aware of the patterns of lan-
guage variation that serve writers’ purposes if they are to help
students in this way. In brief, the issue is not whether language
forms and structures are useful, but whether students can
recognize the relations between language structures and the roles

they play in conveying appropriate meaning. Such awareness of
language-as-a-resource gives students access to what Martin (1989)
calls the discourses of power - ways of writing by which people
organize and influence the world around them.
Teachers can use students’ writings to develop tasks that will
raise their awareness of structure, organization of information, and
informational flow. A simple activity involves the use of sentence
strips which have to be reconstructed into paragraphs that are
coherent. The same basic activity can be done with paragraphs
from a reading or a longer essay. A second activity is to provide an
incomplete paragraph and have students complete the paragraph
appropriately. Students can also examine cohesive ties by identify-
ing them, filling them in, or considering how different cohesive
ties would change the text and perhaps make the text less effective.
Another type of awareness activity is to examine sentence
arrangements and information ordering. Students can examine
given-new relations in student-written texts and see how informa-
tion is presented. In prepared paragraphs, students can be asked
to fill in the best continuing sentence at given points and discuss
what make a sentence most appropriate. Students can also focus
on first and last sentences of a paragraph; they can either fill in a
first or last sentence, or add additional sentences. In these cases,
students can focus on the best sentences to provide a transition to
a new paragraph or to a new topic. They can also propose the best
transition sentences for the beginning of a new paragraph.
Students can focus on organizational sentences that highlight
the topic and signal text organization; students can discuss what
types of sentences provide useful organizational information and
why. Students can also examine word choices and explore how
word changes influence writing. In particular, students can exam-
ine the effect of changing verb groups to nominalized words and
phrases, and the influence of compressing clause information into
prepositional phrases, participial phrases, relative clauses, and
adjectives and adverbs (and also show how adjectives and adverbs
themselves can be nominalized). Students can compare para-
graphs that are heavily nominalized with paragraphs that retain
the verbal processes as actions and states. Students need to be
made aware of the differences that such changes make in a text
and the conditions under which each type of writing style is more
appropriate.
11.5 Working with writing (themes 11-15)
As students work on various writing tasks and projects, there are a

num ber of supporting activities and skills that students can use
and practise for their writing. In this section, five themes and sup-
porting activities are presented for these purposes: (1) gathering
information, (2) working with content-based material, (3) sum-
mary and response writing, (4) awareness of language uses, and
(5) editing texts effectively. All of these themes provide informa-
tion and resources that allow students to develop more complex
but more controlled writing.
11.5.1 Gathering supporting informationfor uniting (11)

The most obvious way to gather information for writing is from
discussions and books: books from home, from the class library,
from the school library, from a teacher’s special collection for a
project. There are also a num ber of other ways to collect informa-
tion. A popular option is for the students to develop a survey or
simple questionnaire and then hand the instrument out to other
students, to another class, to five (or x) friends, or to parents and
relatives. The surveys and interviews can also be used with people
who work in the school.
Short field trips, in particular, provide useful ways to use ques-
tionnaires and surveys developed by students; the technique
places some responsibility on students for collecting useful infor-
mation (e.g. subsection 11.6.2 below). A simpler variant is to invite
a specialist into the classroom. The invited person can explain a
process or a job, or the person may demonstrate a skill or a pro-
cess to students. In this case, there is an additional emphasis on
note taking since the information would not be constrained by a
survey or questionnaire format.
Writing letters and collecting brochures is another easy way to
assemble a range of information on a topic. An investigation of
diseases and public health might include a letter to the local
health authority, or, in the USA, to an organization such as the
Center for Disease Control. It might also include a letter to the
principal asking for an explanation of school health policies.
Students may also enquire if their parents work in places that have
specific health and safety policies, what those are, and what pur-
poses they serve. Brochure collecting is another useful way to
amass information. Brochures may be collected for future plan-

ning for schools, employment, vacations, or for a study of available
local tourism information. Students might also explore who uses
brochures and flyers to disseminate information, what informa-
tion is included, who the anticipated target audiences are, and
how effectively the information is presented. Such a project may
be combined with letter writing as well as with interviews and sur-
veys.
Observation is another im portant tool for information collec-
tion, though it is often overlooked. Students can observe a group
of people who meet regularly, or observe a person in a public con-
text (e.g. school secretary, postman) and note what these people
do in as much detail as possible. Students can select a person or a
group to observe and consider whether they are doing a good job.
If the student were the boss, would he or she be pleased with the
work done?
11.5.2 Developing content-based instruction (12)

There are virtually unlimited options for creating content-based
units and topics which will engage students and provide for a
range of relevant writing activities. In most cases, teachers can des-
ignate a num ber of general content-units within which a range of
topics and projects can be nominated by the teacher and the stu-
dents for further investigation. Such content-units and topics can
explore personal values, activities and pastimes, world and
national affairs, scientific discoveries and technologies, famous
and im portant people, people and communities, foods, medicine,
etc. Teachers can begin planning by selecting a set of major con-
tent-based units and gathering resources for a class library. These
resources can be supplemented by student resources and library
materials. Teachers and students can explore possible topics and
projects by considering curriculum requirements, student inter-
ests, values clarification exercises, and informational resources
available to the class (interesting video tapes, a person with
unusual expertise, a major current event, etc.).
Content-based units can cover a wide range of issues and topics,
and they can be the source of a set of projects and writing tasks. In
most cases, these content units can be used to bring in informa-
tion from other classes and to incorporate that information into
the writing and information-gathering activities. Moreover, these
content units can be extended across a num ber of classes beyond

the language/language arts school periods if other teachers are
willing to cooperate (e.g. the social studies teacher).
A good example of a content-based unit generating multiple
topics and activities is given in Enright and McCloskey (1988). In
their unit on ‘heroes’, they develop 12 activities and projects
which were selected after brainstorming the content-theme with
the class. Similarly, Tchudi and Huerta (1983) use a content-based
unit on ‘the elderly’ and list 20 possible topics in a num ber of dif-
ferent academic areas which could extend the theme. They
suggest topics on the elderly in the areas of science, mathematics,
art and music, social science, history, civics, vocational education,
and athletics, and they list 33 specific activities and projects
related to the theme of ‘the elderly’. (See also Meinbach et al.
1995, Walmsley 1994.)
11.5.3 Summaries and responses (13)

Unlike the previous two themes which concentrate on generating
information, this theme suggests a means for organizing and con-
trolling the information for later use. Writing summaries and
responses to information are relatively well-accepted practices in
secondary schools and constitute an im portant strategy for acade-
mic learning at higher levels. Summary writing can begin as a
controlled or guided activity. One good beginning exercise is for
students to complete summaries that only need one or two more
sentences. Alternatively, they can write summaries that involve sen-
tence combining of given structures. They can also write a
summary that is parallel to an example summary, or write a sum-
mary with the first and last sentences already given. In groups,
students can trade summaries and note the differences. As stu-
dents gain some sense of the choices made in summary writing,
they can then write short summaries from outlines of readings
they have finished. They can also begin to write summaries from
longer texts that they have read.
Summary writing does not have to be the difficult and challeng-
ing activity that students often consider it to be. If summary
writing is an early goal of a writing curriculum, it can be incorpo-
rated regularly in a non-threatening fashion. In early months,
students can be asked to write out the main idea of a reading in
one sentence; then students can compare these one-sentence
synopses and decide what makes a good summarizing statement.

In the following few weeks, students can summarize in one sen-
tence and then write a second sentence to add some explanatory
information or a second main point to the summary. In two or
three months, students can write out three-sentence summaries. By
the fifth month, students can write four or six sentence summaries
which nominate only the major points, arranged coherendy.
Another basic way to present summaries is for the teacher to use
a reading and line through the least im portant information, find
the topic sentences, and condense/rem ove lists and examples. In
this way, students learn a few simple rules of summary writing: (1)
remove examples, (2) remove dates, (3) remove descriptive details
and explanations, (4) remove concluding comments, (5) combine
sets of points into a single general statement, and (6) find the
topic sentence or create one if none exists. If practised over a
period of time, these rules become relatively easy for students to
control.
As a variant on the basic summarizing activity, students should
expand a teacher’s summary of an existing text by adding exam-
ples, descriptive details, and explanations. Students can then
compare their expanded texts with the original from which the
teacher developed the summary. The basic principle is that the
underlying text should be recoverable to some extent from a suc-
cessful summary. Such an activity would also raise awareness of the
various principles used in summary formation.
In addition to summary writing, students should also be able to
respond to ideas generated by other activities. Any activity which
requires students to take a position on a topic will involve not only
summarizing abilities but also an ability to respond to the infor-
mation and present a point of view. Responses can begin with the
simple task of students writing what they thought, or felt, with
respect to some discussion or information source. As students
become more comfortable with summary writing, they can add an
evaluative paragraph of three or four sentences following a sum-
mary which may represent a critique of the summary information,
a supporting commentary on the summary, or argue a position
noted in the summary. Response writing provides a way into more
complex persuasive and argumentative writing; it does not
demand any complex arrangements of arguments and counter-
arguments because the response is primarily a continuation of the
summary.
Both summary writing and response writing need to be prac-

tised in meaningful contexts so that they are seen as purposeful
writing activities and skills. As students become involved in more
complex writing tasks and as content information increases with
higher grade levels, students will need to use summary writing and
responses to keep track of im portant information and to assess
their views on more complex issues. These writing activities should
be incorporated into writing curricula as a natural extension of
working with academic information and writing critical commen-
tary.
11.5.4 Awareness of language uses (14)

Students, as they progress through grade levels, need to recognize
the uses of print that are relevant to academic purposes. A study of
print in the classroom context and in the home context is a useful
activity for raising student awareness. There are, in fact, many uses
and purposes for print and these should be brought to students’
attention. Uses (and purposes) for academic print include the fol-
lowing: journals, biographies, stories, letters, résumés, histories,
editorials, commentaries, newspapers, case studies, posters, etc.
(cf. Tchudi and Huerta 1983, Frank 1979).
Students also need to be made aware of the role of language in
various less-standard environments. In the process, students can
engage in activities which make use of their expertise and provide
training through cognitive apprenticeship. In a recent set of stud-
ies, Heath (1993, Heath and Mangiola 1991) explores ways to
make students aware of language and reflective towards its uses by
making them experts in language teaching and language-based
projects. Heath (1993) reports on the use of drama and play writ-
ing in community centres as a strong source of oral and written
language development. Students write their own plays and per-
form them for real audiences. In the process, they devote
considerable amounts of time to literate activities and demon-
strate skills which are not reflected in their performance with
school-based literacy activities.
These activities lead students to reflect on oral and literate uses
of language (including conventional editing), and, in the process,
lead students to work much more intensively with written lan-
guage than they would as students in their own class. Heath
and Mangiola (1991) report on three other projects which
transformed students into literacy experts: (1) cross-grade tutor-

ing with note taking and reporting on the younger students’
progress; (2) students writing a book to the parents of the younger
students involved in cross-grade tutoring - these books discuss the
literacy needs of the younger students; and (3) analysing observa-
tions of spoken and written language in their community and
family, and using their data to develop ethnographic analyses of
language use (see also Kutz et al. 1993). Drama, play writing, cross-
age tutoring, and ethnographies of spoken and written language
all provide ways to raise student awareness of appropriate lan-
guage use.
11.5.5 Editing (15)

At some point in writing instruction, and in the writing process,
some emphasis needs to be given to editing, ensuring that the
writing fits conventions and formats appropriately. Editing, unlike
revision feedback, should occur in the final stages of writing; it
represents an effort to make the writing acceptable to a critical
audience. Editing involves checking a wide variety of surface
aspects of the text: punctuation, capitalization, hyphenation, page
formatting, titles and subtitles, spelling, appropriate word choice,
grammatical sentence structure, and a coherent sequence of
information.
Editing should not be an activity that is suddenly sprung on
students long after they have become comfortable with only mini-
mal concern for editing. At a fairly early stage in the educational
curriculum, teachers need gradually to show students basic con-
ventions of writing. Students, after all, are well aware that there
are conventions in writing to which edited texts all conform.
Parents, also, often remind students of the importance of certain
conventions such as spelling, punctuation, and handwriting skills.
A good way to begin attending to writing conventions is to exam-
ine aspects of well-edited texts which are being used for a
content-based unit. Students can be given an altered version of a
paragraph from a text they are using, which has editing errors
throughout. They then play the game of ‘can you find 20 (or x)
differences’ between the original text and the ‘changed’ version.
This can be done on a regular basis with scores recorded through
the year. Over time, students will begin to perceive the acceptable
conventions of edited writing.
Editing can also be demonstrated regularly through teacher

think-aloud protocols on texts that the class produces. Unlike pre-
vious teacher think-aloud demonstrations, the teacher is making
overt various editing strategies. Thus, the teacher can verbalize
the editing skills he or she would normally apply in silence.
Another regular practice is to place two sentences on the board
every day. In each sentence, there is one editing (but not necessar-
ily grammatical) error; students must identify and correct the
sentences. This takes little time; the practice sentences can be
from student papers or related to a content unit currently being
taught, and students gradually come to recognize the critical pat-
terns.
O ther activities can also promote editing skills. Spelling contests
are one good way to focus attention on correct spelling. Group
contests to find the editing errors in a paragraph are another
means to build editing skills. Editing with peer group assistance is
also helpful. Many handbooks have extensive exercises for editing
texts and attending to grammar and vocabulary problems. For the
most part, the ideas are right but the context is wrong. It is
im portant to have students work with texts that are meaningful to
them. For example, writing essays that will be read by other
classes, writing books for younger students, publishing books, writ-
ing a newspaper, writing letters to outside audiences, etc., all place
responsibility on students to produce conventionally appropriate
texts, and these texts should drive home the need for editing
skills.
11.6 Working with different types o f writing

(themes 16-20)
In Chapter 10, four types of texts were considered as part of writ-

ing instruction (telling stories; writing letters; descriptions,
directions and procedures; recording and reporting inform ation).
In this section, five text types are considered: (1) autobiographies
and biographies, (2) surveys and questionnaires, (3) newspaper
writing, (4) informational reporting, and (5) personal writing. In
each case, these extend the types of writing often done by inter-
mediate students, and support the various other themes presented
in this chapter.
11.6.1 Autobiographies and biographies (16)

One effective way to develop fluency in student writing is to have
students explore issues related to themselves and their histories,
and then extend these explorations to other people.
One simple activity is to have students describe and discuss
im portant events in their lives. The events can be displayed as a
group project, as important moments in their lives, and each
event description can be accompanied by a brief autobiography of
three sentences. They can write autobiographies describing who
they are, where they are from, what they are currently doing (e.g.
student in 8th grade), and what their interests and hobbies are.
O ther activities which will result in an autobiography include writ-
ing about an im portant accomplishment in their lives.
The autobiographical activities can also lead into a long autobi-
ography project. Students can ‘collect’ events and people in their
lives: accomplishments, difficult times, vacations, schooling, rela-
tions with family members and friends, etc. Students can ask
parents for pictures, or students can illustrate events. The most
relevant information can then be assembled in a time-line chart
on a poster board. Students then write a num ber of different para-
graphs (or more) on paper which are subsequently fastened to the
board, aligning them with the appropriate place on the time line.
These time lines can be displayed for other students to read.
As an alternative to the project above, students can write biogra-
phies based on the study of biographies of one or two other
persons. Students can choose someone from a content-based unit
and write a brief sketch of that person. This activity can centre on
an im portant person, a person the student admires, a famous per-
son (e.g. a TV actor), or someone in history whom the student
would have liked to have met. In groups, students can decide ways
to organize a person’s biography; for example, discuss what infor-
mation is needed, and why those persons’ lives are important or
relevant to the students. Students can then use biographies to con-
sider what events, information, and organizational goals would be
necessary to write an autobiography in a similar way.
11.6.2 Surveys and questionnaires (17)

Surveys and questionnaires provide important ways to gather
information and serve as useful resources for many other writing
activities. They can be used to collect information about people

and their opinions about events, rules, politics, society, schools,
etc. These sets of information allow students to make comparisons
and draw conclusions. They can also be used to gather informa-
tion on specific individuals, events, and situations for more careful
analyses; thus, students can use this information for writing job
descriptions, biographies, and procedures and techniques.
Surveys and questionnaires, however, require careful planning
and writing if they are to be used successfully. Planning needs to
begin with class discussion of specific goals for information collec-
tion. The students and the teacher then have to design questions
that will be clearly understood and easy to answer. Attention has to
be paid to the num ber of survey items or questions so that the
questionnaire is manageable. The surveys and questionnaires also
have to be formatted in such a way that students using them do
not become lost or find themselves without sufficient space for
recording additional information.
Results of surveys and questionnaires also have to be organized
efficiently and examined carefully if the collected information is
to be useful. The teacher may want to introduce tally sheets or
sheets divided into columns to sort the information. If there is to
be a summarizing of numbers from specific items, sheets should
also be designed to make adding and averaging easy. The results
of the surveys and questionnaires also need to be analysed
through class and group discussion. Students need to consider the
goals of the survey/questionnaire, which of the collected informa-
tion complements the goals, how the information can be
presented, what conclusions can be drawn, and how the survey or
questionnaire could have been done differently. Students should
then summarize the conclusions and write a report describing the
information collected and the possible importance of the infor-
mation for some other project purpose.
11.6.3 Newspapers and media (18)

Writing for student-generated media represents another source of
writing practice. The most common option is writing regularly for
a class newspaper or newsletter. A school or a community newspa-
per also provide viable options. Writing for newspapers, for plays,
for TV episodes and commercials, and for class anthologies all
extend writing into different text types.
The newspaper genre actually comprises a num ber of text types.

Students can write a sports column following their favourite
school or professional teams. They can also write profiles of
famous sports stars for a monthly newspaper. O ther students can
be in charge of an editorial and letters-to-the-editor page. These
writing tasks require students to summarize and critique events of
importance to the class and school. Students can focus on curricu-
lar units, school events, school rules and regulations, food in the
cafeteria, safety problems in the school, how the school money is
budgeted, etc.
O ther students may want to focus on column and feature writ-
ing. Students can write a series of reports on teaching methods,
on the status of the library or the computer lab, on the planning
for school lunches, on choices for textbooks, on school clubs, etc.
Columns can focus on specific individuals in the school, including
behind-the-scenes people; students can conduct interviews with
maintenance people, cafeteria workers, secretaries in the
principal’s office, and even the principal. Students can also inter-
view different teachers each m onth to write up profiles of teacher
backgrounds and interests. They can also interview recently
arrived students. Finally, students can be assigned to be front-page
reporters. They can investigate the most current stories about the
class, specific students, the school, teachers, community decisions
about the school, etc. The newspaper and its various text types can
also be tied to specific content-based units, particularly for writing
editorials, features, and columns.
Other media genres can also provide many useful writing activ-
ities for students. Students can use information from a content
unit to write a play, involving events and people being studied in
other parts of the school curriculum. The play can add a creative
option to exploring characters, personalities, and other informa-
tion. The powers of persuasion can also be tapped through writing
newspaper, magazine, and TV advertisements (using camcorders).
Newspaper advertisements can be published monthly and TV
advertisements can be scripted and videotaped. After a num ber of
TV advertisements have been videotaped, the class can have a TV
advertisement festival and show all the advertisements to other
classes. Students can also plan a TV news show in which all news,
weather, sports, and commentary are scripted and delivered to the
video camera for later ‘broadcast’.
In all cases of media writing, it is important to let students
create the scripts and critique each other. These text formats offer
students responsibility and demand acceptable performances in
front of their peer audiences. These text types combine the need
to demonstrate creativity and the need to report accurately. Since
most of the audience will be aware of the news and features being
reported, there will be a natural constraint on writers and script-
writers to ‘get it right’. All of these media options can also be
incorporated into curricular units and projects. Features can focus
on major student projects; editorials can critique the projects; fea-
tures and columns can explain the goals of projects, the methods
used for finding information, and present interviews of real or fic-
tional personalities; TV shows and advertisements can focus on
outcomes of projects and announce coming events; and front-
page news can report on the latest progress or unexpected
discovery during project work.
11.6.4 Informational reports (19)

Students in secondary schools who need to write for academic
purposes require extensive practice with informational writing.
This type of writing requires students to present information
according to acceptable formats. These would include, for exam-
ple, (1) comparing and contrasting different events, issues,
results, outcomes, or processes; (2) writing about posed or real
problems uncovered during information collection and explain-
ing them in terms of a problem-solution organization; or (3)
writing about a process, object, or concept that is complex and
requires definitions, descriptions, and examples. The essence of
informational reporting is organizing information into patterns of
arrangem ent that are expected and understood by the reader.
Students need to be made aware of the options and possibilities
available to them as they report and synthesize information. Basic
patterns of arrangem ent include: definition, classification, cause
and effect, comparison and contrast, problem-solution, and
analysis.
Definitional writing focuses on the means for delimiting some
object, process, or concept. It provides the information needed
both to relate a concept, for example, to other concepts and, at
the same time, to distinguish it from these other concepts.
Definitions typically include a topic sentence frame such as: ‘An
x is a y that does z .’ This format is then followed by a
variety of strategies: relating and distinguishing, giving examples,

providing behavioural and physical attributes, describing func-
tions and uses, etc. Classification arrangements extend the
definitional notion to include sets of objects, concepts or issues;
the primary goal of this pattern is to highlight relationships and
differences among groups. Classification patterns do not fully
specify the num ber of branches needed to describe the informa-
tion: as a new type is created in the classification, a new branch is
added. Definition and classification are reciprocal processes, the
former placing an individual into a class and the latter dividing a
class to distinguish an individual.
Cause/effect and com parison/contrast patterns of organization
represent two types of logical organization that can, at times, over-
lap, though they are logically distinct. Cause and effect patterns
delimit causes and consequences. The pattern defines a causal
relation that makes a strong claim. Comparison and contrast pat-
terns define relations without necessarily asserting any causation.
This pattern also represents a powerful resource for setting up
argumentative, problem-solution, and synthesis texts.
Analytic patterns of organization provide yet another powerful
means for persuasion, critiquing, and problem-solution. The
force of this arrangem ent pattern comes from the assumption
that the analysis accounts for all of, and only, the im portant vari-
ables. An analysis of how a watch works neither leaves out
im portant parts of the watch nor discusses parts that do not fit in
the watch. An analytic arrangement assumes that all of the critical
components are included; the parts that are presented explain the
larger concept being studied. An analysis of a political debate, like
those which appear on an editorial page, assumes the authority to
claim knowledge of all the critical issues, and an ability to recog-
nize the essence of a political event. Put in these terms, it is easy to
see why students would have difficulty with a pattern of arrange-
m ent that presumes so much authority and expertise.
Nevertheless, such a powerful pattern is important for students if
they are to move beyond a passive participation in their educa-
tion.
The key to using patterns of arrangem ent for information
reporting is not to introduce them in isolation or as decontextual-
ized activities; rather, students working with information through
content-based units and projects will already have a wide range of
information, and one of their major difficulties will involve ways to
present the information accurately and effectively. In this context,

teachers need to show students how this can be done. The teacher
can model thought processes for arranging information as well as
the decision-making processes involved. The teacher can point
out to students how appropriate patterns of arrangement are used
in some texts with which the students have worked, and the
teacher can guide students in groups to practice using patterns of
arrangem ent that fit the information they have collected. Patterns
of arrangem ent need to be seen as resources rather than as goals
for writing; and students need extensive practice in seeing these
patterns in the informational texts they work with, and in realizing
that such patterns represent methods of effective communication.
11.6.5 Personal uniting (20)

Personal writing is not constrained by expectations of information
reporting but involves the views, emotions, and creative resources
of the individual student. It may involve fictional accounts and
narratives; it may involve writing TV advertisements or plays; or it
may simply involve explaining a personal point of view. What
makes personal writing unique is the authority taken by the writer
to be seen as an expert. The expertise lies in the personal
interpretation that must be recognized in this type of writing.
Recognizing expertise, of course, does not mean that any writing
produced is good, only that the idea or concept behind the writ-
ing should not be challenged as inappropriate; after all, it is the
purpose of personal writing to recognize the creative power of the
students.
The most direct form of personal writing which students can
practise are diaries, journals, and learning logs. In the case of
journals and logs, the writing is more responded to than evalu-
ated. Diaries may remain the confidential property of the students
if that is the class policy. These forms of personal writing may be
very useful for generating ideas and topics for other types of per-
sonal writing as well as for finding sources of interest for thematic
units and larger projects. O ther forms of personal writing are
highlighted in a num ber of other themes in this chapter. Plays,
autobiographies, personal information gathering, and free writing
all provide potential avenues for personal writing. Students
involved in a content-based unit or a project can also use personal
writing to decide what is most interesting, most important, or
most disturbing about the information being used to develop the

unit. In writing responses to reading assignments or group work,
students can examine and reflect on the information that was
most striking.
There are several additional ways to focus on personal writing.
Students can work individually from writing tasks in a writing cor-
ner, or can work on class or group assignments with the following
activities:
1. Students can see themselves as time travellers: Where would
they go and why? Whom would they see? What would they do?
When would they return? What would they bring back with
them?
2. Students can decide what they would do with ten million dol-
lars if they were given the money the next day.
3. Students can choose a person whom they admire and describe
the appealing qualities or characteristics of that person.
4. Students can relate an embarrassing moment: What hap-
pened? Why was it so embarrassing? What did the student learn
from the experience?
5. Students can discuss a strong feeling, emotion, or attitude they
have towards some object, event, person, concept, etc.
6. Students can be asked to give spontaneous speeches with one
m inute’s preparation. For one minute they can write down
ideas and phrases that will help them.
7. Students can use an opening sentence from some book, story,
article, or text and continue in their own words in any way they
wish. Sets of starting sentences could be stored in the writing
corner.
8. Students can tell a m odern tale. They may take some fairy tale,
fable, myth, or proverb that is well known and tell a story that is
a modern-day version of the tale, etc. They can also change the
outcome to fit a m odern perspective.
11.7 Extending writing (themes 21-25)
In this section, five themes are discussed which represent useful

extensions for writing practice. Each theme could also be used to
support content-based units and task-based curricula. The five
themes include: (1) sentence combining, (2) portfolios, (3) values
clarification activities, (4) double-entry notebooks, and (5)
student presentations. The themes provide ways to expand stu-

dent resources and to explore the form and the content of their
writing.
11.7.1 Sentence combining (21)

Students need to become aware of the possibilities for varying
structure, combining ideas, and highlighting and backgrounding
different sets of information. It is no simple matter to have stu-
dents recognize these options in writing, and sentence combining
is one of the few recognized ways to accomplish such goals. In
addition, sentence combining activities build writing fluency,
improve tacit knowledge of structure, promote flexibility, and
allow for more complex writing. Sentence combining should not
be treated as a grammatical matter, but should instead be seen in
a discourse context, recognizing the function of the output combi-
nation within the larger text structure. While sentence combining,
by itself, does not guarantee improved writing quality over a short
period of time, despite strong claims by adherents, it does add an
im portant com ponent to writing skills and instruction. Sentence-
combining textbooks are commonly available for use in writing
classes; however, sentence combining, like all other aspects of the
writing curriculum, is best generated out of the writing of the stu-
dents as well as the supporting material used for content units and
curriculum projects. Using this approach within a meaningful
context is more likely to motivate students and lead to greater flu-
ency and complexity in student writing.
There are a num ber of ways to use sentence combining in a
writing programme. The most basic approaches to sentence com-
bining involve lists of heavily redundant sentences which require
the combining of groups of three or four sentences as a single
more complex sentence. The list of sentences, when combined in
this way, should form a reasonably complex paragraph. Another
fairly standard version of sentence combining is to include con-
nector cue words at the end of each sentence which indicate how
one sentence should be combined with the next. Introducing
students to sentence combining can begin with simple exercises
which ask them to combine two and three sentences into one sen-
tence (Linden and Whimbey 1990).
Incorporating sentence combining into the class context is pos-
sible by using material from relevant texts and from paragraphs
written by students themselves. Students can look for the underly-

ing simple sentences in a brief text segment from textbooks read
by all the students. In a five-sentence edited text, for example, stu-
dents look for all the possible simple sentences they can find and
try to recombine the simple sentences in different ways. Students
can also look at their own writing in the same way. They can also
add additional simple sentences to their own texts, making the
final recombination more complex than the original paragraph.
11.7.2 Portfolios (22)

Portfolios have become a major topic of discussion in writing
research and practice (see Chapter 13). It is im portant to recog-
nize that much more research needs to be done to assess the exact
benefits and limitations of using portfolios in different ways and in
different contexts. At the moment, many strong claims are being
made for the benefits of portfolios. Recognizing that some of
these claims will need to be tem pered in the future, the potential
of portfolios for writing instruction is still quite exciting. Portfolios
can be seen as a place to store all writing activities that include cre-
ative input from the student. These portfolios can be visited
regularly by students who are free to work on any piece of writing
already in the portfolio. One student can also share some pieces
of writing with another student and ask for a response. Between
peer feedback and their own later reassessment of pieces in the
portfolio, students learn to become more critical of their own writ-
ing and begin to discriminate between more and less interesting
writing.
Portfolio assessment involves the presentation of final work to the
teacher for evaluation. In this case, students recognize the need to
present their best work and to make final revisions. This type of
writing evaluation gives language minority and at-risk students a
better opportunity to present their best work without the time
pressure of exam writing or of a grade based on a single writing.
11.7.3 Values clarification (23)

As students begin to deal with more complex issues and topics,
many of these topics will include difficult problems and dilemmas
which need to be examined thoughtfully as part of the writing
experience. Initially, such problems can be proposed by the
teacher through specific exercises. The goal, however, is to have

students recognize and raise problematic issues and dilemmas
independently as they work with more complex information. On a
basic level, several types of activities can be used to generate prob-
lem posing and possible resolutions. Straightforward topics for
this purpose could include: reasons for banning some object,
chemical, activity, or habitual practice in a given setting; examin-
ing jobs done in the home and redistributing work expectations;
finding ways to conserve resources such as water used in the home
or school; working through a dilemma that is posed; a n d /o r try-
ing to see a situation from the perspective of another person.
One approach to such activities is to set up a problem which
requires some difficult decisions and ask the students to explain
their decisions in writing. For example, some specific scenarios
could involve the following: (1) with a simple description of ten
(or x) people, students decide which of them to save if they are ill
and there are vaccinations for only four (or y) people; (2) starting
with ten job descriptions, students decide who should keep their
jobs if three jobs have to be cut to save money; (3) starting with
descriptions of six local communities and businesses which need a
regular water supply, students decide which two groups should
receive a 10 per cent cut in water supplies and why; (4) starting
with five locations for building a m odern government-funded hos-
pital, students choose the one site that will best serve a wide range
of patients and communities. (To enable students to answer item
(4), they will need a general geographic description of the region,
a simple explanation of the resources and obstacles existing at
each site, and a brief economic impact statement. Students can
then make a reasoned decision based on the data.)
A second general approach to values clarification is to begin
with students’ listings of their most im portant activities, most val-
ued friends, most interesting TV shows, most frequent work or
chore, favourite food, least favourite activity, etc. From these lists,
students can consider questions such as: What interests do they
seem to have in common with peers? What do these common
interests indicate about underlying values? Where do such values
come from? Are any of the general interests or values potential
sources of problems? Are any worthy of change? Are any worthy of
encouragement and further development?
As a third approach, students can examine other people’s
values and develop surveys and interviews which investigate other
people’s interests, preferences, goals, pastimes, disinterests, etc.

Students can use the results to explore general attitudes and
values and compare group results to their own views.
11.7.4 Double-entry notebooks (24)

When students use journals in classes, one option is to have stu-
dents write only on the right-hand side of the notebook. The
left-hand side is then free for other students (or the teacher) to
write comments after reading the entries. This double-entry
approach can be used for learning logs, note taking, interactive
journals with the teacher, and journals that are shown to other stu-
dents. The extra space allows sufficient room for the responder to
write out a full commentary directly opposite the initiating section
of the original entry.
The double-entry concept can easily be extended to other writ-
ing formats. Rough drafts can be written only on one side of the
page. Responders can then add comments and suggestions in a
way that is immediately accessible. Students can also return to
their own writing and have sufficient space to respond with addi-
tional information, revisions, or critical commentary. In this way,
students learn to return to their own work and develop the habit
of revising their ideas. In a sense, the students can, over time,
engage in a running dialogue with themselves (Berthoff 1981,
Raimes 1992).
11.7.5 Presenting the text (25)

When students have completed their writing tasks and need to use
the texts for additional purposes and different audiences, a good
general approach is to present the writing to others. Presenting
texts can vary: reading a text out loud to a group or to the class,
performing a script, making a video, recording an advertisement,
trying to make a sale, publishing a book and reading from it,
putting on a debate, or providing a public poster session in which
each student has to explain his or her poster to others.
A critical component of making student writing meaningful is
to create realistic applications for the writing. A num ber of these
options have been examined in some detail in previous themes. In
the case of media presentation, the opportunity to present writing
is an important experience. For example, publishing a book
should not be the final step for students; they should also read
from it, either to their own class or to another class (perhaps to a
lower grade). They can also explain to students from lower grades
what they had to do to publish a book, going through the writing
and editing processes.
Reading their own texts aloud to other students in a group or to
the whole class is an im portant way to present their ideas. Once
students get into the habit of reading aloud, they will edit their
own texts quite differently, given appropriate support. One or two
readings with frequent phrase misreadings, hesitations because of
wrong words, and lost clauses in convoluted sentences will moti-
vate students to improve their revising and editing skills. The
regular practice of reading texts aloud to the class also provides an
opportunity for feedback from classmates. Students in the audi-
ence recognize the added pressure of public performance and
usually try to assist the performing student; this assistance leads
students to comment constructively on the text being read. When
this approach becomes a regular feature of writing activities,
students take seriously the task of providing useful feedback since
they all know that they will also be in the position to ask for assist-
ance.
11.8 Conclusion
In this chapter, five overarching principles and 25 corresponding

themes for writing instruction have been considered with the
intermediate audience in mind. Of course, many of these themes
can also be adapted to beginning or advanced writers. In many
ways, the themes on preparing for writing represent a natural
extension of themes that encouraged beginning students to build
fluency and confidence. Similarly, the themes that involve writing
assistance and writing practices are m eant to extend the discus-
sion in Chapter 10.
Of course, this chapter also discusses the increased writing
demands imposed on students as they move towards greater writ-
ing proficiency. Reflecting this transition is the greater emphasis
on types of writing and the concurrent increased attention to the
formal and organizational aspects of writing. In short, while this
chapter builds and extends the discussion begun in Chapter 10,
it also adds new dimensions to writing instruction. This is only
fitting as students move towards more complex academic uses of

writing. The trend towards greater academic demands and more
complex uses of writing is continued in the next chapter, explor-
ing writing instruction for advanced students.
NOTES
1. It is possible to set an arbitrary level for intermediate writing ability,

such as an FSI level or an ACTFL level, but the discussion of writing,
and the range of student writers we consider, is too broad for us to pro-
vide a specific definition characterizing the difference between
beginning and intermediate writers.
2. There is ample evidence that much scientific and technical writing
(including proposals, research reports, etc.) is actually a collaborative
effort. Students need at least to get a sense sof this real-world activity;
thus, the motivation here is twofold - to take advantage of the benefits
of cooperative learning as an educational strategy and to prepare
students for the cooperative writing environment in the real world
beyond the classroom.
3. This discussion assumes a ‘Western’ orientation. In educational
systems heavily dependent on rote learning approaches, such an
approach may in fact be counter-productive.
12
Teaching writing at
advanced levels
12.1 Introduction
Students in advanced academic contexts have needs for writing

which go beyond the basic goals for communication and for learn-
ing new information. For students who are finishing secondary
schools, in post-secondary institutions, and in post-graduate pro-
grammes, the uses of writing evolve to take on new dimensions.
Writing is no longer seen as primarily for personal expression or
presenting school-based information, though they remain impor-
tant issues. Rather, the advanced writer is also often concerned
with analysing and interpreting information critically, synthesizing
disparate sets of information, creating information, arguing alter-
native perspectives, and presenting and promoting research.
While these uses of writing call on many of the skills that have
been discussed in Chapters 10 and 11, some more specialized aca-
demic skills are also required in advanced writing.
The themes in this chapter both expand and contract the focus
of writing instruction. The expansion of themes is seen in terms of
the additional genres discussed and the in-depth exploration of
textual patterns of arrangem ent with more specialized writing
tasks. The narrowing focus is seen in the discussions of more spe-
cialized contexts for writing. As writers take on more complex
writing tasks, these tasks - which reflect the increasing specializa-
tion of purpose and audience, and the processes involved -
necessarily address narrower audiences and require more specific
writing formats and conventions. In this way, discussions of
advanced writing cannot simply be seen as an ever-expanding set
of more complex generalized processes. Instead, advanced writing
must begin to separate into specific streams and channels. Genres
341
become more specialized and academic communities impose dis-

tinct expectations.
For the purposes of this chapter, the term ‘advanced student’
includes tertiary-level undergraduate students and those academi-
cally proficient students in English-speaking countries who are in
their final years of secondary school and have expectations of con-
tinuing for additional academic preparation. These students
require more complex writing skills if they are to be successful in
their studies. At post-graduate levels, the need for advanced writ-
ing skills is even more critical as this becomes the most important
linguistic mode for meaning making and information presenta-
tion. The themes in this chapter address the changing needs of
these writers as they move from the secondary school to the more
demanding expectations of professionalism.
12.2 Themes for advanced writing instruction
The themes in this chapter follow a similar format to that of the

previous two chapters, though the principles themselves are some-
what different, reflecting the changing goals and requirements
involved in advanced writing. The five organizing principles
include:
1. Planning for writing
2. Organizing information and attending to conventions
4. Working with different genres
A first set of themes considers issues which assist planning for
writing and provide information for writing. In particular, there
needs to be an emphasis, at advanced levels, on critical reading
and the ways information and arguments are presented. The sec-
ond set of themes expands upon the concern for organizing
information while working with more complex writing tasks. It
also addresses supporting activities which can maintain and
improve writing skills, making the complexity of the task more
manageable. The third set of themes highlights classroom activi-
ties and writing tasks which assist and guide writing, and thus also
respond to the greater complexity required in advanced writing
tasks. The fourth set of themes considers additional writing gen-
Teaching writing at advanced levels 343 ■
res which expand the range presented in earlier chapters and

represent types of writing that are more im portant for advanced
academic students. The final set of themes considers issues which
extend the uses of writing, address specific contexts for writing,
and highlight the importance of convention, genre constraints,
and writing style.
12.3 Planning for writing (themes 1-5)
The five themes in this section help students plan for writing;
these themes include: (1) building a climate for writing, (2) work-
ing with multiple resources, (3) reading critically, (4) guiding
discussion, and (5) exploratory writing and writing exercises. At
the most basic level, advanced writing requires an appropriate
classroom atmosphere as well as access to more complex sets of
information. This information is gathered from many sources,
especially through guided discussions and critical reading. It is
also im portant not to overlook exercises and activities which allow
students to speculate occasionally in less constrained ways.
12.3.1 Building a climatefor writing (1)

The need to establish a positive atmosphere for writing is as true
of advanced classes as it is for beginning classes. Teachers need to
show enthusiasm and interest in the writing curriculum and in stu-
dents’ efforts. In addition, teachers need to provide students with
the time to write, the encouragement to write often, and opportu-
nities to explore a num ber of ideas without a constant concern for
evaluation and grading. At times, it is im portant for the teacher to
encourage innovative ideas and risk-taking. In all of these
respects, the teacher must assume the role of expert, guide,
resource, and facilitator. The role of teacher as evaluator is a nec-
essary aspect of the writing class, as it is of every academic context.
The teacher needs to participate in writing activities in the class
by modelling expert performance in terms of both the process and
the product. This does not mean that the teacher should also do
the tasks assigned to the students. Instead, the teacher should find
opportunities to brainstorm and compose, not only with the class,
but for the class. Teachers should take a lead role in composing
aloud (thinking aloud) for a class assignment, developing an
outline or some other visual display for a writing task. In addition,

the teacher should think aloud while carrying out the activity in
front of the students. This allows students to see the planning and
decision making that occurs in mature composing. Students can
ask the teacher questions about the strategies and choices made
while writing for the class and the teacher can point out specific
choices made and concepts developed through effective details
and examples.
At other times, the teacher can begin by composing aloud while
writing, using an overhead projector, and then ask students to
contribute. As the activities develop, students take on greater
responsibilities as composers, and the teacher can mention alter-
native choices that can be made. The class composing and the
possible alternatives can then be discussed in terms of reader
expectations, writing organization, information highlighting, and
variations in writer’s perspective. The critical notion through
these teaching-writing activities is that the students observe, then
comment, and assume some of the responsibility for composing.
From a broader perspective, the teacher must not only maintain
a positive climate by giving thought to how the curriculum will ful-
fil course expectations and established criteria, but must also find
time to encourage the free exchange of ideas, and to promote a
willingness by students to assist each other.
12.3.2 Working with multiple resources (2)

Writing from resources involves an important transition from writ-
ing with personal and easily accessible information to writing from
complex sets of information which are collected from multiple
sources. This is a major transition that students need to make if
they are to be successful in advanced academic contexts. Unlike
beginning and intermediate writing activities, which can often call
on personal experiences and impressions as a primary source of
information, advanced academic contexts seldom place great
store in personal knowledge and impressionistic interpretations.
Instead, advanced academic contexts require information from
other sources, and a main goal of advanced complex writing activi-
ties is the analysis, synthesis, and interpretation of information
from a variety of sources.
Readings are the most common and the most productive
resources available for advanced writing situations. Since so much
Teaching writing at advanced levels 345
of academic writing in different disciplines involves working with

readings, it is im portant that students receive as much practice as
possible in working with, and writing from, readings. The writing
course - with direct supervision, guidance, and support - is
usually the students’ main chance to master these complex skills.
A second critical resource for advanced student writing is the
use of the library. Students need to be given an orientation to
library resources and then be expected to use these resources in
their writing. This process can be developed gradually through a
brief sequence of writing activities which require: (1) finding and
using a particular resource in the library, (2) including a source of
their own choosing in their writing, and (3) referring to two dif-
ferent sources in the same writing assignment. Students need to
recognize the possibilities that library resources offer when they
write for academic purposes.
Besides extensive reading and the use of library resources, stu-
dents can carry out their own research and data collection to
support a writing activity. This can be done through a case study
m ethod with extended periods of observation and note taking.
Students can decide to interview a relevant group or an individual
expert on their chosen topic. Students can also develop a survey
which can then be distributed to a larger group of people; the tab-
ulation of the survey can provide useful independent evidence for
a position or an argum ent being made. Students can also tele-
phone a variety of offices and request information on a topic they
are investigating. In the USA, for example, various government
agencies are often able to provide a range of resource informa-
tion; such options can extend from census bureau information on
different population groups, to forest service information on hik-
ing trails, walking maps, and policies on tree cutting in national
forest lands.
12.3.3 Reading critically (3)

Students are often asked to read (or find) a text which appears to
be relevant to their needs. However, for a variety of reasons, many
students do not seem to be able to exploit the textual resource to
its full potential. In many cases, they are unable to perceive under-
lying assumptions guiding the text, or follow the intertextual
relations revealed in the text. A major goal for academic writing,
then, is to make texts truly accessible in terms of the logic of a
complex argument or of the different subtexts running through a

text. This level of text interpretation is not a matter of basic read-
ing comprehension, but is an expectation of successful academic
performance. Moreover, a major goal of insightful academic writ-
ing is the use of texts in sophisticated and non-obvious ways, in
some ways a peeling back of layers of a text; this is particularly true
in cases of a text viewed as central to an academic course or used
as the basis for interpretive writing.
Students should not be expected to know how to read critically;
rather, they need to be taught explicitly how to do so. They also
need considerable practice in this activity, since it is a complex
skill which draws on many interpretative resources. One way to
help students with critical reading is for the teacher to read a text
with students, or to students, and, through a think-aloud process,
reveal to students what the text invokes through teacher interpre-
tation. A teacher can read an im portant but difficult text and stop
every few lines to discuss the various ideas that come to mind: for
example, the authority the writer has, who the writer draws on for
theoretical authority, the strengths or weaknesses of the argu-
ments, the standard alternative views, a text that interlinks with
this text in support or in opposition, etc. This process of inquiry is
useful for students to witness since students seldom ‘see’ a text
directly questioned by an expert reader.
Teachers can also begin to answer the questions that might arise
with respect to a given text; that is, the teacher can name and
describe alternative texts which provide counter-arguments and
the relative degree of authority of the counter-sources. The
teacher can identify the oddity or centrality of the argum ent to
the expectations of a given discipline or field of study. The teacher
can also point out the sub-arguments being made and their rela-
tionship to an overall theoretical position taken by the author.
Such interpretation takes considerable expertise and students
are seldom even aware that others read in this way, but practice in
doing so and writing with this knowledge gives students an author-
ity and voice which can significantly change their writing
practices. After students have seen a teacher ‘demonstrate’ the
critical reading process a num ber of times, they can be encour-
aged to engage in some critical commentary of their own. They
will be in a better position to do this if they have already been
exposed to a num ber of texts on a similar topic, if they have dis-
cussed the various motivations and theories guiding different
authors, and if they have practised commenting on the strengths

of arguments, sources of background evidence, and linkages
between a given text and various other im portant texts.
12.3.4 Guiding discussion (4)

Another im portant resource that can change the way students
plan their writing, and read and comment on their own texts, is
through carefully guided discussions. Discussions can easily be
sidetracked or be only exploratory in nature, and careful analyses
of reading and writing require them to be specifically focused.
Even a persistent questioning by the teacher of the ways that stu-
dents are interpreting various texts can lead students to a different
perspective on the textual information as well as on the writing
activity itself.
In discussions of readings, a major goal for the teacher is to lead
students to reassess sources of evidence, examine more carefully
the logic of certain polemic arguments, reconsider the consis-
tency of a theoretical position, and question the interpretative
conclusions drawn from a complex set of data/evidence. Students
can also be asked to outline a complex argum ent (either orally or
in writing), permitting them to examine points where the argu-
m ent is weak or requires a great degree of inference. Students can
also be given a short list of analytical questions which they can
learn to use themselves as they listen to a classroom discussion of a
reading or to some other source of information (e.g. a video, a
guest speaker, a radio talk-show).
The guidance needed for critical analysis of discussions, again,
is not easy, and the process requires continual practice. Students
have to learn how to challenge each other but to do so in a way
that is not personally threatening. Students have to be willing to
be persuaded that another position is, at least for the moment,
better supported; they have to recognize that most positions can
be strengthened or weakened by referring to further resources or
bringing in additional perspectives. They also need to recognize
polemic arguments and opinions masquerading as evidence. The
goal of the writing course, and of the teacher, is to place these
skills and their development in the context of writing activities
and content-topics that are central to the writing curriculum.
Guided discussions as well as critical reading should not be
divorced from the central content and topics proposed for the
course; rather, guided discussions, as well as critical reading, need

to be seen as central to the inquiry that constitutes successful aca-
demic writing.
One im portant set of resources for supporting these critical
skills is through a regular sequence of writing responses to
assigned readings and classroom discussions. These responses can
be in the form of fairly standard prose assignments which call for
a summary and reaction to the materials. Alternatively, students
can work through schematic representations of the information.
For example, teachers can provide students with an outline frame
only partially filled in, and students can complete the outline. The
students can also represent information graphically by means of
flow charts, tables, figures, time-lines, etc. These resources are a
good check on comprehension of the information; they lead stu-
dents to interpret the information in ways that the teacher sees as
useful; they become resources which students can share and dis-
cuss; they provide a means for students to synthesize and evaluate
the information; and they create a framework for using the infor-
mation for other writing tasks.
12.3.5 Exploratory writing and writing exercises (5)

The converse of carefully guided analyses of writing resources lies
in providing opportunities to explore ideas speculatively. Rather
than always demanding a controlled assessment of writing
resources, students, at some points, need to be allowed to con-
sider less clearly supported alternative positions, to explore their
own views in writing, or to take an opposing point of view and see
ideas from a new perspective.
Just as beginning writers employ various activities to initiate
writing, advanced writers sometimes need to return to basic exer-
cises to break free of comfortable positions. Since advanced
writers often feel affinities towards given views - the result of many
years of schooling - the ability to explore other views is not a sim-
ple matter. One way to see other possibilities is for students to
generate lists of words on a topic, or to argue for a position oppo-
site to their own, or to examine ideas that some students have
rejected.
Students can also write speculative comments on ideas that are
im portant to them but have not been discussed or have been of
little interest to the class previously. Some students may wish to
write about a problem that has concerned them in the treatm ent
of a given topic, even if they have difficulty explaining why the
problem deserves more attention, or why it has been overlooked.
The major goal of these writing exercises is to train students to
speculate and adopt alternative points of view; the aim is to
encourage students to explore additional ideas, make room for
speculative thinking, and recognize that there may be a num ber
of alternative positions or perspectives, even if they have not all
been given equal attention in the main activities of the class.
12.4 Organizing information and attending to

conventions (themes 6-10)
The themes in this section point out the importance of organizing

information and attending to reader expectations in academic
writing. The five themes include: (1) outlining and structured
overviews, (2) using heuristics for organizing writing, (3) using
patterns of essay organization, (4) paragraphing and text format-
ting, and (5) starting writing and ending writing. The advanced
writer needs to be aware of the many ways to organize writing and
the constraints on writing imposed by expected forms of organiza-
tion.
12.4.1 Outlining and structured overviews (6)

Writing at advanced levels requires greater attention to the organi-
zation of the writing, the logic of the arguments, and the
sequencing of information. All of these concerns suggest that
mechanisms such as outlining and structured overviews are useful
supports for writing. It is important to note, however, that the use
of outlining in writing does not necessarily refer back to tradi-
tional outlining practices in which a writing topic had to be fully
dissected in an elaborate formal outline before writing began.
Instead, the goal of outlining is to work out the most appropriate
logic for assembling information and arguments. For some writ-
ers, this may mean writing out a fairly unstructured set of notes
which are then reassembled more carefully as a useful working
outline. For other students, a careful outline is an important pre-
liminary step in writing; if the formal process works for these
students, it should not be discouraged because it is too strict for
others. Still other students gain most from outlining when they
develop a careful outline after writing the first draft. This practice
allows students to see weaknesses in the structuring of the infor-
mation and problems in the logic of the writing.
Outlining - to whatever degree of formality is effective for a
given set of students - is a skill that cannot simply be explained
once or twice and then assumed. If it is to be an effective support,
some type of outlining activity should be included in any larger
writing task. One of the easiest ways to incorporate outlining is to
work from semantic maps and other forms of organized brain-
storming. Often semantic mapping will provide a central theme,
much terminology and a hierarchy of information one or two
levels deep.
Outlining is also a useful activity for peer revisions. Asking the
other members of a group to construct an outline that reflects a
given piece of writing can be very revealing not only for the stu-
dents writing the essay but also for the other cooperating students
in the group. It also allows peer students to offer suggestions for
revision which should become readily apparent to the student
whose paper is being reviewed/outlined.
O ther options for displaying the structure of a text are possible.
Rather than working from the same outlining format every time,
students may want to experiment with other visual displays
(graphs, charts, other types of figures, etc.) which reflect the logi-
cal organization of the text. A tree-branching display for a
classification paper may be an effective way to fill in gaps or to
consolidate information. A time-line may be a useful display for
historical cause and effect analyses. A flow chart with several
branching options at different points may capture the complexi-
ties of interacting concepts or processes better than an outline
format because it better depicts the dynamics of options at differ-
ent points. These alternative formats can be as complex as the
detailed formal outline in displaying the order of information and
the logic of the writing.
12.4.2 Using heuristics for organizing uniting (7)

In writing informative and argumentative texts, there are several
stages at which guidelines of various types can assist the writer.
These guidelines, often in the form of sets of questions, can reveal
for the writer options that should be explored and extensions that
Teaching uniting at advanced levels 351
should be considered. In the more formal terminology of

rhetoric, these guidelines are term ed ‘topics’; in less formal termi-
nology, they can be seen as schemes for organizing writing and
considering options at various stages in the process.
Raimes (1992) provides a useful set of these topics for writing.
She provides schemes for question forming, schemes for generat-
ing ideas, topic development schemes, purpose schemes, schemes
for supporting a thesis, ‘openings’ schemes, drafting schemes, and
editing schemes. For example, in her scheme for generating ideas
and information on a topic, Raimes refers to the standard journal-
istic format: What?, When?, Where?, Who?, Why?, How? She then
follows up with an extensive sequence of exploratory questions
which can be applied to any type of writing. For persuasive writ-
ing, she sets up the following scheme:
What do you know a b o u t...........?

What do others need to know a b o u t...........?
Who believes...........now?
Who is affected b y ...........?
How is everyday life affected b y ...........?
How will things change if people change their minds about
...........p
How can people’s opinions a b o u t...........be changed?
When did people first begin to think the way they do about
........... ?
When d id /d o views a b o u t...........change?
Who feels strongly a b o u t...........?
Where is support f o r ...........found?
Where is opposition t o ...........found?
Why is it important to convince people a b o u t...........?
(Raimes 1992: 41-2)
In another set of schemes, Raimes provides guidelines for review-
ing a first draft and supporting the thesis statement. These are
presented as ways that writers support a thesis:
Provide facts or Statistics

Provide examples or illustrations (tell stories)
Provide sensory details
Use chronological or spatial sequences to describe events or
scenes
Compare or contrast two persons, things, or ideas
Analyze or classify people, objects, or ideas

Examine causes and effects or problems and solutions
Give definitions
Examine what others have said.
(Raimes 1992: 65)
It is no accident that these guidelines for thesis development

reflect basic logical patterns for organizing informational texts in
general, and highlight the need to develop a thesis along the lines
of a framework that structures the writing as a whole.
There are a num ber of other im portant schemes for organizing
ideas for writing. Perhaps the best known among these is the set
developed for writing by Young et al (1970), based on a tagmemic
framework for linguistic theory and a set of Aristotelian concepts.
12.4.3 Using patterns of essay organization (8)

One of the most basic topics for writing instruction centres on the
set of patterns of organization which underlie much of expository
and argument writing. As discussed in Chapter 11, these patterns
are im portant for the logical development of informational writ-
ing. Recognizing these patterns in writing, however, is not enough
for advanced writers; they must develop facility with the many pos-
sible variations on these patterns that will lead to effective writing
and assist the reader of the text.
These patterns of organization, as described in many writing
texts, are intended to provide the writer with an efficient yet con-
strained way to address different types of issues and concepts. If
the goal is to describe a situation, setting, or process, the patterns
for organizing texts will define, describe, classify, and develop a
setting/procedure. It would make less sense to present a descrip-
tion in terms of a com parison/contrast or a cause/effect relation
unless the writer intends to signal more than a description
through the text. In this way, writers’ arrangements are signalled
and reader expectations are activated based on patterns that con-
ventionally structure information. The patterns are typically
classified as follows: definition (by taxonomy, by function, by
example, by analogy), description (spatial, temporal, process),
classification, comparison and contrast, problem and solution,
cause and effect, analysis, and synthesis.
These patterns of arrangem ent do not represent major genres
of writing; rather they are means for organizing information

within a num ber of genres and text types (e.g. narrative, argu-
ment, report writing, letters, sermons, lectures). Patterns of
organization need to be made explicit through critical readings,
outlines, visual displays, model texts, and group analysis of student
texts. Rather than employ decontextualized lessons and exercises,
teachers should look for ways to exploit reading material, make
use of semantic mapping and information generation, and focus
on analyses of student texts as ways to raise student awareness.
Peer feedback and teacher responses to writing can also enquire
about the organizational logic of students’ writing.
12.4.4 Paragraphing and textformatting (9)

At a certain point in advanced academic writing, the adherence to
conventions and the appropriate use of specific writing formats
become an im portant part of the presentation of the text for the
reader.
One particularly im portant convention is that of the paragraph.
While historically the paragraph has been perceived as a visual
entity intended to improve the physical symmetry of the text and
make reading easier, the notion of the paragraph has taken on a
new purpose in the latter half of the twentieth century. It is now
expected that paragraphing, aside from matters of display and
reading ease, reflects the psychological units of textual informa-
tion. The paragraph, as currently used in informational writing, is
intended to signal a coherent set of ideas, typically with a main
theme and supporting information. W hether or not an explicit
topic sentence is provided, readers expect the paragraph, as a
unit, to convey a sense of information which can be summarized
in a single sentence, and which contributes to the organization of
the total text. This m odern view of the paragraph fits psychologi-
cal models of language comprehension as developed by Kintsch
and van Dijk (1978, van Dijk and Kintsch 1983) and Singer
(1990).
The development of organized and logical paragraphs cannot
be assumed, even for advanced writers. The point of instruction is
not to engage in an endless series of paragraph writing exercises,
but to raise awareness of the role of paragraphing in writing, to
discuss its relation to careful outlining and other visual displays of
information organization, and to examine the effective use of
paragraphing in model texts being explored for writing assign-

ments and projects. Of course, if students do have difficulties with
appropriate paragraphing, it is im portant to address those prob-
lems directly rather than hope that the advanced writer will
resolve the difficulties naturally. Also, it cannot be assumed that
advanced writers are fully aware of the power and purpose of the
paragraph, since this aspect of writing is often ignored or treated
in a simplistic way at lower levels of writing instruction.
12.4.5 Starting writing and ending uniting (10)

A major problem for many writers, including advanced writers,
centres on appropriate ways to begin and end writing. As students
learn to work with new genres and new types of academic tasks,
the simpler notions of openings and closings often are no longer
adequate. It is not sufficient for students to recognize that open-
ings state a thesis and indicate an organizational plan; students
must also recognize that openings for different academic pur-
poses will provide specific types of information and references to
theoretical positions or concepts that ground a particular piece of
writing.
A good way to raise student awareness of openings and closings
is to collect sets of openings from various sources, particularly
from readings used for the writing class and from student writing,
showing effective and less-effective openings. The openings from
edited texts can be examined for the consistent features common
to many openings; they can also be examined for features that
vary from topic to topic, from purpose to purpose, from genre to
genre, and from audience to audience. These patterns of consis-
tent and varying features can, in turn, be used to construct
guidelines for looking at student openings in writing for various
purposes.
The ending of a writing assignment is as important as its open-
ing. In many cases, a useful essay strategy in timed writing exams is
to stop near the end of time, no matter how seemingly incoherent
the break might be, and compose a clear two-sentence conclusion
that can then be fitted to the text as well as possible. A clear end-
ing will make up for many inconsistencies along the way. And
since the conclusion is indeed the last information read by the
reader, the text is likely to give the reader a better impression.
However, even outside of examination contexts, the ending pro-
Teaching uniting at advanced levels 355
vides a useful check for the reader and re-establishes the major
issues and goals of the text.
In working with more complex writing tasks and assignments, stu-

dents need to consider more than information sources and the
general organization of the text; they also need to consider how
arguments and positions are presented and how they can be made
more effective. There are two im portant issues in this regard: (1)
using persuasive information and arguments with factual support
(rather than unsubstantiated opinion), insightful examples, and
illustrative details; (2) providing students with feedback from
expert sources that will make them more effective writers. The five
themes in this section consider both of these concerns as they
address: (1) distinguishing fact from opinion, (2) developing
appropriate detail, (3) responding to student writing, (4) present-
ing texts for class revision, and (5) teacher-student conferencing.
12.5.1 Distinguishingfactfrom opinion (11)

Effective academic writing typically marshals a variety of evidence
to support the main points of a text and the veracity of the writer’s
arguments. A major change in advanced academic writing is the
recognition of weak or spurious arguments, or of information and
arguments that are supported by opinions rather than by factual
information. While inappropriate reporting of information is a
serious concern for academic argumentation, the problem of sep-
arating fact from opinion is a broader and more basic concern.
All students, whether LI or L2, need to develop a clear sense of
the difference between fact and opinion. Arguments and informa-
tion supported by a belief system, a set of personal experiences, or
the assertions of other students do not provide evidence that
would be considered factual. Rather, students need to recognize
and use factual information as illustrated in the following: (1)
carefully controlled case study information; (2) survey and ques-
tionnaire data which includes some information on the
methodology of data collection; (3) quantitative data reported
from reliable sources (e.g. government documents, census infor-
mation, major reference publications); (4) empirical experimental
results which include some explanation of the methodology

employed; (5) visual display resources such as maps, charts, dia-
gram, figures, tables, etc., derived from recognized reference
sources; (6) carefully presented ethnographies and participant-
observer discussion; and (7) theoretical arguments which com-
bine informational resources with support from multiple sources
of evidence.
Having suggested useful resources for factual information, it is
im portant to acknowledge that all information and all facts can be
seen as ‘constructed’ in some sense. However, there are clearly dif-
ferent degrees of control over the ‘constructing’ of fact, as was
discussed in Chapter 6 (referring to Bazerman’s (1983, 1988) dis-
cussion of ‘passive constraints’ on theorizing and actual
information). It is also true that all academic writing retains a
degree of rhetorical framing (even in the most seemingly informa-
tional of texts, such as professional science articles or annual
business reports to stockholders). Nevertheless, it is im portant to
see the use of factual information and the rhetorical framing of
the information, or a set of arguments, as two distinct issues, and
students need to be made aware of this difference.
There are also situations in which L2 students will have pro-
found difficulties with the distinction between opinion and fact.
In particular, students from certain cultures have difficulty under-
standing this distinction since polemic presentation may be, in
their cultures, a valued m ethod for organizing an argum ent and
providing information. For such students, teachers need to
address explicitly the issue of fact versus opinion, analyse differ-
ences between fact and opinion in various readings, and explore
student composing processes in the development of argumenta-
tive or persuasive writing.
Teachers are aware of a num ber of well-known techniques
which can raise awareness of the uses of factual information, and
which can help students work more effectively with facts. For
example, teachers and students can make lists of causes, evidence,
examples, data, and theories which relate to a specific issue. These
resources can then be organized in terms of their factual empha-
sis, and this listing may become a resource as students develop a
visual display or outline of their draft and begin to write.
As a second example, teachers and students can explore a num -
ber of texts that they have read for a thematic unit or a project
and examine the sorts of evidence used. These sets of evidence
can then be classified as empirical fact, theory, inference, or opin-

ion. Discussion of the choices should help to raise students’ levels
of awareness.
12.5.2 Developing appropriate detail (12)

In addition to a concern for using factual information appropri-
ately in writing, there is a need for writers to provide vivid detail -
the striking analogy and the effective illustration. Used in con-
junction with a variety of factual resources, these details build a
relation between the writer and the reader, a sense of involve-
ment, as Tannen (1989) would argue.
The use of figures of speech includes similes and metaphors -
ways of making less familiar information more understandable
through example and illustration - and vivid detail can also be
drawn directly from the data resources being reported. Typically,
when there is extensive data display and discussion, the detail that
might make the information more manageable is missing. Good
writers, in contrast, know when to focus on a small segment of the
data and when to make its appropriateness and importance clear
for the reader. Thus, the detail can serve as the most persuasive
aspect of a complex and demanding written presentation.
There are a num ber of ways to develop a student’s sense of
effective detail in academic writing. For a writing project, students
can begin by making lists of details, examples, and illustrations
that they can remember from various sources and these lists can
then be combined into a master list. Within a few weeks, this list
should be reviewed two or three times to allow new items to be
added. The teacher can also rem ind students to look for details in
their readings and point them out to the class. The teacher and
students can also explore ways in which details, examples, etc.,
vary from one academic genre to another. As another general
activity, students can review readings used in class for the variety of
details in these texts. These can then contribute the basis for a tax-
onomy of effective details to be used in writing.
12.5.3 Responding to student uniting (13)

Teachers need to provide helpful and informative feedback to stu-
dents at a variety of points during the writing process; these
feedback mechanisms will also need to vary with the teacher’s
purpose and the individual student’s progress. Students need

both informal and formal feedback from teachers, whether the
feedback be assistance with brainstorming, informal feedback dur-
ing class discussions, or more specific and formal feedback on
written drafts.
From the beginning of a writing task, the teacher can provide
assistance with brainstorming and organizing, reviewing heuris-
tics, and exploring the extent of content information available.
Teachers can provide feedback during class discussions, in reac-
tion to free-response writing on a topic that students might
pursue, in response to initial outlines prepared in groups, and in
reaction to class lists of student-raised topics and informational
resources. Teachers can remind students of the various ways they
should consider organizing information, and the various places
they can look for more information. Much of this initial feedback
will be informal and will not involve specific evaluation mechan-
isms; however, the teacher can create activities which assess early
efforts to develop drafts and give formal evaluative feedback.
Given the typically evaluative nature of most academic writing, the
pressures of formal evaluations might best be postponed until
later in the process unless students need extrinsic motivation.
The feedback process commonly involves reading the first
drafts for content and organization (e.g. factual information, logic
of presentation). In many classes, student peer-revision occurs in
conjunction with the first draft; however, the first draft provides a
better opportunity for the teacher, as expert writer, to help stu-
dents consider fundamental issues in the organization and
content of the writing without becoming too committed to the
original writing plan. During feedback on the first draft, and dur-
ing subsequent class discussions, the teacher can provide
supporting comments so that students may attend to important
features of their writing (e.g. information structure, and details of
grammar, logic, and fact). Students can then form groups, follow-
ing class discussions and individual feedback on rough drafts, and
identify three or four concerns which should be given priority
while working on the second draft.
The teacher needs to decide how feedback can be provided
most effectively for the final draft (or the draft that will go into a
portfolio). The teacher has the option to give both instructional
and evaluative feedback on final drafts. The instructional feed-
back might point out important issues that were discussed but that
were not addressed adequately in a given final draft. The teacher

can also ask students to write a reflection page on their experi-
ences writing the paper, on the feedback from the teacher on the
final draft, and on changes that the student might make in a later
revision or for another paper. A variety of feedback options are
examined in detail in Chapter 13.
12.5 A Presenting textsfor class revision (14)

A common, and very effective, technique for teacher feedback is a
whole-class critique of one or two texts written by students in the
class (or from another class). The technique also provides oppor-
tunities for quick reviews of previous issues and allows the teacher
to think-aloud through ideas for revision in front of the class.
The most im portant concern when using student papers for
whole-class revision sessions is to design the activity in a positive
way. Students need to be convinced that feedback from the class
offers additional benefits to feedback from a peer group. The
teacher needs to keep the authorship confidential, preferably pre-
senting a retyped version for the class. In addition, the students
need to be assured that they will all have some piece of their work
reviewed in this way, so all students will have the same experience
and no student will be singled out inequitably. In cases in which
emphasis is to be given to some particular aspect of writing, such
as opening paragraphs, the teacher can choose to present three or
four introductory paragraphs for class work. In this way, all mem-
bers of the class know that three or four students have a part in
the exercise, and no one student is receiving all the revision
responses.
Teacher-fronted class revision activities can focus on a wide vari-
ety of topics and issues. First, the teacher can focus on ways to
develop a rough draft and on concerns for overall organization.
The teacher can read through a student’s text, then put the text
aside and outline how he or she would do it differently using the
same information. This think-aloud process allows students to see
alternative approaches and to ask questions about choices made
by the teacher. The teacher, in turn, can comment on choices
made and explain motivations guided by audience, purpose, con-
tent, and task. Students can take notes while the teacher is
thinking aloud and then meet in groups to decide on the two or
three most im portant changes or considerations proposed by the
teacher. The class discussion to follow will allow the teacher to

emphasize important points for drafting and revising. This feed-
back should help students not only in writing but also in
developing their own feedback abilities during peer revision ses-
sions.
The teacher can also guide a class discussion of a student’s
paper (cf. Beach 1989). Students may begin by reading the paper
silently, noting places in the text at which a comment may be
made, and selecting two points to make. The teacher can read
the paper aloud so that all students hear the rhythm of the text
and recognize problem spots that they might have overlooked.
The teacher can also offer guidelines for the critiques so that the
comments are taken constructively. For example, the teacher can
ask all reviewers to begin with some positive comment or some
interesting idea that is suggested even if it does not actually
appear in the text. Students can then describe something that
they think would improve the text. The goal is to focus more on
larger issues such as organization, audience, content, purpose,
coherence, detail, illustration, etc. If a student would like to com-
m ent on structure, then the student should propose an
alternative structure. A final five minutes should be provided to
enable students to write out their critiques on their copies of the
text which the teacher collects and later gives, as an entire class
set of ideas for revision, to the student whose paper was cri-
tiqued.
These class revision techniques provide opportunities for stu-
dents to work with meaningful texts and to engage in constructive
commentary in peer writing (see also Chapter 13).
12.5.5 Teacher-student conferencing (15)

Teacher-student conferences are an effective, if time-consuming,
alternative to class feedback and written feedback on individual
papers (Beach 1989, Freedman 1987). Such conferences can be
used as a regular feature of the writing curriculum, with a num ber
of regular conference times scheduled, or they can be conducted
two to three times during a semester.
Conferencing, when it is possible and feasible in a given instruc-
tional context, should provide students with unique learning
opportunities. Recent research on cooperative learning suggests
that intensive interaction with teachers will improve learning
(Fathman and Kessler 1993, Slavin 1990). Similarly, a Vygotskean

interpretation of conferencing would argue that conferences will
lead an apprentice writer to perform at a level not possible with-
out expert assistance (Newman et al 1989, Rogoff 1990).
Collaborative learning research in composition also suggests that
the writing conference is a productive way for students to learn
(Flynn and King 1993, Golub 1988). Finally, research on tutoring
in educational contexts suggests that intensive interaction is a
powerful means for learning complex higher-order skills (Wasik
and Slavin 1993).
Conferencing, however, is no panacea for writing success. There
are a num ber of issues which have to be attended to if conferenc-
ing is to be effective. First, the teacher needs to know how to share
the power relation in the conferences and not be overly directive.
Second, the teacher must listen and give the student time to con-
sider answers. Third, the teacher needs to know when particular
information and assistance will be most useful. Fourth, the
teacher must structure the conference so that the ground rules
are clear for the students. Fifth, the teacher has to make students
feel at ease and confident, and also have students recognize that
the conference is a time for work. To have a conference work well,
the teacher should discuss conference methods with students,
describe expectations and goals for conferencing, explain why it
will improve students’ writing, and provide a set of written guide-
lines for what should happen in conferences.
In conferences held on a regular basis, students should recog-
nize the shared responsibility for this interaction. Students should
bring in-class and homework papers, questions from assignments,
papers from other classes, and, more generally, be willing to dis-
cuss their progress in the writing class. For conferences held once
or twice during a writing course, the dynamics of the conference
will also include discussion of progress in the class, feelings and
attitudes towards the class and the writing tasks, and expectations
for learning to write from the class.
Despite the greater complexity and time requirements of con-
ferences, there are a num ber of advantages that make them
serious options for writing instruction. Students receive, and give
themselves, detailed feedback in situations in which they feel free
to ask more questions and in which the teacher can probe in more
detail. Teachers can also provide a range of important feedback
for students:
1. Carefully evaluate writing progress

2. Examine past writing assignments
3. Provide help with brainstorming, topics, details, etc.
4. Examine and discuss portfolio progress
5. Introduce more detailed heuristic schemes with carefully
guided practice
6. Practise those skills that students should be using in class peer-
revision sessions.
12.6 Working with different types o f writing

(themes 16-20)
At advanced levels, it is im portant that students expand their

range of writing abilities into additional genres and also develop
specific skills useful for more demanding academic contexts. This
section examines five genres which are im portant for the develop-
m ent of advanced writing abilities: (1) critical reviews, (2)
argum ent writing, (3) case study writing, (4) interpretive essays,
and (5) the research paper.
The research paper represents, in some respects, the final stage
of advanced student writing. While the research paper is some-
times viewed as an abstract and time-consuming exercise in
writing classes, it is, in fact, the best opportunity that students have
to grapple with the full complexity of academic writing and acade-
mic expectations. Genres which are even more complex or
specialized than the ones discussed in this section typically only
emerge in a graduate student-faculty or apprentice-m entor rela-
tionship that represents initiation as an active member of an
academic discourse community. The more complex genres typi-
cally indicate sufficient expertise in writing that the term ‘writing
student’ is no longer appropriate and mastering these advanced
genres does not typically include the need to develop general writ-
ing skills further.
12.6.1 Critical review writing (16)

The ability to analyse a text and provide a critical review is a major
academic task in most fields of study. In advanced learning con-
texts, it is not assumed that course lectures provide all the
information that students are expected to learn. The most com-

mon source of additional information is through reading other
text material. The additional content gained through reading is
not expected to be inert, but to be used and integrated with other
academic content. The most efficient way to assess the reading
material is through a critical review. It provides a summary of the
information, and forces the reader to approach the material with
a sceptical and enquiring manner. Many courses, particularly at
the graduate level, make regular use of this genre, and it is an
im portant one for students to have practised.
Students are also expected to evaluate longer readings and
books through the critical review essay. In simple form, the review
essay first summarizes the author’s work as neutrally as possible.
This is followed by the author’s position and intentions, both
overt and covert. The review then typically describes the text’s
strengths and weaknesses, which usually indicates acceptance or
rejection of the text. The review often ends with a reviewer’s inter-
pretation of the value of the text and its appropriateness for other
readers.
Before students can develop skills with this genre, they must
read books and longer articles which raise important issues and
discuss these issues with some degree of complexity. One good
way to introduce students to critical reviews is to have them read a
book and find a set of reviews of that book. The teacher can also
provide a commentary on a published review that is handed out to
the class: the teacher can include an outline of the organization of
the review, a discussion of the highlighted issues that motivate the
review, a critique of the review’s summary as well as the shaping of
the summary to meet the author’s intentions, and an evaluation of
the reviewer’s rhetorical strategies.
Students can also find a review of a book they have read and
outline the organization of the review. Students can point out
places in which the reviewer indicates a positive or negative evalu-
ation, issues which guide the review, the summary of the book,
and the analysis of the book in terms of the issues highlighted.
Students can then work in groups or as a class to write a review of
a book read by the entire class. In a brainstorming discussion, stu-
dents can outline a plan for the review, list im portant topics in the
book, offer genuine criticisms of the book, and nominate issues
which students could use to focus the review.
12.6.2 Argument uniting - taking a position (17)

A standard genre that students are expected to master is the argu-
m ent paper, in which they take a position on an issue. Argument
papers typically propose a thesis and then support the thesis
through a set of evidence (facts, logic, examples, etc.). Writing
assignments in which students are to take a position require them
to learn how to consider both sides of an argument, describe the
alternative positions effectively, note the strengths and weakness
of both positions, and provide a persuasive explanation for choos-
ing one position over another.
The first requirem ent for having students write argumentation
essays is the availability of a set of information which genuinely
establishes alternative positions or points of view. This is easiest to
do in the context of a content-based unit, a project, or a set of read-
ings on a topic, as these formats raise a num ber of issues which
should allow different perspectives. These alternatives should be
explored in group discussion with students taking notes and the
teacher summarizing the main points of the discussion on the
board. Students should then seek additional information which
can support the positions for which they would like to argue.
Information collected should be reformatted as a set of sup-
porting arguments. This set should be discussed in class, and the
teacher should provide optional frameworks for presenting the
arguments. Teachers also need to point out that a strong argu-
m ent is one which includes discussion of weaknesses with the
preferred position, but which responds to these weaknesses with
reasonable counter-arguments.
Argument papers can be the source for, or the consequence of,
a class debate on a particular topic. In either case, the work done
collecting information for the essay becomes an important
resource for the debate. Argument papers can also be combined
into a more elaborate report which can be sent to an appropriate
outside institution, for example, a local authority.
12.6.3 Case study reports and simulations (18)

Students are often asked to explore an issue or a situation for
which there is no simple solution (see also Chapter 11, subsection
11.7.3). These situations typically are presented as a set of prob-
lems or decisions to be made on the basis of information available.
This is a common teaching approach in a num ber of academic
disciplines and it can be a powerful means for introducing stu-

dents to the goals and methods of a discipline.
Case studies and simulations can involve a range of possible sce-
narios: the building of a hospital or a freeway; the use of pesticides
or fertilizers; water rights or dam construction; government fund-
ing decisions; the costs and benefits of business relocation into the
local community. These and many other scenarios can be
designed to explore issues associated with the content of an acade-
mic course, including a content-based writing course.
Such case study reports and simulations force students to take a
position in writing based on specific sets of data, specific m ethod-
ologies for analysis, and specific alternative options, all of which
typically have advantages and disadvantages. The writing of simu-
lation and case study reports constrains the options available to
students and teaches students to work with methods of analysis
which are preferred by some academic disciplines (i.e. some dis-
course communities).
12.6.4 Creative and interpretive writing (19)

A further im portant type of writing, or perhaps a num ber of types
of writing, could be classified here under the label of creative and
interpretive writing. Students at advanced levels might engage in
at least four types of writing which might be considered creative or
interpretive: (1) creative fiction and poetry, (2) creative non-fic-
tion, (3) the interpretive critical essay, and (4) the interpretive
theoretical synthesis.
Creative fiction and poetry is, of course, of interest to students
who wish to pursue writing creatively. While students seeking
degrees in English come immediately to mind, there are, in fact,
many other students who would like to develop some abilities to
write short stories, poetry, screen plays, etc. These students are
interested in learning to write in new genres. Moreover, the work
of Heath (1993, Heath and Mangiola 1991), studying perfor-
mance-arts writing of disadvantaged youth, shows that creative
writing can be an im portant resource for developing writers, par-
ticularly for students who are considered to be at-risk.
Creative non-fiction provides another option for writing. As a
growing field of professional writing, creative non-fiction typically
refers to travel writing, to reflections on personal experiences, or
to interviews with persons who are not explicitly named for a
variety of reasons. Both forms of writing typically include a narra-

tion but are not fictional. Students may also be encouraged to
write creative non-fiction as part of certain academic courses (e.g.
anthropology, sociology) as well as in a writing course. In some
sense, creative non-fiction is the mature version of expressive writ-
ing done at earlier stages of writing instruction.
The interpretive essay is a genre which is common in humani-
ties disciplines. In many courses, students are expected to read
literature or philosophy and provide an insightful interpretation
of what the reading means to them. This is typically a difficult
genre to master since it often presumes a wide range of back-
ground knowledge, a finely honed sense of intertextuality, an
ability to read on multiple levels and to recognize symbolic struc-
ture in a text, and a knowledge of ways that critical theory is used
in interpretation. While it can be seen as a genre that extends
beyond the writing student, in fact students are often asked to
work in this genre whenever there is an emphasis on literary, cul-
tural, religious, or philosophical texts. Unfortunately, it is very
difficult to teach students this genre without an intensive appren-
ticeship and attendant practise.
The interpretative theoretical synthesis is a different type of
writing, one which synthesizes information and arguments from a
num ber of sources and infers a larger theoretical foundation for
the sources. Often this type of writing summarizes a num ber of
research studies and interprets them in a way that attempts to
unify the various results. This type of writing is fundamental to
academic activities which look to find patterns and generalizations
in sets of information, perhaps the most fundamental purpose
underlying any academic enterprise. It is not uncommon for stu-
dents in writing classes to look for patterns and generalizations
across sets of readings and, in effect, build a theory. Such an
approach to writing is the foundation also for Bartholomae and
Petrosky’s (1986) writing curriculum, Facts, artefacts, and counter-
facts. It is also a basic goal for writing curricula which seek to
problematize commonly held cultural assumptions, and, as such,
are seen as Freirean approaches to writing instruction.
12.6.5 The research paper (20)

The research paper is a well-known option for advanced writing,
and one which has moved in and out of favour at different times
in writing pedagogy. For advanced writing, students need to learn

how to work with sets of information, how to synthesis knowledge
from multiple sources, how to use and cite resources appropri-
ately, and how to explain a complex issue in a manageable format.
The research paper (or research project) can provide a reason-
able way to promote these goals, and promote exploration and
enquiry on some topic of interest to students.
The research paper provides a manageable type of writing
which can incorporate other genres presented in this section. At
the same time, the research paper is sufficiently general that stu-
dents can carry it out without the task becoming overly difficult.
Students can be expected to explore some topic of interest related
to the course curriculum and teachers should provide guidelines
and assistance for focusing the topic, for information collection,
and for using an appropriate format for the paper itself.
In most cases, the research paper involves an extended investi-
gation and collection of outside resources on a topic. Teachers
should offer guidelines throughout the process of writing the
research paper, and can discuss various possible topics and varia-
tions on a theme. They should schedule library activities which
teach students how to make use of library resources and should
also provide a time-table of assignments which will keep students
on task. Many of the brainstorming, drafting, and writing activities
that are useful for shorter writing activities are also im portant for
writing the research paper.
The research paper can assume various formats and conven-
tions; nevertheless, it is a widely recognized genre in academic
courses and in advanced writing classes. It is also an im portant way
to introduce students to synthesizing and information reporting,
basic activities in all academic classes.
12.7 Extending writing (themes 21-25)
This section raises a num ber of issues which reflect differing

demands of writing in different disciplines. How these various
concerns can be incorporated into a writing curriculum, or the
extent to which they can be incorporated, is the focus of the final
five themes; these include: (1) content-based writing and writing
across the curriculum, (2) discourse communities, (3) working
with visual displays, (4) working with media, and (5) style and the
individual writer.
12.7.1 Content-based uniting courses and uniting across

the curriculum (21)
Throughout the previous three chapters, as well as in a num ber of
the other chapters, a basic orientation to writing has been in
terms of content-centred instruction. Such an approach does not
guarantee successful instruction, nor does it make the issues and
problems in writing instruction any less complex. It does, however,
offer genuine academic learning experiences in the writing class-
room and give writing instruction a realistic content: while writing
itself could be seen as content, and it often is in writing textbooks,
most students are not persuaded by such rationalizations. For
advanced writing instruction to be effective, writing itself must be
used in much the same way that it would be used in any academic
context. This realistic context is what content-centred courses can
bring to writing instruction.
There are many ways to carry out content-based writing instruc-
tion. One common option at advanced writing levels is to use a
major theme with which students can connect. The thematic cur-
riculum by Bartholomae and Petrosky (1986) presents a detailed
discussion and rationale. Their curriculum, centred on adoles-
cence, allowed students to bring expertise and authority to the
issues; at the same time, the curriculum emphasizes dialogue and
discussion, extensive reading, many writing activities, a thematic
progression from self to other’s views of adolescence, and a pro-
gression in writing from personal to academic theory building. A
num ber of other topics can be used in similar ways, giving stu-
dents some amount of expertise, while also sequencing writing
towards more complex academic tasks. Such themes might
include education, technology, the future, culture, academic cul-
ture, etc. (See also Kutz et al. 1993.)
A second curricular format is to develop subtopics around
loosely related sets of content material. In this format, the curricu-
lum is centred around three to five topics, some of which may be
connected. The connections need not be made across all topics,
and topics can, in part, be the choice of the students, assuming
that the teacher has collected information on a num ber of
options for units. Aside from the need to develop a motivating
sequence of writing and reading activities for each topic, the
teacher should also find ways to strengthen relations across topics.
This can be done by links across main topics at times, but it is also
possible to connect topics in other ways. For example, a set of
subtopics (apart from the main topic) may focus on a famous per-
son, an im portant set of consequences from an event, an issue that
cannot be easily resolved, or some abstract concept such as
responsibility, technology, etc. The recurrence of sub-issues across
topics can be exploited to build coherence into the curriculum,
and a set of connections can bind the content of the various top-
ics. This option should not, however, be seen as a rationale for a
composition ‘reader’. Such composition texts do not allow for the
open development of a topic, nor do they lead to student input or
motivation. Many ‘readers’ tend to be frozen in concept and liter-
ary-classical in orientation: the spontaneity and creativity needed
for interesting content/topic development is generally missing
from these texts.
A further option for thematic writing curricula is the often dis-
cussed notion of Writing Across the Curriculum (WAC). The goal
in this case is to bring the writing instruction to the content class
rather than the reverse. While there are a num ber of good ratio-
nales for WAC, there is no clear evidence that it promotes
writing/learning in ways beyond writing-class instruction
(Ackerman 1993). Of course, the more opportunities that stu-
dents have to write, the better they should become as writers in a
discipline. While WAC, in theory, is a good idea, it is not guaran-
teed to improve writing instruction. In WAC contexts, introducing
sets of writing techniques, exercises, and activities which are not
integral to the content curriculum will not lead to any significant
improvement in academic writing. In fact, content-area teachers
often resent the intrusions that many writing techniques represent
when they are not well-integrated into a curriculum, and when
suggested by the well-meaning writing specialist without careful
consideration of the goals of the content-area courses (Kaufer and
Young 1993, Spack 1988).
In spite of these real concerns by content-area teachers, the
ideal of students engaging more intensively in writing across a
range of courses is a goal that should be a priority of academic
programmes everywhere. Many content teachers are well aware of
the genre-specific writing demands of their disciplines. The atten-
tion to writing, while also focusing on the real content of a
discipline, provides opportunities for meaningful writing activities
which should motivate students. An effective WAC programme
could have a profound effect on writing development, but it is
difficult to implement and requires considerable negotiation and
coordination (McLeod and Soven 1992).
12.7.2 Discourse communities (22)

The significance of the concept of discourse communities for
writing development has been discussed in Chapter 5.
Nevertheless, it is im portant to point out the practical implica-
tions of this concept for writing instruction. Academic discourse
communities are typically discipline-specific groups of academics
who share sets of theoretical, methodological, and conventional
notions about their discipline. One im portant issue is whether or
not such communities see initiating students into their discourses
as a goal when teaching advanced students. In a recent book
examining writing in various undergraduate disciplines,
Walvoord and McCarthy (1991), in fact, found this to be a pri-
mary consideration of many faculties. Many faculties are
motivated to have students undertake writing tasks which model
real writing activities carried out by professionals in the outside
world. Thus, the notion of the discourse community is an impor-
tant one for advanced writing instruction.
Discourse communities provide a resource for real and realistic
writing activities. Specific discourse communities make extensive
use of different genres and conventions which will be important
for students to master. Some of these genres and conventions will
be specialized, but others will generalize across a num ber of disci-
plines. For example, the ability to form a coherent and
well-supported set of arguments from multiple, and sometimes
conflicting, resources is a skill that is valued in a num ber of disci-
plines. Similarly, the ability to write a critical review of a major text
is widely valued. A num ber of disciplines also value the ability to
construct a set of valid recommendations from case study scenar-
ios and simulations of problems. In these respects, the study of
writing uses in different discourse communities can have a strong
impact on the design of writing courses and the integration of
WAC with various academic discourse communities.
Writing specialists must recognize the genres and conventions
of writing in different disciplines and look for ways to support
writing activities which both improve writing and prepare students
to carry out activities valued by a discipline.
12.7.3 Working with visual displays (23)

A skill in advanced writing that is valued in many academic con-
texts is the ability to write with visual displays (e.g. incorporating a
table of data into the prose discussion). From earlier sections, the
notion of a visual display can mean a visual representation of the
text organization used to generate an essay. In this section, the
concept of visual display is extended to mean also: (1) an alterna-
tive (e.g. visual) interpretation of the data discussed in a text, or
(2) non-prose displays of information in the writing which add
information and which may be partly explained in the accom-
panying writing. Both situations represent im portant writing
issues for advanced students.
On one level, students in writing classes are often asked to write
essays, informational reports, and arguments based on visual mater-
ial: this sort of writing is sometimes done as a writing activity and
sometimes as a test and evaluation of knowledge. Less formally,
students write from displays when they write from semantic maps,
time-lines, self-constructed charts and tables, and draft outlines to
help organize their own information. There are no simple guide-
lines for these types of activities since various visual displays can
invoke the full range of organizational options in writing. It is
im portant to give students practice with assignments of these
types, and such activities can be integrated into project work and
summaries of content-based units. The writing from various visual
displays can also be used to show the options available to students
when converting outlines into connected prose.
On another level, students are sometimes expected to integrate
visual material into their own prose writing. This type of integra-
tion between visual and linear prose is a different m atter from
using a display to provide information for writing. Visual displays
accompanying writing can be used in a num ber of ways. The
visual material can add further information not explained in the
text. Alternatively, visual displays can be coordinated with the
prose of texts and selected aspects of the display can be referred
to explicitly. A third option occurs when the visual display sum-
marizes the information in the text. In this case, the display
completely overlaps with the text and functions primarily as an
additional comprehension support for the reader. All three
options may be required of academic students in certain discip-
linary contexts.
Students may also be expected to create their own visual sup-
ports for their writing. This is another reason to incorporate visual
displays in brainstorming, planning, and organizing writing.
These planning and organizing formats then become potential
visual displays to accompany texts. Again, there is no simple recipe

for learning to use visual displays in writing, but consistent prac-
tice helps. In many cases, these writing activities can be
incorporated when advanced students know that the skills are val-
ued in their future academic disciplines. At the same time, it is
im portant to recognize that the adept use of accompanying dis-
plays in writing often makes a notable impression on readers.
12.7.4 Working zvith media (24)

In the previous chapters on writing instruction, a num ber of writ-
ing techniques and activities have made use of media sources and
used media activities primarily as the impetus for writing (Chapter
10, subsections 10.5.6, 10.7.1; Chapter 11, subsections 11.5.4,
11.6.3, 11.7.5). The uses of writing for media purposes may also be
an im portant extension for advanced writing. As students develop
writing skills and are asked to work with more specialized tasks,
they may be asked to write text for a media project or write prose
in conjunction with a media resource as a multimedia project.
At a fairly basic level of advanced writing, students can be asked
to script a video production or an act from a play. Students can
script and produce a debate or interviews with famous people
from the past. Students can produce a TV talk show in which two
‘correspondents’ take a conservative view, and two other ‘corre-
spondents’ take a more liberal view. A news feature report is
another project which can follow an extended content-based unit
and which can draw together different sets of information into a
coherent documentary.
At more advanced levels, students can also produce a news-
paper or a newsletter with information about the school, local
events, class and group projects, and biographies of class mem-
bers. The class can produce a book or an extended report which
can be distributed to other classes and to other schools. For exam-
ple, in Bartholomae and Petrosky’s (1986) writing curriculum on
adolescence, the class produces a set of student autobiographies
which become the text data for discovering general patterns in
adolescent experiences. Students can also write a report or a news
feature which can be submitted to the school newspaper. It is
im portant to recognize that, while students are at an advanced
level, they still have much to gain by combining academic writing
with other forms of media.
12.7.5 Style and the individual uniter (25)

Advanced students will need to develop a strong sense of style in
their writing. This individual style will include a recognition of
how strongly a claim should be asserted, a knowledge of how and
when to speak as an expert and to invoke authority, and a sense of
appropriateness of form and convention in writing which matches
the genre and situation of writing. The ability of students to find
their own ‘voice’, and to develop a voice that is appropriately aca-
demic is a difficult task; it requires extended practise with
expository and argument writing, and a long-range time commit-
ment. The development of an appropriate academic voice is also
likely to require some amount of discipline-specific apprentice-
ship. While many writing texts talk about students developing a
voice in their writing, this usually means a personal expressive
style that is similar to the way students relate ideas orally. This
notion of voice will not typically translate into an individual acade-
mic style, demonstrating authority and confidence while writing
for academic purposes.
The development of an individual academic style requires both
a sense of stylistic options available to writers and a clear sense of
valued academic and formal conventions of writing. One way to
raise awareness of academic style would be to compare two writers
on the same topic and explore ways in which the two texts differ
stylistically. Two or three of these awareness exercises will allow
students to see a few advanced options in writing such as the
extensive use of appositive parenthetical clauses and absolutive
clauses, variation in modality for modifying assertions, modifica-
tion with final participial clauses, complex patterns of
nominalization, the use of a variety of punctuation options (e.g.
dash, colon, semicolon), and the use of passives and impersonal
structures. Certain of these features can then be explored in the
revision process with students’ texts. Another option for raising
student awareness lies in comparing an academic written version
of a text to an oral version (see Christie 1992, Kutz et al 1993,
Martin 1993).
Students need to become aware that the formal conventions of
writing are not only patterns which match academic constraints
and the expectations of informed readers, they also signal a will-
ingness to work within disciplinary expectations and to project an
academic persona. While the need to edit texts carefully and
produce ‘clean’ texts is im portant for evaluation purposes, this

goal only indirectly reflects the real impetus for careful editing
and appropriate formatting. Academic writers are careful about
formats, conventions, and error-free texts because they know that
other academics view such carefully produced text as a sign of
respect and, for better or worse, a mark of careful scholarship.
The impressions given by the form of a text are not too impor-
tant in lower levels of writing instruction, but they are critical in
advanced writing. If, indeed, a student is an advanced writer, and
is aware of the conventions and expectations of readers, then a
poorly edited and stylistically weak text indicates some degree of
contem pt and disrespect for the reader. In the final analysis, there
is no avoiding the need to combine insightful organization and
stimulating content with a careful attention to the form of the
text. It is the combination of these features of text which carries
the ideas, intentions, and attitudes of the writer.
12.8 Relating writing instruction to writing theory
The themes addressed in this chapter, while focusing primarily on

the advanced student writer, also, in many cases, are adaptable to
students at lower writing levels. This and each of the previous two
chapters present, as an organizing format, 25 themes for writing.
However, many themes are applicable across all levels of writing
instruction, for example:
■ the writing process, in various forms
■ writing for various purposes, for different audiences, and for
different genres
■ different ways to incorporate informal, personal, and creative
writing
■ the need to focus on conventions and formal aspects of writing.
Students at all levels need to generate and organize a large
amount of information for writing, and they need to recognize
appropriate patterns for organization. Students must also be able
to combine reading and writing, and use many sources of informa-
tion.
A num ber of general curricular guidelines, suggested in
Chapter 9, provide the foundation for the discussion in this and
the previous two chapters. Some of the guidelines are noted
above; others, which can be observed across the various themes

and instructional discussions, include an equal emphasis on prod-
uct and process, the need for students to develop strategies for
generating content and organizing information, the need for
many types of feedback, an awareness of the uses of writing and
the various purposes that writing serves, and a willingness by the
teacher to provide many opportunities for writing - some of which
will be non-evaluative (see Chapter 9).
In these three chapters on writing instruction, one can see an
interdependence among activities, plans, intentions, and formats
supporting overall guidelines for writing curricula. The critical
notion underlying the many themes presented is that they are
guided by theoretical concerns and research, though not in any
direct or simplistic way. One cannot translate theory directly into a
neat package of activities and exercises; the multiple worlds of stu-
dent writers are much too disparate and complex for neat
solutions. At the same time, it should be clear that the generaliza-
tions suggested in Chapters 8 and 9 are reflected here and in
Chapters 10 and 11. The fit between general curricular guidelines
and instructional practices suggests that theory and instruction
can be integrated, though one should not expect a simple proce-
dural blueprint for instruction. There are many interesting issues
and problems associated with writing instruction, but there are no
simple solutions.
12.9 Conclusion
This chapter, together with Chapters 10 and 11, has discussed a

wide range of ideas for writing instruction. They could be read
independently of the rest of the book, but that is not our inten-
tion. As the previous section suggested, many of the instructional
concepts follow directly from theoretical perspectives on writing.
Moreover, Chapter 9 represents an effort to bridge theory and
instruction, allowing the ideas of these three instructional chap-
ters to be incorporated into a principled philosophy of instruction
and a well-reasoned curriculum plan.
This chapter, in particular, offers many suggestions for
advanced writing instruction which should apply across many aca-
demic contexts, assuming instructional settings from the end of
secondary schooling to the beginning of post-graduate education.
It is also true that advanced writing instruction begins to require

specializations that are less amenable to general curricular plans;
but this is an ongoing dilemma in many tertiary-level writing
courses that are not embedded within a specific discipline. The
tertiary-level writing course represents that point of partition into
more specific disciplinary demands, and it produces pedagogical
tensions in that the writing course cannot by itself meet all the
needs of advanced students. The movement into writing instruc-
tion in specific disciplinary contexts and for specific purposes
also, perhaps, represents the point at which a generalized notion
of writing instruction comes to an end (see Swales 1990).
13
Responding to writing and

writing assessment
13.1 Introduction
In this chapter we review research and practice in responding to

writing and in writing assessment. In recent years, these areas have
received much more attention than they had previously, both with
respect to research and to practical implementation. At the same
time, however, the issues surrounding responses to writing and
writing assessment are still discussed more in terms of the ques-
tions raised and the concerns to be addressed than in terms of
clear answers to difficult and important issues.
Responding to writing and writing assessment have im portant
consequences for students as well as for teachers. Responding to stu-
dents ’ xjuriting can greatly influence student attitudes to writing and
their motivation for future learning. Students can be easily con-
fused by unclear, vague, or ambiguous responses and can become
frustrated with their writing progress. Alternatively, students can
be positively motivated to explore many areas of knowledge and
personal creativity through supportive and constructive responses
to their writing. Unfortunately, there is no clear set of universal
guidelines that will guarantee such a supportive and positive
experience for all students. In any given context for writing
instruction, students will differ, teachers will differ, institutional
expectations will differ, and tasks, topics, and resources will differ.
Nevertheless, we do believe that certain results of research and
practice provide im portant insights; these represent strong
options and resources for teachers who are looking for ways to
provide feedback on student writing.
The impact of writing assessment on students is apparent to any
person involved in academic learning contexts. Writing assessment,
377
whether as in-class assessment of student progress or as standard-

ized proficiency assessment, is a major determ inant of students’
future academic careers. Writing is commonly used to assess not
only students’ language skills but also their learning in many acad-
emic content-areas. For this reason, among others, the ability to
provide students, teachers, and administrators with fair and sup-
portable assessment approaches is a serious issue. Not only do
many decisions rest with writing assessments, but assessment
processes have a great impact on student attitudes and their moti-
vation for future work. All concerned in writing instruction need
to have confidence in assessment approaches - approaches that
are designed to perm it students to perform to the best of their
abilities and that are appropriately interpreted by decision makers
and test-score users.
13.2 Response to student writing
Until the emergence of the process movement in writing instruc-

tion, much feedback to students on their writing appeared in the
form of a final grade on a paper, often accompanied by much red
ink throughout the essay. This practice assumed that students would
take to heart all the mistakes pointed out, infer the reasoning
behind the grade, and be motivated to avoid the multiplicity of mis-
takes on their next writing tasks. It is now apparent to most teachers
and researchers that such an approach left many bewildered and
confused students unable to work constructively on their writing
skills. One of the major positive impacts of the writing process
approach has been the thorough rethinking of responses to student
writing. A direct outcome of multiple drafts and pre-writing activi-
ties has been the exploration of ways in which teachers can assist
students most effectively in their writing. Student revision and
teacher response has become central at all stages of the writing
process: pre-writing, first drafting, revising, and final-draft writing.
In addition to exploring the role of responses to writing at all
stages of composing, research has examined many alternative
methods for responding to students. These options run from
teacher-student conferences and response journals to peer group
responses of various types, to differing teacher responses on
students writing, to analytic-guide questions and lists, to teacher-
class discussions, and to using other audiences for feedback.
These options all offer ways to promote effective responses to
Responding to writing and writing assessment 379
student writing, recognizing that different students may react dif-

ferently to alternative types of responses.
13.2.1 Peer group responses

In the last ten years, the most popular means for providing stu-
dent responses in English LI contexts is through peer group
interactions. As an example, students would enter class with their
first or second draft completed. Students would then get together,
or be assigned in groups of two, three, or four. The students would
exchange or pass around the papers and receive comments from
the other students in the group. This process would take 40 to 60
minutes. Students would subsequently use these comments and
suggestions from their peers to write the next draft. Some peer
sessions would involve the free exchanges of reactions to a given
student’s writing. O ther peer sessions might require students to
work with specific peer-review guidelines, and peer groups might
be organized so that each student would be responsible for some
specific task or set of tasks. Alternatively, students might be
required to present a written response to each text read and to
complete response sheets to be evaluated at some later point.
In fact, there are many variations on peer group responses, and
much professional debate centres on deciding the forms that are
most effective. These debates cover such issues as the amount and
type of student training for such group work, the amount of peer
group response (per session and per semester), the m anner of
peer group response, the num ber of students involved per group,
the amount of teacher intervention, the goals set for peer group
response, the types of follow-up procedures used, etc. The specific
issues are numerous, and there is much research that remains to
be done in understanding the effectiveness and dynamics of peer
group interactions (see, e.g., Anson 1989, Carson and Nelson
1994, Connor and Asenavage 1994, DiPardo and Freedman 1988,
Freedman 1987, Gere 1987, Golub 1988, Lawson et al. 1989,
Nelson and Murphy 1993, Spear 1988, 1993, Zhu 1994).
The motivations for using peer response groups derive from a
num ber of sources. From a practical perspective, writing process
teachers and researchers recognize the need for audiences other
than the teacher, and for alternative feedback on student writing.
Responding to peer work will also involve students in each other’s
writing; in that process, students can learn the language of, and
uses for, responding to texts. From this experience, students
would then begin to see similar problems and weaknesses in their

own writing as well as use innovative and creative ideas to which
they are exposed. Students also would need to learn to accept rea-
sonable reader responses and to use them constructively in their
revisions. A num ber of arguments are offered to support the prac-
tice of peer group feedback.
Theoretical research in rhetoric, literacy criticism, and social-
construction theory converge to argue that knowledge itself is
constructed through the negotiations of a community. Knowledge
does not reside outside of a community of scholars, and knowledge
can change as the community reconstructs itself. This argument is
used to promote peer groups in writing - students discover new
ways of reading other texts and their own texts. This perspective is
also compatible with recent reader-response and post-structuralist
approaches to text analysis and interpretation: the coherence of
texts is created and recreated by readers rather than being inher-
ent in texts. The three fields, together, provide a strong rhetorical
argument for effective ways for students to read texts and receive
meaningful feedback (Bruffee 1986, Lawson et al. 1989).
A second theoretical argument for peer group responses
derives from a combination of sociolinguistics and cognitive psy-
chology. Socio-cognitive approaches to learning argue that
knowledge is best acquired through negotiated interaction. The
issue is not whether knowledge is created in this way (as claimed
in social constructionist arguments), but that knowledge is best
acquired and used through negotiated interaction (Flower 1994,
Langer 1987, Nystrand 1990). According to this view, students will
develop as writers more effectively as they engage in transactions
over their own texts and the texts of others - negotiating real
intentions to communication with real audience expectations.
A different line of support for peer group responses is one that
is only now receiving attention among many teachers and compo-
sition researchers. This line of argum ent follows from research on
cooperative learning as it is applied in K-12 educational contexts
rather than in university writing courses (Fathman and Kessler
1993, Kagan 1992, Slavin 1990). In cooperative learning approaches
to instruction, groups are used in various ways to collaborate on
tasks and projects that promote learning and typically involve
extended writing.
Peer group approaches are typically carried out in a num ber of
distinct ways. In many cases, students are given little overt direc-
tion except to read each other’s papers and comment. In other

cases, teachers provide occasional training sessions with the whole
class on how to take part in an effective peer group. In many peer
group sessions, students can either bring their own papers and
pass them around, or they can bring copies for each member of
the group so that other students can read silently and write com-
ments on the copies. In some cases, students read their own
papers aloud while others listen; in other cases, a member of the
group reads another student’s paper aloud. Responding students
can provide comments orally after the reading, or they can give
written comments to the student who wrote the paper.
A more structured alternative is to provide each group member
either with a worksheet for analytic feedback or with a guideline
containing options for feedback. In the former case, the students
work through the sheet and typically focus on mechanics, gram-
mar, and word choice, as well as on the coherence of the main
argument. Guidelines with options for feedback may include sets
of questions that any member of the group can ask other mem-
bers; these questions cover organization, planning, the frequency
and distribution of main ideas, examples used, good points of the
papers, perceived weak points of the paper, introductions, conclu-
sions, etc. Specific sheets for recording student responses may be
handed out by the teacher, or students can write out their own
feedback based on questions and comments they offer during the
group work. Students typically use these sheets in their revisions
and attach the feedback from peer group members to the back of
the next draft handed in to the teacher. The teacher then sees the
work done by the student and the amount and quality of the feed-
back that other group members provided.
There are a num ber of composition teachers and researchers
who object to the use of explicit guidelines in peer revision. They
argue that students often use these guidelines as mechanical
check-off sheets rather than as a way to provide meaningful sup-
port. It is certainly true that students may be tempted to use
guidelines as simple checklists to minimize their roles as commen-
tors. Guidelines also cannot act as a substitute for the teacher or
for coaching on how to be an effective responder. At the same
time, there is little evidence that peer groups without coaching
and response guidelines perform better than groups with
guidelines, or even as well as supported peer groups. Teachers
and students need to recognize the uses and limitations of guide-
lines but, used effectively by the teacher, many guidelines offer

students useful options for peer responses.
Guidelines for feedback with peer group responses can involve
any of the following example formats from sources such as
Copeland and Lomax (1988), Elbow (1981), Frank (1979), Leki
(1992), Raimes (1992), Reid (1993), Spear (1988), and Tompkins
(1990):
1. Writing group response

Listener’s Compliments
I like the part where ...
I’d like to know more a b o u t...
I think your main idea is ...
You used some powerful words, like ...
I like the way you described ...
I like the way you explained ...
Your writing made me feel ...
Writer’s Questions
What did you learn from my writing?
What do you want to know more about?
What part doesn’t make sense?
Is there a part I should throw away?
Can you tell what my main idea is?
Did I use some words I need to change?
What details can I add?
Listeners Comments and Suggestions
What is your favorite part?
What part are you having trouble with?
Do you need a closing?
I got confused in the part a b o u t...
Could you leave this part out because ...
Could you add more to this part because ...
Is this paragraph on one topic?
Could you combine some sentences?
What do you plan to do next?
(Excerpted from Tompkins 1990: 86)
2. Written peer criticism: getting the reader to respond
Sayback: Ask readers: ‘Say back to me in your own words what
you hear me getting at in my writing.’
Movies of the Reader's Mind: Get readers to tell you frankly what
happens inside their heads as they read your words.
Pointing: Ask readers: ‘Which words or phrases stick in your
mind? Which passages or features do you like best? Don’t
explain why. ’
What's Almost Said or Implied: Ask readers: ‘What’s almost said,
implied, hovering around the edges? What would you like to
hear more about?’
Voice, Point of View, Attitude toward the Reader, Language, Diction,
Syntax: Ask readers to describe each of these features or
dimensions of your writing.
Center of Gravity: Ask readers: ‘What do you sense as the source
of energy, the focal point, the seedbed, the generative center
for this piece [not necessarily the main point]?’
Believing and Doubting: ‘Believe (or pretend to believe) every-
thing I have written. Be my ally and tell me what you see.
Give me more ideas and perceptions to help my case. Then
doubt everything and tell me what you see. What argum ents
can be made against what I say?’
(Holt 1992: 385; excerpted from Elbow and
Belanoff (1989) Peer Response Exercises)
3. Response questions
What is the main idea that the writer is trying to express in this
draft?
Can you find any parts that do not relate to the main idea?
Underline them.
Which part of the piece of writing do you like best?
Find two or three places where you would like more explana-
tions, examples, or details. Write questions about them.
Did you at any point lose the flow of the writing or find places
where the writer seemed to jum p too suddenly from one idea
to another? Were there any places that seemed unclear to
you as a reader?
Did the beginning capture your attention and make you want
to read on? Why or why not?
Can you summarize in one sentence the main idea of each para-
graph? For each paragraph, complete the following statement:
Paragraph 1 says t h a t ...........
Paragraph 2 says th a t ...........
Proceed in the same way for the remaining paragraphs.
(Raimes 1992: 64)
4. A process model for reading and revising

Read through the draft. Really read it. Understand what the
author m eant to say. Don’t rely on one quick reading. Read it
once to get a sense of the piece as a whole, again to look at
details like what it says and how it is put together. Don’t be
too concerned with the exact words. Go underneath them to
understand the ideas they represent.
Now take the draft apart. List what you think are the major
ideas and state what seems to you to be the author’s purpose.
(This can be done either orally, going around the group, or
in writing, during the class or prior to it.)
Under each major idea, note what the writer does to back it up.
Considering the purpose, the major ideas, and supporting
information as you see them, what ideas affect you most
strongly? What else might you need to know? What might
you do differently? What would you like to hear more about?
Next, compare the outlines and comments prepared by each
member of your group. See if the readers are consistent with
each other in identifying the purpose, main points, and sup-
porting information. See if their constructions are
acceptable to the writer. What does the writer like and dislike
about his or her draft? What does the writer regard as its
strengths and weaknesses? If your group has differences of
opinion, share them with the writer. Consider especially the
additions and changes you were thinking about. Discuss what
effects these changes might have and what the writer might
do about them. (Remember, writers can’t incorporate all
suggestions: the idea is to use them to simulate thoughts on
specific ways to revise the draft.)
(Spear 1988: 110-11)
5. Reader response worksheets
Worksheet: Reader-Writer Response
1. Writer: What one question would you like your reader to
answer, or what one problem did you need a second opin-
ion about?
2. Reader: Answer the question. Be specific. Then complete
the following statements:
a. The best part of this paper was ...
b. When I finished the essay I th o u g h t/fe lt...
c. One place I disagreed was where you said ...
d. One experience or idea I had that was similar to this
was ...
e. When you s a id ..........., I thought a b o u t...

f. One suggestion I want to make to improve this paper is
Worksheet: Group Response

1. Reader: Ask the writer: ‘How can we help you?’
2. Writer (who comes to class with notes that anticipate the
question): Indicate specific areas in which you need help.
3. Reader: Listens and take notes, then offer verbal and writ-
ten feedback and suggestions.
4. Writer: Listens and takes notes. (S/he retains full author-
ity to evaluate the advice and make the final decisions.)
Worksheet: Descriptive Response
1. Tfy'e subject/topic of this paper is ...
2. The intended audience for this paper is ...
3. The main ideas of this paper, in order, are ...
4. This essay h a s ...........paragraphs (sentences).
5. This piece of writing is written from the point of view of a
person who i s ............ (Describe the w riter/narrator - this
may be a persona, not the student herself.)
(Reid 1993: 211-12)
A final type of guideline resource for peer group feedback pre-

sents each group member with a long checklist of points to
inspect in conjunction with the essay under consideration. This
checklist, often improperly used as the only peer feedback mecha-
nism, may play some role in an editing support group. The
checklists would cover such topics as:
Topic and thesis
Composition as a whole
Reasoning and evidence
Sentence structure
Paragraph structure
Diction (word choice)
Mechanics and form
(From an example checklist in Spear 1988: 48-50)
Research on peer group approaches presents a complex set of
findings. To summarize a few main points noted in previous chap-
ters, Clifford (1981) and Nystrand (1986) provide empirical
evidence which suggests that the employment of peer groups
improves student writing. At the same time, Flower (1994),

George (1984), Harris (1992), and Spear (1988) all point out a
lack of improvement, or a variety of problems resulting from the
use of peer group approaches.
Despite these mixed results for improved writing quality, various
other benefits are claimed for the use of peer group approaches.
There are, for example, results across a num ber of studies which
show that peer groups promote collaboration, develop a better
sense of audience, improve confidence, help develop a sense of
community, lead students to consider alternative strategies, and
expose students to a variety of writing styles (Gere 1987, Harris
1992, Nelson and Murphy 1992, Spear 1988). Many of these
claims, however, rest on limited empirical support. In a particu-
larly influential study, Freedman (1987) found that students were
less satisfied with the results of peer group feedback than were
teachers. Teachers, despite being associated with the US National
Writing Project, were ambivalent over the benefits of peer group
approaches.
The results to date suggest that the success of peer group
approaches is heavily dependent on student variables, on the role
of teachers, on the institutional goals, and on the classroom con-
text. In research on ESL students, the diversity of student
populations, differences in prior instructional socializations, and
differing L2 command of English lead to different results for peer
group approaches. For example, Carson (1992) notes that EFL
students from China and Japan do not respond well to the educa-
tional assumptions underlying peer group work. In fact,
sociolinguistic and educational expectations create problems for
many L2 students (Allaei and Connor 1991, Fathman and Whalley
1990, Leki 1990, 1992, Nelson and Murphy 1992, Stanley 1992).
A somewhat different approach to peer group responses has been
suggested recently in research by Flower (1994). In this alternative
approach, students are asked to form groups for collaborative plan-
ning Using audiotaped sessions, responders encourage the writer to
discuss his or her plans, and to develop them so that they address
many rhetorical concerns in writing. Student-writers then reflect on
the planning session, listen to the taping of the session, and write up
a plan for their assignment. The object of these collaborative plan-
ning sessions is to raise student awareness of the rhetorical issues
that are involved in writing and to develop more appropriate plan-
ning strategies for writing. As Flower (1994: 142) states:
Writers are shown how to encourage each o th e r to go beyond infor-

m ation-driven plans ‘to say’ som ething in o rd er to create rhetorical
plans ‘to d o ’ som ething in writing (1) by focusing on purposes, key
points, audience, and textual conventions, (2) by trying to consoli-
date these goals, and (3) by reflecting on their thinking. These three
moves reflect the long-term goal o f collaborative planning: to help
... students develop strategies for constructive planning an d to [help
students] gain m ore awareness o f their own choices.
This approach seeks to move peer-group interaction beyond the

assumption that simple conversational interaction will be benefi-
cial to students; rather, the goal is to promote negotiation of
rhetorical planning for writing. In this respect, Flower draws on
Vygotskean notions of activity in the ‘zone of proximal develop-
m ent’ and the concepts of task appropriation and the
intersubjective construction of goals (Collins et al. 1989, Newman
et al 1989, Rogoff 1990). Apart from recognizing the importance
of task negotiation and conversational interaction, and the need
for direct guided support in training students, this approach
focuses on students goals and planning before they produce the
text rather than after they have written a first draft. To what extent
the initial results of Flower’s (1994) work will be replicated in vari-
ous contexts remain to be seen. It does, however, suggest an
interesting integration of peer group responses with rhetorical
strategy development and Vygotskean interpretations of learning.
Overall, the strongest conclusion which can be drawn from
current research is that peer group approaches vary in their effec-
tiveness, depending on the extent to which:
■ students are persuaded that such approaches will lead to writing
improvement;
■ students are trained to provide peer group feedback effectively;
■ students have clear goals and guidelines for peer group work;
and
■ peer group members are held accountable for their feedback.
Peer groups appear to be most effective when students are moti-
vated by the approach, when they are trained carefully to carry
out the group work, when they are given many suggestions and
guidelines for supportive feedback, when they are assisted in giv-
ing appropriate feedback, and when the feedback provided by
them is reviewed by the teacher (see, e.g., Elbow and Belanoff
1989, Golub 1988, Holt 1992, Leki 1992, Reid 1993, Spear 1988,
1993, Stanley 1992). The most reasonable position to take with

respect to the use of peer groups is that they can be effective; how-
ever, they should not be seen as an answer to all of the students’
needs (Whitworth 1988).
13.2.2 Teacher-students responses

While teacher-student feedback is often seen as a more tradi-
tional overall format for feedback, the range of possibilities for
feedback are as broad as the variations with peer group feedback.
Teacher-student feedback may involve:
■ teacher whole-class discussion of major points for revision;
■ teacher demonstrations of revisions with specific student essays;
■ teacher mini-conferences in class;
■ one-on-one conferences away from the class;
■ written comments on essay drafts.
Each of these options includes many variations which can be
adjusted to meet the needs of particular groups of students. Since
they are often treated as distinct issues in responding to student
writing, one-on-one conferences outside the class and written
comments on drafts will be discussed in the next two section.
Teachers are often able to provide feedback to students not
only on their essay drafts but also on their pre-writing and idea-
generating activities. Teachers should demonstrate the types of
skills that they want students to use, and should make strategies
for brainstorming or invention real to students through think-
aloud demonstrations. This is often done efficiently in a
whole-class context with the teacher writing on an overhead
projector and discussing the strategic choices made while brain-
storming or composing. As students become more comfortable
with the process, they can provide more of the information and
text for such cooperative writing as they learn to appropriate the
teacher’s composing strategies.
The teacher can, in this way, provide feedback on a num ber of
forms of writing that are used in the class. For example, students
may be asked to do free writing from time to time. A teacher can
demonstrate this by free writing aloud in front of the students.
Students and teacher can then discuss some of the strategies used
by the teacher and some of the ways that ideas were developed.
Teachers can also provide similar feedback on reaction-writing
to readings or other assignments. If students are asked to read a

text in front of the class, the teacher may begin the next class by
asking students to write a brief (e.g. 15 minute) reaction to the
reading or to some aspect of the reading. The teacher can then
demonstrate appropriate strategies and techniques for such an
activity by composing aloud while writing such a reaction assign-
ment. Again, students and teacher can discuss what students
thought they saw and what the teacher was doing. Teachers can
also provide feedback on such tasks by reading and discussing the
students’ response writings, not in terms of their writing mechan-
ics, but in terms of idea generation and missed opportunities for
ideas to use.
Another type of teacher feedback can be provided by having
students write evaluations of their essay drafts and then discussing
these evaluations. This type of discussion can also lead the teacher
into mini-lessons on such issues as controlling the topic, rethink-
ing the organization of a paper, considering more carefully who
the readers might be and what they expect as readers, etc. In this
way, topics and issues in writing and rhetorical planning can be
addressed within the context of the students’ own interests and
concerns rather than as decontextualized lessons.
Teacher feedback with the whole class can also involve the use
of specific students’ essays or essay excerpts (see also Chapter 12).
The use of student writing, while requiring care and sensitivity to
students’ feelings, offers many options for teacher feedback.
Examining students’ writing can lead to discussions of certain
problems that a num ber of students share in their writing. The
teacher can bring in a num ber of similar excerpts that indicate a
common problem and the class can discuss these problems
together. For example, many students may not be using clear
extended illustrations of complex ideas, and thus, their writing
may require too much inferencing on the part of the reader.
Students and the teacher can then work with a few excerpts from
students’ writing to develop strategies for generating clear and
vivid examples.
A second major form of teacher feedback on students’ writing
occurs when the teacher works with one volunteer student’s essay.
This procedure has a num ber of benefits: the individual student
receives feedback from the entire class; the teacher can give multi-
ple readings and reader responses to the paper for the entire class
to see and hear (e.g. a descriptive response, a personal response,
an evaluative response), and the teacher can demonstrate a variety

of feedback guidelines/issues and a set of appropriate strategies
for revising. If this activity is done in a constructive way, other stu-
dents may be willing to volunteer their papers for future sessions.
Teacher feedback can also be provided by engaging students in
language awareness activities that are based on students’ writings.
These activities could include scrambling sentences in student
essays to highlight cohesion devices, removing paragraph bound-
aries to determine useful logical breaks (and problems in trying to
find them ), looking for attitudinal and affective signals in student
texts and discussing their effectiveness, highlighting opinion and
argum ent markers and noting their relative effectiveness (see,
e.g., White and Arndt 1991).
Finally, teachers often have opportunities for brief conferences
with individual students in class while the rest of the students are
involved in group work, generating ideas with pre-writing, writing
first drafts, or carrying out other tasks or projects. These mini-con-
ferences should last no more than 5-10 minutes and should
involve feedback on a specific piece of writing by the student, or
feedback on a writing journal or portfolio. The mini-conference
should always balance critiques with positive commentary on some
aspect of student writing. These brief conferences should
conclude with the student stating what he or she will do next.
Teachers should keep a record of mini-conferences and note the
im portant points covered.
13.2.3 Conferencing outside of class

Conferencing with students either in a one-on-one situation or
with a small group represents another major innovation in writing
practice over the past 15 years. (This technique assumes that stu-
dents and teachers are also meeting for regularly scheduled
classes.) In typical conferences, a few students meet with the
teacher for 15-30 minutes and discuss writing progress being
made by those students. A conference can review previous writing
in a portfolio, review the current draft of an essay assignment, or
explore ways of improving specific aspects of class performance
(practising peer responding, new strategies for planning, strate-
gies for content generation and elaboration, or reviewing a
self-assessment form for student progress in writing, etc.)
Research on the impact of conferencing on student writing is
not extensive. While there are many accounts of case studies by

teachers who discuss their use of conferences, there is as yet little
empirical evidence which examines the role of conferences in stu-
dent writing (cf. Calkins 1986, Graves 1983). Carefully controlled
research on writing conferences focuses on two issues: the evalua-
tions of teachers and students after conferences, and the nature of
teacher-student interaction in sets of teacher-student confer-
ences. Results of teacher and student evaluations support the
notion that students receive more focused and comprehensible
feedback during conferences than they do through written feed-
back (Carnicelli 1980, Sokmen 1988, Zamel 1985, Zhu 1994).
Research on the nature of teacher-student interaction reveals that
conferences can vary considerably in the extent to which they are
helpful in improving student writing. Evidence suggests that con-
ferences in which students participate actively and negotiate
meaning (confirmation checks, clarification requests, revision
confirmations, revision queries, etc.) are more effective than
those in which students are passive recipients of teacher com-
ments (Goldstein and Conrad 1990). The research on negotiation
in writing conferences suggests that conferences must be well
planned, and that both teacher and student must be prepared to
negotiate aspects of the student’s writing (Freedman and Sperling
1985, Walker and Elias 1987).
Conferences with students have certain obvious advantages and
disadvantages. The teacher is able to work with students inten-
sively, on a one-on-one basis or in small groups, and the teacher
gets to know the students better on a personal level. Students have
a more informal atmosphere in which to ask questions without
embarrassment, and are able to receive more immediate and
more elaborate feedback than they would through exclusively
written comments. The teacher and students can work carefully
and thoroughly through important strategies for improving stu-
dent performance at all stages of the writing process and students
are better motivated by the personal attention received. The
teacher can help the students to establish clear and efficient goals
for writing assignments and for overall writing improvement. The
major disadvantages for the teacher are the much greater
demands on time and the need to become skilled as an interactive
negotiator. One way to lessen the disadvantages is to have students
prepare a guideline which specifies certain objectives - changing
for each round of meetings - for the conference ahead of time.
This format keeps the conference on task and may reduce the
time needed for each meeting.
Conferences are typically described as following certain com-
mon sequences. Reid (1993), for example, describes the stages of
a typical conference as follows:
Openings
Student-initiated comments
Teacher-initiated comments
Reading of the paper
Closings.
Reid also strongly recommends the use of conference planning
worksheets which should vary according to the overriding goals of
each particular conference. She provides specific examples of con-
ference worksheets for initial planning, essay drafting, and
revision planning; for example, the following is a revision plan-
ning guideline:
Revision Planning Conference

1. I thought the best part of my essay w as...........
2. I thought the weakest part of my essay w as...........
3. According to your [i.e. the instructor’s] comments, the
strengths and problems in the essay draft are as follows:
STRENGTHS PROBLEMS
a..................... ........................a......................
b .............................................b .....................
c..................... ........................c......................
4. Based on the feedback, here is my plan for revising this essay
(list specific steps you intend to take and specific paragraphs
you intend to revise):
a.................................................................................
b .................................................................................
c.................................................................................
5. Three questions I want to ask you [i.e. the instructor] are:
a.................................................................................
b .................................................................................
c.................................................................................
(Reid 1993: 222-3)
In addition, Reid (1993: 222-3) offers a set of metacognitive ques-

tions/statem ents that students should draw on for any conference
session. She also suggests that conferences should end with stu-
dents committing to 1 or 2 most important changes to make so that they
can form an immediate plan of action. Similar conference guide-
lines are provided by Taylor (1993), Tompkins (1990), and White
and Arndt (1991).
Overall, the use of conferences is well supported by logical argu-
ments, observations of teachers and researchers, and a small set of
empirical studies. The conferencing approach can be applied flex-
ibly to address pre-writing, drafting, revising, editing, and
assessment issues in student writing (Tompkins 1990). From a
research perspective, much additional work is needed to examine
the impact of various formats and aspects of writing conferences.
Until the results of such additional research becomes available, it
would seem prudent to use writing conferences as one effective
support for writing development, but not to expect it to be a
panacea for improving student writing.
13.2.4 Responding in uniting

A common format for responding to student writing is through
written comments on drafts and final versions of student papers.
Comments can be offered at any point in the writing process from
planning, to initial drafting, to revisions, to final text. Teachers
can also respond in writing to other shorter tasks such as free writ-
ings, journal entries, reaction assignments, and post-essay
evaluations written by students. Not only does a teacher have
access at many points to offer written response to student writing,
but there are also a num ber of approaches and formats that a
teacher can adopt while responding. The most common choices
include descriptive written commentary (what the teacher
thought while reading the essay), comments for revising major
content and organization of the essay, and prose-editing responses
on major editorial weaknesses.
While a num ber of researchers and practitioners feel that writ-
ten commentary may be the least effective form of feedback for
students, there is also persuasive counter-evidence to suggest that
written commentary is a viable and effective approach when used
appropriately (Fathman and Whalley 1990, Ferris 1995). Rather
than make sweeping claims for or against written commentary,
our goal is to understand the ways in which such feedback can be

used effectively. One of the primary problems commonly noted
with written feedback is that teacher comments are often vague,
confusing, and provide little specific direction for students when
they attempt revision. At the other extreme, teachers sometimes
provide students with detailed editing comment on the surface
form with minimal attention to major organizational and content
issues. The idea is to find the middle road between these
extremes.
Good responses to student writing work in much the same way
as effective peer group feedback. The teacher should find some
positive things to say about any essay, raise a num ber of specific
questions which will allow students to carry out revisions, make
suggestions for changing the organization or elaborating parts of
the essay, and provide a small set of concrete suggestions for
improving the structural and mechanical aspects of the text.
Teacher feedback can be given on the student’s draft, in a
response note attached to the text, or on a tape cassette. Editorial
comments should be presented clearly, whether as prose com-
ments or as symbol notation (though there are drawbacks to the
use of symbol notation). One effective preparation for teacher
response is to have students mark and comment on sections of
their texts before handing them in: this serves as a useful self-eval-
uation mechanism and alerts the teacher to problems and
concerns experienced by the students.
Comments on a text should avoid exclusive attention to surface
conventions (e.g. spelling, punctuation, paragraphing, hyphen-
ation), and should also avoid vacuous commentary or notation
(e.g. AWK). Comments should not overwhelm the students with a
sense of failure (i.e. by noting absolutely every flaw), but should
offer positive support (by praising what is genuinely good), ask
what is m eant at points where ambiguities occur (rather than
putting words in their m ouths), and address organizational issues
by suggesting options rather than dictating solutions.
Specific guidelines and worksheets for teacher feedback on stu-
dent writing should be used cautiously. Without careful use, they
can be impersonal and intimidating. Training should accompany
the use of such analytic guidelines so that students can take con-
crete steps for revision based upon such worksheets. Such
guidelines are often used in final editing stages and for assessment
feedback; however, such guides can also be used for informal
teacher response without grading. Guidelines of this type are

discussed in Hughey et al. (1983), Reid (1993), and Tompkins
(1990).
13.2.5 Other audiences

Peer students and the teacher are not the only potential sources
of feedback. Students should be encouraged to seek out other
audiences and responders. Students can, for example, give papers
to colleagues for feedback. The goal of this activity is to receive
supportive feedback from additional audiences; and the addi-
tional readers should not engage in extensive editing and
rewriting of the student’s paper. An obvious concern is that other
readers try to rewrite the essay, and the essay no longer represents
the student’s ability. Getting feedback from additional readers
constitutes an appropriate part of the writing process for many
advanced writers, and all students should be encouraged to view
such opportunities similarly.
Students should also make use of writing centres and writing labs
where those entities are available. Tutors in these centres can pro-
vide a distinct and honest reader interpretation, and they are able
to provide useful assistance for revisions. Occasionally, tutors are
available to individual students through class cooperation or learn-
ing-assistance centres; these teacher-in-training tutors are earning
credit for teacher preparation programmes and often need tutor-
ing experience. These tutors provide another audience resource
for student writers which should be used whenever possible.
Outside readers can provide much the same type of responses
as peers and teachers. Teachers should work with writing tutors
and suggest ways for tutors to provide effective feedback. Teachers
can provide tutors with various guidelines and options for feed-
back by showing them worksheets, response forms, suggestions for
appropriate interactional negotiations, and various activities
which focus on specific writing issues.
13.3 Writing assessment
Writing assessment involves both teacher responses and more for-

mal mechanisms for student evaluation. The most common
mechanisms involve grading of various types, though evaluation of
writing does not necessarily require grading. Discussions of writ-

ing assessment typically occur in two contexts: the classroom
context and the standardized testing context. The former context
usually involves achievement assessment (both formative and sum-
mative) while the latter involves proficiency assessment. In the
classroom context, any given assessment can be formative
(emphasis on the process) or summative (emphasis on the prod-
uct) . Classroom assessment can also be used for diagnostic
purposes and placement purposes, though this rarely happens. In
contrast, standardized assessment is primarily used to make profi-
ciency judgem ents (and, at times, placement decisions). It is often
not a sufficiently fine-tuned tool for achievement testing in spe-
cific classroom contexts, nor is it typically useful for diagnosing
student difficulties.
While it is possible to view responses to student writing and writ-
ing assessment along a continuum of formality, this chapter treats
the two somewhat independently. Such an approach does not
deny the similarities between responding to writing and more for-
mal assessment, but it does recognize that certain issues in writing
assessment are distinct from issues that are central to feedback on
student writing. In this section, indirect, direct, and portfolio
assessment will be discussed. To the extent that these categories
include standardized assessment and in-class assessment contexts,
both will be discussed under each main subheading.
133.1 Indirect uniting assessment

Indirect writing assessment, typically multiple-choice measures
that test students’ grammar, vocabulary, and written expression
knowledge, is largely limited to standardized writing assessment,
i.e. any assessment situation which is not classroom centred and
involves groups larger than the single classroom in a common
assessment approach. Until 40 or 50 years ago, standardized
assessment of writing was carried out by experts - the individual
teachers and groups of teachers who subjectively assessed student
writing on admission and placement writing samples. With
increasing university enrolments, and a greater demand for
accountability in terms of reliability, psychometricians argued for,
and developed, indirect writing assessments (Camp 1993, Perkins
1983, White 1993).
The fundamental commonality across the large majority of indi-
rect measures results from their attempts to assess writing ability

by testing a subset of skills assumed to constitute components of
writing ability. In most cases, this subset consists of vocabulary,
sentence grammar, and points of writing usage. Most of these
measures were developed during the 1950s and 1960s - the hey-
day of discrete-point testing and of psychometric priorities
favouring reliability over other testing concerns. A num ber of
these indirect measures continue in use today, though many of
these are being retired in favour of direct measures of writing abil-
ity.
While now generally out of favour, these tests did establish two
benchmarks for which all current assessment efforts must
account: statistical reliability and some measure of validity. There
is no question that indirect measures produced a strong reliability
statistic when the tests were carefully developed. The effort to
establish validity with indirect measures centred primarily around
the concepts of predictive validity and concurrent validity. Indirect
measures of writing, when designed carefully, did, to some extent,
predict success in later academic work. Further, indirect measures
exhibited high correlations with direct measures of writing and,
therefore, were assumed to be accounting for much of what direct
assessments measured (Breland et al 1987, Camp 1993, Carlson et
al 1985, Perkins 1983, Stansfield 1986). The fundamental prob-
lem with these arguments is that writing assessment now gives
much greater weight to content validity and construct validity, two
issues which indirect measures cannot easily address.
Among the well-known instruments employing indirect mea-
sures were two early University of Michigan tests. Early versions of
the Michigan English Language Assessment Battery (MELAB) for
English L2 writers included a multiple-choice writing component.
Similarly, the University of Michigan composition placement
examination (for English LI writers), in early versions, included
multiple-choice questions. The Michigan Composition Board
placement examination, required of all entering freshmen, con-
sists now, and has for some time consisted, only of a direct writing
sample holistically scored.
A second source of indirect writing assessment is the Test of
Standard Written English (TSWE) for English LI writers.
Developed by the Educational Testing Service (ETS), it originated
as a companion to the Scholastic Aptitude Test (SAT), the most
basic and common pre-university assessment measure in the USA.
The TSWE is a 30-minute multiple-choice test consisting of 50

items. Two types of questions appear on the test: written usage
questions, and questions on sentence grammar (Perkins 1983).
Among native English-speaking populations, it has high reliability
and some degree of predictive validity as well as concurrent valid-
ity with direct writing samples.
A third common indirect measure of writing ability is repre-
sented by section 2 of the Test of English as a Foreign Language
(TOEFL). This section consists of 40 items to be completed in 25
minutes. The section comprises two types of questions: sentence
grammar and written expression (usage). The section is highly
reliable, and correlates well with holistic measures of writing
samples (Carlson et al. 1985, DeMauro 1992). Despite high corre-
lations, however, DeMauro (1992) expressly states that the TOEFL
section 2 and the Test of Written English (TWE; see below), a
direct measure of writing ability, are not parallel statistically.
There are a num ber of other indirect measures of writing ability
that are noted from time to time. Best known among these are the
Test of Ability to Subordinate (TAS) for English LI writers, and
the Michigan Test of English Language Proficiency (MTELP) for
English L2 writers. The TAS is a 50-item sentence-combining test
which appears to have some predictive and concurrent validity,
but there is minimal documentation (Mullen 1987). The MTELP
does not specifically include a measure of writing ability. It is, how-
ever, inappropriately used from time to time as a measure of
concurrent validity. It should not be considered as either a valid or
reliable indirect measure of writing ability.
Indirect measures of a different kind have been occasionally dis-
cussed as potential measures of writing ability. In particular, cloze
tests, when properly designed and appropriately used, have
demonstrated a high correlation with direct writing measures. For
example, Fotos (1991) argued that cloze tests correlated strongly
with essay writing for a population of English L2 university stu-
dents in Japan. Similarly, Hanania and Shikhani (1986) argued
that cloze tests and writing samples are both good measures of
higher-order language abilities. In both cases, however, the
authors refrained from proposing cloze tests as direct substitutes
for essay texts; rather, they suggested that both essay tests and
cloze tests are good integrative measures of overall language profi-
ciency at advanced levels. It remains to be seen whether these
relationships will be treated in greater depths in future research.
For the moment, the use of cloze testing is best viewed as a useful
supplement to essay writing, providing a more effective overall
measure of language ability (see alsojonz 1990, Oiler 1983, Oiler
andjonz 1994).
The use of standardized indirect measures of writing ability has
decreased markedly in the past ten years, and indirect measures
are not likely to regain popularity at any time in the near future
(cf. White 1995). The increasing emphasis on construct and con-
tent validity - whether the test reflects what research understands
writing to be, and what is normally covered by writing practices -
will push future writing assessment further towards direct assess-
m ent approaches. Past concerns with the reliability of holistic
scoring in direct writing assessment no longer represent a threat
to direct assessment when such scoring is carried out appropri-
ately. Since there is a strong general sense that good writing tests
should involve students producing writing, indirect measures of
writing ability are not likely to remain viable options in the fore-
seeable future.
13.3.2 Direct xvriting tests

Increasing concerns in testing theory with validity, particularly
with content and construct validity, led many researchers to ques-
tion the ability of indirect writing tests to assess students’ writing
abilities accurately. Beginning in the 1970s, many assessment mea-
sures of writing turned towards the production of controlled
student writing samples. While this approach addressed the need
for the assessment of a real instance of writing, it also raised a
whole host of issues concerning the methods for collecting and
evaluating a given writing sample as a true indicator of overall
writing ability.
In the past ten years, standardized assessment with direct writ-
ing samples has focused research attention in a num ber of validity
issues. In particular, research has focused on improving proce-
dures for obtaining a valid writing sample, one which accounts for,
or is at least reflective of, issues such as the impact of audience,
task type, topic type, nature of the rater, rater training, assessment
rubric for grade assignment, and the types of grading scales used.
Many of these issues are also critical in discussions of holistic
assessments in classroom contexts, but they take on greater impor-
tance in large-scale assessment situations since the consequences
of the outcome scores are more serious for the students and for
institutions.
In-class contexts for writing assessment, whether limited to a sin-
gle class or as a school-wide activity, have moved even further from
indirect measures than large-scale assessment. This reflects the
less constrained environment of classroom-based assessment as
well as the greater impact of instructional innovations on assess-
m ent approaches. Direct assessment issues will first be addressed
in the context of large-scale assessment, followed by direct assess-
m ent in the classroom.
13.3.2.1 S tan dardized d irect a ssessm en t
Most standardized direct measures of writing ability involve writing

on a single topic for a relatively brief time. The essays are then
scored holistically by a group of trained raters. In most reliable large-
scale assessment environments, all essays are read by two raters and
discrepancies are handled through a third reading by a ‘head rater’.
There is considerable debate concerning the validity and reliability
of single-item essay writing samples and their holistic scoring proce-
dures. This debate will be addressed briefly at the end of this section.
One of the largest direct measures of writing ability is the
recently developed Test of Written English (TWE); TWE volume
has exceeded a quarter of a million candidates per year for the
last five years (ETS 1992, Test of Written English Guide 1992).
The test is a single-topic essay prom pt on which test takers write
for 30 minutes. The test has perhaps undergone the most exten-
sive set of validity and reliability studies of any essay-writing
measure, and many of the results of these studies have directly
contributed to improvements in prom pt development, task and
topic comparison, time for writing, essay rater training, and essay
scoring (Carlson et al. 1985, Golub-Smith et al. 1993, Hale 1992,
Henning 1992, Stansfield 1986, Stansfield and Ross 1988).
A commonly used British examination with a holistically scored
writing sample is the International English Language Testing
System (IELTS) examination, a joint product of The British Council
and the University of Cambridge Local Examination Syndicate
(UCLES). This test is representative of direct writing assessment on
a number of British examinations for English L2 subjects (e.g.
University of Cambridge Local Examination Syndicate examina-
tions, Royal Society of Arts examinations, University of Oxford
Delegacy of Local Examinations; see reviews in Alderson et al.

1987). In the IELTS test, students write on essay prompts which
reflect either ‘academic writing’ or ‘general training writing’ (as
described in 1995 IELTS Handbook). In the earlier (ELTS) ver-
sion, test takers wrote a single essay; in the revised International
English Language Testing System (IELTS), test takers write two
essays. In earlier tests, test takers were given subject-area options;
most recent versions have dropped the subject-area options in
favour of a choice between academic or general writing (cf.
Alderson 1993b).
In almost all cases of British tests that include a component for
L2 writer proficiency, there is a greater emphasis on communica-
tive language use and expert judgem ent in test development and
less emphasis on psychometric concerns for quantitative reliability
and validity specifications (Alderson 1987, Alderson and Buck
1993). For test users with an orientation towards statistical reliabil-
ity, this difference is unfortunate, since many of the tests
developed in the UK include innovative test-taking options and
interesting prompts. If future efforts are made to address issues of
validity and reliability (e.g. further development of the IELTS),
these tests would likely receive even wider acceptance as options
for large-scale writing assessment.
Large-scale writing assessment also takes place in many other
countries, providing measures of writing abilities for both English
LI and English L2 students (see also Blok and De Glopper 1992).
For example, The Australian Second Language Proficiency Ratings
(ASLPR) is a mzyor exam in Australia for English language profi-
ciency which includes an essay examination (Ingram 1990, Ingram
and Wylie 1984). A major examination in Hong Kong is the Hong
Kong Certificate of Education Examination (HKCEE). Annually, in
Hong Kong, 150,000 16-18-year-old students take this examination
as a qualification for tertiary-level admissions and for many
employment opportunities (Milanovic 1987).
In Australia, the responsibility for language curricula and assess-
m ent rests with the individual states, and different states have
experimented with different solutions to the assessment issue. At
the same time, research is being carried on through the National
Languages and Literacy Institute of Australia (NLLIA), headquar-
tered in Canberra but supporting specialist centres in various
parts of the country. Im portant work in various contexts of
language assessment is being undertaken at the Centre for
Applied Linguistics and Languages at Griffith University in

Brisbane, at Macquarie University in Sydney, at the University of
Melbourne, and at Monash University.
In New Zealand, the process is centralized; the Ministry of
Education is in the process of developing a revised National
Curriculum in which English and other language instruction is
central. At the same time, the New Zealand Qualifications
Authority, separate from the Ministry but working cooperatively
with it, is developing appropriate assessment instruments. In both
instances, developments are of such recent date that very little
information is readily available in published form.
In the USA, major English LI testing efforts that have devoted
considerable attention to reliability and validity include certain
US universities and university systems. Two exemplary pro-
grammes are the Michigan Composition Board and the California
State University Composition Board. The University of Michigan
Composition Board, begun in 1979, assesses the writing abilities of
all incoming undergraduates to the university (approximately
6000 per year). Students write a 50-minute essay and, based on the
results, are placed in one of three levels of the writing pro-
gramme. More recent versions of this test have employed
multiple-trait scoring procedures, and these changes have been
seen as successful (Hamp-Lyons and Condon 1993).
The California State University English Composition Board
administers a 2-hour writing test that consists of two objective mea-
sures and a writing sample. The writing sample consists of a single
essay prompt, and students are allowed 40 minutes to write a
response. The writing samples are sent from the 20 campuses of
the California State College and University System to a central
scoring location, and beginners are placed in the English writing
programmes on each of the various campuses based on these
results. Prompt designs and scoring procedures are carefully con-
trolled, and the test achieves high reliability measures (Greenberg
1988).
A different type of large-scale assessment approach is carried
out in the USA by the National Assessment of Educational
Progress (NAEP) programme for writing assessment (Applebee et
al. 1986, 1990a, 1990b). This US-based programme (housed at
ETS in Princeton) is funded by the federal government to moni-
tor changes in educational progress across the range of student
groups in elementary and secondary education contexts. Since its
inception in 1969, the NAEP has carried out five national assess-
ments (1969-70, 1973-74, 1978-79, 1983-84, and 1987-88). Over
these five assessments, randomly sampled groups of students
write on one to four writing prompts (out of 12 options) ranging
across informative, persuasive, and personal/narrative tasks.
Students participating in the NAEP receive a booklet and are
given one hour to complete a set of background questions and a
set of brief reading and writing tasks. The most recent NAEP
assessment, in 1992, has expanded the time allotted; students are
now given more time to respond to writing prompts (25- and 50-
minute periods) (Gentile 1992). Overall, more than 1,300,000
elementary and secondary students have participated in the pro-
gramme over the past 25 years. Based on the results, the NAEP
provides reports which extrapolate from the sample to statements
about the progress that US students make in writing develop-
ment.
These writing samples are scored though a primary trait rating
procedure, scoring each writing on the extent to which it carries
out the task. Prompts are reviewed for bias and validity criteria,
and the reliability measures are high (most in the 0.87 to 0.95
range). While there are serious criticisms of this programme in
terms of testing design and procedures used, it does produce a
large volume of data for assessing writing, for developing scoring
procedures, and for interpreting the results. Moreover, many of
the criticisms of the NAEP programme are no different from
recent criticisms addressed to most large-scale direct assessment
single-item approaches. (Other large-scale assessment pro-
grammes at Educational Testing Service (USA) which involve
writing samples include the Advanced Placement (AP) testing pro-
gramme for high school students and the General Education
Development (GED) programme which awards equivalency for
high school graduation.)
A final major international effort to understand writing devel-
opm ent across a range of countries has used direct writing
samples; carried out by the International Association for the
Evaluation of Educational Achievement (IEA), the project is
known as the IEA Study of Written Composition (carried out pri-
marily in 1982-83). The testing design for the project required
students at three different grade levels in 14 different countries to
write on eight different writing prompts, ranging across pragmatic
tasks (bicycle description, letter of advice) and essay-type tasks
(persuasive essay, narrative essay). The prompts were extensively

reviewed and piloted by an international review board, and the
study developed strong criteria for validity and reliability of scor-
ing. Results of this research have appeared in various places,
though most are not easily accessible (Degenhart 1987, Gorman et
al. 1988, Purves 1984, 1988, Purves et al. 1984, Vahapassi 1988).
13.3 .2 .2 S corin g ap p roach es
There are a num ber of approaches to grading writing samples in

large-scale assessment. The most commonly used type of scoring is
a holistic score representing the overall quality of the writing.
O ther types of scoring include primary-trait scores and combined
multiple-trait scores. Writing assessment in classroom contexts
might alternatively include analytic scales and non-quantitative
assessment. These latter two options will be discussed in a later
section.
Holistic rating is the most common form of scoring for large-
scale writing assessment, though in-class writing assessment can
also use this approach. Its popularity is a direct reaction to the use
of multiple-choice writing measures (Perkins 1983, White 1993).
Holistic scoring is achieved by reading a text and deciding on a
general, and subjective, score based on a numerical scale ranging
anywhere from 1-4 to 1-9. The numbers on the scale are often
described briefly as specifications on a scoring rubric (see, e.g.,
Test of Written English Guide 1992). When holistic scores are
used, efforts should be made to socialize a rater to the ways other
raters assign grades and their rationales for doing so. This
‘anchoring’ constitutes an im portant support for generating con-
sistently reliable scoring (White 1994, 1995).
Primary-trait scoring is a grading scheme that is not so commonly
used in either in-class or large-scale writing assessments (except in
experimental studies). The use of primary-trait scoring is not a
very economical approach since primary-trait guidelines should
be rewritten for every writing task (White 1993). Because the goal
is to match the salient characteristics of the task requirem ent to
the primary trait that is evaluated, it is necessary to assess before-
hand the potential of the task to highlight fairly that trait in the
students’ writing. Questionable validity is a further problem with
primary-trait scoring. As Hamp-Lyons (1991c) argues, it is seldom
the case that raters of essays rely on the visibility of single traits in
an essay to arrive at the score. In observed studies of essay rating,

it is apparent that raters rely on multiple traits as they are high-
lighted at various points in the students’ essays (see also Perkins
1983, White 1993). For this reason, primary-trait scoring is not a
preferred option for student writing assessment.
The one well-known consistent use of primary-trait scoring with
large-scale assessment is its use in the US NAEP testing. Writing
assignments are designed to reflect specific tasks that students
must carry out. Students are rated as successfully completing a
persuasive task, a story-telling task, or a letter-writing task. Scoring
is designed specifically to reflect success in carrying out the task
(Applebee et al 1990a, 1990b).
The use of multiple-trait scores is seen as a preferred option for
writing assessment that requires more refined profiles of student
writers, and it is particularly useful for diagnostic purposes if the
scoring instrum ent is designed and used appropriately (Hamp-
Lyons 1991c). In multiple-trait scoring, a rubric is designed which
allows the rater to consider a num ber of aspects of the essay; some
rubrics may be more general in nature and some may be designed
more specifically (e.g. a general rubric might require a holistic
score for accuracy, one for overall coherence, one for content,
etc.).
The potential danger with multiple-trait scoring lies in the fact
that it is difficult to design and use rubrics appropriately. In cases
in which multiple trait scoring has been used without careful scor-
ing rubrics, there are serious questions about the validity and
reliability of results (e.g. Canale et al. 1988, Cumming 1990a,
Hamp-Lyons and Henning 1991). Raters have difficulty in scoring
student essays along independent dimensions in a reliable man-
ner if they do not have clearly defined rubrics to support their
rating. While this issue is more a matter of concern for research
and for large-scale assessment, it nevertheless suggests caution
when providing weakly detailed and specified multiple holistic
scores on in-class student essays (such as for accuracy, style, and
content).
1 3.3.2.3 P er fo r m a n ce assessm en t
Performance assessment is often referred to as a type of testing

procedure which stresses authentic and direct measures of abili-
ties. It is also typically associated with complex problems, higher
order skills, context sensitive strategies, and student choice (see

Baker et al. 1993, Messick 1994 for current overviews). In many
contexts, such as science learning, engineering, or architecture
certification, performance assessment may be carried out in a
num ber of different ways.
For writing, this type of assessment has been discussed in two
contexts: in classroom and institutional assessments (including
portfolios; e.g. Dunbar et al 1991, Faigley et al. 1985), and in
national assessment by the National Assessment of Educational
Progress programme. In both of these contexts, performance
assessment employs a variation on primary- and multiple-trait
scoring approaches in that it is designed to assess a specific type of
writing ability for a specific purpose or set of purposes rather than
to assess general writing proficiency (Faigley et al. 1985). The dis-
tinct specification arises from a carefully designed prom pt which
determines the type of writing to be evaluated. It is like a primary-
trait assessment in that one specific facet is being assessed; the
difference is that the specific facet is not a measurement criterion
(e.g. coherence) but the type of writing that is to be assessed. For
example, a performance assessment would have students write a
classification paper, and assessment might focus specifically on the
extent to which the paper meets the demands of the rhetorical sit-
uation, establishes appropriate categories for classification, and
classifies items in these categories. (A primary-trait score, on the
other hand, would only measure the extent to which the overall
task was carried out.)
Performance assessment is most likely to be carried out using a
carefully constructed multiple-trait scoring rubric, one which
addresses specific aspects of the writing and which should be
appropriate for placement and diagnosis. Such a scoring
approach may not be very useful in most large-scale assessment
programmes, but it is likely to have potential in institutional-level
testing such as in a university placement programme for subse-
quent writing instruction. Efforts to develop performance
assessment for different purposes are discussed in Allaei and
Connor (1991), Baker et al (1993), Faigley et al. (1985).
1 3.3 .2 .4 C lassroom -b ased d irect assessm en t
Students’ in-class writing abilities are determined almost entirely

on assessments of student writing samples. However, classroom
contexts for direct assessment is primarily concerned with provid-

ing the student with evaluative feedback of some kind which
should lead to improved performance. If class assessment is not
used in this way, its primary function as a learning tool is being
missed. There are a num ber of specific wavs that in-class writing
assessment can be carried out, and a num ber of options for assess-
m ent exist which can be matched to student, teacher, and
institutional goals.
In-class assessment, for example, can be based on a single writ-
ing task, a series of related writing tasks, a long-term writing
project, or a writing portfolio. Assessment can be ongoing or final;
it can be based on in-class writing or out-of-class writing; it can be
focused on formal, rhetorical, or content aspects of the writing;
and it. can be focused on writing as an independent language skill
or integrated with other language tasks (primarily reading).
Assessment can be carried out via a conference, a negotiated con-
tract, a specific written sample, or an ongoing record of teacher
observations. Assessment can be based on teacher evaluation, stu-
dent self-evaluations, or a negotiation between the two. Finally,
assessment in class can be oriented towards mastery of certain
skills and strategies (process and criterion-referenced) or towards
student performance compared with other student performances
(norm-referenced). Grading can be based on a holistic score or
on a set of analytic scales which reflect course emphases.
Accordingly, teachers and institutions have numerous options for
carrying out appropriate assessment of student writing.
The most traditional formats for writing assessment (typically
summative) often involve a single essay based on a relatively sim-
ple prom pt and administered over a relatively limited time. Such
measures of student writing for in-class contexts are now consid-
ered, by themselves, to be of questionable validity. A second, more
acceptable, option resulting in a single essay involves a longer
period of time preparing students to develop a given topic, and
then giving students more time, and multiple-writing sessions, for
the essay itself. For example, a common format uses the last week
of a semester to introduce students to a topic, to discuss a set of
readings, and to develop a topic through pre-writing activities, dis-
cussions, and information drawn from the reading material. After
a week’s study of the topic, students are given a prom pt and
allowed one to two hours to write a rough draft. The next day, stu-
dents return to revise for an hour. While this format still preserves
the single essay evaluation format, it does allow the students to

engage in the writing process while preparing for the final essay,
and it gives students adequate time and resources to write to the
best of their capabilities under time constraints.
The latter option, an extended period of class time for writing
evaluation purposes, is more consistent with current instructional
practices. Teachers spend much time convincing students that
good writing involves pre-writing, brainstorming, planning, and
revising. It is im portant that summative writing evaluation be con-
sistent with the ongoing instruction. Having students work
through the various writing processes as part of the evaluation
preparation is one reasonable option for maintaining similar goals
across instruction and assessment.
Ongoing (formative) assessment of essays written out of class -
as the result of multiple drafting, feedback, and revision - repre-
sents a different, though still traditional, context for evaluation,
one which is more compatible with the practices of a writing class
than is an in-class time-constrained evaluation. In this case, both
ongoing and final evaluations of student writing are based on
assessments of the final versions of a set of essay assignments writ-
ten throughout the course. A variation on this collective
assessment option lies in including other types of writing in the
evaluation: summaries of readings, critical reviews, reaction
papers, etc., as additional evidence of writing development and
improvement.
A third writing assessment option lies in having students write a
series of shorter assignments during the final week of a course and
then deriving an assessment over the combined performance on
all of the writing tasks. These tasks might include a reading and
writing summary, a position paper, a written self-assessment, and a
synthesis of a first reading with a second reading. The advantage
of such an approach is that the assessment is not dependent on a
single writing sample or a single genre/task type. The disadvan-
tages include the development of appropriate multiple prompts,
the possibility of student (and teacher) fatigue, and a more com-
plex weighting of grades for the various writing samples.
In all of these cases, assessment procedures must consider the
development of appropriate and engaging prompts for writing,
specific tasks which allow students to perform to the best of their
abilities, evaluation schemes which are fair and readily inter-
pretable, and constructive feedback based on the evaluation.
Recent research would suggest that the matter of appropriate

prompts, tasks, and evaluation schemes is not simple. Hamp-Lyons
(1991b) and Horowitz (1991), for example, argue that different
writing prompts are likely to produce different results, as are dif-
ferent writing tasks. For this reason, Horowitz suggests that
prompts and tasks for evaluation be situated in topics that are
familiar to students and also be appropriate to the students’ spe-
cific fields of study (see also Kroll and Reid 1994, Reid and Kroll
1995).
13.3.2.5 A ssessm en t o p tio n s in cla ssro o m co n tex ts
Classroom contexts for writing assessment open up a num ber of

scoring options which are not used in large-scale assessments. In
addition to a relatively easier adaptation of performance assess-
m ent and portfolio assessment (see below), classroom contexts
also perm it the effective use of analytic rating scales, non-quantita-
tive assessment, and alternative assessment.
Analytic scale rating has been, and remains, a popular way to
assign a grade to students’ writing and to provide feedback on spe-
cific aspects of the writing. The best of these analytic rubrics is
that developed by Diederich (1974), though with most analytic
scales there are validity problems (Faigley et al 1985, White 1993).
Without clear specifications for grades given with each scale, it is
not entirely clear whether the scales provide appropriate feedback
or genuinely represent students’ writing abilities. For example, a
decision to give a grade of 80 per cent on a scale of vocabulary use
may not provide either reliable or usable information for students
(Hamp-Lyons 1991c). Nevertheless, such analytic scales for assess-
m ent are common (see examples in Fagan et al 1985, Reid 1993).
One of the earliest and best-known scoring guides of this type
for ESL is that developed by Jacobs et al (1981; see also Hughey et
al 1983). Unlike the more typical analytic scales noted above,
these scales are well supported by content and construct validity,
they have been field tested, and they are designed specifically for
classroom use rather than for large-scale assessment. They remain
among the most popular analytic guidelines for L2 writing use;
moreover, because of their specificity, they are also useful for in-
class multiple-trait scoring procedures.
Non-quantitative assessment of student writing is an option that
some teachers adopt because they do not feel the numerical
scoring of student writing provides a useful mechanism for feed-

back to students. Further, some teachers do not feel that such
scoring focuses the student’s attention on the writing itself. A non-
quantitative orientation would focus on recommendations for
further changes to the student essays, and teacher assessment
would be in the form of prose responses to the essays. These
responses would highlight strong and weak points of the essays,
suggest areas in which the student could extend the writing, and
even indicate ways that the teacher might have written about the
same assignment. Non-quantitative assessment would also ask the
student to write reflective responses on the experience of writing
the essay, including the following:
■ a review of the process
■ particular problems that were difficult to resolve
■ strong points of the writing
■ other information that might have been included, and
■ two or three points that will receive more attention during the
next writing task.
Alternative approaches to writing assessment suggest relatively
uncommon options for assessing student performance that
extend to the writing process for a specific essay. All of the grading
approaches discussed above provide evaluative feedback at some
designated end-point (the product) in the student writing process
(Calfee 1992, Hiebert and Calfee 1992). Alternative assessment
approaches do not always have such a rigidly fixed evaluation
point (Pierce and O ’Malley 1992).
One alternative that is becoming more common is the use of an
assessment conference. In this situation, student-teacher conferences
are set up to assess students’ progress, to plan future goals for writ-
ing, to examine and assess a particular writing task or set of tasks,
a n d /o r to discuss problems and strong points that have been
observed in the student’s writing. This approach can also incorpo-
rate the notion of negotiated assessment. In this case, the student
and the teacher negotiate the assessment to be recorded for the
student’s work, as well as plan future goals and decide how the
new goals should be assessed. The use of conferences for assess-
m ent purposes gives students some degree of ownership and
responsibility over the assessment process; it also allows for stu-
dents’ insights into the assessment process and permits them to
reflect on course goals and individual progress. This approach
may, however, be inappropriate for teachers who have reservations

about students participating in their own assessment.
If assessment conferences are used, the goals should be set
before the conference begins. Students should assemble relevant
writing material and records of performance in the class. A guide-
line written by the teacher on assessment conference expectations
may be useful to students; a form on which students list relevant
material and writing assignments to be discussed during the con-
ference is also a way to focus students and have well organized
conferences. Students’ letters of agreement and understanding
also constitute useful follow-up to the conference.
A somewhat different version of student involvement in assess-
m ent occurs through the use of self-assessment procedures. Under
this option, teachers ask students to work through a guideline
form that asks the student to account for the work he or she has
done to date. On such a form, each student reviews: his or her
recent written work; his or her accomplishments in the class;
strong performances or improvements that he or she wishes to
claim as accomplishments; weaknesses that still give problems; and
one or two areas that will represent further improvement if these
areas are addressed appropriately in future work. Self-assessment
procedures can also be used for specific essay assignments
(Spandel and Stiggins 1990, White 1994). Students are asked to
reflect on an essay recently completed, pointing out the strong
and weak points of the essay and of the writing process. In this
process, students should also point out their improvements
reflected in this task over previous writing assignments. Based on
this self-analysis, students assign themselves a grade for the essay.
The teacher then tells the students how the teacher would evalu-
ate the assignment in terms of improvement and grade. This pair
of assessments can then be used to negotiate a grade and an
agreed-upon report of student progress in the class. Alternatively,
such a process can form only a percentage of the overall grade
assigned to a paper.
Still another alternative type of assessment involves both a nego-
tiated assessment and a self-assessment. At the beginning of the
course, the teacher outlines possible assessment contracts from
among which students can choose to fulfil the requirements for
the course. Students are free to select different options and even
to negotiate minor individual variations. These contracts are then
used as the basis for assessing student writing progress during the
course of the term. Such contracts will usually stipulate a certain

quantity of writing, the types of assignments that will be com-
pleted, additional types of writing to be done, writing
improvement goals for the student’s progress, and the grade to be
earned on the basis of a satisfactory completion of the contract.
These contracts can also serve as the basis for self-assessment,
negotiated assessment, and assessment conferences.
A more teacher-focused form of non-graded assessment is one
based on teacher observation. Assessment in this format involves
ongoing record keeping of teacher observations. Teachers regu-
larly make notes of students’ progress, of consistent weaknesses, of
growing strengths. Teachers also make notes of students’ writing
assignments and their salient aspects. Additionally, teachers keep
records of informal conferences and discussions with students.
These records and observations then form the basis of student
writing assessment. They may take the form of a set of grades, but
more commonly they occur as written evaluations of student
progress, focusing equally on strong points and perceived weak-
nesses. This approach is well suited for teachers who are obliged
to report student progress but who do not put much faith in
grades. Teacher observations can also be noted on guideline
sheets that assist in principled observation. These sheets should
provide space for all of the important concerns that an assessment
of student writing might need to consider. Finally, such an on-
going assessment approach can be combined with a num ber of
the other assessment options discussed above, including both con-
ventional and alternative procedures.
A final approach to writing assessment is through informal assess-
ment, in some respects a seeming contradiction in terms. Informal
assessment involves teacher observation, and it stresses non-grading
discussion and negotiation with students over their progress. This
approach is partly predicated on the very optimistic assumption
that students will bring strong self-motivations for working on
their writing abilities. This approach is compatible with a curricu-
lum that emphasizes engaging instructional activities and projects
- providing a natural avenue for real or realistic writing (e.g. writ-
ing and presenting a play or a TV newscast, students reading their
writing publicly to student audiences, discussion of various possi-
ble student journals and learning logs). It is well suited for
whole-language classrooms and other elementary-level curricula
which seek to encourage language skills development based on
Responding to uniting and uniting assessment 413
positive and meaningful learning experiences through language.

It is perhaps a less viable assessment option for more advanced stu-
dents and in institutional structures which require some
formalized mechanism for student assessment of progress.
13.3.2.6 P ro b lem s w ith d irect a ssessm en t
In most assessment situations, simpler assumptions about the

development of writing tests have been replaced by the recogni-
tion of the many variables which are likely to have an impact on
student performance: student backgrounds and interests, topics
chosen, writing tasks used, prom pt wordings, presence of visual
resources in the prom pt (e.g. tables, diagrams, charts, etc.),
am ount of time, and scoring procedures. Valid and reliable writ-
ing assessment will attempt to control a num ber of these variables,
or at the least to be cognizant of their potential impact on assess-
m ent outcomes.
The rise of direct writing assessment, and in particular the sin-
gle essay prom pt that is holistically scored, has generated a
num ber of criticisms, both in its early use and more recently,
though for different reasons. Earlier criticism dealt primarily with
the problems of establishing inter-rater reliability across essay
raters. This criticism has been directly addressed by a num ber of
studies and ongoing assessment programmes which have consis-
tently demonstrated high inter-rater reliabilities (cf. Cherry and
Meyer 1993).
Newer challenges to direct assessment of writing with holistic
scoring have raised questions about the content and construct
validity of such approaches. The questions concerning validity
centre around the match between research on writing develop-
m ent and the writing assessment procedure. These concerns are
particularly serious for large-scale assessment, since in-class con-
texts provide a num ber of responses to the difficulties with direct
assessment of writing. Among the concerns typically discussed are
the following:
1. Can assessment be achieved in a decontextualized one-time
writing sample? (Is the format too artificial?)
2. Can a brief timed sample represent a student’s ability to write
in various contexts and for various purposes?
3. Can a single general rating reflect a student’s writing ability?
(Is a global measure possible?)
4. Can assessment of a writing product provide information about

how the writing process has been carried out?
5. Can prompts be generalized across different topics?
6. Can an arbitrary task and topic generate equal interest and
motivation across large groups of students?
7. Can a brief writing sample provide feedback and diagnostic
information for student’s later learning?
8. Can raters actually provide a holistic score or are they concen-
trating on specific parts of an essay?
As can be seen from the concerns above, the challenge to direct

assessment is not that it requires a writing sample, but that the
common practice of a single sample of a student’s writing is insuf-
ficient for a valid assessment (Camp 1993, Hamp-Lyons 1991c,
Horowitz 1991, White 1993, 1995, Williamson 1993). Many of
these questions can be addressed in various ways, but they never-
theless pose real issues that must be considered by anyone
choosing to employ direct assessment. One direction that has
developed as a result of these questions is the movement towards
portfolios for large-scale assessment.
13.3.3 Portfolios
A portfolio approach to writing assessment constitutes a major
recent direction in assessment, both in large-scale and in class-
room contexts. While there is not, as yet, an extensive research
literature on this movement, there are a num ber of sources which
describe several projects, their methods, and the assessment
results to date (see, e.g., Belanoff and Dickson 1991, Tierney et al
1991; cf. Hamp-Lyon and Condon 1993). The discussion to follow
will first review a num ber of attempts to use portfolio assessment
in large-scale assessment contexts.
13.3.3.1 P o r tfo lio s in large-scale a ssessm en t
The increasing exploration of portfolios for assessing the writing

abilities of students beyond the classroom context has generated
both excitement and a fair amount of scepticism. The excitement
is due in part to the opportunity to break away from artificial
‘objective’ writing samples and in part to a num ber of interesting
pilot efforts that have been reasonably successful in differing con-
texts. The scepticism is due to the emphatic and perhaps overzeal-

ous claims made by portfolio proponents and to the recognition
that assessment itself is a somewhat artificial undertaking and will
always encounter limitations, irrespective of format and proce-
dures (Dunbar et al 1991, Messick 1994).
Perhaps the best-known example of portfolio use in advanced
writing in the USA is the effort to develop portfolio assessment
over the past decade at the State University of New York (SUNY)
at Stonybrook. In the English Freshman Composition programme
at SUNY, students must assemble a portfolio of their course writ-
ing and receive a grade of C or better to pass. Every portfolio is
graded by the classroom teacher and by one other instructor, and
the outside reader must also give the portfolio a grade of C or
higher for the portfolio to receive a passing grade. Student portfo-
lios consist of three revised essays, one in-class essay, and cover
sheets for each essay describing the process of the writing and the
purpose of each essay. The three revised pieces include: (1) a nar-
rative, descriptive, or expressive essay, (2) an academic essay, and
(3) an essay that analyses and critiques another essay (Elbow and
Belanoff 1991).
Raters are trained to give a ‘pass’ or ‘no-pass’ grade to the port-
folios; individual teachers decide at a later point on higher grades
for a passing portfolio. While the programme is effective in get-
ting students to write more and perhaps write better, a num ber of
problems have been noted with this approach. Teachers are asked
to do more work in evaluating student writing; students may have
more opportunities to cheat; some teachers feel that the portfolio
overwhelms the course; and some teachers feel that repeated
opportunities for revising constitute too lenient a system.
Conversely, the strengths of this programme are that students
must take the writing process seriously, that they must take
account of audience as a realistic concern, that they must be
responsible for a wider array of their writing and must reflect on
this array, and that they are assessed on their real writing produc-
tion for the course - not on a remote final essay (Elbow and
Belanoff 1991).
A num ber of other universities have developed portfolio assess-
m ent programmes which operate almost similarly to the
Stonybrook model. Minor variations involve the different uses of
outside raters, different mid-term evaluations of the portfolios,
various schemes for rating the portfolios, various options for types
and numbers of essays included, various options for revising the

submitted essays, and different types of cover information
required from the students. Such programmes include those at
the City University of New York, Christopher Newport College
(Virginia), Kansas State University, the University of Michigan,
Northern Arizona University, and Miami University of Ohio,
among others (Hamp-Lyons and Condon 1993, Rosenberg 1991,
Smit et al 1991).
Portfolios have also been used as a university writing compe-
tency requirem ent in lieu of an upper-level writing proficiency
exam. In these cases, universities typically decide on a subset of
courses which would be used to generate writings that could be
placed in the portfolio. For example, at the University of Alaska
Southeast, portfolios are used during the junior year as an upper-
level writing proficiency requirem ent instead of a writing
examination (Wauters 1991). Students are expected to submit
writing that covers expressive, explanatory, and persuasive modes
as well as a cover letter explaining the pieces in the portfolio.
Another variation has been put in place at Southeastern
Missouri State University (Holt and Baker 1991). In this case, stu-
dents who fail an upper-division writing requirement, or who
choose to opt for a portfolio alternative, must submit four papers
from courses as well as a cover letter describing the papers. The
papers submitted include a persuasive essay, a more general
expository essay, and two other pieces of academic writing (letters,
poetry, and short stories are not acceptable). These portfolios are
then graded on a six-point scale. Three of the five pieces must
receive a score of 4 or better for the portfolio to pass.
Portfolio assessment is also being examined as an assessment
option for the NAEP programme (Gentile 1992). In a 1990 pilot
study, NAEP collected approximately 2000 portfolios assembled by
4th and 8th graders. These portfolios included writing samples
primarily of informative and narrative pieces and were rated on
scales similar to those used for the NAEP writing prompts. Results
of the pilot study indicated patterns of writing similar to the types
elicited by the typical assessment prompts used by NAEP. In 1992,
a more complete collection procedure was implemented but the
results of this assessment of portfolios have not yet appeared pub-
licly.
There are undoubtedly other uses of portfolios in various con-
texts around the world, and more reports of their use will appear
in the coming years. In the USA, for example, a num ber of state
Departments of Education are using portfolios for state-wide
writing assessment. Vermont, in particular, has been the object of
reports recently and, as might be expected with beginning efforts,
is encountering a num ber of difficulties (Koretz 1993). Neverthe-
less, portfolios provide an approach which addresses many of the
construct-validity problems noted with indirect and single-sample
direct measures of writing ability. At the same time, portfolio
assessment faces many of the same problems that confront direct
measures of writing, and it also creates a few additional problems
that will have to be addressed in the coming years (Camp 1993,
Hamp-Lyon and Condon 1993, White 1994, 1995).
Among the limitations noted for large-scale portfolio assess-
m ent are the following:
1. Portfolios really refer to a means of writing collection rather
than to a means of assessment.
2. The problems with establishing a single score or scale for essay
prompts is even more complex for portfolios. How will a single
score capture the variation in writing ability indicated in a port-
folio?
3. There are serious problems with reliability. As more leeway is
given to choice of writings in the portfolio, it becomes more
difficult to establish grading equivalence.
4. Portfolios take more time to grade and thus are much costlier
as an assessment option.
5. A system for portfolio assessment must also establish the
authenticity of the writing. How will the portfolio raters know
that the students actually wrote all the pieces in the portfolio,
and when is editing and revising assistance from others too
extensive to represent the student’s own writing abilities?
6. A portfolio consisting of different writing tasks assigned by dif-
ferent teachers will lead to differences simply because some
writing tasks are inherently more interesting and better con-
structed. How will such variations be controlled?
These and many other questions provide the challenges for the
future of portfolio assessment in the coming decade.
While there are serious obstacles to portfolio assessment, just as
there are to most other forms of large-scale assessment, there are
also strengths to the portfolio approach which should not be over-
looked:
1. It allows for assessment of multiple writing samples across a

range of topics and task types.
2. It allows the possibility for students’ reflection on their writing
and their writing progress.
3. It assesses the best that a student is capable of producing.
4. It creates a more realistic audience for student writing.
5. It gives students responsibility for choosing the writing on
which they want to be evaluated.
6. It provides for a strong link between instruction and assess-
ment, a highly desirable quality for any assessment mechanism.
These potential benefits may make portfolios a powerful assess-
m ent approach in the years to come. The fact that there are
serious issues to be addressed in making portfolio assessment
work effectively does not detract from its potential as a strong
alternative to single prom pt writing evaluation.
13.3.3.2 P o r tfo lio s fo r in-class a ssessm en t
Portfolio uses in a classroom context perm it many more options

for assessment of student progress and have greater potential both
for diagnostic uses and future placement uses. In a classroom con-
text, a portfolio may be designed as a controlled format, rather
like the way it is used in large-scale assessment, or it may become a
larger ongoing collection that best represents a student’s writing
abilities and most successful writings. In this respect, portfolios fol-
low directly from professional adult uses of portfolios as a way to
sell one’s skills and talents. In yet other contexts, portfolios are
not merely a matter of students choosing their best work; rather,
students include all of their major writing as well as much
exploratory writing which may serve as a resource for future
essays. Portfolios can also include writing resources such as
instructional guidelines, semantic maps, readings, pictures,
poems, and other possible items that might be associated with spe-
cific writing tasks and assignments.
One of the best qualities of portfolio assessment in in-class con-
texts lies in the fact that it allows for a combination of standard
assessment approaches together with many alternative assessment
options (as described in subsection 13.3.2.5 above). Portfolios, for
example, constitute an ideal mechanism for conferencing with
students. Students have in one place the im portant writings on
which they are working, and the teacher and the student can both
refer directly to various parts of those writings. The portfolio pro-
vides a natural agenda for the conference, and allows the student
to do most of the talking about his or her writing. Not only can
the portfolio allow both teacher and student to examine the
strengths and weaknesses of writing in detail, but it can also allow
the conference to range across the student’s various efforts at revi-
sion and future plans for revision, identifying those pieces of
writing that are likely to become part of his or her final portfolio
assessment.
A second way that portfolios tie in with alternative assessment is
that they provide a springboard for self-evaluation, whether as in-
class practice or as part of any assessment conference. Tierney et
al (1991) offer a persuasive argument to the effect that portfolios
assist self-evaluation because they emphasize a num ber of valued
qualities for self-assessment approaches: student ownership, stu-
dent centredness, non-competitiveness, individual customizing, a
more objective (or reader’s) view of the writing, student self-selec-
tion, and student involvement in establishing evaluation criteria.
In classes with younger students, portfolios can also become
repositories of all the student’s writings, including non-formal
writing, writing across a range of subjects, and pre-writing notes,
lists, semantic maps, charts, tables, etc. In a sense, such a portfolio
becomes a knowledge resource for future writing activities as well
as a record of past and ongoing writing tasks. Towards the end of
the school year, portfolios can be examined to create a more
focused representation of a student’s writing progress through the
year. At that point, the student and the teacher can work through
choices for the portfolio, decisions on organizing the portfolio,
explanations for items in the portfolio, and means for evaluating
the portfolio. This process can also include sending the portfolio
to parents so that they may see the student’s progress together
with what it represents from the teacher’s perspective.
With older students, and in more advanced academic contexts,
portfolios become a more personal documentation of student
development. Portfolios become increasingly the student’s
responsibility, and the student and teacher typically work together
to organize the material it contains. Students, however, write their
own introductions to their portfolio material and explain the rele-
vance of each piece they possess. While portfolios are a focal point
for teacher-student negotiation and conferencing, students
recognize that their portfolios for writing take on a larger assess-

m ent purpose for the course; portfolios become the means for
students to represent themselves in the best possible light - the
writing portfolio becomes more like the professional portfolio in
this context.
Overall, the use of portfolios in classroom contexts appears to
add im portant assessment dimensions that extend beyond the
positive features noted for large-scale assessment:
1. Portfolios allow feedback into instruction and curriculum
development, linking assessment and instruction.
2. They allow parents, administrators, and others to see accurate
profiles of student abilities: their strengths, weaknesses, and
progress.
3. They can be planned over the course of a curriculum cycle or
academic year, well in advance of assessment time.
4. They allow for a collaborative approach to evaluation.
5. They base assessment on what students actually write in the class-
room rather than on an artificially imposed external measure.
6. They allow for continuous, ever-expanding, and systematic
assessment.
7. They position the teacher as the natural ‘expert’ evaluator.
Teachers are called on to make im portant professional judg-
ments.
(Drawn from Tierney et al 1991)
With portfolio assessment, as with any approach, there is the dan-
ger of a myopic view, presenting only the positive aspects of a
challenging and interesting assessment approach. Moreover, a
procedure that emphasizes work over the extended course of time
will be challenged by some who feel that assessment should value
a student’s proficiency level at the end-point rather than empha-
size the ongoing process of writing and revising over time.
Furthermore, the same dilemmas impact on portfolios as on other
writing assessment. In many cases, students must receive grades,
and ultimately portfolios are evaluated for the quality of writing
more than for effort and student cooperation (White 1993).
13,4 The limitations o f assessment

The greater emphasis on writing assessment in the past ten years,
and the parallel increase in research, has made both researchers
and teachers much more aware of the complexity involved in fair

and appropriate assessment of student writing abilities. It is now
well recognized that no approach to writing assessment is without
problems. Recent research has also pointed out that different
types of assessment are better suited to distinct instructional con-
texts, student abilities and goals, teacher preferences and
purposes, and institutional expectations.
f Wilting assessment research and practice has certainly been
confronting the limitations inherent in the evaluation of a com-
plex advanced set of cognitive processes. Earlier exploration of
writing skills was, perhaps, too confident that writing development
could be readily assessed. The current discussion of assessment,
and particularly of large-scale assessment, suggests a reasoned
humility in the face of an extremely difficult task (assuming it is to
be done fairly, appropriately, and reliably). Future work on writing
assessment will continue to explore ways to meet teacher and insti-
tutional needs and, at the same time, to provide students with
feedback and information that will instruct as well as assess (Camp
1993, White 1994).
Moreover, much practical assessment research now reflects
recent changes in assessment theory. Construct validity is now
seen by most assessment specialists as the driving force behind
choices for appropriate assessment; that is, writing assessment
should match its criteria to the writing that students are expected
to carry out after they have completed writing courses, to reflect
current notions of writing ability, and to do so in a socially respon-
sible m anner (Camp 1993, White 1993).
For issues surrounding the development and implementation of
large-scale assessment, notions such as reliability, content validity,
and concurrent validity are now seen as aspect of construct validity
rather than as independent considerations (Camp 1993, Hamp-
Lyons 1991c, Messick 1989). A num ber of consequences follow
from this emphasis on construct validity: (1) there is a greater
effort to investigate carefully the types of writing that are important
for given groups of students; (2) there is greater interest in the atti-
tudes of teachers towards types of writing skills; and (3) there is
greater attention to the relations between writing instruction, writ-
ing assessment, and later uses of writing skills. Writing assessment
must be seen not only as reflecting appropriate instructional goals
but also as an active component in writing instruction and an inte-
gral part of the students’ learning processes.
14
Conclusions: Writing in English
In the chapters which precede this authorial last gasp, we have

tried to summarize at considerable length the various theoretical
strands that have been explored by applied linguists and others in
the recent past, and to draw these strands together into a coher-
ent picture of where we have arrived in trying to understand the
nature of a written text. Given this summary of what written text is
and how it is created, we have tried to suggest ways in which writ-
ing might be taught that are consistent with what is known about
the nature of text, the composing process, and the surrounding
social context.
In a sense, we have not really dealt with the universal theory
and practice of the teaching of writing; rather, we have looked at
these issues in the context of the English language. We believe
that what is true of the nature of text in English, and what applies
to the teaching of writing in English, is probably also true in other
languages, but we do not know that to be so, and therefore we
make no claims for what we have said beyond the limits of English.
We welcome commentary from scholars working in other lan-
guages to provide confirmation or disconfirmation of our findings
and recommendations.
At this point, a final recapitulation of the volume would, we
fear, tem pt us to add yet another full chapter, so we shall refrain.
We would like, however, to reaffirm a small set of conclusions
which deserve special consideration. From our survey, we can
claim that:
■ the teaching of writing is separate and distinct from the teach-
ing of syntactic accuracy and the teaching of various text
conventions (e.g. spelling, punctuation);
422
Conclusions: Writing in English 423
■ the learning of writing is more likely to succeed in conditions in

which students find themselves wanting to communicate real
messages to real audiences;
■ the learning of writing is more likely to succeed in conditions in
which extensive support is offered to the learner from prior
text, in conditions in which there is peer involvement in the
preparatory stages for writing, and in conditions in which there
is also support from the teacher.
Teachers need not only offer the learner support and encourage-
m ent but must also serve as master craftsmen (and women) to
whom the learner is apprenticed - the learner may then, over
time, become a functioning member of a community in which the
literate discourse function serves a real and viable purpose.
We have claimed that learners who come from communities
which practise other sorts of discourse, or in which literate dis-
course served either no viable function or an entirely different
real purpose, are at a disadvantage - not because they are igno-
rant or stupid, but rather because they have not realized or
accepted the registers and purposes of discourse in the academic
community. An important aspect of teaching such learners is sim-
ply to make them aware of the purposes served by academic
discourse, but there is no guarantee that such awareness will make
learners wish to participate in this particular type of literate dis-
course. At least, if learners choose not to participate, they will
understand what they have chosen not to do and will be able to
make an informed decision.
We have claimed that there are im portant differences within lit-
erate academic discourse practices between tasks which require
learners to draw only upon their internal resources and tasks
which require learners to go beyond themselves, to draw upon
other text, and to use the writing process as a heuristic for the dis-
covery of new meaning.
Having provided and supported these various research views, we
have tried to develop a series of suggestions for ways in which the
objectives deriving from theory can be put into practice in the
classroom. We have divided this discussion into suggestions applic-
able to the beginning, intermediate, and advanced levels of
writing instruction. We recognize that such a division is arbitrary
and ultimately unsatisfactory; there are almost infinite sets of con-
ditions that falsify such a division of the learning world. Indeed,
we understand that, in a system that is age-graded, individuals may

enter at any point at the proper age level but be entirely unpre-
pared to benefit from instruction.
In another sense, the learning world can be divided into three
other groupings: kinesic, tactic, and legeric. This separation, not
based on age-grading, may help us to understand why certain indi-
viduals are not served by the system and may provide insights
which will perm it the system to adapt to the needs of misplaced
individuals.
Kinesic learning involves physical coordination of various types,
ranging from basic hand/eye coordination to the complex
process involved in wrapping one’s tongue around an unfamiliar
phonology and the equally complex problem of manipulating
some sort of writing implement to shape the graphemic symbols
of an unknown orthographic system. Educational systems are rela-
tively proficient at this sort of teaching.
Tactic learning can best be illustrated by an analogy. If one goes
to a strange city and hires a taxi to go from the airport to a central
hotel, one expects the taxi driver to go through a fairly complex
navigational analysis, considering available routes, traffic patterns
in relation to time of day, temporary obstacles (e.g. caused by con-
struction) along the way, and one expects the driver to arrive at
the hotel in an efficient and economical way. In working this navi-
gational problem, the taxi driver functions within a closed system;
there are, after all, a limited num ber of ways to move between the
airport and the hotel, and there are limited choices with respect
to traffic patterns and obstacles. In language, this compares to
learning the syntactic system and the lexicon of a language.
Educational systems are quite good at inculcating tactic learning.
Legeric learning, too, can best be illustrated by an analogy. If one
goes to see a physician, complaining of some more or less local-
ized pain, one expects the physician to diagnose the illness and to
prescribe an appropriate treatment. In order to do so, the physi-
cian must function in an open-ended system in which the
alternatives are virtually infinite. Employing laboratory tests, phys-
ical examinations, intuition, experience, and perhaps other skills,
the physician can accomplish the objective and the patient can
become well. In language, legeric learning implicates the mastery
of the (educationally valued) rhetorical system - its discourses and
its pragmatics. Educational systems are not very expert in teaching
legeric skills; on the contrary, they tend to assume that students
Conclusions: Writing in English 425
come fully equipped with such skills, or they do not even recog-
nize the nature of such skills. As a result, students who are
equipped with appropriate legeric skills tend to succeed while stu-
dents who are not tend to fail, to be stigmatized by the system.
A society can be described in terms of the relative value it
places on these three learning strands. Most educational systems
begin paying attention to the kinesic skills at the earliest possible
moment. It is assumed that all individuals within the normative
ranges will easily acquire such skills. Except in the USA where
professional athletes are outrageously rewarded, these skills are
not normally highly valued, and it does not take long to acquire
them - most elementary school students can be said to have mas-
tered these skills. In most societies, tactic skills - the skills
involved in driving a taxi - are not highly regarded, and it does
not take long to acquire them. Taxi drivers are not too well com-
pensated; indeed, they tend to fall towards the lower end of the
socioeconomic scale. (This is not to say that there are no soci-
eties in which navigational skills are highly valued; in some
Pacific Island societies, the navigator is among the most valued
members of society.) On the other hand, in most societies, leg-
eric skills are highly valued. Physicians take many years to acquire
the licensure necessary to practise their trade and the skills req-
uisite to the acquisition of a licence. They are among the most
highly valued members of the society and are rewarded with
handsome incomes. (This has not always been the case; in the
seventeenth and eighteenth centuries, surgery in the USA was
undertaken by barbers, and physicians did not have high social
status. As the paradigms of medicine changed, as the disciplines
of medicine redefined themselves, the social value of physicians
increased.)
We would want to argue that the teaching of writing in many
classrooms remains primarily focused on mastery at the kinesic
and tactic levels; that is, instead of being assisted to acquire
advanced skills involved in writing, students are taught the skills
involved in producing the orthographic system (whether with a
pen or with a word processor) and skills implicating surface struc-
ture features like grammar and spelling. Students are, in short,
instructed essentially in the conventions associated with writing.
We would want to argue further that the teaching of writing as we
mean to discuss it implicates the legeric level - a level not well
understood by researchers let alone by teachers. It is at the legeric
level that experience and intuition combine, and the processes

become a heuristic for the discovery of new knowledge.
To the extent that classes concerned with writing limit student
performance to kinesic and tactic skills, and restrict students to
operating within closed systems depending entirely on their inter-
nal resources, such classes are actually teaching a very limited
form of writing. We have attempted in this volume to provide the
rationale for moving the teaching of writing to the legeric level
and then to provide the pedagogical means for doing so.
We understand that students begin at different stages; if a stu-
dent enters the system with only kinesic skills, at whatever level,
then the system has the responsibility to orient instruction suffi-
ciently to perm it that student to acquire tactic skills as part of
acquiring legeric skills. Many students come equipped with tactic
skills; after all, writers of graffiti, for example, are well equipped
with tactic skills, but may be lacking legeric skills. In order for stu-
dents to acquire legeric skills, they must want to do so, and they
will want to do so only if they see that such skills have value in
their society and their lives.
We understand further that the task we have undertaken is not
complete. As our knowledge of the structure of written texts
expands, and as our awareness of the means of inculcating legeric
skills increases, the teaching of writing will change. To the best of
our knowledge, we have provided a synthesis of where scholarship
has arrived. We look forward to the next cycle of research which
will expand or alter the underlying paradigms and will result in
still more effective means for the teaching of writing. We under-
stand that each generation of scholars stands on the shoulders of
its parent generation and thus is able to see farther. We welcome
this probability of greater insight through continued research on
writing.
Seventy-five themes for writing
instruction
This list is provided tofacilitate easy access to a major part of this book.
C hapter 10: T each in g w riting at b eg in n in g le v els
1 Using print in the classroom 269

2 Writing daily 270
3 Prompting discussion prior to writing 270
4 Tapping student interest 271
5 Providing reasons for writing 271
6 The use of sheltered instruction 272
7 Literacy scaffolding 274
8 Attention to vocabulary development 274
9 Responding and feedback 275
10 Helping those who cannot get started 276
11 Using big books and Language Experience Activities 277
12 Generating word lists 278
13 Working with phrases and sentences 281
14 Copying and dictation 284
15 Working with poems 284
16 Working with pictures 286
17 Telling stories 289
18 Writing letters 290
19 Providing descriptions, directions, and procedures 291
20 Keeping records and writing reports 293
21 Using a dialogue journal 295
22 Publishing completed work 296
23 Establishing a writing corner 296
24 Including extensive reading 297
25 Providing early access to word processing 298
428 General editor's preface
C hapter 11: T each in g w riting at in term ed ia te lev els
1 Cooperative learning and group work 306

2 Content-based instruction 308
3 Audience awareness 309
4 Free writing and brainstorming 310
5 Semantic mapping and graphic organizers 311
6 Strategy instruction 313
7 Peer response groups 315
8 The writing process 316
9 Controlled, guided, parallel, and free writing 317
10 Awareness of language and genre structure 318
11 Gathering supporting information for writing 321
12 Developing content-based instruction 322
13 Summaries and responses 323
14 Awareness of language uses 325
15 Editing 326
16 Autobiographies and biographies 328
17 Surveys and questionnaires 328
18 Newspapers and media 329
19 Informational reports 331
20 Personal writing 333
21 Sentence combining 335
22 Portfolios 336
23 Values clarification 336
24 Double-entry notebooks 338
25 Presenting the text 338
C hapter 12: T each in g w riting at ad van ced lev els
1 Building a climate for writing 343

2 Working with multiple resources 344
3 Reading critically 345
4 Guiding discussion 347
5 Exploratory writing and writing exercises 348
6 Outlining and structured overviews 349
7 Using heuristics for organzing writing 350
8 Using patterns of essay organization 352
9 Paragraphing and text formatting 353
10 Starting writing and ending writing 354
11 Distinguishing fact from opinion 355
Seventy-five themes for writing instruction 429
12 Developing appropriate detail 357

13 Responding to student writing 357
14 Presenting texts for class revision 359
15 Teacher-student conferencing 360
16 Critical review writing 362
17 Argument writing - taking a position 364
18 Case study reports and simulations 364
19 Creative and interpretive writing 365
20 The research paper 366
21 Content-based writing courses and writing across the
curriculum 368
22 Discourse communities 370
23 Working with visual displays 370
24 Working with media 372
25 Style and the individual writer 373
Page Intentionally Left Blank
Bibliography
A b rah am son , D . 1985. Creatures of invention. National Wildlife 2 3 , 2. 25-8.

A ck erm an , I. 1993. T he prom ise o f writing to learn. Written Communication
10, 334-70.
A d a m s , M. 1989. Beginning to read: Thinking and learning about print.
L ondon and Cam bridge, MA: MIT Press.
A l d e r s o n , C. 1987. An overview o f ESL/EFL testing in Britain. In
C. A l d e r s o n , K . K r a h n k e an d C. S t a n s f i e l d (eds) Reviews of English
language proficiency tests. W ashington, DC: TESOL Publications. 3-4.
A l d e r s o n , C . 1993a. Judgm ents in language testing. In D . D o u g l a s and
C . C h a p e l l e (eds) A new decade in language testing research. Alexandria,
V A : TESOL Publications. 46-57.
A l d e r s o n , C . 1993b. T he relationship between gram m ar an d reading in
an English for academ ic purposes test battery. In D . D o u g l a s and
C . C h a p e l l e (eds) A new decade in language testing research. Alexandria,
V A : TESOL Publications. 203-19.
A l d e r s o n , C. and G. B u c k . 1993. Standards in testing: A study of the practice
of UK exam ination boards in EFL/EL testing. Language Testing 10, 1-26.
A l d e r s o n , C., K K r a h n k e an d C. S t a n s f i e l d (eds). 1987. Reviews of
English language proficiency tests. W ashington, DC: TESOL Publications.
A l l a e i , S . and U. C o n n o r . 1991. Using perform ative assessment instru-
m ents with E S L student writers. In L . H a m p -L y o n s (ed.) Assessing second
language uniting in academic contexts. N orw ood, NJ: Ablex. 227-40.
A lle n , R.V. 1976. Language experiences in communication. Boston:
H oughton Mifflin.
A m n e s t y I n t e r n a t i o n a l . 1984. Exiled South Korea leader calls for
Amnesty’s continued h um an rights support. Amnesty International
Newsletter. (December, 1984) 1.
A n d e r s o n , J. 1990. Cognitive psychology and its implications, 3rd edn. O xford
and New York: Freem an.
A n d e r s o n , P. 1985. W hat survey research tells us about writing at work. In
L. O d e l l and D. G o s w a m i (eds) Writing in nonacademic settings. New
York: Guilford. 3-83.
431
432 Bibliography
A n son , C. (ed.). 1989. Writing and response: Theory, practice, and research.
U rbana, IL: N ational Council o f Teachers o f English.
A p p l e b e e , A . 1981. Writing in the secondary schools. U rbana, IL: National
Council of Teachers of English.
A p p l e b e e , A . 1984. Contexts for learning to unite. Norw ood, NJ: A b le x .
A p p l e b e e , A . 1986. Problem s in process approaches: Toward a reconcep-
tualization o f process instruction. In A . P e t r o s k y and D. B a r t h o l o m a e
(eds) The teaching of writing. Chicago, IL: National Society of the Study
of E ducation (University o f Chicago Press). 95-113.
A p p l e b e e , A ., J . L a n g e r , L . J e n k i n s , I. M u l l i s and M . F o e r t s c h . 1990a.
Learning to write in our nation's schools: Instruction and achievement in 1988
at grades 4, 8, and 12. Princeton, N J: Educational Testing Service.
A p p l e b e e , A , J. L a n g e r and I. M u l l i s . 1986. Writing: Trends across the
decade, 1974-84. Princeton, NJ: Educational Testing Service.
A p p l e b e e , A ., J . L a n g e r , I. M u l l i s and L . J e n k in s . 1990b. The writing report
card, 1984-88. Princeton, N J: Educational Testing Service.
A p p l e b e e , A ., J. L a n g e r , I. M u l l i s , A . L a t h a m and C. G e n t i l e . 1994. NAEP
1992: Writing report card. W ashington, DC: Office o f Educational
Research and Im provem ent, US D epartm ent of Education. [R eport
#23-W01.]
A r m b r u s t e r , B. 1991. Framing: A technique for im proving learning from
science texts. In C. S a n t a an d D. A l v e r m a n n (eds) Science learning:
Processes and applications. Newark, DE: In ternational Reading
Association. 104-13.
A r m b r u s t e r , B., T. A n d e r s o n an d J. M e y e r . 1991. Im proving co n ten t area
reading using instructional graphics. Reading Research Quarterly 26,
393-416.
A r n o v e , R. and H. G r a f f (eds). 1987. National Literacy campaigns. L ondon
and New York: Plenum .
A r o n s o n , E., N. B l a n e y , C. S t e p h a n , J. S ik e s and M. S n a p p . 1978. The
figsaw classroom, Beverley Hills, C A /L ondon: Sage.
A t k i n s o n , D. 1 9 9 1 . Discourse analysis and w ritten discourse conventions.
In W. G r a b e et al. (eds) Annual review of applied linguistics, 11.
Cam bridge, E ngland and New York: Cam bridge University Press.
57-76.
A t k i n s o n , D. 1993. A historical discourse analysis o f scientific research
writing from 1675 to 1975: T he case o f the Philosophical Transactions of
the Royal Society of London. Los Angeles: University o f S outhern
California. Ph.D. diss.
A u , K., D. C r o w e l l , C . J o r d a n , K. S l o a t , G. S p e i d e l , T . K l e i n and
R. T h a r p . 1986. D evelopm ent an d im plem entation o f the KEEP
reading program . In J . O r a s a n u (ed.) Reading comprehension: From
research to practice. Hove, Sussex an d Hillsdale, NJ: L. Erlbaum.
235-52.
A u , K. and C. J o r d a n . 1981. Teaching reading to Hawaiian children:
Bibliography 433
Finding a culturally appropriate solution. In H. T r u e b a , G . G u t h r i e ,

and K. Au (eds) Culture and the bilingual classroom: Studies in classroom
ethnography. Rowley, MA: Newbury House. 139-52.
B a c h m a n , L. 1990. Fundamental considerations in language testing. O xford
and New York: O xford University Press.
B a k e r , E ., H . O ’N e i l a n d R. L in n . 1 9 9 3 . P o l ic y a n d v a lid it y p r o s p e c t s f o r
p e r f o r m a n c e - b a s e d a s s e s s m e n t . American Psychologist 48, 1 2 1 0 - 1 8 .
B a k h t i n , M. 1981. The dialogic imagination. (C. Em erson an d M. Holquist,
Trans.) Austin, TX: University of Texas Press.
B a l d a u f , R. 1986. Linguistic constraints on participation in psychology.
American Psychologist 41, 220-4.
B a l d a u f , R. and B . J e r n u d d . 1983. Language o f publications as a variable
in scientific com m unication. Australian Review of Applied Linguistics 6,
97-108.
B a l d a u f , R. and B . J e r n u d d . 1987. Academic com m unication in a foreign
language: T he exam ple o f Scandinavian psychology. Australian Review of
Applied Linguistics 10, 98-117.
B a m b e r g , B . 1983. W hat makes a text coherent? College Composition and
Communication 34, 417-29.
B a n g e r t - D r o w n s , R. 1993. T he word processor as an instructional tool:
A meta-analysis of word processing in writing instruction. Review of
Educational Research 63, 69-93.
B a r b e r , C. 1962/1985. Some m easurable characteristics o f m o d ern sci-
ence prose. In J. S w a l e s (ed.) Episodes in ESP. H em el H em psted and
Englewood Cliffs, NJ: Prentice Hall. 1-14.
B arsalou, L. 1992. Cognitive psychology: An overview for cognitive scientists.
Hove, Sussex and Hillsdale, NJ: L. Erlbaum.
B a r s a l o u , L . 1993. Frames, concepts, and conceptual fields. In A. L e h r e r
and E. K i t t a y (eds) Frames, fields, and contrasts. Hove, Sussex and
Hillsdale, N J: L . Erlbaum. 21-74.
B a r t h o l o m a e , D. 1985. Inventing the university. When a xvriter can’t write.
New York: Guilford. 134-165.
B a r t h o l o m a e , D. and A. P e t r o s k y . 1986. Facts, artifacts and counterfacts:
Theory and method for a reading and writing course. U pper Montclair, NJ:
B oynton/C ook.
B a r t o n , D. and R. I v a n ic (eds). 1991. Writing in the community. L ondon
and Newbury Park, CA: Sage.
B a s s o , K. 1974. T he ethnography of writing. In R. B a u m a n and J. S h e r z e r
(eds) Explorations in the ethnography of speaking. Cam bridge, E ngland and
New York: Cam bridge University Press. 425-32.
B a t e s o n , G. 1979. M ind and nature: A necessary unity. L ondon and New
York: Bantam Books.
B a z e r m a n , C. 1983. Scientific writing as a social act. In P A n d e r s o n ,
J. B r o c k m a n and C. M i l l e r (eds) New essays in technical writing and
communication. Farm ingdale, NY: Baywood. 156-84.
434 Bibliography
B azerm an , C. 1985. Physicists writing physics: Schem a-laden purposes and

purpose-laden schema. Written Communication 2, 3-23.
B a z e r m a n , C. 1988. Shaping written knowledge. Madison, WI: University o f
Wisconsin Press.
B a z e r m a n , C. 1991. Discourse analysis and social construction. In
W. G r a b e et al. (eds) A nnual review of applied linguistics, 11. Cam bridge,
E ngland and New York: Cam bridge University Press. 77-83.
B a z e r m a n , C. 1 9 9 3 . Constructing experience. Carbondale, IL: S outhern
Illinois University Press.
B e a c h , R. 1 9 8 9 . Showing students how to assess: D em onstrating tech-
niques for response in the writing conference. In C. A n s o n (ed.) Writing
and response. Urbana, IL: National Council of Teachers of English. 1 2 7 - 4 8 .
d e B e a u g r a n d e , R. 1980. Text, discourse and process. N orw ood, NJ: Ablex.
d e B e a u g r a n d e , R. 1982. Psychology and com position: Past, present,
future. In M. N y s t r a n d (ed.) What writers know: The language, process,
and structure of written composition. L ondon and N e w York: Academic
Press. 211-67.
d e B e a u g r a n d e , R. 1984. Text production: Toward a science of composition.
Norw ood, NJ: Ablex.
d e B e a u g r a n d e , R. 1985. Writing step by step. L ondon an d New York:
H arcourt Brace.
d e B e a u g r a n d e , R. and W. D r e s s l e r . 1981. Introduction to text linguistics.
L ondon and New York: Longm an.
B e c k , I., M . M c K e o w n , G . S i n a t r a an d J . L o x t e r m a n . 1 9 9 1 . Revising
social studies text from a text-processing perspective: Evidence of
im proved com prehensibility. Reading Research Quarterly 26, 2 5 1 - 7 6 .
B e l a n o f f , P. and M. D i c k s o n (eds). 1991. Portfolios: Process and product.
Portsm outh, NH: Boynton/C ook.
B e n s o n , J. and W. G r e a v e s (eds). 1985. Systemic perspectives on discourse.
2 Volumes. Norw ood, NJ: Ablex.
B e r e i t e r , C. 1990. Aspects o f an educational learning theory. Review of
B e r e i t e r , C . and M. S c a r d a m a l i a . 1985. Cognitive coping strategies
and the problem of ‘in ert know ledge’. In S . C h ip m a n , J. S e g a l and
R. G l a s e r (eds) Thinking and learning skills: Volume 2. Research and open
questions. Hove, Sussex an d Hillsdale, NJ: L. Erlbaum. 65-80.
B e r e i t e r , C. and M. S c a r d a m a l i a . 1987. The psychology of -written composi-
tion. Hillsdale, NJ: L. Erlbaum.
B e r e i t e r , C. and M. S c a r d a m a l i a . 1989. Intentional learning as a goal of
instruction. In L. R e s n i c k (ed.) Knowing, learning, and instruction: Essays
in honor of Robert Glaser. Hillsdale, NJ: L. Erlbaum . 361-91.
B e r e i t e r , C. and M. S c a r d a m a l i a . 1993. Surpassing ourselves: An inquiry
into the nature and complications of expertise. Chicago: O pen C ourt Press.
B e r k e n k o t t e r , C. and T. H u c k i n . 1993. R ethinking genre from a
sociocognitive perspective. Written Communication 10, 475-509.
Bibliography 435
Be r k e n k o t t e r , C. and T. H u c k i n . 1995. Genre knowledge in disciplinary com-

munication. Hillsdale, NJ: L. Erlbaum.
B e r l i n , J. 1984. Writing instruction in nineteenth century American colleges.
C arbondale, IL: Southern Illinois University Press.
B e r l i n , J. 1987. Rhetoric and reality: Writing instruction in American colleges,
1900-1985. C arbondale, IL: Southern Illinois University Press.
B e r m a n , R. and D. S l o b i n . 1994. Relating events in narratives: A crosslinguis-
tic and developmental study. Hove, Sussex and Hillsdale, NJ: L. Erlbaum.
B e r n h a r d t , E. 1994. A co n ten t analysis o f reading m ethods texts: W hat
are we told about the nonnative speaker o f English? Journal of Reading
Behavior 26, 159-89.
B e r n s t e i n , B . 1972a. Social class, language and socialization. In P.P.
G i g l i o l i (ed.) Language and social context. H arm ondsw orth, England:
Penguin. 157-78.
B e r n s t e i n , B . 1972b. A critique of the concept of com pensatory educa-
tion. In C . C a z d e n , V. J o h n and D. H y m e s (eds) Functions of language in
the classroom. New York: Teachers College Press. 135-51.
B e r n s t e i n , B . 1 9 9 0 . Class, codes and control 4: The structuring of pedagogic
discourse. L ondon and New York: Routledge.
B e r t h o f f , A. 1981. The making of meaning. U pper Montclair, NJ:
B oynton/C ook.
B e s n i e r , N. 1 9 8 8 . T he linguistic relationships o f spoken and written
N ukulaelae registers. Language 64, 707-36.
B h a t i a , V .K . 1 9 9 3 . Analysing Genre: Language use in professional settings.
L ondon and New York: Longm an.
B i b e r , D. 1988. Variation across speech and writing. Cam bridge, E ngland and
New York: Cam bridge University Press.
B i b e r , D. 1989. A typology o f English texts. Linguistics 27, 3-43.
B ib e r , D . 1992. O n t h e c o m p l e x i t y o f d i s c o u r s e c o m p le x it y : A m u l t i -
d i m e n s i o n a l a n a ly s is . Discourse Processes 15, 133-63.
B i b e r , D. 1994. An analytic framework for register studies. In D. B i b e r
and E. F i n e g a n (eds) Sociolinguistic perspectives on register. O xford and
New York: O xford University Press. 31-56.
B i b e r , D. 1995. Cross-linguistic patterns of register variation: A multi-dimen-
sional comparison of English, Tuvaluan, Korean, and Somali. Cambridge,
E ngland and New York: Cam bridge University Press.
B i b e r , D. and E. F i n e g a n . 1988. Adverbial stance types in English.
Discourse Processes 11, 1-34.
B ib e r , D. and E. F i n e g a n . 1989. Styles o f stance in English: Lexical and
gram m atical m arkings o f evidentiality and affect. Text 9, 93-124.
B i c k n e r , R. and P. P e y a s a n t i w o n g . 1988. Cultural variation in reflective
writing. In A. P u r v e s (ed.) Writing across languages and cultures. L ondon
and Newbury Park, CA: Sage. 160-74.
B i s s e x , G. 1980. Gyns at work: A child learns to read and xvrite. Cambridge,
MA: H arvard University Press.
436 Bibliography
B iz z e ll, P. 1982. College com position: Initiation into the academ ic

discourse community. Curriculum Inquiry 12, 191-207.
B i z z e l l , P. 1986a. W hat happens when the basic writer com es to college?
College Composition and Communication 37, 294-310.
B i z z e l l , P. 1986b. Foundationalism an d anti-foundationalism in com posi-
tion studies. Pre/Text 7, 37-56.
B i z z e l l , P. 1993. Academic discourse and critical consciousness. Pittsburgh:
University of Pittsburgh Press.
B l a i r , H. 1783/1965. Lectures on rhetoric and belles lettres (H. H arding, ed.).
C arbondale, IL: S outhern Illinois University Press.
B l o k , H. and K. D e G l o p p e r . 1992. Large scale writing assessment. In
L. V e r h o e v e n and J.H.A.L. D e J o n g (eds) The construct of language
proficiency. Philadelphia: J. Benjamins. 101-11.
B l o o m e , D. and J. G r e e n . 1992. E ducational contexts o f literacy. In
W. G r a b e et al. (eds) Annual review of applied linguistics, 12. Literacy.
Cambridge, England and New York: Cambridge University Press. 49-70.
B o g g s , S. 1985. Speaking, relating and learning. Norw ood, NJ: Ablex.
B o n k . W. 1990. A synthesis o f social cognition and writing research.
Written Communication! , 136-63.
B o o m e r , G. 1985. Fair dinkum teaching and learning: Reflections on literacy
and power. U pper Montclair, NJ: B oynton/C ook.
B r a c e y , G. 1995. T he assessor assessed: A ‘revisionist’ looks at a critique of
the Sandia rep o rt .Journal of Educational Research 88, 136-44.
B r e l a n d , H., R. C a m p , R . J o n e s , M . M o r r i s a n d D. R o c k . 1987. Assessing
writing skill. N e w Y ork: C o l l e g e E n t r a n c e E x a m i n a t i o n B o a r d . [ R e s e a r c h
M o n o g r a p h # 1 1 .]
B r in to n , D., M.A. Snow and M. W esch e (eds). 1989. Content-based second
language instruction. New York: Newbury House.
B r itto n , B. and J. B l a c k . 1985a. U nderstanding expository text: From
structure to process and world knowledge. In B . B r i t t o n and J. B l a c k
(eds) Understanding expository text. Hove, Sussex and Hillsdale, NJ: L.
Erlbaum . 1-9.
B r i t t o n , B . a n d j. B l a c k (eds). 1985b. Understanding expository text. Hove,
Sussex and Hillsdale, NJ: L. Erlbaum.
B r i t t o n , B . and S. G u l g o z . 1991. Using K intsch’s com putational m odel
to improve instructional text: Effects o f repairing inference calls on
recall and cognitive structures. Journal of Educational Psychology 83,
329-45.
B r i t t o n , B ., A. W o o d w a r d and M. B i n k l e y (eds). 1993. Learning from
textbooks: Theory and practice. Hove, Sussex and Hillsdale, NJ: L. Erlbaum.
B r i t t o n , J. 1983. Shaping at the p o in t o f utterance. In A. F r e e d m a n ,
I. P r i n g l e and J. Y a l d e n (eds) Learning to unite: First language/second
language. L ondon and New York: Longm an. 13-19.
B r i t t o n , J., T. B u r g e s s , N. M a r t i n , A. M c L e o d , and H. R o s e n . 1975. The
development of writing ability (11-18). L ondon: Macmillan.
Bibliography 437
Br o w n, A. 1994. T he advancem ent o f learning. Educational Researcher 2 3 ,

4-12.
Br o w n , A. and J. C a m p i o n e . 1994. G uided discovery in a com m unity of
learners. In K. M c G i l l y (ed.) Classroom lessons: Integrating cognitive
theories. Cam bridge, MA: MIT Press. 229-70.
B r o w n , A. and A. P a l i n c s a r . 1989. G uided, cooperative learning and
individual knowledge acquisition. In L. R e s n i c k (ed.) Knowing learning,
and instruction: Essays in honor of Robert Glaser. Hove, Sussex and
Hillsdale, NJ: L. Erlbaum . 393-451.
B r o w n , G. and G. Y u l e . 1983. Discourse analysis. Cam bridge, England and
New York: Cam bridge University Press.
B r u c e , B., J.K. P e y t o n , an d T. B a t e s o n (eds.). 1993 Network-based class-
rooms: Promises and realities. Cam bridge an d New York: C am bridge
University Press.
B r u c e , B . and A. R u b i n . 1993. Electronic quills: A situated evaluation of using
computers for writing in classrooms. Hove, Sussex and Hillsdale, NJ:
L. Erlbaum.
B r u f f e e , J . 1986. Social construction, language, and the authority o f
knowledge: A bibliographic essay. College English 48, 773-90.
B r u n e r , J. 1983. Child’s talk: learning to use language. Oxford: O xford
University Press.
B r u t h i a u x , P. 1993. Child’s talk: learning to use language. Oxford: O xford
Univesity Press.
B r y s o n , M . and M . S c a r d a m a l i a . 1991. Teaching writing to students at
risk for academ ic failure. In B . M e a n s , C . C h e l e m e r an d M . K n a p p
(eds) Teaching advanced skills to at-risk students. San Francisco: Jossey-
Bass. 141-67.
B u l l o c k C o m m i t t e e . 1975. A language for life. London: D epartm ent of
Education and Science, H er Majesty’s Stationery Office.
C a l f e e , R. 1992. A uthentic assessment of reading and writing in the
elem entary classroom. In M. D r e h e r and W. S l a t e r (eds) Elementary
school literacy: Critical issues. Norw ood, MA: C hristopher G ordon.
211-39.
C a l k i n s , L. 1 9 8 3 . Lessons from a child. L ondon and Portsm outh, NH:
H einem ann.
C a l k i n s , L. 1986. The art of teaching writing. L ondon and Portsm outh, NH:
H einem ann.
C a m p , R. 1993. C hanging the m odel for the direct assessment o f writing.
In M. W i l l i a m s o n and B. H u o t (eds) Validating holistic scoringfor writing
assessment. Cresskill, NJ: H am pton Press. 45-78.
C a m p b e l l , C . 1990. W riting with o th ers’ words: Using background reading
text in academ ic com positions. In B. K r o l l (ed.) Second language writ-
ing. Cam bridge and New York: Cam bridge University Press. 211-230.
C a m p b e l l , G. 1776/1963. The philosophy of rhetoric (L. Bitzer, ed.).
Carbondale, IL: Southern Illinois University Press.
438 Bibliography
C a n a le , M. 1983. From com municative com petence to com municative

language pedagogy. In J. R i c h a r d s and R . S c h m i d t (eds) Language and
communication. L ondon and New York: Longm an.
C a n a l e , M., M. B e l a n g e r and N. F r e n e t t e . 1988. Evaluation o f m inority
student writing in first and second languages. In J. F i n e (ed.) Second
language discourse: A textbook of current research. N orw ood, NJ: Ablex.
147-65.
C a n a l e , M. and M. S w a in . 1980. T heoretical bases of com m unicative
approaches to second language teaching and testing. Applied Linguistics
1, 1-47.
C a n t o n i , G. 1987. Content-area language instruction. W okingham, E ngland
and Reading, MA: Addison-Wesley.
C a r l i s l e , R. 1989. T he writing o f Anglo an d H ispanic elem entary school
students in bilingual, subm ersion, an d regular program s. Studies in
Second Language Acquisition II, 257-80.
C a r lso n , S. 1988. Cultural differences in writing and reasoning skills. In
A. Purves (ed.) Writing across languages and cultures. L ondon and
Newbury Park, CA: Sage. 227-60
C a r l s o n , S., B . B r i d g e m a n , R. C am p and J. W a a n d e r s . 1985. Relationships
of admission test scores to zvriting performance of native and nonnative speakers
of English. P rinceton, NJ: Educational Testing Service. [TOEFL
Research R eport #19.]
C a r n ic e lli, T. 1980. T he writing conference: A one-on-one conversation.
In T. D o n o v a n and B. M c C l e l l a n d (eds) Eight approaches to teaching
composition. Urbana, IL: National Council of Teachers of English. 101-31.
C a r r e l l , P. 1982. Cohesion is n o t coherence. TESOL Quarterly 16,
479-88.
C a r r e l l , P. 1987. Text as interaction: Some im plications o f text analysis
and reading research for ESL com position. In U. C o n n o r and R.B.
K a p l a n (eds) Writing across languages: Analysis of L 2 text. Wokingham,
E ngland and Reading, MA: Addison-Wesley. 47-56.
C a r s o n , J. 1992. Becom ing biliterate: First language influences. Journal of
Second Language Writing 1, 37-60.
C a r s o n , J., P. C a r r e l l , S . S i l b e r s t e i n , B. K r o l l and P. K u e h n . 1990.
Reading-writing relationships in first and second language. TESOL
Quarterly 24, 245-66.
C a r s o n , J. and G. N e l s o n . 1994. W riting groups: Cross-cultural issues.
Journal of Second Language Writing 3, 17-30.
C a r t e r , M. 1990. T he idea o f expertise: An exploration o f cognitive and
social dim ensions of writing. College Composition and Communication 41,
265-86.
C a r t e r , R. 1990. W hen is a rep o rt n o t a report? O bservations from acade-
mic and nonacadem ic settings. In W. N a s h (ed.) The writing scholar:
Studies in academic discourse. L ondon an d Newbury Park: Sage. 171-91.
C a z d e n , C . 1988. Classroom discourse: The language of teaching and learning.
O xford and Portsm outh, NH: H einem ann.
Bibliography 439
Ca z d e n , C. and B. G r a y . 1992. Australian approaches to genre-based

writing instruction. Paper p resented at the 27th A nnual TESOL
Convention, Vancouver, BC, M arch 1992.
C e l c e -M u r c i a , M ., Z. D o r n y e i a n d S. T h u r e l l . 1994. C o m m u n i c a t i v e
c o m p e t e n c e : A p e d a g o g ic a lly m o tiv a te d fr a m e w o r k . P a p e r p r e s e n t e d a t
1994 A r i z o n a R o u n d T a b le o n L a n g u a g e . T u c s o n , A Z , F e b r u a r y 1994.
C e r n i g l i a , C ., K. M e d s k e r and U. C o n n o r . 1990. Im proving coherence
using com puter-assisted instruction. In U. C o n n o r and A. J o h n s (eds)
Coherence in writing. W ashington, D C : TESOL Publications. 229-41.
C h a f e , W. 1982. Integration and involvement in speaking, writing, and
oral literature. In D. T a n n e n (ed.) Spoken and written language.
Norw ood, NJ: Ablex. 35-54.
C h a f e , W. 1985. Linguistic differences pro d u ced by differences between
speaking and writing. In D. O l s o n , N. T o r r a n c e and A. H i l y a r d (eds)
Literacy, language, and learning. Cam bridge, England an d New York:
Cam bridge University Press. 105-23.
C h a f e , W. and J. N i c h o l s (eds). 1986. Evidentiality: The linguistic coding of
epistemology. Norw ood, NJ: Ablex.
C h a p e l l e , C ., W. G r a b e and M. B e r n s . 1993. Communicative language
proficiency: Definitions and implications for TOEFL 2000. [ETS In tern al
R eport.] Princeton, NJ: Educational Testing Service.
C h a r n e y , D. an d R. C a r l s o n . 1995. L earning to write in a genre: W hat
student writers take from m odel texts. Research in the Teaching; of English
29, 88-125.
C h e n g , P. 1985. An analysis o f contrastive rhetoric: English an d Chinese
expository prose, pedagogical implications, an d strategies for the ESL
teacher in a nin th grade curriculum . U npublished Ph.D. diss., T he
Pennsylvania State University.
C h e r r y , R. and P. M e y e r . 1993. Reliability issues in holistic assessment. In
M . W i l l i a m s o n and B. H u o t (eds) Validating holistic scoring for writing
C h o i , S. and M. B o w e r m a n . 1991. L earning to express m otion events in
English and Korean: T he influence o f language-specific lexicalization
patterns. Cognition 44, 83-121.
C h r i s t i e , F. 1985. Some c u rren t issues in first language writing develop-
m ent. Australian Review of Applied Linguistics 8, 27-54.
C h r i s t i e , F. 1989. Language developm ent in education. In R. H a s a n and
I. M a r t i n (eds) Language development: Language learning, learning culture.
N orw ood, NJ: Ablex. 152-98.
C h r i s t i e , F. 1992. Literacy in Australia. In W. G r a b e et al. (eds) A nnual
review of applied linguistics, 12. Literacy. Cam bridge, E ngland and New
York: Cam bridge University Press. 142-55.
C h r i s t i e , F. (ed.). 1990. Literacy for a changing world. H aw thorn, Victoria:
A ustralian Council for Educational Research.
C h r i s t i e , F. et al. 1991. Language as a resource for meaning. Report booklets 1 -4
and teacher manual. Sydney: H arco u rt Brace Jovanovich.
440 Bibliography
C h r is tie , F. et al. (eds). 1989. Writing in schools: Study guide. Geelong,

Victoria: Deakin University Press.
C l a n c y , P. 1986. T he acquisition o f com m unicative style in Japanese. In
B. S c h i e f f e l i n and E . O c h s (eds) Language socialization across cultures.
L ondon and New York: Cam bridge University Press. 213-50.
C l a p h a m , C . 1993. Is ESP testing justified? In D . D o u g l a s and
C . C h a p e l l e (eds) A new decade in language testing research. A lexandria,
VA: TESOL Publications. 257-71.
C la y , M. 1975. What did I zvrite: Beginning writing behavior. O xford and
Auckland, New Zealand: H einem ann Educational Books.
C l i f f o r d , J. 1981. Com posing in stages: T he effects o f a collaborative
pedagogy. Research in the Teaching of English 15, 37-53.
C l y n e , M. 1981. C u lt u r e a n d d i s c o u r s e s t r u c t u r e . Journal of Pragmatics 5,
61-6.
C l y n e , M. 1983. Linguistics and w ritten discourse in particular languages:
Contrastive studies: English and G erm an. In R.B. K a p l a n et al. (eds)
A nnual review of applied linguistics, 3. Cam bridge, E ngland and New
C l y n e , M. 1985. Language and society in the German-speaking countries.
C l y n e , M. 1987. Cultural differences in the organization o f academic
texts: English and G erm an. Journal of Pragmatics 11, 211-47.
C l y n e , M. 1991. T he sociocultural dim ension: T he dilem m a o f the
Germ an-speaking scholar. In H. S c h r o d e r (ed.) Subject-oriented texts.
New York: W alter deGruyter. 49-67.
C o c h r a n - S m i t h , M. 1991. Word processing and writing in elem entary
classrooms: A critical review of related literature. Review of Educational
Research 61, 107-55.
C o e , R. 1986. Teaching writing: T he process approach, hum anism , and
the context of ‘crisis.’ In S. D e C a s t e l l , A. L u k e and K. E g a n (eds)
Literacy, society and schooling: A reader. Cam bridge, E ngland and New
C o e , R. 1994. ‘An arousing and fulfillm ent of desires’: T he rhetoric of
genre in the process era - and beyond. In A, F r e e d m a n an d P. M e d w a y
(eds) Genre and the new rhetoric. L ondon and New York: Taylor 8c
Francis. 181-90.
C o h e n , A. 1990. Language learning. New York: Newbury House.
C o h e n , A. and M. Cavalcanti. 1990. Feedback on com position: Teacher
and student verbal reports. In B. K r o l l (ed.) Second language uniting.
Cam bridge, England and New York: Cam bridge University Press.
155-77.
C o h e n , G. 1983. The psychology of cognition, 2nd edn. L ondon and New
York: Academic Press.
C o l l e r s o n , J. (ed.). 1990. Writing for life. Rozelle, NSW: Prim ary English
Teaching Association.
Bibliography 441
C o llin s , A., J.S. B r o w n and S. N e w m a n . 1989. Cognitive apprenticeship:

Teaching the craft of reading, writing, and m athem atics. In L. R e s n i c k
(ed.) Knowing , learning, and instruction: Essays in honor of Robert Glazer.
Hove, Sussex and Hillsdale, N J: L. Erlbaum. 453-94.
C o l l i n s , H. 1985. Changing order: Replication and induction in scientific
practice. L ondon and Beverly Hills, C A : Sage.
C o l l i n s , J. and M. W i l l i a m s o n . 1984. Assigned rhetorical context and
semantic abbreviation in writing. In R. B e a c h and L. B r i d w e l l (eds)
New directions in composition research. New York: Guilford. 285-96.
C o m p r o n e , J. 1987. Literary theory and com position. In G. T a t e (ed.)
Teaching composition. Fort W orth, TX: Texas Christian University Press.
291-330.
C o n l e y , J. and W. O ’B a r r . 1990. Rules versus relationships: The ethnography
of legal discourse. Chicago: University o f Chicago Press.
C o n n o r , U. 1987. Research frontiers in writing analysis. TESOL Quarterly
21, 677-96.
C o n n o r , U. 1995. Contrastive rhetoric: Cross-cultural aspects of second language
writing. Cam bridge, E ngland and New York: Cam bridge University
Press.
C o n n o r , U. and K. A s e n a v a g e . 1994. Peer response groups in ESL writing
classes: How m uch im pact on revision. Journal of Second Language
W riting 3, 257-76.
C o n n o r , U. and M. F a r m e r . 1990. T he teaching o f topical structure
analysis as a revision strategy for ESL writers. In B. K r o l l (ed.) Second
language writing. Cam bridge, England and New York: Cam bridge
University Press. 126-39.
C o n n o r , U . a n d J . L a u e r . 1988. C r o s s - c u lt u r a l v a r ia t io n in p e r s u a s iv e
s t u d e n t w r it in g . I n A . P u r v e s ( e d . ) Writing across languages and cultures.
L o n d o n a n d N e w b u r y Park, CA: S a g e. 138-59.
C onnor, U. and P. M c C a g g . 1987. A contrastive study o f English exposi-
tory prose paraphrases. In U. C o n n o r an d R.B. K a p l a n (eds) Writing
across languages: Analysis of L 2 text. W okingham, England and Reading,
MA: Addison-Wesley. 73-86.
C o o k - G u m p e r z , J. (ed.). 1986. The social construction of literacy. Cam bridge,
C o o p e r , C . and S. G r e e n b a u m (eds). 1986. Studying uniting: Linguistic
approaches. L ondon and Beverly Hills, C A : Sage.
C o o p e r , C . and A. M a t s u h a s h i . 1983. A theory o f the writing process. In
M . M a r t l e w (ed.) The psychology of written language: A developmental
approach. Chichester, Sussex and New York: Jo h n Wiley. 3-39.
C o o p e r , M. 1986. T he ecology of writing. College English 48, 364-75.
C o o p e r , M. 1989. Why are we talking about discourse com munities? Or,
foundationalism rears its ugly h ead once m ore. In M. C o o p e r and
M. H o l z m a n (eds) Writing as social practice. Portsm outh, N H :
B oynton/C ook. 202-20.
442 Bibliography
Co o pe r , R. 1979. Language planning, language spread, and language

change. In J. A l a t i s and G.R. T u c k e r (eds) Language in public life.
G eorgetown University R ound Table on Languages and Linguistics
1979. W ashington, DC: Georgetown University Press. 23-50.
C o p e , B. and M. K a l a n t z i s (eds). 1993. The powers of literacy: A genre
approach to teaching writing. Pittsburgh: University o f Pittsburgh Press.
C o p e l a n d , J. and E. L o m a x . 1988. Building effective student writing
groups. In J. G o l u b (ed.) Focus on collaborative learning. U rbana, IL:
N ational Council of Teachers of English. 99-104.
C o r b e t t , E. 1971. (1) Classical rhetoric for the modern student, 2nd edn.
O xford and New York. O xford University Press.
C o u l t h a r d , M. (ed.). 1994. Advances in written text analysis. L ondon and
New York: Routledge.
C o u t u r e , B. (ed.). 1 9 8 6 . Functional approaches to writing: Research perspec-
tives. Norw ood, NJ: Ablex.
C o u t u r e , B. and J. R y m e r . 1993. Situational exigence: Com posing
processes on the jo b by w riter’s role and task value. In R . S p i l k a (ed.)
Writing in the workplace. Carbondale, IL: S outhern Illinois University
Press. 4-20.
Cox, B., T. S h a n a h a n and E. S u l s b y . 1990. G ood and p o o r elem entary
readers’ use of cohesion in writing. Reading Research Quarterly 25,
47-65.
Cox, B., T . S h a n a h a n and M. T i n z m a n n . 1991. C hildren’s knowledge of
organization, cohesion, and voice in written exposition. Research in the
Teaching of English 25, 179-218.
C r a n d a l l , J. 1993. C ontent-centered learning in the U nited States. In
W. G r a b e et al. (eds) Annual review of applied linguistics, 13. Issues in
second language teaching and learning. Cam bridge, England an d New
C r e s s y , D. 1980. Literacy and the social order. Cam bridge an d New York:
C am bridge University Press.
C r o w h u r s t , M. 1987. Cohesion in arg u m en t and narratio n at th ree grade
levels. Research in the Teaching of English 21, 185-201.
C r o w h u r s t , M. 1990. T he developm ent o f persuasive/argum entative
writing. In R. B e a c h and S. H y n d s (eds) Developing discourse practices in
adolescence and adulthood. Norw ood, NJ: Ablex. 200-23.
C r o w l e y , S. 1989. A n introduction to deconstruction. U rbana, IL: National
Council of Teachers o f English.
C r o w l e y , S. 1990. The methodical memory. Carbondale, IL: S outhern
C r o w l e y , S. 1995. C om position’s ethic o f service, the universal o f require-
m ent, and the discourse o f student need. Journal of Advanced
Composition 15, 227-39.
C u m m i n g , A. 1989. W riting expertise and second language proficiency.
Language Learning 39, 81-141.
Bibliography 443
Cu m m in g , A. 1990a. Expertise in evaluating second language com posi-

tions. Language Testing 7, 31-51.
C u m m i n g , A . 1990b. Metalinguistic and ideational thinking in second
language com posing. Written Communication! , 482-511.
C z u b a r o f f , J. 1989. T he deliberative character o f strategic discourse
debates. In H. S i m o n s (ed.) Rhetoric in the human sciences. L ondon and
N ewbury Park, CA: Sage. 28-47.
D a i k e r , D ., A. K e r e k and M . M o r e n b e r g (eds). 1985. Sentence combining:
A rhetorical perspective. C arbondale. IL: S outhern Illinois University
Press.
D a n e s , E 1974. Functional sentence perspective an d the organization of
text. In F. D a n e s (ed.) Papers on functional sentence perspective. T he
Hague: M outon. 106-28.
D ’A n g e l o , F. 1975. A conceptual theory of rhetoric. Cam bridge, MA:
W inthrop.
D ’A n g e l o , F. 1987. Aims, modes, and form s o f discourse. In G. T a t e (ed.)
131-54.
D a n t a s -W h i t n e y , M. and W . G r a b e . 1989. English an d Brazilian
Portuguese editorial prose. Paper p resented at the 24th A nnual TESOL
Convention, San A ntonio, TX, April 1989.
D a v i s , A., M. C l a r k e and L. R h o d e s . 1994. E xtended text an d the writing
proficiency of students in urb an elem entary schools. Journal of
Educational Psychology 86, 556-66.
D e g e n h a r t , R.E. (ed.). 1987. Assessment of student uniting in an interna-
tional context. Jyvaskyla: Institute for Educational Research, University of
Jyvaskyla.
D e M a u r o , G . 1 9 9 2 . A n investigation of the appropriateness of the TOEFL test as
a matching variable to equate TWE topics. Princeton, NJ: Educational
Testing Service. [TOEFL Research Reports #37.]
D e m u t h , K. 1990. M aturation and the acquisition o f the Sesotho passive.
Language 65, 56-81.
D e r e w i a n k a , B. 1990. Exploring how texts work. Rozelle, NSW: Prim ary
English Teaching Association (dist. by H einem ann, Portsm outh,
NH ).
D i e d e r i c h , P 1974. Measuring growth in English. U rbana, IL: N ational
D i l l o n , G. 1981. Constructing texts. Bloom ington, IN: Indiana University
Press.
D i l l o n , G. 1983. Interpersonal functions o f textual analysis. Lecture p re-
sented at the University o f Southern California. March, 1983.
D i P a r d o , A. and S. F r e e d m a n . 1988. Peer response groups in the writing
classroom: T heoretical foundations and new directions. Review of
D i x o n , C. and D . N e s s e l . 1 9 8 3 . Language experience approach to reading (and
writing). Hayward, CA: Alemany Press.
444 Bibliography
D o b r in , D . 1986. Protocols once m ore. College English 48, 713-25.

D o b r in , D . 1989. Writing and technique. U rbana, IL: N ational Council of
Teachers of English.
D o h e n y - F a r i n a , S. 1986. W riting in an em erging organization: An eth n o -
graphic study. Written Communication 3, 158-85.
D u d l e y - E v a n s , T. 1989. An outline o f the value o f genre analysis in LSP
work. In C. L a u r e n and M. N o r d m a n (eds) Special language. Clevedon,
England and Philadelphia: M ultilingual Matters. 72-9.
D u n b a r , S ., D . K o r e t z a n d H . H o o v e r . 1 9 9 1 . Q u a l it y c o n t r o l in t h e
Applied Measurement
d e v e lo p m e n t a n d u se o f p e r fo r m a n c e a sse ssm e n ts.
in Education 4, 2 8 9 - 3 0 3
D y s o n , A. 1989. Multiple worlds of child writers: Friends learning to xmite. New
York: Teachers College Press.
D y so n , A. 1993. Social worlds of children learning to xmite. New York:
Teachers College Press.
E a g l e t o n , T. 1983. Literary theory: An introduction. Minneapolis: University
of M innesota Press.
E d e , L . and A. L u n s f o r d . 1984. A udience ad d ressed /au d ien ce invoked:
T he role of audience in com position theory and pedagogy. College
Composition and Communication 35, 155-71.
E d e ls k y , C. 1982. W riting in a bilingual program : T he relation of LI and
L2 texts. TESOL Quarterly 16, 211-28.
E d e ls k y , C. 1986. Writing in a bilingual program: Habia una vez. N orw ood,
NJ: Ablex.
E d e l s k y , C., B. A l t w e r g e r and B. F l o r e s . 1 9 9 1 . Whole language: What’s the
difference? O xford and Portsm outh, NH: H einem ann.
E g g i n g t o n , W. 1987. W ritten academ ic discourse in Korean: Im plications
for effective com m unication. In U. C o n n o r an d R.B. K a p l a n (eds)
Writing across languages: Analysis of L2 text. W okingham, England and
Reading, MA: Addison-Wesley. 153-68.
E i s e n s t e i n , M. 1979. The printing press as an agent of change, communications,
and cultural transformations in early modern Europe. 2 Volumes.
E i s e n s t e i n , M. 1985. O n the printing press as an agent o f change. In D.
O l s o n , N. T o r r a n c e and A. H i l y a r d (eds) Literacy, language and learning.
E l b o w , P. 1973. Writing without teachers. O xford and New York: O xford
University Press.
E l b o w , P. 1 9 8 1 . Writing with power. O xford and New York: O xford
University Press.
E l b o w , P. and P. B e l a n o f f . 1989. Sharing and responding. L ondon and
New York: Random House.
E l b o w , P. and P. B e l a n o f f . 1991. State University o f New York at
Stonybrook portfolio-based evaluation program . In P. B e l a n o f f and
M. D i c k s o n (eds) Portfolios: process and product. Portsm outh, NH:
B oynton/C ook. 3-16.
Bibliography 445
E m i g , J. 1971. The composing processes of twelfth graders. U rbana, IL: National

E m i g , J. 1983. The web of meaning. U pper Montclair, NJ: B oynton/C ook.
E m p s o n , W. 1961. Seven types of ambiguity. H arm ondsw orth: Penguin.
E n r i g h t , D.S. and M. M c C l o s k e y . 1988. Integrating English. W okingham,
England and Reading, MA: Addison-Wesley.
E r i c s s o n , K. and H. S i m o n . 1984. Protocol analysis: Verbal reports as data.
Cam bridge, E ngland an d Cam bridge, MA: MIT Press.
E v e n s o n , L. 1990. Pointers to superstructure in standard writing. In
U . C o n n o r and A . J o h n s (eds) Coherence in writing. W ashington, DC:
TESOL Publications. 169-83.
F a g a n , W., J . J e n s e n and C . C o o p e r . 1985. Measures for research and evalua-
tion in the English language arts, Vol. 2. U rbana, IL: N ational Council of
Teachers o f English.
F a i g l e y , L. 1979. Problem s in analyzing m aturity in college and adult
writing. In D . D a i k e r , A. K e r e k and M . M o r e n b e r g (eds) Sentence
combining and the teaching of writing. Akron, OH: L&S Books. 94-100.
F a i g l e y , L. 1985. N onacadem ic writing: T he social perspective. In
L. O d e l l and D. G o s w a m i (eds) Writing in nonacademic settings. New
York: G uilford Press. 231-48.
F a i g l e y , L. 1986. C om peting theories o f process: A critique and a
proposal. College Composition and Communication 48, 527-42.
F a i g l e y , L., R. C h e r r y , D. J o l i f f e and A. S k i n n e r . 1985. Assessing writers’
knowledge and processes of composing. N orw ood, NJ: Ablex.
F a i r c l o u g h , N. 1992a. Discourse and text: Linguistic an d intertextual
analysis within discourse analysis. Language and Society 3, 193-217.
F a i r c l o u g h , N. (ed.). 1 9 9 2 b . Critical language awareness. L ondon and New
York: Longm an.
F a l t i s , C. 1992. Joinfostering: Adapting teaching strategies for the multilingual
classroom. New York: Merrill.
F a t h m a n , A. and C. K e s s l e r . 1993. Cooperative language learning in
school contexts. In W. G r a b e et al. (eds) A nnual review of applied linguis-
tics, 13. Issues in second language teaching and learning. Cam bridge,
F a t h m a n , A. and E. W h a l l e y . 1990. T eacher response to student writing:
Focus on form versus content. In B. K r o l l (ed.) Second language writing.
178-90.
F e i t e l s o n , D. 1988. Facts and fads in beginning reading. Norw ood, NJ:
Ablex.
F e r r i s , D. 1994. Rhetorical strategies in student persuasive writing:
Differences between native an d non-native English speakers. Research in
the Teaching of English 28, 45-65.
F e r r i s , D . 1995. Student reactions to teacher response in m ultiple-draft
com position studies. TESOL Quarterly 29, 33-53.
446 Bibliography
F ir b a s, J. 1986. O n the dynamics o f written com m unication in light of

the theory of functional sentence perspective. In C . C o o p e r and
S. G r e e n b a u m (eds) Studying writing: Linguistic approaches. Beverly Hills,
CA: Sage. 40-71.
F l e c k , L. 1935. Genesis and development of a scientific fact, (trans. F. Bradley
a n d j. T renn). Chicago: University o f Chicago Press.
F l o w e r , L. 1979. W riter-based prose: A cognitive basis for problem s in
writing. College English 41, 19-37.
F l o w e r , L. 1988. T he construction o f purpose in writing an d reading.
College English 50, 528-50.
F l o w e r , L. 1989. Cognition, context, an d theory building. College
F l o w e r , L. 1994. The construction of negotiated meaning: A social cognitive
theory of writing. Carbondale, IL: S outhern Illinois University Press.
F l o w e r , L. and J. H a y e s . 1977. Problem-solving strategies and the writing
process. College English 39, 449-61.
F l o w e r , L. and J. H a y e s . 1980a. T he cognition of discovery: Defining a
rhetorical problem . College Composition and Communication 31, 21-32.
F l o w e r , L. and J. H a y e s . 1980b. T he dynamics of composing: Making
plans and juggling constraints. In L. G r e g g an d E. S t e i n b e r g (eds)
Cognitive processes in writing. Hove, Sussex and Hillsdale, NJ: L. Erlbaum.
31-50.
F l o w e r , L. and J. H a y e s . 1981a. A cognitive process theory o f writing.
F l o w e r , L. and J. H a y e s . 1981b. Plans th at guide the com posing process.
In C. F r e d r i k s e n and J. D o m i n i c (eds) Writing: The nature, development,
and teaching of written communication, Vol. 2. Hove, Sussex an d Hillsdale,
NJ: L. Erlbaum. 39-58.
F l o w e r , L. and J. H a y e s . 1984. Images, plans an d prose: T he representa-
tion of m eaning in writing. Written Communication 1, 120-60.
F l o w e r , L., V. S t e i n , J. A c k e r m a n , M . K a n t z , K . M c C o r m i c k and
W. P e c k . 1990. Reading-to-write: Exploring a cognitive and social process.
O xford and New York: O xford University Press.
F l y n n , T. and M. K in g (eds). 1993. Dynamics of the writing conference.
U rbana, IL: N ational Council of Teachers o f English.
F o t o s , S. 1991. T he cloze test as an integrative m easure o f EFL
proficiency: A substitute for essays on college entrance examinations?
Language Learning 41,313-36.
F o w l e r , R. 1986. Linguistic criticism. O xford and New York: O xford
University Press.
F r a n k , M . 1979. If you ’re trying to teach kids how to write, you ’ve gotta have this
bookl Nashville, TN: Incentive Publications.
F r a w l e y , W. 1993. Linguistic semantics. Hove, Sussex and Hillsdale, NJ:
L. Erlbaum.
F r e e b o d y , P. and A. W e l c h (eds). 1993. Knowledge, culture & power:
Bibliography 447
International perspectives on literacy as policy and practice. Pittsburgh:

University of Pittsburgh Press.
F reedm an, A. 1 9 9 3 . Show an d tell? T he role of explicit teaching in the
learning o f new genres. Research in the Teaching of English 27, 2 2 2 - 5 1 .
F r e e d m a n , A. and P. M e d w a y (eds). 1994. Genre and the new rhetoric. Bristol,
PA: Taylor 8c Francis.
F r e e d m a n , S. (ed.). 1985. The acquisition of written language: Response and
revision. N orw ood, NJ: Ablex.
F r e e d m a n , S. 1987. Response to student writing. U rbana, IL: National
Council of teachers of English. [NCTE Research R eport No. 23.]
F r e e d m a n , S . and M. S p e r l i n g . 1985. W ritten language acquisition: T he
role of response and the writing conference. In S. F r e e d m a n (ed.)
Acquisition of written language: Response and revision. Norw ood, NJ: Ablex.
106-30.
F r e e m a n , Y. and D. F r e e m a n . 1992. Whole language far second language
learners. O xford and Portsm outh, NH: H einem ann.
F r e i d l a n d e r , A. 1990. Com posing in English: Effects o f a first language
on writing in English as a second language. In B. K r o l l (ed.) Second
language writing. Cam bridge, England and New York: Cam bridge
F r e i r e , P. 1985. The politics of education. South Hadley, MA: Bergin &
Garvey.
F r e i r e , P. 1994. Pedagogy of hope. New York: C ontinuum .
F r e i r e , P. and D. M a c e d o . 1987. Literacy: Reading the word and the world.
South Hadley, MA: Bergin 8c Garvey.
F r i e s , P. 1994. O n them e, rhem e and discourse goals. In M. C o u l t h a r d
(ed.) Advances in written text analysis. O xford an d New York: Routledge.
229-49.
G a ie s , S. 1980. T-unit analysis in second language research: Applications,
problem s and limitations. TESOL Quarterly 14, 53-60.
G a r d n e r , H. 1985. The m ind’s new science: A history of the cognitive revolution.
New York: Basic Books.
G a r n h a m , A. 1985. Psycholinguistics: Central topics. L ondon an d Baltimore,
MD: Edward Arnold.
G a s k i l l , W. 1986. Revising in Spanish and English as a second language:
A process oriented study o f com position. Los Angeles: UCLA. Ph.D.
diss.
G a s k in s , I. 1994. Creating optim um learning environm ents: Is m em ber-
ship in the whole language com m unity necessary? In F. L e h r and J.
O sb orn (eds) Reading , language, and literacy. Hove, Sussex and
Hillsdale, NJ: L . Erlbaum . 115-30.
G e e , J. 1986. Orality and literacy: From the Savage mind to Ways with words.
TESOL Quarterly 20, 717-46.
G e e ,J . 1 9 9 0 . Social linguistics and literacies. New York: Falm er Press.
G e e r t z , C. 1973. The interpretation of cultures. New York: Basic Books.
448 Bibliography
Geer t z, C. 1983. Local knowledge. New York: Basic Books.

G e is l e r , C. 1994. Academic literacy and the nature of expertise. Hove, Sussex
and Hillsdale, NJ: L. Erlbaum.
G e n e s e e , F. 1994. Some holes in whole language. TESOL Matters. 4.3. 3.
G e n t i l e , C. 1992. Exploring new methods for collecting students’ school-based
writing: NAEP’s portfolio study. Princeton, NJ: Educational Testing
Service.
G e o r g e , D. 1984. Working with p eer groups in the com position class-
room . College Composition and Communication 35, 320-6.
G e r e , A. 1987. Writing groups: History, theory and implication. Carbondale,
IL: S outhern Illinois University Press.
G i l b e r t , G . and M . M u l k a y . 1984. Opening Pandora’s box: A sociological
analysis of scientists’ discourse. Cam bridge, England an d New York:
Cam bridge University Press.
G i r o u x , H. 1988. Schooling and the struggle for public life. M inneapolis:
University of M innesota Press.
G i v o n , T. (ed.). 1983. Topic continuity in discourse. Philadelphia: Jo h n
Benjamin.
G i v o n , T. 1985. Quantified studies in discourse. [Special issues o f Text. 5.
1/2.]
Go l d be r g , A. 1985. Groves, philharm onic at C handler Pavilion. Los
Angeles Times. (March 16, 1995) IV. 1.
G o l d m a n , S . and H. T r u e b a (eds). 1987. Becoming literate in English as a
second language. Norwood, NJ: Ablex.
G o l d s t e i n , L. and S. C o n r a d . 1990. S tudent in p u t and negotiation of
m eaning in ESL writing conferences. TESOL Quarterly 24, 443-60.
G o l u b , J. (ed.). 1988. Focus on collaborative learning. U rbana, IL: National
Golub-Smith, M., C. Reese and K. Steinhaus. 1993. Topic and topic type
comparability on the test of written English. Princeton, NJ: Educational
Testing Service. [TOEFL Research Reports #42.]
G o o d m a n , K. 1986. What’s whole in whole language. O xford and
P ortsm outh, NH: H einem ann.
G o o d m a n , Y. 1985. Kidwatching: O bserving children in the classroom. In
A. J a g g e r and M.T. S m i t h -B u r k e (eds) Observing the language learner.
Newark, DE: International Reading Association. 9-18.
G o o d y , J. 1977. The domestication of the savage mind. Cam bridge, England
and New York: Cam bridge University Press.
G o o d y , J. 1987. The interface between the written and the oral. Cam bridge,
England and New York: Cam bridge University Press.
G o o d y , J. and I. W a t t . 1963. T he consequences o f literacy. Comparative
Studies in History and Society 5, 304-45.
G o r m a n , T .P ., A. P u r v e s an d R.E. D e g e n h a r t (eds). 1988. The inter-
national writing tasks and scoring scales. International study of achievement in
writing: Vol. 5. Oxford: Pergam on Press.
Bibliography 449
Go u g h , K. 1968. Im plications o f literacy in traditional China and India.

In J. G o o d y (ed.) Literacy in traditional societies. Cam bridge, England
and New York: Cam bridge University Press. 69-84.
G r a b e , W. 1987. Contrastive rhetoric and text-type research. In U.
C o n n o r and R.B. K a p l a n (eds) Writing across languages: Analysis of L2
text. W okingham, E ngland and Reading, MA: Addison-Wesley. 115-35.
G r a b e , W. 1988a. Assessing the term ‘interactive’. In P. C a r r e l l , J. D e v i n e
and D . E s k e y (eds) Interactive approaches to second language reading.
56-70.
G r a b e , W. 1988b. English, inform ation access, and technology transfer:
A rationale for English as an international language. World Englishes 7,
63-72.
G r a b e , W. 1990. C u rren t developm ents in w ritten discourse analysis.
Lenguas Modernas 17, 35-56.
G r a b e , W. 1992. A pplied linguistics and linguistics. In W. G r a b e an d R.B.
K a p l a n (eds) Introduction to applied linguistics. W okingham, E ngland and
Reading, MA: Addison-Wesley. 35-58.
G r a b e , W. and D. B i b e r . 1987. Freshm an student writing and the con-
trastive rhetoric hypothesis. Paper p resented at the 7th SLRF
Conference. Los Angeles, F ebruary 1987.
G r a b e , W. and C . C h a p e l l e . 1995. Com municative com petence, strategic
com petence, and procedural com petence. Paper presented at the
annual m eeting of the A m erican Association for Applied Linguistics.
Long Beach, CA, March 1995.
G r a b e , W. and R.B. K a p l a n 1986. Science, technology, language and
inform ation: Im plications for language- and language-in-education
planning. InternationalJournal of the Sociology of Language 59, 41-79.
G r a b e , W. and R .B . K a p l a n . 1989. W riting in a second language:
Contrastive rhetoric. In D. J o h n s o n an d D. R o e n (eds) Richness in
writing. L ondon and New York: Longm an. 263-83.
G r a f f , H. 1987. The legacies of literacy. Bloom ington, IN: University of
Indiana Press.
G r a v e s , D. 1983. Writing: Teachers and children at work. L ondon and
Portsm outh, NH: H einem ann.
G r a v e s , D. 1984. A researcher learns to write. L ondon an d Portsm outh, NH:
H einem ann.
G r a v e s , R. 1993. Com position in C anadian universities. Written
G r a y , B. 1990. N atural language learning in aboriginal classrooms:
Reflections on teaching an d learning. In C. W a l t o n and W . E g g i n g t o n
(eds) Language: Maintenance, power and education in Australian Aboriginal
contexts. Darwin: N orth ern T erritories University Press. 105-39.
G r e e n b e r g , K. 1988. Review o f Assessing Writing Skills. College Composition
and Communication 39, 478-9.
450 Bibliography
H aas, C. and L . F l o w e r . 1988. Rhetorical reading strategies and the

construction of m eaning. College Composition and Communication 39,
167-83.
H a i r s t o n , M. 1982. Winds o f change: Thom as Kuhn and the revolution in
the teaching of writing. College Composition and Communication 33, 76-88.
H a l e , G. 1992. Effects of amount of time allowed on the test of written English.
P rinceton, NJ: Educational Testing Service. [TOEFL Research Reports
#39.]
H a l l i d a y , M.A.K. 1973. Explorations in the functions of language. London
and Boston: Edward Arnold.
H a l l i d a y , M.A.K. 1975. Learning how to mean: Explorations in the development
of language. L ondon and Boston: Edward Arnold.
H a l l i d a y , M.A.K. 1978. Language as a social semiotic: The social interpretation
of language and meaning. L ondon and Boston: Edward Arnold.
H a l l i d a y , M.A.K. 1985. An introduction to functional grammar. London and
Boston: Edward A rnold.
H a l l i d a y , M.A.K 1989. Spoken and zvritten language. O xford an d New York:
O xford University Press.
H a l l i d a y , M.A.K. 1993a. Language in a changing world. Deakin, AUS:
Applied Linguistics Association o f Australia. [Occasional P aper #13.]
H a l l i d a y , M.A.K. 1993b. Toward a language-based theory o f learning.
Linguistics and Education 5, 93-116.
H a l l i d a y , M.A.K 1994. T he construction o f knowledge an d value in the
gram m ar of scientific discourse, with reference to Charles D arw in’s The
origin of species. In M. C o u l t h a r d (ed.) Advances in written text analysis.
L ondon and New York: Routledge. 136-56.
H a l l i d a y , M.A.K. and R. H a s a n . 1976. Cohesion in English. L ondon and
New York: Longm an.
H a l l i d a y , M.A.K. and R. H a s a n . 1989. Language, context, and text: Aspects of
language in a social semiotic perspective. O xford and New York: O xford
University Press.
H a l l i d a y , M .A .K a n d j . M a r t i n . 1993. Writing science: Literacy and discur-
sive power. Pittsburgh: University o f Pittsburgh Press.
H a m p -L y o n s , L . 1990. Second language writing: Assessment issues. In
B. K r o l l (ed.) Second language writing. Cam bridge, E ngland and New
H a m p -L y o n s , L . 1991a. Issues an d directions in assessing second language
writing in academ ic contexts. In L . H a m p -L y o n s (ed.) Assessing second
language writing in academic contexts. N orw ood, NJ: Ablex. 323-9.
H a m p -L y o n s , L . 1991b. Pre-text: Task related influences on the writer. In
L . H a m p -L y o n s (ed.) Assessing second language uniting in academic contexts.
H a m p -L y o n s , L . 1991c. Scoring procedures for ESL contexts. In L. H a m p -
L y o n s (ed.) Assessing second language -writing in academic contexts.
Bibliography 451
(ed.). 1991 d. Assessing second language writing in academic

H a m p -L y o n s , L .
contexts. Norw ood, NJ: Ablex.
H a m p -L y o n s , L . and W. C o n d o n . 1993. Q uestioning assum ptions about
portfolio-based assessment. College Composition and Communication 44,
176-90.
H a m p -L y o n s , L . and G. H e n n i n g . 1991. Com m unicative writing profiles:
An investigation of the transferability o f a m ultiple-trait scoring instru-
m ent across E S L writing assessment contexts. Language Learning 41,
337-73.
H a n a n i a , E. and M. S h i k h a n i . 1986. Interrelationships am ong th ree tests
of language proficiency: Standardized ESL, cloze and writing. TESOL
H a r l e y , B., P. A l l e n , J. C u m m in s and M. S w a in (eds). 1990. The develop-
ment of second language proficiency. Cam bridge, England and New York:
H a r r i s , D. 1990. T he use o f ‘organizing sentences’ in the structure o f
paragraphs in science textbooks. In U. C o n n o r and A. J o h n s (eds)
Coherence in writing. A lexandria, VA: TESOL Publications. 69-86.
H a r r i s , M. 1992. C ollaboration is n o t collaboration is n o t collaboration:
W riting center tutorials vs. peer-response groups. College Composition and
H a r s t e , J., V. W o o d w a r d and C . B u r k e . 1984. Language stories and literacy
lessons. L ondon and Portsm outh, N H : H einem ann.
H a s a n , R. 1989. Semantic variation and sociolinguistics. Australian Journal
of Linguistics 9, 221-75.
H a s a n , R. and J. M a r t i n (eds). 1989. Language development: Learning
language, learning culture. N orw ood, NJ: Ablex.
H a t i m , B. 1991. T he pragm atics o f argum entation in Arabic: T he rise and
fall of a text type. Text 11, 189-99.
H a v e l o c k , E. 1976. The origins of western literacy. Toronto: O ntario Institute
for Studies in Education.
H a y e s , J. an d L. F l o w e r . 1983. Uncovering cognitive processes in writing:
An introduction to protocol analysis. In P. M o s e n t h a l , L. T a m o r and
S. W a l m s l e y (eds) Research in writing: Principles and methods. L ondon
and New York: Longm an. 206-19.
H a y e s , J., L. F l o w e r , K. S c h r i v e r , J. S t r a t m a n , an d L. C a r e y . 1987. Cog-
nitive processes in revision. In S. Rosenberg (ed.) Advances in applied
psycholinguistics: Vol 2. Reading, writing, and language learning. Cambridge,
England and New York: Cam bridge University Press. 176-240.
H e a t h , S.B. 1982. W hat no bedtim e story means: Narrative skills at hom e
and in school. Language in Society 11, 49-76.
H e a t h , S.B. 1983. Ways with words. Cam bridge, E ngland an d New York:
H e a t h , S.B. 1985. Literacy o r literate skills? Considerations for ESL/EFL
learners. In P. L a r s e n , E. J u d d , and D. M e s s e r s c h m i t t (eds) On TESOL
’84. W ashington, DC: TESOL Publications. 15-28.
452 Bibliography
Heat h, S.B. 1986a. Critical factors in literacy developm ent. In S.

D e Ca s t e l l , A. L u k e and K. E g a n (eds) Literacy, society and schooling:
A reader. Cam bridge, England an d New York: Cam bridge University
Press. 209-29.
H e a t h , S .B . 1986b. S o c i o c u l t u r a l c o n t e x t s o f l a n g u a g e d e v e l o p m e n t . I n
C a l i f o r n i a O f f i c e o f B i l i n g u a l E d u c a t i o n , Beyond language: Social and
cultural factors in schooling language minority children. L o s A n g e l e s , C A :
E v a lu a t i o n , D is s e m in a tio n and A s s e s s m e n t C e n te r , C a lifo r n ia S ta te
143-86.
U n iv e r s it y , L o s A n g e l e s .
H e a t h , S.B. 1993. In n er city life th rough dram a: Im agining the language
classroom. TESOL Quarterly 27, 177-92.
H e a t h , S .B . and A. B r a n d s c o m b e . 1985. Intelligent writing in an audi-
ence community: Teacher, students, and researcher. In S . F r e e d m a n
(ed.) The acquisition of written language: Response and revision. Norw ood,
NJ: Ablex. 3-32.
H e a t h , S.B. and L. M a n g i o l a . 1991. Children of promise: Literate activity in
linguistically and culturally diverse classrooms. W ashington, DC: N ational
E ducation Association.
H e d g e , T. 1 9 8 8 . Writing. O xford and New York: O xford University Press.
H e i m l i c h , J. and S. P i t t e l m a n 1986. Semantic mapping: Classroom applica-
tions. Newark, DE: International Reading Association.
H e n n i n g , G. 1992. Scalar analysis of the test of written English. Princeton, NJ:
Educational Testing Service. [TOEFL Research R eport #38.]
H i e b e r t , F. and R. C a l f e e 1992. Assessment o f literacy: From standard-
ized tests to perform ance and portfolios. In A. F a r s t r u p and S.J.
S a m u e l s (eds) What research has to say about reading instruction. Newark,
DE: International Reading Association. 70-100.
H i l g e r s , T. an d J. M a r s e l l a . 1992. M aking your writing program work: A
guide to good practices. L o ndon an d New bury Park, CA: Sage.
H i l l o c k s , G. 1986. Research on written composition. U rbana, IL: National
Council of Research in English.
H i n d s , J. 1983a. Linguistics an d w ritten discourse in particular languages:
Contrastive studies: English and Japanese. In R.B. K a p l a n et al. (eds)
Annual review of applied linguistics, 3. Cam bridge, E ngland an d New
H i n d s , J. 1983b. Contrastive rhetoric: English and Japanese. Text 3,
183-95.
H i n d s , J. 1987. R eader vs. w riter responsibility: A new typology. In U.
C o n n o r and R.B. K a p l a n (eds) Writing across languages: Analysis of L2
text. W okingham, England and Reading, MA: Addison-Wesley. 141-52.
H i n d s , J . 1990. I n d u c t i v e , d e d u c t i v e , q u a s i- in d u c t iv e : E x p o s i t o r y w r it in g in
J a p a n e s e , K o r e a n , C h i n e s e , a n d T h a i. I n U . C o n n o r a n d A . J o h n s ( e d s )
Coherence in writing. W a s h in g t o n , D C : TESOL P u b l i c a t i o n s . 87-109.
E.D. 1987. Cultural literacy. Boston: H oughton Mifflin.
H ir s c h ,
H o e y , M. 1983. On the surface of discourse. L ondon: Allen 8c Unwin.
Bibliography 453
Hoey, M. 1986. O verlapping patterns o f discourse organization an d their

im plications for clause relational analysis o f problem -solution texts. In
C . C o o p e r and S. G r e e n b a u m (eds) Studying uniting: Linguistics
approaches. L ondon and Beverly Hills, CA: Sage. 187-214.
H o e y , M. 1991. Patterns of lexis in text. O xford an d New York: O xford
University Press.
H o e y , M. 1994. Signalling in discourse: A functional analysis o f a com -
m on discourse pattern in w ritten an d spoken English. In M.
C o u l t h a r d (ed.) Advances in urritten text analysis. L ondon and New
York: Routledge. 26-45.
H o l t , D . and N . B a k e r . 1991. Portfolios as a follow-up option in a profi-
ciency-testing program . In. P. B e l a n o f f and M. D i c k s o n (eds) Portfolios:
Process and product. Portsm outh, N H : B oynton/C ook. 37-45.
H o l t , M. 1992. T he value o f w ritten p ee r criticism. College Composition
and Communication 43, 384-392.
H o r n e r , W. (ed.). 1983. The present state of scholarship in historical and
contemporary rhetoric. Columbia, MO: University o f Missouri Press.
H o r o w i t z , D. 1991. E S L writing assessments: C ontradictions and resolu-
tions. In L . H a m p -L y o n s (ed.) Assessing second language uniting in
academic contexts. Norw ood, NJ: Ablex. 71-85.
H o u s t o n , R. 1988. Literacy in early modern Europe. L ondon and New York:
Longm an.
H u d e l s o n , S. 1984. ‘Kan yu ret an rayt en ingles’: Children becom e literate
in English as a second language. TESOL Quarterly 18, 221-38.
H u d e l s o n , S. 1989a. W riting in a second language. In R.B. K a p l a n et at.
(eds) A nnual review of applied linguistics, 9. Cam bridge, E ngland and
H u d e l s o n , S. 1989b. Write on: Children writing in ESL. Englewood Cliffs,
NJ: Prentice H a ll.
H u d s o n , R. 1980. Sociolinguistics. Cam bridge an d New York:
H u g h e y , J ., D. W o r m u t h , V.E H a r t f i e l an d H . J a c o b s . 1983. Teaching
ESL composition: Principles and techniques. Rowley, MA: Newbury
House.
H u n t , E. and F. A g n o l i . 1991. T he W horfian hypothesis: A cognitive
psychological perspective. Psychological Review 98, 377-389.
H u n t , K. 1965. Grammatical structures written at three grade levels. U rbana,
IL: N ational Council of Teachers of English.
H u n t , K. 1983. Sentence com bining an d the teaching o f writing. In
M . M a r t l e w (ed.) The psychology of urritten language: A developmental
approach. New York: J. Wiley. 99-125.
H u t c h i n s o n , T. and A. W a t e r s . 1987. English for specific purposes: A learning
centered approach. Cam bridge, England and New York: Cam bridge
University Press.
H y m e s , D. 1972. O n com municative com petence. In J. P r i d e and
454 Bibliography
A. H o l m e s (eds) Sociolinguistics. H arm sw orth and New York: Penguin.

269-93.
IELTS H a n d b o o k . 1995. Cambridge: University o f Cam bridge Local
Exam ination Syndicate.
I n d r a s u t r a , C. 1988. N arrative styles in the writing o f T hai an d A merican
students. In A. P u r v e s (ed.) Writing across languages and cultures.
Newbury Park, CA: Sage. 206-26.
I n g r a m , D. 1990. T he Australian second language proficiency ratings
(ASLPR). I n J.H.A.L. De Jo n g (ed.) Standardization in language testing.
Amsterdam: Free University Press. 46-61. [AILA Review , 7]
I n g r a m , D. and E. W y l i e . 1984. Australian second language proficiency
ratings. Canberra: Australian G overnm ent Publishing Service.
J a c k e n d o f f , R. 1972. Semantic interpretation in generative grammar.
Cam bridge, MA: MIT Press.
J a c o b s , H ., S. Z i n g r a f , D . W o r m u t h , V. H a r t f i e l and J . H u g h e y . 1981.
Testing ESL composition: A practical approach. Rowley, MA: Newbury
House.
J a g g e r , A. an d M.T. S m i t h - B u r k e (eds.). 1985. Observing the language
learner. Neward, DE: In tern atio n al R eading Association.
J e n k i n s , S. and J . H i n d s . 1987. Business letter writing: English, French
and Japanese. TESOL Quarterly 21, 327-49.
J o h n s , A . 1986. C oherence and academ ic writing: Some definitions and
suggestions for teaching. TESOL Quarterly 20, 247-65.
J o h n s , A. 1990. LI com position theories: Im plications for developing the-
ories of L2 com position. In B. K r o l l (ed.) Second language writing.
24-36.
J o h n s , A. and T. D u d l e y - E v a n s . 1991. English for specific purposes:
International in scope, specific in purpose. TESOL Quarterly 25,
297-314.
J o h n s o n , D. and D. R o e n (eds). 1989. Richness in writing: Empowering ESL
students. L ondon and New York: Longm an.
J o h n s o n , D.W. and R. J o h n s o n . 1991. Learning together and alone:
Cooperative, competitive, and individualistic learning. Edina, MN:
Interaction Book Company.
J o h n s o n - L a i r d , P. 1983. Mental Models. Cam bridge, MA: H arvard
University Press.
J o n e s , S. 1985. Problem s with m onitor use in second language com pos-
ing. In M. R o s e (ed.) When a writer c a n i write. New York: Guilford.
96-118.
J o n e s , S. and J . T e t r o e . 1987. Com posing in a second language. In
A. M a t s u h a s h i (ed.) Writing in real time. Norw ood, NJ: Ablex. 34-57.
J o n z , J . 1990. A nother tu rn in the conversation: W hat does cloze
measure? TESOL Quarterly 24, 61-83.
Jo o s, M. 1967. The five clocks. L ondon an d New York: H arco u rt Brace
Jovanovich.
Bibliography 455
J u st, M. and P. C a r p e n t e r . 1992. A capacity theory of com prehension:

Individual differences in working memory. Psychological Review 99,
122-49.
K a c h r u , B. 1985. Institutionalized second language varieties. In S.
G r e e n b a u m (ed.) The English language today. Oxftm l and New York:
O xford University Press. 211-26.
K a c h r u , B. 1992. Why applied linguistics leaks. Plenary address presented
at the annual m eeting o f the A m erican Association for Applied
Linguistics. Seattle, WA, February, 1992.
K a c h r u , Y. 1983. Linguistics and written discourse in particular lan-
guages: Contrastive studies: English an d H indi. In R.B. K a p l a n et al.
(eds) A nnual review of applied linguistics, 3. Cam bridge, E ngland and
K a c h r u , Y. 1987. Cross-cultural texts, discourse strategies an d discourse
interpretation. In L. S m it h (ed.) Discourse across cultures: Strategies in
world Englishes. Englewood Cliffs, NJ: Prentice Hall. 87-100.
K a c h r u , Y. 1988. Writers in H indi and English. In A. P u r v e s (ed.) Writing
across languages and cultures. L ondon and Newbury Park, CA: Sage.
109-37.
K a e s t l e , C., H. D a m o n - M o o r e , L. S t e d m a n , K. T i n s l e y and W.
T r o l l i n g e r . 1991. Literacy in the United States. New Haven, CT: Yale
University Press.
K a g a n , S. 1992. Cooperative learning. San Ju a n Capistrano, CA: Resources
for Teachers.
K a p l a n , R.B. 1966. Cultural th o u g h t p atterns in intercultural education.
Language Learning 16, 1-20.
K a p l a n , R.B. 1972. The anatomy of rhetoric: Prolegomena to a functional theory
of rhetoric. Philadelphia: C enter for C urriculum D evelopm ent (distrib-
uted by H einle & H einle, New York).
K a p l a n , R.B. 1983. An introduction to the study o f written texts: T he
‘discourse com pact.’ In R.B. K a p l a n et al. (eds) A nnual review of applied
linguistics, 3. Cam bridge, E ngland and New York: Cam bridge University
Press. 138-51.
K a p l a n , R.B. 1987. Cultural th o u g h t pattern s revisited. In U. C o n n o r and
R.B. K a p l a n (eds) Writing across languages: Analysis of L2 text.
W okingham, E ngland and Reading, MA: Addison-Wesley. 9-22.
K a p l a n , R.B. 1988. Contrastive rhetoric and second language learning:
N otes toward a theory o f contrastive rhetoric. In A. P u r v e s (ed.)
Writing across languages and cultures. L ondon and Newbury Park, CA:
Sage. 275-304.
K a p l a n , R.B. 1991. Concluding essay: O n applied linguistics and
discourse analysis. In W. G r a b e et al. (eds) A nnual Review of Applied
Linguistics, 11. Cam bridge, E ngland and New York: Cam bridge
K a p l a n , R.B. 1992. Applied linguistics and language policy and planning.
456 Bibliography
In W. G r a b e and R.B. K a p la n (eds.) Introduction to applied linguistics.

W okingham, England and Reading, MA: Addison-Wesley. 143-165.
K a p l a n , R.B. 1993. T he hegem ony o f English in science and technology.
Journal of M ultilingual and Multicultural Development 14, 151-72.
K a p l a n , R.B. (ed.). 1995. The teaching of uniting around the Pacific Basin.
[Special issue of Journal of Asian Pacific Communication. 6]
K a p l a n , R.B. and W. G r a b e . 1991. T he fiction in science writing. In
H. S c h r o d e r (ed.) Subject-oriented texts: Language for special purposes and
text theory. New York: W alter De Gruyter. 199-217.
K a p l a n , R.B. and R S h a w . 1983. Exploring academic English. Rowley, MA:
Newbury House.
K a p l a n , R.B. et al. (eds). 1983. A nnual review of applied linguistics, 3.
Rowley, MA: Newbury House.
K ä u f e r , D. and R. Y o u n g . 1993. W riting in the co n ten t areas: Some
theoretical complexities. In L. O d e l l (ed.) Theory and practice in the
teaching of uniting: Rethinking the discipline. Carbondale, IL: S outhern
Illinois University Press. 71-104.
K e l l o g g , R. 1994. The psychology of writing. O xford and New York: O xford
University Press.
K in n e a v y , J. 1971. A theory of discourse: The aims of discourse. Englewood
Cliffs, NJ: Prentice Hall.
K in n e a v y , J. 1983. A pluralistic synthesis o f four contem porary models for
teaching com position. In A. F r e e d m a n , I. P r i n g l e and J. Y a l d e n (eds)
Learning to unite: First language/second language. L ondon an d New York:
Longm an. 121-38.
K in n e a v y , J. 1987. W riting across the curriculum . In G. T a t e (ed.)
353-77.
K i n t s c h , W. and T. v a n D ijk . 1978. Toward a m odel o f text com p reh en -
sion and production. Psychological Review 85, 363-94.
K i r s c h , G. and D. R o e n (eds). 1990. A sense of audience in written communi-
cation. L ondon and Newbury Park, CA: Sage.
K o r e t z , D. 1993. New rep o rt on V erm ont portfolio project docum ents
challenges. National Council on Measurement in Education Quarterly
Newsletter. 1.4. 1-2.
K r a p e l s , A. 1990. An overview of second language writing process
research. In B. K r o l l (ed.) Second language uniting. Cam bridge,
K r a s h e n , S. 1984. Writing: Research, theory and applications. O xford and
New York: Pergam on Press.
K r e s s , G. 1989. Linguistic processes in sociocultural practice. O xford and New
York: O xford University Press.
K r e s s , G . 1991. Critical discourse analysis. In W. G r a b e et al. (eds) Annual
review of applied linguistics, 11. Cam bridge, E ngland an d New York:
Bibliography 457
Kr e s s , G. 1994. Learning to write. 2nd ed. L ondon: R oudedge.

Kr u e g e r , M. and F. R y a n (eds). 1993. Language and content: Discipline and
content-based approaches to language study. Lexington, MA: D.C. H eath.
K r o l l , B a r b a r a , (ed.). 1990. Second language writing. Cam bridge, England
K r o l l , B a r b a r a and J . R e i d . 1994. Guidelines for designing writing
prom pts: Clarifications, caveats, and cautions. Journal of Second
Language Writing 3, 231-55.
K r o l l , B a r r y . 1981. D evelopm ental relationships betw een speaking and
writing. In B K r o l l an d R. V a n n (eds) Exploring speaking-writing rela-
tionships. U rbana, IL: N ational C ouncil o f T eachers o f English. 32-54.
K r o l l , B a r r y . 1984. W riting for readers: T hree perspectives on audience.
K u h n , T. 1970. The structure of scientific revolutions, 2nd edn. Chicago:
University of Chicago Press.
K u t z , E., S. G r o g a n and V. Z a m e l . 1993. The discovery of competence:
Teaching and learning with diverse student writers. L ondon and
Portsm outh, NH: H einem ann.
L a c k s t r o m , J., L . S e l i n k e r and L . T r i m b l e . 1973. Technical rhetorical
principles and gram m atical choice. TESOL Quarterly 7, 127-36.
L a d u s a w , W. 1988. Semantic theory. In F. N e w m e y e r (ed.) Linguistics: The
Cambridge survey, Vol. 1. Cambridge, England and New York:
L a n g a c k e r , R. 1995. Raising and transparency. Language. 1-62.
L a n g e r , J. 1987. A sociocognitive perspective on literacy. In J. L a n g e r
(ed.) Language, literacy, and culture: Issues of society and schooling.
Norw ood, NJ: Ablex. 1-20.
L a p o n c e , J. 1987. Languages and their territories. Toronto: Univesity of
T oronto Press.
L a r e a u , A. 1989. Home advantage. New York: Falm er Press.
L a r g e , J.A. 1983. The foreign language barrier: Problems in scientific communi-
cation. London: A ndre Deutsch.
L a r s e n -F r e e m a n , D. 1978. An E S L index o f developm ent. TESOL Quarterly
12, 439-48.
L a t o u r , B. 1987. Science in action. Cam bridge, England and New York:
L a t o u r , B. and S. W o o l g a r . 1979. Laboratory life: The social construction of
scientific fact. L ondon and Beverly Hills, CA: Sage.
L a u e r , J. a n d j. A s h e r . 1988. Composition research: Empirical designs. O xford
and New York: O xford University Press.
L a u t a m a t t i , L . 1987. O bservations on the developm ent o f the topic of
sim plified discourse. In U. C o n n o r and R.B. K a p l a n (eds) Writing
across languages: Analysis of L2 text. Wokingham, England an d Reading,
MA: Addison-Wesley. 87-114.
L a w s o n , B., S. R y a n and R . W i n t e r o w d (eds). 1989. Encountering student
texts. U rbana, IL: N ational Council of Teachers o f English.
458 Bibliography
L ay, N. 1982. Com posing processes o f adult ESL learners: A case study.
TESOL Quarterly 16, 406.
L e k i, I. 1990. Coaching from the margins: Issues in w ritten response. In
B. K r o l l (ed.) Second language uniting. Cam bridge, E ngland an d New
L e k i, I. 1991. Twenty-five years o f contrastive rhetoric: Text analysis and
writing pedagogies. TESOL Quarterly 25, 123-43.
Leki, I. 1992. Understanding ESL uniters: A guide for teachers. London and
Portsmouth, NH: Heinemann.
L e m k e , J. 1995. Textual politics: Discourse and social dynamics. New York:
Taylor 8c Francis.
L e y , C., B. S c h a i r and B. D is m u k e s . 1994. L ongitudinal study o f the
reading attitudes and behaviors o f m iddle school students. Reading
Psychology 15, 11-38.
L ib e r m a n , Y. and I. L ib e r m a n . 1990. W hole language vs. code emphasis:
Underlying assumptions and their implications for reading instructions.
Annals of Dyslexia 40, 51-76.
L in d e m a n n , E. 1987. A rhetoric for uniting teachers. O xford and New York:
L i n d e n , M. and A. W h im b e y . 1990. Why Johnny can’t unite: How to improve
-writing skills. Hove, Sussex and Hillsdale, N J: L . Erlbaum.
L o b a n , W. 1976. Language development: Kindergarten through grade twelve.
U rbana, IL : National Council o f Teachers o f English.
L o n g , R . 1990. T he w riter’s audience: Fact or fiction? In G . K i r s c h and
D. R o e n (eds) A sense of audience in written communication. L ondon and
Newbury Park, CA: Sage. 73-84.
L u c y , J. 1992. Language diversity and thought: A reformulation of the linguistic
relativity hypothesis. Cam bridge, E ngland and New York: Cam bridge
University Press.
Lux, P. 1991. Discourse styles of Anglo and Latin Am erican college
student writers. Tem pe, AZ; A rizona State University. Ph.D. diss.
Lux, P. and W. G r a b e . 1991. M ultivariate approaches to contrastive
rhetoric. Lenguas Modernas 18, 133-60.
Lynch, M. 1985. A rt and artifact in laboratory science: A study of shop work and
shop talk at a research laboratory. London and Boston: Routledge 8c Kegan
Paul.
M a c K i n n o n , J. 1993. Becom ing a rhetor: Developing writing ability in a
m ature, writing-intensive organization. In R. S p i l k a (ed.) Writing in the
workplace. C arbondale, IL: Southern Illinois University Press. 41-55.
M a c l e a n , R. 1989. Com posing processes. In F. C h r i s t i e et al. (eds) Writing
in schools: Study guide. Geelong, Victoria: Deakin University Press. 131-49.
M a c r o r i e , K. 1970. Uptaught. Rochelle Park, NJ: Hayden.
M a c r o r i e , K. 1980. Searching writing. U p p er Montclair, NJ: B oynton/
Cook.
M a h e r , J. 1987. T he role o f English as the international language of
m edicine. Applied Linguistics 7, 206-18.
Bibliography 459
Ma h e r , J. and D. R o k o s z . 1992. Language use and the professions. In

W. G r a b e and R .B . K a p l a n (eds) Introduction to applied linguistics.
W okingham, E ngland and Reading, MA: Addison-Wesley. 231-53.
M a n n , W. and S. T h o m p s o n . 1988. Rhetorical structure theory. Text 8,
243-81.
M a n n , W. and S. T h o m p s o n (eds). 1992. Discourse description: Diverse
linguistic analyses of a fund-raising text. Philadelphia: J. Benjamin.
M a n n i n g , M ., G. M a n n i n g and R. L o n g . 1994. Theme immersion: Inquiry-
based curriculum in elementary and middle schools. L ondon and
M a r l a n d , M . 1977. Language across the curriculum. L ondon: H einem ann.
M a r t i n , J. 1985. Process and text: Two aspects o f semiosis. I n J . B e n s o n
and W. G r e a v e s (eds) Systemic perspectives in discourse. Vol. 1. Norw ood,
NJ: Ablex. 248-74.
M a r t i n , J. 1989. Factual uniting: Exploring and challenging social reality.
O xford and New York: O xford University Press.
M a r t i n , J. 1992. English text: System and structure. Philadelphia: J.
Benjamin.
M a r t i n , J. 1993. G e n r e a n d lit e r a c y - M o d e l i n g c o n t e x t in e d u c a t i o n a l
li n g u i s t i c s . I n W. G r a b e et al. ( e d s ) A nnual review of applied linguistics, 13.
Issues in second language teaching and learning. C a m b r id g e , E n g l a n d a n d
N e w Y o rk : C a m b r i d g e U n iv e r s it y P r e s s . 141-72.
M a r t i n , J. and J. R o t h e r y . 1986. W hat a functional approach to the writ-
ing task can show teachers about ‘good w riting’. In B. C o u t u r e (ed.)
Functional approaches to uniting: Research perspectives. Norw ood, NJ: Ablex.
241-65.
M a t a l e n e , C . 1985. C o n t r a s t iv e r h e t o r ic : A n A m e r i c a n w r it in g t e a c h e r in
C h in a . College English 47,789-807.
M a t a l e n e , C. 1995. O f the people, by the people, for the people: Texts in
public contexts. In J. R e y n o l d s , C. M a t a l e n e , J. M a g n o t t o , D. S a m s o n ,
J r . and L. S a d l e r (eds) Professional uniting in context: Lessons from teaching
and consulting in worlds of work. Hove, Sussex and Hillsdale, NJ: L.
Erlbaum . 33-67.
M c G i n l e y , W. and R. T i e r n e y . 1989. Traversing the topical landscape.
Written Communication 6, 243-69.
M c K e n n a , M ., R . R o b i n s o n and J. M i l l e r . 1993. W hole language and
research: T he case for caution. In D. L e u et al. (eds) Examining central
issues in literacy research, theory, and practice. 42nd Yearbook o f the
N ational Reading Conference. Chicago, IL : N ational Reading
Conference. 141-52.
M c K e n n a , M ., S . S t a h l and D. R e i n k i n g . 1994. A critical com m entary on
research, politics, and whole language. Journal of Reading Behavior 26,
211-33.
M c L e o d , S. and M . D o v e n (eds.). 1992. Writing across the curriculum: A
guide to developing programs. L ondon and Newbury Park, CA: Sage.
460 Bibliography
M ech, D. 1985. How delicate is the balance of nature? National Wildlife

23, 2. 54-8.
M e d g y e s , P. and R.B. K a p l a n . 1992. Discourse in a foreign language: T he
exam ple of H ungarian scholars. International Journal of the Sociology of
Language 98, 67-100.
M e i n b a c h , A., L. R o t h l e i n and A. F r e d e r i c k s . 1995. The complete guide to
thematic units: Creating the integrated curriculum. N orw ood, MA:
C hristopher-G ordon.
M e r t o n , R. 1973. The sociology of science. Chicago: University of Chicago
Press.
M e s s ic k , S. 1989. Validity. In R. L in n (ed.) Educational measurement, 3rd
edn. L ondon and New York: Macmillan. 13-104.
M e s s ic k , S. 1994. T he interplay o f evidence and consequences in the
validation of perform ance assessments. Educational Researcher 23, 13-24.
M e y e r , B. 1975. The organization of prose and its effects on memory.
Amsterdam: N orth-H olland.
M e y e r , B. 1984. O rganizational aspects o f text: Effects on reading
com prehension and applications for the classroom. In J. F l o o d and
D. L a p p (eds) Promoting reading comprehension. Newark, DE:
International Reading Association. 113-38.
M e y e r , B . 1985. Prose analysis: Purposes, procedures, and problem s. In
B . B r i t t o n a n d j. B l a c k (eds) Understanding expository text. Hove, Sussex
and Hillsdale, NJ: L. Erlbaum. 11-64.
M e y e r , B. 1987. Following the a u th o r’s top-level organization: An im por-
tant skill for reading com prehension. In R. T i e r n e y , P. A n d e r s and
J. M i t c h e l l (eds) Understanding readers’ understanding: Theory and
practice. Hillsdale, NJ: L. Erlbaum. 59-76.
M i l a n o v i c , M . 1987. Large-scale language testing. In R. L o r d and
H. C h e n g (eds) Language education Hong Kong. H ong Kong: T he
Chinese University Press. 123-32.
M i l l e r , C. 1984. G enre as social action. Quarterly Journal of Speech 70,
151-67.
M i l l e r , T. (ed.). 1995. Functional approaches to written texts: Classroom
applications. Paris: TESOL France.
M o f f e t t , J. 1968. Teaching the universe of discourse. B o s t o n : H o u g h t o n
M if f l in .
M ohan, B. 1986. Language and content. W okingham, E ngland and
Reading, MA: Addison-Wesley.
M o h a n , B. 1990. LEP students and the integration o f language and
content: Knowledge structures an d tasks. In C. S i m i c h - D u d g e o n (ed.)
Proceedings of the first research symposium on limited English proficient student
issues. W ashington, DC: Office o f Bilingual Education an d Minority
Language Affairs. 113-60.
M o ll, L. (ed.). 1990. Vygotsky and education: Instructional implications and
applications of socio-historical psychology. Cam bridge, E ngland and New
York: Cam bridge University Press.
Bibliography 461
M o n t a ñ o -H a r m o n , Μ. 1988. Discourse features in the com positions of

Mexican English as a second language, M exican-A m erican, Chicano
and Anglo high school students: C onsiderations for the form ulation of
educational policy. Los Angeles: University of S outhern California.
Ph.D. diss.
M o n t a ñ o - H a r m o n , M. 1991. Discourse features o f w ritten Mexican
Spanish: C u rren t research in contrastive rhetoric an d its implications.
Hispania 74, 417-25.
M o s e n t h a l , P. and R. T ie r n e y . 1984. Cohesion: problem s with talking
about text. Reading Research (Quarterly 19, 240-4.
M u c h i r i , Μ., N . M u la m b a , G. M y e r s an d D . N d o l o i . 1995. Im porting
com position. College Composition and Communication 46, 175-98.
M u lk a y , M . 1979. Science and the sociology of knowledge. L o n d o n : U n w i n &
A lle n .
M u lle n , 1987. Test o f ability to subordinate. In C. A l d e r s o n , K .
K.
K rahnke and C. S t a n s f i e l d (eds) Reviews of English language profidency
tests. W ashington, DC: TESOL Publications. 73-6.
M u r r a y , D. 1968. A writer teaches writing. Boston: H oughton Mifflin.
M u r r a y , D . 1980. W riting as process: How writing finds its own m eaning.
In T. D o n o v a n and B. M c C l e l l a n d (eds) Eight approaches to teaching
composition. U rbana, IL: N ational Council o f Teachers o f English.
3-20.
M u r r a y , D. 1985. A writer teaches writing , 2nd edn. Boston: H oughton
M if f l in .
M yers, G. 1985. T he social construction o f two biologists’ proposals.
Written Communication 2, 219-45.
M yers, G. 1986. W riting research an d the sociology o f scientific
knowledge: A review of three new books. College English 48, 595-610.
M y e r s , G. 1989. T he pragm atics o f politeness in scientific articles. Applied
Linguistics 10, 1-35.
M y e r s , G. 1990. Writing biology: Texts in the sodal construction of knowledge.
M adison, WI: University o f Wisconsin Press.
N a g y , W. 1988. Teaching vocabulary to improve reading comprehension.
U rbana, IL: N ational Council o f teachers o f English.
N a t i o n a l C o u n c i l o f T e a c h e r s o f E n g l i s h ( N C T E ) ( e d s ) . 1983. Idea
exchange for English teachers. U r b a n a , IL: N a t i o n a l C o u n c i l o f t e a c h e r s o f
E n g lis h .
N e e d e ls , M. and M. K napp. 1994. T e a c h i n g w r it in g t o c h i l d r e n w h o a r e
u n d erserved . Journal of Educational Psychology 86, 339-49.
N e e l , J. 1988. Plato, Derrida, and writing Carbondale, IL: S outhern Illinois
University Press.
N e ls o n , G. and J. M u r p h y . 1992. An L2 writing group: Task and social
dim ensions. Journal of Second Language Writing 1, 171-93.
N e l s o n , G. and J. M u r p h y . 1993. P eer response groups: Do L2 writers use
p eer com m ents in revising th eir drafts? TESOL (Quarterly 27, 135-42.
462 Bibliography
N e w e ll, A. and H. S im o n . 1972. Human problem solving. Englewood Cliffs,

N J: Prentice Hall.
N e w k ir k , T. 1987. More than stories. L ondon and Portsm outh, N H :
H einem ann.
N e w k ir k , T. (ed.). 1993. Nuts and bolts: A practical guide to teaching college
composition. L ondon and Portsm outh, NH: H einem ann.
N e w m a n , D., P. G r i f f i n and M. C o l e . 1989. The construction zone.
N o l d , E. and S. F r e e d m a n . 1977. An analysis o f read ers’ responses to
essays. Research in the Teaching of English 11, 164-77.
N o r t h , S. 1987. The making of knowledge in composition. L ondon and
N y stra n d , M . 1 9 8 6 . The structure of written communication: Studies in
reciprocity between writers and readers. L o n d o n a n d N e w Y ork : A c a d e m i c
P ress.
N y stra n d , M. 1989. A social-interactive m odel of writing. Written
N y stra n d , M. 1990. Sharing words: T he effects o f readers on developing
writers. Written Communication 7, 3-24.
N y s t r a n d , M. and D. B r a n d t . 1989. Response to writing as a context for
learning to write. In C. A n s o n (ed.) Writing and response. U rbana, IL:
N ational Council o f Teachers o f English. 209-30.
O a k h i l l , J. and A. G a r n h a m . 1988. Becoming a skilled reader. O xford and
New York: Basil Blackwell.
O ’B a r r , W. 1982. Linguistic evidence: Language, power, and strategy in the
courtroom. L ondon and New York: Academic Press.
O c h s , E. 1979. P lanned and u n p la n n ed discourse. In T. G i v o n (ed.)
Discourse and syntax. [Syntax and Semantics 12.] L ondon an d New York:
Academic Press. 51-80.
O c h s , E. 1988. Language and cultural development. Cam bridge, E ngland
O d e l l , L. 1985. Beyond the text: Relations between writing an d social
context. In L. O d e l l and D. G o s w a m i (eds) Writing in nonacademic
settings. New York: Guilford. 249-80.
O d e l l , L. and D. G o s w a m i (eds). 1985. Writing in nonacademic settings.
New York: Guilford Press.
O ’H a r e , F. 1973. Sentence combining: Improving student writing without formal
grammar instruction. U rbana, IL: N ational Council o f Teachers of
English.
O l l e r , J ., J r . 1983. Evidence for a general language proficiency factor: An
expectancy gramm ar. In J . O l l e r (ed.) Issues in language testing research.
Rowley, MA: Newbury House. 3-10.
O l l e r , J ., J r . a n d j . J o n z (eds). 1994. Cloze and coherence. Lewisburg, PA:
Bucknell University Press.
O l s e n , L. 1993. Research on discourse com m unities: An overview. In
Bibliography 463
R. S p i l k a (ed.) Writing in the workplace. Carbondale, IL: S outhern

Illinois University Press. 181-94.
O l s o n , D. 1977. From utterance to text: T he bias o f language in speech
and writing. Harvard Educational Review 47, 257-81.
O l s o n , D. 1991. Literacy as metalinguistics. In D. O l s o n and
N. T o r r a n c e (eds) Literacy and orality. Cam bridge, E ngland an d New
O l s o n , D. 1994. The world on paper. Cam bridge, E ngland and New York:
O n g , W. 1982. Orality and literacy: The technologizing of the word. London:
M ethuen.
O s t l e r , S. 1987. English in parallels: A com parison o f English an d Arabic
prose. In U. C o n n o r and R.B. K a p l a n (eds) Writing across languages:
Analysis of L2 text. W okingham, England and Reading, MA: Addison-
Wesley. 169-85.
P a i n t e r , C. 1989. L earning language: A functional view o f language
developm ent. In R. H a s a n an d J. M a r t i n (eds) Language development:
Learning language, learning culture. Norw ood, NJ: Ablex. 18-65.
P a i v i o , A. 1986. Mental representations. O xford and New York: O xford
University Press.
P a l i n c s a r , A. and A. B r o w n . 1984. Reciprocal teaching o f com p reh en -
sion-fostering and m onitoring activities. Cognition and Instruction 1,
117-75.
P a l i n c s a r , A. and L. K l e n k . 1991. Dialogues prom oting reading com pre-
hension. In B. M e a n s , C . C h e l e m e r an d M . K n a p p (eds) Teaching
advanced skills to at-risk students. San Francisco: Jossey Bass. 112-30.
P a n d h a r i p a n d e , R. 1983. Linguistics and written discourse in particular
languages: Contrastive studies: English and Marathi. In R.B. Ka p l a n et
al. (eds) Annual review of applied linguistics, 3. Rowley, MA: Newbury
House. 118-38.
P a r a d i s , J. 1991. Text and Action: The o p era to r’s m anual in context and
in court. In C. B a z e r m a n an d J. P a r a d i s (eds) Textual dynamics of the
professions. M adison, WT: University of Wisconsin Press. 256-78.
P a r a d i s , J., D . D o b r i n and R . M i l l e r . 1985. W riting at Exxon I T D : Notes
on the writing environm ent o f an R & D organization. In L. O d e l l and
D . G o s w a m i (eds) Writing in nonacademic settings. New York: Guilford.
281-307.
P a r k , D. 1982. T he m eaning o f audience. College English. 44, 247-57.
P e a r s o n , P .D . and L. F i e l d i n g . 1991. C om prehension instruction. In
R. Barr et al. (eds) Handbook of reading research. Vol. 2. L ondon an d New
York: Longm an. 815-60.
P e n n i n g t o n , M. 1993. A critical exam ination o f word processing effects
in relation to L2 writers. Journal of Second Language Writing2, 227-55.
P e r e g o y , S. and O. B o l y e . 1993. Reading, writing, and learning in ESL:
A resource book for K-8 teachers. L ondon and New York: Longm an.
464 Bibliography
P erera, K. 1984. Children’s uniting and reading: Analysing classroom language.

O xford and New York: Basil Blackwell.
P e r f e t t i , C . a n d D. M c C u t c h e n . 1987. S c h o o l e d l a n g u a g e c o m p e t e n c e :
L in g u is tic a b il it ie s in r e a d in g and w r it in g . In S. R o sen b erg (e d .)
Advances in applied psycholinguistics: Volume 2. Reading , uniting, and lan-
guage learning C a m b r i d g e , E n g l a n d a n d N e w Y ork: C a m b r id g e
105-41.
U n iv e r s it y P r e s s .
P e r k in s , K. 1983. O n the use o f com position scoring techniques, objective
m easures, and objective test to evaluate ESL writing ability. TESOL
P e r l , S. 1979. T he com posing processes of unskilled college writers.
Research in the Teaching of English 13, 317-36.
P e y t o n , J.K. (ed.). 1990. Students and teachers uniting together. W ashington,
DC: TESOL Publications.
P e y t o n , J.K., J. S t a t o n , G. R i c h a r d s o n and W. W o l f r a m . 1990. T he
influence of writing task on E S L students’ w ritten production. Research
in the Teaching of English 24, 142-71.
P h e l p s , L. 1988. Composition as a human science. O xford and New York:
P h i l i p s , S. 1983. The invisible culture: Communication in classroom and
community on the Warm Springs indian reservation. L ondon an d New York:
L ongm an.
P h i l l i p s o n , R. 1992. Linguistic imperialism. O xford and New York: O xford
University Press.
P i e r c e , L. and M . O ’M a l l e y . 1992. Performance and portfolio assessment for
language minority students. W ashington, DC: N ational C learinghouse for
Bilingual Education. [NCBE Program Inform ation G uide Series
#9.]
P o g n e r , K.-H. (ed.). 1994. More on uniting. O dense, Denmark: Institute of
Language and C om m unication. [O dense W orking Papers in Language
and Com m unication #6.]
P o o l e , D. 1991. Discourse analysis in ethnographic research. In W. G r a b e
et al. (eds) Annual review of applied linguistics, 11. Cam bridge, England
and New York: Cam bridge University Press. 42-56.
P r e l l y , L. 1989. T he rhetorical construction o f scientific ethos. In
H. S im o n s (ed.) Rhetoric in the human sciences. L ondon and Newbury
Park, CA: Sage. 48-68.
P r e s s l e y , M. et al. 1989. Strategies that improve m em ory and com p reh en -
sion of w hat is read. Elementary SchoolJournal 90, 3-32.
P r e s s l e y , M . et al. 1991. Teaching literacy strategies across the curriculum :
A case study at benchm ark school. In J. Z u t e l l and S. M c C o r m i c k
(eds) Learner factors/teacher factors: Issues in literacy research and
instruction. 40th Yearbook o f the National Reading Conference.
Chicago, IL: N ational Reading Conference. 219-28.
P r e s s l e y , M. e ta l 1992. Beyond direct explanation: Transactional instruction
Bibliography 465
in reading com prehension strategies. Elementary SchoolJournal 92, 513-55.

P r e s s le y , M. et a l 1994. Transactional instruction o f com prehension
strategies: T he M ontgom ery County, M aryland, SAIL program . Reading
& Writing Quarterly 10, 5-20.
P r i e s t l e y , J. 1777/1965. A course of lectures on oratory and criticism, [ed.
V.M. Bevilacqua and R. Murphy.] Carbondale, IL: Southern Illinois
University Press.
P r i n c e , E. 1981. Toward a taxonom y o f given-new inform ation. In P. C o l e
(ed.) Radical pragmatics. L ondon and New York: Academic Press.
223-55.
P r i n c e , E. 1992. T he ZPG letter: Subjects, definiteness, and inform ation
status. In W. M a n n and S. T h o m p s o n (eds) Discourse description: Diverse
linguistic analyses of a fund-raising text. Philadelphia: J. Benjamins.
295-325.
P u r v e s , A. 1984. In search o f an internationally-valid schem e for scoring
com positions. College Composition and Communication 35, 426-38.
P u r v e s , A. 1991. The scribal society. L ondon an d New York: Longm an.
P u r v e s , A. ( e d . ) . 1988. Writing across languages and culture. L o n d o n a n d
N e w b u r y P a r k , CA: Sage.
P u r v e s , A. and G. H a w i s h e r . 1990. Writers, judges, and text models. In
R. B e a c h and S. H y n d s (eds) Developing discourse practices in adolescence
and adulthood. N orw ood, NJ: Ablex. 183-99.
P u r v e s , A. and W. P u r v e s . 1986. Culture, text models, and the activity of
writing. Research in the Teaching of English 20, 174-97.
P u r v e s , A., A. S o t e r , S . T a k a l a and A. V a h a p a s s i. 1984. Toward a
dom ain-referenced system for classifying com position assignments.
Research on the Teaching of English 18, 385-416.
P u r v e s , A. and S. T a k a l a (eds). 1982. An international perspective on the
evaluation of written composition: Evaluation in education: An international
review series, Vol. 3, No. 5. Oxford: Pergam on.
Q u i r k , R. et a l 1985. A comprehensive grammar of English. L ondon an d New
York: Longm an.
R a b in , C . 1986. T he discourse status o f com m entary. In C . C o o p e r and
S. G r e e n b a u m (eds) Studying writing: Linguistic approaches. L ondon and
Beverly Hills, C A : Sage. 215-25.
R a d e c k i , P. and J. S w a l e s . 1988. ESL student reaction to w ritten com -
m ents on their written work. System 16, 355-65.
R a f o t h , B. 1988. Discourse community: W here writers, readers, and texts
com e together. In B. R a f o t h an d D. R u b in (eds) The social construction
of written communication. N orw ood, NJ: Ablex. 131-46.
R a f o t h , B. 1990. T he concept o f discourse community: Descriptive and
explanatory adequacy. In G. K i r s c h and D. R o e n (eds) A sense of
audience in written communication. L ondon and Newbury Park, CA: Sage.
140-52.
R a im e s , A. 1983a. Anguish as a second language? Remedies for com posi-
tion teachers. In A. F r e e d m a n , I. P r i n g l e and J. Y a l d e n (eds) Learning
466 Bibliography
to unite: First language/second language. L ondon and New York:

Longm an. 258-72.
R a i m e s , A. 1983b. Tradition an d revolution in ESL teaching. TESOL
R a i m e s , A. 1983c. Techniques in teaching uniting. O xford an d New York:
R a i m e s , A. 1985. W hat unskilled ESL students do as they write: A
classroom study of com posing. TESOL Quarterly 19, 229-58.
R a i m e s , A. 1987. Language proficiency, writing ability and com posing
strategies: A study of ESL college stu d en t writers. Language Learning 37,
439-68.
R a i m e s , A. 1991. O ut of the woods: Em erging traditions in the teaching o f
writing. TESOL Quarterly 25, 407-30.
R a i m e s , A. 1992. Exploring through uniting: A process approach to ESL composi-
tion, 2nd edn. New York: St M artins Press.
R a y n e r , K. and A. P o l l a t s e k . 1989. The psychology of reading. H em el
H em pstead and Englewood Cliffs, NJ: Prentice-Hall.
R e a d a n c e , J., T. B e a n and R .S . B a l d w i n . 1989. Content area reading: A n
integrated approach, 3rd edn. D ubuque, IA: Kendall H unt.
R e i d , J . 1 9 8 8 . Quantitative differences in English prose written by Arabic,
Chinese, Spanish, and English students. Ft Collins, CO: Colorado State
University. Ph.D. diss.
R e i d , J. 1993. Teaching ESL uniting. Englewood Cliffs, NJ: Regents Prentice
Hall.
R e i d , J. and B. K r o l l . 1995. Designing an d assessing effective classroom
writing assignments for NES an d ESL students. Journal of Second
Language Writing4, 17-41.
R e i t h e r , J . 1985. W riting an d knowing: Toward redefining the writing
process. College English 47, 620-8.
R e i t h e r , J. 1993. Bridging the gap: Scenic motives for collaborative
writing in workplace an d school. In R . S p i l k a (ed.) Writing in the
workplace. Carbondale, IL: S outhern Illinois University Press.
195-206.
R e p p e n , R . 1995a. V ariation in elem en tary stu d e n t language: A m ultidi-
m ensional perspective. Flagstaff, A2: N o rth e rn A rizona University.
Ph.D. diss.
R e p p e n , R . 1995b. A genre-based approach to co n ten t writing instruction.
TESOL Journal 4, 2. 32-5.
R e p p e n , R . and W. G r a b e . 1993. Spanish transfer effects in the English
writing o f elem entary school students. Lenguas Modernas 20,
113-28.
R e s n i c k , D. and L. R e s n i c k . 1977. T he nature o f literacy: A historical
explanation. Harvard Educational Review 47, 370-85.
R e y n o l d s , J., C . M a t a l e n e , J. M a g n o t t o , D, S a m s o m , J r . and L. S a d l e r .
1995. Professional writing in context: Lessons from teaching and consulting.
Hove, Sussex and Hillsdale, NJ: L. Erlbaum.
Bibliography 467
R ic h a r d s , J. 1990. From m eaning into words: W riting in a second or

foreign language. In J. R i c h a r d s , The language teaching matrix. Cambridge,
R i c h g e l s , D ., L . M c G e e , R . L o m a x an d C. S h e a r d . 1987. Awareness o f '
four text structures: Effects on recall o f expository prose. Reading
Research Quarterly 22, 177-96.
R ie b e n , L. and C. P e r f e t t i (eds). 1991. Learning to read. Hove, Sussex and
Hillsdale, NJ: L. Erlbaum.
R o b b , T., S. Ross and I. S h o r t r e e d . 1986. Salience o f feedback on erro r
and its effect on ESL writing quality. TESOL Quarterly 20, 83-93.
R o b i n s o n , G. 1992. ESP today. H em el H em psted and Englewood Cliffs,
NJ: Prentice Hall.
R o g o f f , B. 1990. Apprenticeship in thinking. O xford and New York: O xford
University Press.
R o r t y , R . 1979. Philosophy and the mirror of nature. Princeton, NJ:
P rinceton University Press.
R o s e n b e r g , R . 1991. Using the portfolio to m eet state-m andated assess-
m ent: A case study. In P. B e l a n o f f and M. D i c k s o n (eds) Portfolios:
Process and product. P ortsm outh, NH: B oynton/C ook. 69-79.
R o t h e r y , J. 1989. L earning about language. In R . H a s a n a n d j . M a r t i n
(eds) Language development: Learning language, learning culture.
S a ’A d e d d i n , M. 1989. Text developm ent an d Arabic-English negative
interference. Applied Linguistics 10, 36-51.
S a n t a , C. and D. A l v e r m a n n (eds). 1991. Science learning: Processes and
applications. Newark, DE: International Reading Association.
S c a r d a m a l i a , M. and C. B e r e i t e r . 1986. Research on written com posi-
tion. In M. W i t t r o c k (ed.) Handbook of research on teaching, 3rd edn.
New York: M acmillan Education Ltd. 778-803.
S c a r d a m a l i a , M. and C. B e r e i t e r . 1987. Knowledge telling and knowl-
edge transform ing in written com position. In S . R o s e n b e r g (ed.)
Advances in applied psycholinguistics, Volume 2: Reading, uniting, and
language learning. Cam bridge, E ngland and New York: Cam bridge
S c a r d a m a lia , M. and C. B e r e i t e r . 1991. Literate expertise. In K.A.
E r i c s s o n and J. S m i t h (eds) Toward a general theory of expertise.
S c h i e f f e l i n , B. and E. O c h s (eds). 1986. Language socialization across cul-
tures. Cam bridge, E ngland and New York: Cam bridge University Press.
S c h i f f r i n , D. 1987. Discourse markers. Cam bridge, England and New York:
S c h i f f r i n , D. 1991. Conversational analysis. In W. G r a b e et al. (eds)
A nnual review of applied linguistics, 11. Discourse analysis. Cambridge,
S c h i f f r i n , D. 1994. Approaches to discourse. O xford and New York:
Blackwell.
468 Bibliography
S c h o le s , R. (ed.). 1993. Literacy and language analysis. Hove, Sussex and

Hillsdale, NJ: L. Erlbaum.
S c h r i v e r , K. 1992. Teaching writers to anticipate read ers’ needs: A
classroom-evaluated pedagogy. Written Communication 9, 179-208.
S c h r o d e r , H. (ed.). 1991. Subject-oriented texts: Language for special purposes
and text theory. New York: W alter de Gruyter.
S c o l l o n , R. 1991. Eight legs an d one elbow: Stance and structure in
C hinese-English com positions. Paper p resented at the In ternational
Reading Association, Second N orth A m erican C onference on A dult
and Adolescent Literacy. Banff, Canada, M arch 1991.
S c o l l o n , R. and S . S c o l l o n . 1981. Narrative, literacy and face in interethnic
communication. N orw ood, NJ: Ablex.
S c o l l o n , R . and S . S c o l l o n . 1983. Face in interethnic com m unication.
In J. R i c h a r d s and R . S c h m i d t (eds) Language and communication.
L ondon and New York: Longm an. 156-88.
S c r i b n e r , S . and M. C o l e . 1981. The psychology of literacy. Cam bridge, MA:
H arvard University Press.
S e l f e , C. 1 9 8 1 . T he com posing processes o f four high an d four low writ-
ing apprehensives: A case study. Austin, TX: University o f Texas. Ph.D.
diss.
S e l f e , C . 1984. T h e p r e d r a f t i n g p r o c e s s e s o f f o u r h i g h - a n d f o u r lo w -
a p p r e h e n s i v e w r ite r s . Research in the Teaching of English 18, 45-64.
S e l f e , C. and S . H i l l i g o s s (eds). 1994. Literacy and computers: The complica-
tions of teaching and learning with technology. New York: T he M odern
Language Association.
S h a p i r o , M. 1987. Language an d Politics. In R.B. K a p l a n et al. (eds)
A nnual review of applied linguistics, 7. Cam bridge, E ngland an d New
S h a r a n , Y. and S . S h a r a n . 1992. Expanding cooperative learning through
group investigation. New York: Teachers College Press.
S h a u g h n e s s y , M. 1977. Errors and expectations: A guide for the teacher of basic
writing.O xiord and New York: O xford University Press.
S h i h , M. 1986. Content-based approaches to teaching academ ic writing.
TESOL Quarterly 20, 617-48.
S h or, I. (ed.). 1987. Freire for the classroom. Portsm outh, NH:
B oynton/C ook.
S h o r t , D. 1991. How to integrate language and content instruction.
W ashington, DC: C enter for Applied Linguistics.
S h o r t , D. 1994. Expanding m iddle school horizons: Integrating language,
culture, and social studies. TESOL Quarterly 28, 581-608.
S h u c h m a n , H. 1981. Information transfer in engineering. W ashington, DC:
Futures Group.
S h u y , R. 1987. Language and the law. In R.B. KAPLAn et al. (eds) Annual
review of applied linguistics, 7. Cam bridge, E ngland an d New York:
C am bridge University Press. 50-63.
Bibliography 469
Sh u y , R. 1991. Language crimes. O xford an d Cam bridge, MA: Blackwell.

S ilv a , T. 1990. Second language com position instruction: Developments,
issues, and directions in ESL. In B. K r o l l (ed.) Second language writing.
Cam bridge, E ngland an d New York: Cam bridge University Press.
11-23.
S i l v a , T. 1992. Research agendas for ESL writing: Differences in ESL and
NES writing. P aper presen ted at the 27th A nnual TESOL Convention.
Vancouver, BC, March.
Silva, T. 1993. Toward an u nderstanding o f the distinct n atu re o f L2 writ-
ing: T he ESL research an d its implications. TESOL Quarterly 27, 657-77.
S i n c l a i r , J. 1991. Corpus, concordance, collocation. O xford an d New York:
S i n c l a i r , J. 1 9 9 4 . T rust the text. In M. C o u l t h a r d (ed.) Advances in
written text analysis. L ondon an d New York: Routledge. 1 2 - 2 5 .
S i n g e r , M. 1990. Psychology of language: An introduction to sentence and
discourse processes. Hove, Sussex and Hillsdale, NJ: L. Erlbaum .
S l a v i n , R. 1990. Cooperative learning: Theory, research, and practice. L ondon
and Englewood Cliffs, NJ: Prentice Hall.
S l o b i n , D. 1990. T he developm ent from child speaker to native speaker.
In J . S t i g l e r , R. S h w e d e r an d G. H e r d t (eds) Cultural psychology: Essays
on comparative human development. Cam bridge, England an d New York:
C am bridge University Press. 233-56.
S l o b i n , D. and A. B o c a z . 1988. L earning to talk ab o u t m ovem ent
through tim e and space: T he developm ent o f narrative abilities in
Spanish and English. Lenguas Modernas 15, 5-24.
S m a g o r in s k y , P. 1992. How reading m odel essays affects writers. In
J. I r w i n and M. D o y l e (eds) Reading/writing connections: Learning from
research. Newark, DE: Intern atio n al Reading Association. 160-76.
S m a g o r in s k y , P. (ed.). 1994. Speaking about writing: Reflections of research
methodology. L ondon and T housand Oaks, CA: Sage.
S m it, D ., P. K o l o n s k y and K . S e l t z e r . 1991. Im plem enting a portfolio
system. In P. B e l a n o f f and M. D i c k s o n (eds) Portfolios: Process and
product. Portsm outh, NH: B oynton/C ook. 46-56.
S m i t h , N. 1 9 8 9 . The twitter machine. O xford an d New York: Blackwell.
S n o w , C ., W. B a r n e s , J. C h a n d l e r , I. G o o d m a n a n d L. H e m p h i l l . 1991.
Unfulfilled expectations: Home and school influences on literacy. C a m b r id g e ,
MA: H a r v a r d U n iv e r s it y P r e s s .
Sn yder, I. 1993. W riting with w ord processors: A research overview.
Sokm en, A. 1988. T a k in g a d v a n ta g e o f c o n fe r en ce -c e n te r ed w r it in g .
TESOL Newsletter 2 2 , 1 , 5 .
S om m er, N. 1980. Revision strategies o f student writers an d experienced
adult writers. College Composition and Communication 3 1 , 378-88.
S o t e r , A. 1988. T h e s e c o n d l a n g u a g e l e a r n e r a n d c u lt u r a l t r a n s f e r in
n a r r a t i o n . I n A. P u r v e s ( e d . ) Writing across languages and cultures.
L o n d o n a n d N e w b u r y P a r k , CA: S a g e . 177-205.
470 Bibliography
Spa c k , R. 1988. Initiating ESL students into the academ ic discourse

community: How far should we go. TESOL Quarterly 22, 29-51.
S p a n d e l , V. and R. S t i g g i n s . 1990. Creating writers: Linking assessment and
writing instruction. L ondon an d New York: Longm an.
S p e a r , K. 1 9 8 8 . Sharing writing: Peer response groups in English classes.
P ortsm outh, NH: B oynton/C ook.
S p e a r , K. (ed.). 1993. Peer response groups in action: Writing together in
secondary schools. P ortsm outh, NH: B oynton/C ook.
S p e r b e r , D. and D. W i l s o n . 1986. Relevance. O xford: Blackwell Publishers
an d Cam bridge, MA: H arvard University Press.
S p i e g e l , D. and J. F i t z g e r a l d . 1991. Textual cohesion and coherence
in children’s writing revisited. Research in the Teaching of English 24, 48-66.
S p i l k a , R. 1993a. Influencing workplace practice: A challenge to pro-
fessional writing specialists in academia. In R. S p i l k a (ed.) Writing in the
workplace. C arbondale, IL: S outhern Illinois University Press. 207-19.
S p i l k a , R. (ed.). 1993b. Writing in the workplace. Carbondale, IL: S outhern
S p i n d l e r , G . and L. S p i n d l e r . 1987a. Teaching and learning how to do
the ethnography of education. In G . S p i n d l e r and L. S p i n d l e r (eds)
Interpretive ethnography of education. Hove, Sussex and Hillsdale, NJ: L.
Erlbaum . 17-33.
S p i n d l e r , G. and L. S p i n d l e r (eds). 1987b. Interpretive ethnography of educa-
tion. Hove, Sussex and Hillsdale, NJ: L. Erlbaum.
S p i r o , R. a n d B. T a y l o r . 1987. O n i n v e s t i g a t i n g c h i l d r e n ’s t r a n s i t i o n f r o m
n a r r a t iv e to e x p o s ito r y d is c o u r s e : T h e m u ltid im e n s io n a l n a tu r e of
p s y c h o lo g ic a l te x t R. T i e r n e y , P. A n d e r s a n d J.
c la s s if i c a t io n . In
M i t c h e l l ( e d s ) Understanding readers’ understanding. H o v e , S u s s e x a n d
H i l l s d a l e , NJ: L. E r l b a u m . 77-93.
S p i r o , R., W. V i s p o e l , J. S c h m i t z , A. S a m a r a p u n g a v a n and A.E. B o e r g e r .
1987. Knowledge acquisition for application: Cognitive flexibility and
transfer in com plex co n ten t dom ains. In B . B r i t t o n an d S . G l y n n
(eds) Executive control processes in reading. Hove, Sussex an d Hillsdale, NJ:
L. Erlbaum . 177-99.
S r i d h a r , S .N . 1994. A reality check for SLA theories. TESOL Quarterly 28,
800-5.
S t a h l , R. (ed.). 1 9 9 4 . Cooperative learning in social studies. W okingham,
England an d Reading, MA: Addison-Wesley.
S t a h l , S . 1994. W hole language in kindergarten and first grade:
Separating the rhetoric from the effects. In F. L e h r an d J. O s b o r n
(eds) Reading language, and literacy. Hove, Sussex an d Hillsdale, NJ:
L . Erlbaum . 101-14.
S t a h l , S . and P. M i l l e r . 1989. W hole language and language experience
approaches for beginning reading: A quantitative research synthesis.
Review of Educational Research 59, 87-116.
Bibliography 471
S t a n le y , J. 1992. Coaching student writers to be effective p eer evaluators.

Journal of Second Language Writing 1, 217-33.
S ta n o v ic h , K. 1980. Toward an interactive-com pensatory m odel of
individual differences in the developm ent o f reading fluency. Reading
Research Quarterly 16, 32-71.
S t a n o v i c h , K. 1986. Matthew effects in reading: Some consequences of
individual differences in the acquisition o f literacy. Reading Research
S t a n o v i c h , K. 1992. T he psychology o f reading: Evolutionary and revolu-
tionary developm ents. In W. G r a b e et a l (eds) A nnual review of applied
linguistics, 12. Literacy. Cam bridge, England and New York: Cam bridge
S t a n s f i e l d , C. 1986. A history o f the Test o f W ritten English: T he
developm ental years. Language Testing 3, 224-34.
S t a n s f i e l d , C and J. Ross. 1988. A long-term research agenda for the Test
of W ritten English. Language Testing 5, 160-86.
S t a t o n , J., R . S h u y , J.K. P e y t o n and L. R e e d . 1988. Dialogue journal
communication: Classroom, linguistic, social and cognitive views. Norwood,
NJ: Ablex.
S t e d m a n , L. 1994. T he Sandia rep o rt and U.S. achievement: An assess-
m ent. Journal of Educational Research 87, 133-146.
S t e d m a n , L. an d C. K a e s t l e . 1987. Literacy an d read in g p erfo rm an ce in
the U nited Stated from 1880 to the present. Reading Research Quarterly
22, 8-46.
S t e i n e r , G. 1978. On difficulty and other essays. O xford an d New York:
S t e w a r t , M.E and C. G r o b e . 1979. Syntactic maturity, mechanics of writing,
and teachers’ quality ratings. Research in the Teaching of Writing 13, 207-15.
S t o l a r e k , E. 1994. Prose m odeling and m etacogition: T he effect o f m od-
eling on developing a metacognitive stance toward writing. Research in
the Teaching of Writing 28, 154-74.
S t r e e t , B. 1984. Literacy in theory and practice. Cam bridge, E ngland and
New York: C am bridge University Press.
S t r e e t , B. (ed.). 1993. Cross-cultural approaches to literacy. Cam bridge,
England an d New York: Cam bridge University Press.
S t u b b s , M. 1980. Language and literacy: The sociolinguistics of reading and
writing. L ondon and Boston: Routledge 8c Kegan Paul.
S t y g a l l , G. 1991. Texts in oral context: T he ‘transm ission’ o f ju ry in-
structions in an Indiana trial. In C. B a z e r m a n an d J. P a r a d i s (eds)
Textual dynamics of the professions. Madison, WI: University o f Wisconsin
Press. 234-253.
S w a l e s , J. 1981. Aspects of article introductions. Birm ingham : Language
Studies U nit, University o f Aston. [Aston ESP Research Reports 1.]
S w a l e s , J. (ed.). 1985. Episodes in ESP. O xford: Pergam on.
S w a l e s , J. 1990. Genre analysis: English in academic and research settings.
Cam bridge, England an d New York: Cam bridge University Press.
472 Bibliography
S w a l e s , J. 1991. Discourse analysis in professional contexts. In W. G r a b e et

al (eds) Annual review of applied linguistics, 11. Cam bridge, E ngland and
S w a l e s , J. 1993. G enre and engagem ent. Revue Beige de Philologie et
d ’H istoire71, 687-98.
S w a l e s , J. and H. N a j j a r . 1987. T he writing o f research article intro-
ductions. Written Communication 4, 175-91.
Downloaded by [University of California, San Diego] at 23:05 19 June 2017
T a d r o s , A. 1994. Predictive categories in expository text. In M.

C o u l t h a r d (ed.) Advances in written text analysis. L ondon an d New
York: Routledge. 69-82.
T a n g , G. 1992. T h e e f f e c t s o f g r a p h i c r e p r e s e n t a t i o n o f k n o w l e d g e
s t r u c t u r e s o n E S L r e a d i n g c o m p r e h e n s i o n . Studies in Second Language
Acquisition 14, 177-95.
T a n n e n , D. 1982. O ral and literate strategies in spoken and w ritten n arra-
tives. Language 58, 1-21.
T a n n e n , D. 1985. Relative focus on involvement in oral and w ritten
discourse. In D. O l s o n , N. T o r r a n c e an d A. H i l y a r d (eds) Literacy,
language, and learning: The nature and consequence of reading and writing.
124-47.
T a n n e n , D. 1987. T he orality o f literature and the literacy o f conversa-
tion. In J. L a n g e r (ed.) Language, literacy and culture: Issues in society and
schooling. N orw ood, NJ: Ablex. 67-88.
T a n n e n , D. 1989. Talking voices. Cam bridge, England an d New York:
T a r o n e , E., S. D w y e r , S, G i l l e t t e an d V. I c k e . 1981. O n the use o f the
passive in two astrophysics jo u rn a l papers. ESP Journal 1, 123-40.
T a t e , G. (ed.). 1987. Teaching composition. Fort Worth, TX: Texas Christian
University Press.
T a y l o r , D. 1993. A counseling approach to writing conferences. In T.
F l y n n and M. K i n g (eds) Dynamics of the writing conference. U rbana, IL:
N ational Council of Teachers o f English. 24-33.
T a y l o r G. and T . C h e n . 1991. Linguistic, cultural, and subcultural issues
in contrastive discourse analysis: Anglo-American and Chinese
scientific texts. Applied Linguistics 12, 319-36.
T c h u d i , S. an d M. H u e r t a . 1983. Teaching writing in the content areas:
Middle school/junior high. W ashington, DC: N ational Education
Association.
T e s t o f W r i t t e n E n g l i s h (T W E ) G u i d e . 1992. P r in c e t o n , NJ: E d u c a t io n a l
T e s t in g S e r v ic e .
T h a r p, R. and R. G a l l i m o r e . 1988. Rousing minds to life. Cam bridge,
T i e r n e y , R., M. C a r t e r and L. D e s a i . 1991. Portfolio assessment in the
reading-writing classroom. N orw ood, MA: C hristopher G ordon.
Bibliography 473
T ie r n e y , R. and R Mo sent h a l . 1983. Cohesion and textual coherence.

Research in the Teaching ofEnglish 17, 215-29.
T ie r n e y , R. and RD. P e a r s o n . 1983. Toward a com posing m odel o f
reading. Language Arts 60, 568-80.
T i e r n e y , R ., J. R e a d a n c e and E. D i s h n e r . 1990. Reading strategies and
practices: A compendium. 3rd edn. Boston: Allyn 8c Bacon.
T o m p k i n s , G. 1990. Teaching writing: Balancing process and product. New
York: Merrill.
T r i m b l e , L. 1985. English far science and technology. Rowley, MA: Newbury
H ouse.
T s a o , F. 1983. Linguistics an d w ritten discourse in particular languages:
Contrastive studies: English and Chinese. In R.B. K a p l a n et al. (eds)
Annual review of applied linguistics, 3. Rowley, MA: Newbury House.
99-117.
T u m a n , M. (ed.). 1992. Literacy online. Pittsburgh, PA: University o f
Pittsburgh Press.
U n iv ersity o f S o u th e r n C a lifo r n ia (U SC), UCC.USM. 1985. University
o f Southern California, University C om puter Center, U sers Manual,
A A 02. page 8.
Urzua, C. 1987. ‘You stopped too soon’: Second language children
composing and revising. TESOL Quarterly 21, 279-304.
V a c h e k , J . 1966. The linguistic school of Prague. Bloom ington, IN: University
of Indiana Press.
V a h a p a s s i , A. 1988. T he problem o f selection o f writing tasks in cross-
cultural study. In A. P u r v e s (ed.) Writing across languages and cultures.
L ondon and Newbury Park; CA: Sage. 51-78.
V a n d e K o p p l e , W. 1982. T he given-new strategy o f com prehension and
some natural expository paragraphs. Journal of Psycholinguistic Research
11,501-20.
V a n d e K o p p l e , W. 1983. Som ething old, som ething new: Functional
sentence perspective. Research in the Teaching ofEnglish 17, 85-99.
V a n d e K o p p l e , W. 1986. Given and new inform ation and some aspects o f
the structures, semantics, an d pragm atics o f w ritten texts. In C . C o o p e r
and S. G r e e n b a u m (eds) Studying writing: Linguistic approaches. L ondon
and Newbury Park, C A : Sage. 72-111.
V a n d e K o p p l e , W. 1990. Them es, them atic progressions, and some impli-
cations for understandin g discourse. Written Communication 8, 311-47.
v a n Dijk, T. 1985. Handbook of discourse analysis, 4 Vols. L ondon and New
York: Academic Press.

v a n D i j k , T. 1988. News as discourse. Hillsdale, NJ: L. Erlbaum.
v a n D i j k , T. and W. K i n t s c h . 1983. Strategies of discourse comprehension.
L ondon and New York: Academic Press.

v a n N a e r r s e n , M. and R.B. K a p l a n . 1987. Language and science. In R.B.
K a p l a n et al. (eds) Annual review of applied linguistics, 7. Cam bridge,

England and New York: C am bridge University Press. 86-104.
V e n t o l a , E. and A. M a u r a n e n . 1991. Non-native an d native revising of
474 Bibliography
scientific articles. In E. V e n t o l a (ed.) Functional and systemic linguistics:

Approaches and uses. Berlin: M outon de Gruyter. 457-92.
V y g o t s k y , L. 1962. Thought and language (E. H aufm ann and G. Vadar, eds
and trans.). Cam bridge, MA: MIT Press.
V y g o tsk y , L. 1978. M ind in society (M. Cole et al. eds an d trans.).
Cam bridge, MA: H arvard University Press.
V y g o t s k y , L. 1983. T he prehistory o f written language. In M . M a r t l e w
(ed.) The psychology of urritten composition. New York: J. Wiley. 279-91.

W a l d , B. 1987. T he developm ent o f writing skills am ong H ispanic high
school students. In S. G o l d m a n and H. T r u e b a (eds) Becoming literate in
English as a second language. Norw ood, NJ: Ablex. 155-185.
W a l k e r , C. and D. E l i a s . 1987. W riting conference talk: Factors
associated with high- an d low-rated writing conferences. Research in the
Teaching of English 2 1, 266-85.
W a l s h e , R. 1981a. Every child can unite! Victoria, NSW: Prim ary English
T eacher’s Association.
W a l s h e , R. 1981b. Donald Graves in Australia: Children want to unite ...
Victoria, NSW: Prim ary English T eacher’s Association.
W a l s m l e y , S. 1994. Children exploring their world: Theme teaching in elemen-
tary school. L ondon and Portsm outh, NH: H einem ann.
W a l v o o r d , B. and L. M c C a r t h y . 1991. Thinking and uniting in college: A
naturalistic study of students in four disciplines. U rbana, IL: National
W a s ik , B. and R. S l a v i n . 1993. Preventing early reading failure with one-
to-one tutoring: A review of five program s. Reading Research Quarterly 28,
179-200.
W a u t e r s , J. 1991. Evaluation for em pow erm ent: A portfolio proposal for
Alaska. In P. B e l a n o f f and M. D i c k s o n (eds) Portfolios: Process and prod-
uct. Portsm outh, NH: B oynton/C ook. 57-68.
W e a s e n f o r t h , D. 1995. Rhetorical abstration as a facet o f expected
response: A structural equation m odeling analysis. Los Angeles:
University of S outhern California. Ph.D. diss.
W e l l s , G. 1986. The meaning makers: Children learning language and using
language to learn. L ondon and Portsm outh, NH: H einem ann.
W e l l s , G. 1994. T he com plem entary contributions o f Halliday and
Vygotsky to a ‘language-based theory o f learn in g ’. Linguistics and
Education 6, 41-91.
W e l l s , G. and G. C h a n g - W e l l s . 1 9 9 2 . Constructing knowledge together.
L ondon and Portsm outh, NH: H einem ann.
W e r t s c h , J. (ed.). 1985. Culture, communication and cognition. Cam bridge
W e r t s c h , J. 1991. Voices o f the m ind: A sociocultural ap p ro ach to m edi-
ated action. C am bridge, MA: H arvard University Press.
W h i t e , E. 1993. Holistic scoring: Past trium phs, future challenges. In
M. W i l l i a m s o n and B. H u o t (eds) Validating holistic scoring for uniting
Bibliography 475
W h ite , E. 1994. Teaching and assessing uniting, 2nd edn. San Francisco:
Jossey Bass.
W h i t e , E. 1995. An apologia for the tim ed im prom ptu essay test. College
W h i t e , R. and V . A r n d t . 1991. Process uniting . L ondon an d New York:
Longm an.
W h i t w o r t h , R. 1988. Collaborative learning an d o th er disasters. In

J. G o l u b (ed.) Focus on collaborative learning. U rbana, IL: N ational
Council of teachers o f English. 13-20.
W h o r f , B.L. 1941. G ram m atical categories. Language 21, 1-11.
W i d d o w s o n , H.G. 1978. Language teaching as communication. O xford and
New York: O xford University Press.
W i d d o w s o n , H.G. 1979. Explorations in applied linguistics. O xford and New
York: O xford University Press.
W i d d o w s o n , H.G. 1980. C onceptual an d com municative functions in
written discourse. Applied Linguistics 1, 234-43.
W i d d o w s o n , H.G. 1983. Learning purpose and language use. O xford and
New York: O xford University Press.
W i l d , A . 1985. P erform ance review. L A . Times. (March 16, 1985) V . l .
W i l l e y , R . 1990. Pre-classical roots o f the addressed/invoked dichotom y
of audience. In G . K i r s c h and D . R o e n (eds) A sense of audience in
written communication. L ondon and Newbury Park, CA: Sage. 25-39.
W i l l i a m s , F. 1976. Explorations of the linguistic attitudes of teachers. Rowley,
MA: New bury H ouse.
W i l l i a m s o n , M. 1993. An introduction to holistic scoring: T he social,
historical, and theoretical context for writing assessment. In M.
W i l l i a m s o n and B. H u o t (eds) Validating holistic scoring for writing
W i n s o r , D . 1990. E ngineering w riting/w riting engineering. College
W i t t e , S. 1983a. Topical structure an d revision: An exploratory study.
W i t t e , S. 1983b. Topical structure and writing quality: Some possible text-
based explanations for read ers’ ju d g m en ts o f student writing. Visible
Language 17, 177-205.
W i t t e , S. 1992. Context, text, intertext: Toward a constructivist semiotic
o f writing. Written Communication 9, 237-308.
W i t t e , S. and R. C h e r r y . 1986. W riting processes and w ritten products in
com position research. In C . C o o p e r and S. G r e e n b a u m (eds) Studying
writing: Linguistics approaches. L ondon an d Newbury Park, CA: Sage.
112-53.
W i t t e , S. and L. F a i g l e y . 1981. C oherence, cohesion, and writing quality.
W o d a k , R. 1989. Topic developm ent, syntax, and social class. Discourse
Processes 11, 267-86.
476 Bibliography
Wo dak, R. 1990. Discourse analysis: Problems, findings, perspectives. Text

10, 125-32.
Wo l f so n, N. 1989. Perspectives: Sociolinguistics and TESOL. New York:
Newbury House.
Y o u n g , R. 1987. R ecent developm ents in rhetorical invention. In G. T a t e
(ed.) Teaching composition. Fort W orth, TX: Texas Christian University
Press. 1-38.
Y o u n g , R., A. B e c k e r and K. P i k e . 1970. Rhetoric: Discovery and change.

London and New York: H arcourt, Brace and World.
Z a m e l , V. 1982. Writing: T he process o f discovering m eaning. TESOL
Z a m e l , V . 1983. T he com posing processes o f advanced ESL students: Six
case studies. TESOL Quarterly 17, 165-87.
Z a m e l , V. 1985. R esponding to student writing. TESOL Quarterly 19,
79-101.
Z a m e l , V. 1987. R ecent research on writing pedagogy. TESOL Quarterly
21, 697-715.
Z a m e l , V. 1992. W riting o n e ’s way into reading. TESOL Quarterly 26,
463-85.
Z e l l e r m e y e r , M. 1988. An analysis o f oral and literate texts: Two types o f
reader-w ri ter relationships in H ebrew an d English. In B. R a f o t h and
D. R u b i n (eds) The social construction of written communication. N orw ood,
NJ: Ablex. 287-303.
Z h u , W. 1994. Effects of training for p eer revision in college freshm an
com position classes. Flagstaff, AZ: N o rth ern A rizona University. Ph.D.
diss.
Subject Index
Assessment: Discourse Community: 21, 106-11,

Classroom-Based Approaches, 370
406-13,418-20
Limitations of, 420-1 Editing: 326-7
Portfolios, 414-20 English as a World Language:
Scoring Approaches, 404-20 156-8
Audience: 20, 34, 207-12, 241, 264, ESL/EFL: 24-5, 247-50,257-8
309-10 ESP (English for Special Purposes):
27-8, 147, 155-61, 256-7
Bereiter and Scardamalia’s Writing Ethnographic Research: 19, 28, 30-1,
Process theory: 94, 117-28, 145, 90, 96-102, 105
214, 229, 243-6, 314 Ethnography of Writing: 203-16
Expressive Approaches to Writing:
Cognitive Approaches to Writing 88-9
Research: 19, 89-94, 113-32, Extensive Reading: 297-8, 345
213-14, 223-35, 240-1, 243-6
Coherence: 20, 28, 59, 61-2, 65-76, Flower and Hayes Model of the
80-1 Writing Process: 91-4, 114-17,
Cohesion: 20, 28, 53, 56-9, 65, 70, 126-7, 145, 214, 243^·
80-1,239 Freirien Approaches: 32, 110-11,
Communicative Competence: 224-34 263, 274
Computer Assisted Writing Freshman Composition: 22
Instruction: 21, 298
Content-Based Instruction: 21, 32, Genre: 20, 108, 1 3 2 ^ 0 , 160, 172,
262, 291-3, 301, 305, 308-9, 179-80, 199, 205-6, 209-12,
322-3,368-9 242, 260, 264, 289-94, 318-20,
Contrastive Rhetoric: 26, 172, 327-34, 362-7
176-201,239
Corpus Research: 16-17, 46-9, 195-7 Hallidayan Approaches: 56-9, 103-4,
Curricula for Writing Instruction: 133-7, 260
32-3, 110-11,261-5 Hermeneutics: 20, 38
477
478 Subject Index
Heuristics for Organizing Writing: Revision: 32, 262, 275, 326-8,

270-1,276, 278-9,310-13, 357-60, 379-93
317-18, 350-2
Sentence Combining: 43-5, 335-6
Information Structuring: 20, 28, 33, Social Construction: 20-1,147,
42-56, 61, 75, 204 161-71, 174
Focus-Presupposition, 20, 51-2 Social Contexts of Writing: 19,24-5,

Functional Sentence Perspective, 29, 34, 95-112, 203-16, 241-3
49 Social Semiotic Approaches: 103—4
Given-New, 20, 49-51, 53-5, 61, Style: 373-4
75, 204 Summary Writing: 323-5
Theme-Rheme, 20,49-51, 204 Syntactic Analyses: 43-9, 62-5,
Topic-Comment, 20, 49-51, 53-5, 77-8, 80-1, 159
61,204
Instructional Activities: 266-376 Taxonomy of Writing Skills :
Intertextuality : 20-1,111 216-23
Technical Writing: 147-51
Journal Writing: 125, 295-6, 338 Text Construction; 32,42, 60-83,
204-5, 238-9
Knowledge Telling: 118-22 Textlinguistics: 20, 28, 33, 36-83,
Knowledge Transforming: 4-5, 118, 204-5
122-8 Topic Continuity: 53^l·, 56
Topical Sentence Structure: 52-3
Language Experience Approach Topical Structure Analysis: 54-5
(LEA): 18, 273, 277-8, 309,
318 Vygotskean Approaches: 18, 242-3,
Lexico-Syntactic Features: 28, 33, 259, 262,315, 387
43-9, 77-8, 80-1
Literacy Whole Language: 18, 27, 29, 31-3,
Cognitive differences, 12-14 99, 103-5, 258-60
History of, 5-14 Word Processing: 21, 298
Social Contexts of, 12-14, 17, Writing
34-5 A Model of, 223-34
Across the Curriculum, 32, 368-69
Oral versus Written Language: 15-18, As Communication, 41, 224-34,
4 0,61,77, 102, 180, 325-6 266
Outlining: 349-50 As Composing, 4-5, 17, 32,40,
89-100, 113-28, 224-34,
Peer Feedback: 32, 264, 275-6, 343-4
315-16, 379-88 As Technology, 6-7, 17
Portfolios: 336, 414-20 Assessment, 21, 34,264, 377-8,
Publishing: 296-7 395-421
History of, 5-11, 14, 85-112
Responding to Writing: 264, 275-6, In a Second Language, 22-31,
315-16, 357-62, 377-95 140-5, 238-41, 247-58
Subject Index 479
In Non-Academic Settings: 147, 224-34, 240-1, 243-6, 262, 276,

152-5 316-17
In Science: 161-73 Writing Comer: 296-7
Process, 3 2 ^ , 84-146, 213-14,
Author Index
Aaronson, E., 307 Bartholomae, D., 32, 107, 109-10,

Abrahamson, D., 42, 57 261,308, 366, 368, 372
Ackerman, J., 152, 369 Barton, D., 7
Adams, M., 36, 105, 270 Basso, K., 203
Agnoli, F., 179 Bateson, G., 80
Alderson, C., 401 Bazerman, C., 107, 162-3, 165-70,
Allaei, S., 386, 406 172, 185, 386
Allen, R. V.,318 Beach, R., 360
Alvermann, D., 308 de Beaugrande, R., 39, 43^4-, 60, 67,
Anderson, P., 151-3, 155 75,80, 93, 135,202
Anson, C., 379 Beck, I., 20, 61,69-70
Applebee, A., 31, 84, 88, 120, 125, Belanoff, P., 383, 387,414-15
132, 258, 316, 402, 405 Benson, J., 51
Armbruster, B., 312-13 Bereiter, C., 5, 19, 21, 60, 80, 89-90,
Amdt, V., 269, 390, 393 93—4, 98, 114, 116-19, 122,
Amove, R., 9 124-9, 131-2, 134, 141, 199,
Asenavage, Κ., 379 202, 214, 229, 242-6, 314
Asher, J., 91, 98-9 Berkenkotter, C., 106-7, 165, 179,
Atkinson, D., 161, 165, 199, 204, 206, 238
206, 242 Berlin, J., 11, 30, 39, 88, 112, 139,
Au, Κ., 25, 101 200
Berman, R., 179, 184
Bachman, L., 216, 224-5, 228 Bernhardt, E., 140
Baker, E., 406 Bernstein, B., 14, 100
Baker, N., 416 Berthoff, A., 338
Bakhtin, M., 164 Besnier, N., 187
Baldauf, R., 156 Bhatia, V., K., 140, 159, 163, 172,
Bamberg, B., 67 206
Bangert-Drowns, R., 21, 213 Biber, D., 15-16,40,46,48, 63,
Barber, C., 159 77-8, 163, 177, 179, 186-7, 195,
Barsalou, L., 68, 226-8 198,203, 208
Author Index 481
Bickner, R., 186, 191 Carter, R., 21, 125

Bissex, G., 105 Cavalcanti, M., 143-4
Bizzell, P., 109, 165 Cazden, C., 19, 95,100,238
Black, L, 19, 39 Celce-Murcia, M., 216, 225, 228
Blair, H., 11 Cemiglia, C., 33, 53, 186
Blok, H., 401 Chafe, W., 15-16,77-8
Bloome, D., 15, 19, 102 Chang-Wells, G., 15, 19, 84
Bocaz, A., 184 Chapelle, C., 216, 225, 232

Boggs, S., 13, 19,96, 100-1 Chamey, D., 143
Bonk, W., 242 Chen, T., 180-1
Boomer, G., 84 Cheng, P., 189
Bowerman, M., 184 Cherry, R , 44,413
Boyle, O., 29, 32, 269, 274 Choi, S., 184
Bracey, G., 11 Chomsky, C., 105
Brandt, D., 315 Christie, F., 19, 33, 59, 84,102, 125,
Branscombe, A., 102 132-4, 136, 242, 260, 319, 370
Breland, H., 397 Clancy, P., 185
Brinton, D., 32, 107 Clay, M., 97, 105
Britton, B., 19, 39, 61,69-71, 75 Clifford, J., 385
Britton, J., 5,21 ,3 0 , 84, 118, 139 Clyne, M., 186-7
Brown, A., 116, 131, 136, 314 Cochran-Smith, M., 21, 213
Brown, G., 39-40, 55-6, 58,60, Coe, R., 11, 132,138
67-70 Cohen, A., 1 4 3 ^ , 244
Bruce, B., 21 Cohen, G., 79
Bruffee, J., 164-5, 185, 380 Collerson, J., 136, 260, 319
Bruner, J., 131, 136 Collins, A., 116, 131, 136, 387
Bruthiaux, P., 204 Collins, H., 163, 165
Bryson, M., 244, 246 Collins, J., 242
Buck, G., 401 Comprone, J., 39
Condon, W., 402, 414, 416-17
Calfee, R., 410 Conley, J., 173
Calkins, L., 18, 90, 96-7, 316, 391 Connor, U., 54-5, 181, 186-7, 197,
Camp, R , 396-7, 414, 417, 421 379,386,406
Campbell, C., 240 Conrad, S., 391
Campbell, G., 11 Cook-Gumperz, J., 9, 13, 15, 19, 101
Campione, J., 314 Cooper, C., 94, 108, 178, 202
Canale, M., 216, 224, 405 Cooper, M., 106, 108-09
Cantoni, G., 318 Cooper, R., 203
Carlisle, R., 142 Cope, B., 19, 100, 132, 137, 145, 260,
Carlson, S., 195, 397-8, 400 319
Camicelli, T., 391 Copeland, J., 382
Carpenter, P., 226-7 Corbett, E., 20
Carrell, P., 70 Coulthard, M., 20, 39, 51, 178
Carson, J., 143, 379, 386 Couture, B., 51,152
Carter, M., 116, 129, 132, 245 Cox, B., 20, 58-9
482 Author Index
Crandall, J., 32, 308 Faigley, L., 45, 58, 65, 88, 90, 94,
Cressy, D., 7 111-12, 154, 164,406, 409
Crowhurst, M., 140, 242 Fairclough, N., 109
Crowley, S., 11,39, 139, 185 Faltis, C„ 29, 32
Cumming, A., 241, 244, 405 Farmer, M., 186
Czubaroff, J., 169,185 Fathman, A., 142-3, 238, 306-7, 361,
380, 386, 393

D ’Angelo, F., 135, 139 Feitelson, D., 105
Daiker, D., 43 Ferris, D., 142-3
Dantas-Whitney, M., 195 Fielding, L., 313
Davis, A., 107 Finegan, E., 77, 177
Degenhart, R. E., 404 Firbas, J., 49
De Glopper, K., 401 Fitzgerald, J., 20,58-9
DeMauro, G., 398 Fleck, L., 167
Demuth, K., 184 Flower, L., 19, 34, 80, 90-5, 99,
Derewianka, B., 136,140, 260, 319 106-7,113-17, 126-7, 145, 151,
Dickson, M., 414 202,214, 224, 229,244, 380,
Diederich, R, 409 386-7
Dillon, G., 60, 76-7, 80 Flynn, T„ 361
DiPardo, A., 379 Fotos, S., 398
Dixon, C., 274 Fowler, R., 39
Dobrin, D., 92, 148, 151—2 Frank, M., 269, 275-6, 316, 325, 382
Doheny-Farina, S., 151 Frawley, W., 64
Dressier, W., 39, 60, 67 Freebody, P., 100
Dudley-Evans, T., 29, 78, 158-9 Freedman, A., 21, 106-7, 137, 206
Dunbar, S., 406, 415 Freedman, S., 45, 142, 360, 379, 386,
Dyson, A., 18, 29, 32, 145 391
Freeman, D., 28
Eagleton, T., 38-9 Freeman, Y., 28
Ede, L., 107 Freidlander, A., 205, 241
Edelsky, C., 28,103-4, 141, 241, 259 Freire, P., 110
Educational Testing Service (ETS), Fries, P., 51
400
Eggington, W., 186, 188 Gaies, S., 43
Eisenstein, M., 10 Gallimore, R., 19, 25, 96, 101, 243
Elbow, P., 88-9, 316, 382-3, 387, 415 Gardner, H., 86
Elias, D., 391 Gamham, A., 36, 67, 71, 76, 80
Emig, J., 90 Gaskill, W„ 141
Empson, W., 41 Gaskins, I., 105
Enright, D. S., 29, 105, 260, 308, Gee, J., 12, 185
323 Geertz, C., 164
Ericsson, K., 91 Geisler, C., 150, 162
Evenson, L., 199 Genesee, F., 259
Gentile, C., 403,416
Fagan, W., 409 George, D., 386
Author Index 483
Gere, A., 379, 386 Havelock, E., 13

Gilbert, G., 163, 165, 167, 169 Hawisher, G., 196
Giroux, H., 110 Hayes, J., 80, 90-4, 99, 113-17, 126,
Givon, T., 39, 5 3 ^ 145, 151,214, 229, 233
Goldberg, A., 56 Heath, S. B., 13-14, 19, 25, 95, 98,
Goldman, S., 28 101-2, 137, 185,241,325,365
Goldstein, L., 391 Hedge, T., 269
Golub, J., 361, 379, 387 Heimlich, J., 312

Golub-Smith, M., 400 Henning, G., 400, 405
Goodman, Κ., 27, 99, 103-5 Hiebert, E, 410
Goodman, Y., 104 Hilgers, T., 265
Goody, J., 8, 12-14 Hilligoss, S., 21
Gorman, T. P., 196, 404 Hillocks, G., 18-19,43-5, 90, 98-9,
Goswami, D., 152
112, 142-3,241,319
Gough, Κ., 13
Hinds, J., 186,188-90
Grabe, W., 33, 39, 47-8, 77-8, 93,
Hirsch, E. D., 110
142, 150, 156-7, 163, 177,
Hoey, M., 20, 56, 58-9, 65, 71, 75,
185-7, 194-6, 200, 207, 225
159
Graff, H., 7-10
Holt, D., 416
Graves, D., 18, 90, 95-100, 105, 120,
Holt, M., 383, 387
316,391
Gray, B., 238, 242 Homer, W., 20
Greaves, W., 51 Horowitz, D., 409, 414
Green, J., 15, 19, 102 Houston, R., 7
Greenbaum, S., 178 Huckin, T., 106-7, 165,179, 206, 238
Greenberg, Κ., 402 Hudelson, S., 28, 84, 105, 141
Gulgoz, S., 61, 69, 71, 75 Hudson, R., 39
Huerta, M., 308, 323, 325
Hairston, M., 87-8 Hughey, J., 246, 251, 395, 409
Hale, G., 400 Hunt, E., 179
Halliday, M. A. Κ., 14-15, 20-1, 33, Hunt, K., 43
3 9 ^ 0 , 50-1, 56, 58-60, 65, Hutchinson, T., 29, 171
67-8, 70, 76, 78,100, 1 0 3 ^ , Hymes, D., 40, 224
133-5, 159, 182, 206, 259-60,
266 Indrasutra, C., 186, 191
Hamp-Lyons, L., 142, 402, 404-5, Ingram, D., 401
409,414,416-17,421 Ivanic, R., 7
Hanania, E., 398
Harley, B., 216, 225 Jackendoff, R., 64
Harris, D., 199 Jacobs, H., 409
Harris, M., 386 Jagger, A., 103
Harste, J., 96, 104 Jenkins, S., 190
Hasan, R., 14, 20-1, 39, 51, 56, Jernudd, B., 156
58-60, 65, 67, 70, 78, 100,102, Johns, A., 29, 33, 67, 158, 241
182, 206, 260,319 Johnson, D., 28, 84, 214
Hatim, B., 193^1 Johnson, D. W., 307
484 Author Index
Johnson, R., 307 Latour, B., 165

Johnson-Laird, P., 79 Lauer, J., 91, 98-9, 186
Jones, S., 21, 141-2, 240 Lauttamatti, L., 52, 187
Jonz, J., 399 Lawson, B., 379-80
Joos, M., 77 Lay, N., 241
Jorden, C., 25, 101 Leki, I., 23, 33, 84, 140, 143, 185,
Just, M., 226-7 197, 238-41, 244, 246, 248-9,
251,304, 382, 386-7

Kachru, B., 25 Lemke, J., 172, 206, 208, 211, 215
Kachru, Y., 145,179,186, 192 Ley, C., 258
Kaestle, C., 10 Liberman, I., 5
Kagan, S., 307—80 Liberman, Y., 5
Kalantzis, M., 19, 100, 132, 137, 145, Lindemann, E., 236
260,319 Linden, M., 335
Kaplan, R. B., 28-9, 33-4,71, 75, Loban, W., 43
150,156-8, 163,182, 185-6, Lomax, E., 382
189, 191, 193, 198, 200-1,215, Long, R., 207
318 Lucy, J., 179
Käufer, D., 152, 369 Lunsford, A., 107
Kellogg, R., 19 Lux, P., 186-7, 194-6
Kessler, C., 306, 307, 361, 380 Lynch, M., 165
King, M., 361
Kinneavy, J., 139, 152 Macedo, D., 110
Kintsch, W., 19-20, 69, 71, 353 MacKinnon, J., 149, 151-2
Kirsch, G., 207 Maclean, R., 132
Knapp, M., 107 Macrorie, K., 88
Koretz, D., 417 Maher, J., 29, 165
Krapels, A., 414, 214, 240-1 Mangiola, L., 101-2, 325, 365
Krashen, S., 142 Mann, W., 20, 39, 71, 75
Kress, G., 15, 136-7, 145, 179, 215 Manning, M., 308
Kreuger, M., 32 Marland, M., 47
Kroll, Barbara, 28, 84, 409 Marsalla, J., 265
Kroll, Barry, 90, 107 Martin, J., 21, 33, 51, 59-60, 64, 71,
Kuhn, T., 87,164-5 78, 80, 120, 125,132, 134-5,
Kutz, E., 326, 368, 373 137-8, 140, 159, 199, 206, 224,
242, 260, 319-20, 370
Labov, W., 201 Matalene, C., 149, 190
Lackstrom, J., 159 Matsuhashi, A., 202
Ladusaw, W., 64 Mauranen, A., 186
Langacker, R., 165 McCagg, P., 187
Langer, J., 107, 380 McCarthy, L., 370
Laponce, J., 156 McCloskey, M., 29, 105, 260, 308,
Lareau, A., 100, 102 323
Large, J. A., 156-7 McCutchen, D., 214
Larsen-Freeman, D., 43 McGinley, W., 131
Author Index 485
McKenna, M., 105 Ochs, E., 78, 185

McLeod, B., 369 Odell, L., 151-2, 155
Mech, D., 73 O ’Hare, F., 43
Medgyes, P., 156 Oller, J., 399
Medway, P., 106-7, 206 Olsen, L., 165
Meinbach, A., 323 Olson, D., 12-14, 180, 202-3
Merton, R., 171
O ’Malley, M., 410

Messick, S., 406,415,421 Ong, W., 13, 38
Meyer, B., 71, 77,140 Ostler, S., 186,190,193-4
Meyer, P., 413
Milanovic, M., 401 Painter, C., 133,137
Miller, C., 106 Paivio, A., 79, 228
Miller, J., 105 Palincsar, A., 116, 131, 136, 314
Miller, T., 33 Pandharipande, R., 192
Moffet, J., 89-90, 139 Paradis, J., 154,173
Mohan, B., 137, 308,312-13 Park, D., 107
Moll, L., 18 Pearson, P. D., 73, 313
Montano-Harmon, M., 29,186,194—5 Pennington, M., 21
Mosenthal, P., 58, 70 Peregoy, S., 29, 32,269, 274
Mulkay, M., 163, 165, 169-70, 185 Perera, K., 44, 63
Mullen, K., 398 Perfetti, C., 36, 214
Murphy, J., 379, 386 Perkins, K., 396-7,404-5
Murray, D., 88 Perl, S., 90
Myers, G., 88,165, 169-70, 185 Petrosky, A., 32, 107, 110, 261, 308,
366, 368, 372
Nagy, W., 312 Peyasantiwong, P., 186, 191
Najjar, H., 163 Peyton, J. K., 125, 195
National Council of Teachers of Phelps, L., 18, 39, 67
English (NCTE), 269 Philips, S., 19, 101, 185
Needels, M., 107 Phillipson, R., 157
Neel, J., 39, 185 Pierce, L., 410
Nelson, G., 379, 386 Pittelman, S., 312
Nessel, D., 274 Pollatsek, A., 36, 68, 76
Newell, A., 91 Poole, D., 19, 203
Newkirk, T., 269 Prelly, L., 169
Newman, D., 131,243, 361, 387 Pressley, M., 130-1, 199, 313-14
Nichols, J., 78 Priestley, J., 11
Nold, E., 45 Prince, E., 55
North, S., 18, 20, 31, 37, 88-90, 92, Purves, A., 7, 10, 115, 178-9, 196,
95, 98-9, 112 404
Nystrand, M., 107-8, 110, 315, 380,
385 Quirk, R., 63
Oakhill, J., 36, 67, 76, 80 Rabin, C., 39

Ö ’Barr, W., 173 Radecki, P., 142
486 Author Index
Rafoth, B., 107, 165 Shapiro, M., 185

Raimes, A., 28, 30-1, 33-4, 41, 84, Sharan, S., 307
141-2, 238—41, 244, 261, 269, Sharan, Y., 307
338, 351-52, 382-3 Shaughnessy, M., 90
Rayner, Κ., 36, 68r, 76 Shaw, P., 318
Readance, J., 308 Shih, M., 107
Reid, J., 28, 33, 186, 190, 194, 269,
Shikhani, M., 398

316, 382, 385, 387, 392-3, 395, Shor, I., 110
409 Short, D., 32
Reither, J., 106, 149 Shuchman, H., 157
Reppen, R., 78, 142, 186, 194 Shuy, R., 173
Resnick, D., 9 Silva, T., 23, 137, 1 3 9 ^ 0 , 143^1,
Resnick, L., 9 200, 214, 239-40, 304
Reynolds, J., 149, 151-2, 155 Simon, H., 91
Richards, J., 240, 244, 246 Sinclair, J., 46, 63
Richgels, D., 140 Singer, M., 20, 59, 61, 67-9, 71, 353
Rieben, L., 36 Slavin, R., 306-7, 361,380
Robb, T., 238 Slobin, D., 179, 184
Robinson, G., 158 Smagorinsky, P., 19, 91-2, 143, 199,
Roen, D., 28, 84, 207, 214 238,317
Rogoff, B., 243, 361, 387 Smit, D., 416
Rokosz, D., 29 Smith, N., 68
Rorty, R., 164 Smith-Burke, T., 103
Rosenberg, R., 416 Snow, C., 19
Ross, J., 400 Snyder, I., 21
Rothery, J., 138, 159, 206 Sokmen, A., 391
Rubin, A., 21 Sommer, N., 90
Ryan, F., 32 Soter, A., 186, 191-2
Rymer, J., 152 Soven, M., 369
Spack, R., 369
Sa’Adeddin, M., 193-4, 198 Spandel, V., 411
Santa, C., 308 Spear, K., 306, 315, 379, 382, 384-7
Scardamalia, M., 5, 19, 60, 80, 89-90, Sperber, D., 68, 71
93-4, 98, 114, 116-19, 122, Sperling, M., 391
124-9, 131-2, 134, 141, 199, Spiegel, D., 20, 58-9
202, 214, 229, 242-6, 314 Spilka, R., 149-50, 154
Schieffelin, B., 195 Spindler, G., 95, 98, 185
Schiffrin, D., 40, 178, 203 Spindler, L., 95, 98, 185
Scholes, R., 13 Spiro, R., 131, 137, 140, 245
Schriver, K., 310 Sridhar, S. N., 145
Schroder, H., 210 Stahl, R., 263, 307
Scollon, R., 77, 101, 185, 189 Stahl, S., 105
Scollon, S., 77, 101, 185 Stanley, J., 386-7
Scribner, S., 13 Stanovich, K., 36, 93
Seife, C., 21, 90 Stansfield, C., 397, 400
Author Index 487
Staton, J., 296 Ventola, E., 186

Stedman, L., 10-11 Vygotsky, L., 13, 19, 131, 242-3,
Steiner, G., 41, 212 259
Stiggins, R., 411
Stolarek, E., 199,317 Wald, B., 29
Street, B., 7, 12, 185 Walker, C., 391
Walmsley, S., 308, 323
Stubbs, M., 100

Stygall, G., 173 Walshe, R., 97
Swain, M., 216, 224 Walvoord, B., 370
Swales, J., 29, 78, 107-8, 132, 138, Wasik, B., 361
140, 142, 158-60, 163, 172, 179, Waters, A., 29, 159, 171
199, 205-6, 212, 214, 235, 242, Watt, I., 12
376 Wauters, J., 416
Welch, A., 100
Tadros, A., 65 Wells, G., 14-15, 19, 21, 35, 84, 98,
Takala, S., 196 100, 102
Tang, G., 313 Wertsch, J., 19
Tannen, D., 15-16, 77, 208, 239, 357 Whalley, E., 142-3, 238, 386, 393
Whimbey, A., 335
Tarone, E., 159
White, E., 396, 399, 404-5,409, 411,
Tate, G., 39
414, 417, 419, 421
Taylor, B., 140
White, R., 269, 390, 393
Taylor, D., 393
Whitworth, R., 387
Taylor, G., 180-1
Whorf, B. L., 82
Tchudi, S., 308, 323, 325
Widdowson, H. G., 41, 68, 70,
Test of Written English (TSE) Guide,
159-60, 235
400, 404
Wild, A., 57
Tetroe, J., 141
Willey, R., 207
Tharp, R., 19, 25, 96, 101,243 Williams, J., 19
Thompson, S., 20, 39, 71, 75 Williamson, M., 242, 414
Tiemey, R., 58, 70, 77, 141, 312, 318, Wilson, D., 68, 71
414,419-20 Winsor, D., 148
Tompkins, G., 269, 382, 393, 395 Witte, S., 18, 34, 44, 54, 58, 65, 94,
Trimble, L., 29, 163 106-7, 145, 202, 224, 238
Trueba, H., 28 Wodak, R., 215
Tsao, F., 189 Wolfson, N., 208
Tuman, M., 21 Woolgar, S., 165
Wylie, E., 401
Urzua, C., 141
Young, R., 30, 39, 152, 246, 352, 369
Vahapassi, A., 196, 404 Yule, G., 39-40, 55-6, 58, 70, 67-70
Vande Kopple, W., 20, 49, 52, 54-5,
238 Zamel, V., 84, 141, 214, 238, 240,
van Dijk, T., 19-20, 39, 69, 71, 178, 270, 391
215,353 Zellermeyer, M., 193
van Naerrsen, M., 157-8 Zhu, W., 315,379, 391

Grabe Kaplan (1996)

Uploaded by

Copyright:

Available Formats

Grabe Kaplan (1996)

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Grabe Kaplan (1996)

Uploaded by

Copyright:

Available Formats

Theory and Practice of Writing

APPLIED LINGUISTICS AND LANGUAGE STUDY

Published 2014 by Routledge

Routledge is an imprint o f the Taylor & Francis Group, an informa business

Copyright © 1996, Taylor & Francis.

ISBN 13: 978-0-582-55383-5 (pbk)

British Library Cataloguing-in-Publication Data

Library o f Congress Cataloging-in-Publication Data

Series List vii

1 Issues in writing research and instruction 1

Appendix: Seventy-five themes for xvriting instruction 427

Subject Index 477

Author Index 480

PROFESSOR CHRISTOPHER N. CANDLIN

Macquarie University, Sydney

Second Language Learning: T heory and Practice o f Writing:

Analysing Genre - Language Use Strategies in Learning and Using a

From Testing to Assessment: Teaching an d Language C orpora

Phonology in English Language

Interaction in the Language

M easuring Second Language Literacy in Society

This volume presents an applied linguistics perspective on the

connections between theory and practice in many instructional

educational systems. But as this exposure is certainly not universal

Connor, Tom Miller and Fredricka Stoller. We would also like to

This volume is respectfully dedicated to

the facets of writing sketched above cannot at one level be treated

positions, if as I say, often exploited. Approaches to textual

integrated in a descriptive model of language processing in a man-

Professor Christopher N Candlin

Issues in writing research and

Applied linguistics has concerned itself with the development of

different sorts of writing are done by which different groups of people,

■ include academic writing demands in which a sophisticated

1.2 On the nature o f writing

The need for writing in m odern literate societies - societies

■ writing a novel or a poem may exemplify what is known as aes-

Audience Writing without Writing with

For knowledge For knowledge

Self Shopping list Personal diary ‘J o u rn al’ notes

O ne known Note to milkman Personal letter

Figure 1.1 Patterns of com posing with differing audiences

The planning involves recalling and reiterating. Transforming, on

write to a functionally adequate level, and one-fifth of the world’s

1.3 Literacy and writing

The history of literacy development supports such a writing-as-

types of literacy which have developed historically under very dif-

populace. However, literacy among the classical Greek citizenry

German, Spanish) for purposes of access to the Bible in European

through the spread of literacy, the élite redefined the degree of

etiquette. At the tertiary level, however, writing instruction and

both the individual and society. The consequences of literacy on

demie education. The ‘Great Divide’ view has received support

This current research, in combination with the historical

opment, and provide lessons which should be heeded if the

1.4 Oral and written language

across oral and written language texts differ on a num ber of

all texts are highly complex multidimensional structures, and

1.5 Research on writing in English LI contexts

Generally speaking, writing research in LI contexts has been con-