Grabe Kaplan (1996)
Grabe Kaplan (1996)
Grabe Kaplan (1996)
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
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.
v
vi Contents
Bibliography 431
GENERAL EDITOR
ix
x Foreword
April 1996
William Grabe
Northern Arizona University
Robert B. Kaplan
University of Southern California
‘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
1.1 Introduction
1
2 Theory and Practice of Writing
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
16 Theory and Practice of Writing
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
18 Theory and Practice of Writing
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
26 Theory and Practice of Writing
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.
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.
Issues in writing research and instruction 35
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
36
Textlinguistic research 37
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:
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
48 Theory and Practice of Writing
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
Textlinguistic research 49
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-
52 Theory and Practice of Writing
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 .5 G iven an d n ew in fo rm a tio n
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
Surface j Deep
1
Li
Sentential Syntax El Semantics
XI
11
T I
Cl
Textual Cohesion Ol Coherence
N1
1
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,
Toward a model of text construction 73
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)
74 Theory and Practice of Writing
■ 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
78 Theory and Practice of Writing
The final com ponent of the text construction model provides the
w orld/background knowledge for appropriate interpretation and
Toward a model of text construction 79
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 ......
82 Theory and Practice of Writing
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
Note
4.1 Introduction
84
Writing process approaches 85
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
94 Theory and Practice of Writing
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.
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.
112 Theory and Practice of Writing
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
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.
113
114 Theory and Practice of Writing
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)
Notes
6.1 Introduction
147
148 Theory and Practice of Writing
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.
Writingfor professional purposes 155
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
158 Theory and Practice of Writing
7.1 Introduction
176
Writing across cultures: contrastive rhetoric 177
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
Writing across cultures: contrastive rhetoric 179
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
180 Theory and Practice of Writing
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
Writing across cultures: contrastive rhetoric 183
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)
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
Writing across cultures: contrastive rhetoric 189
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.
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
192 Theory and Practice of Writing
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
Writing across cultures: contrastive rhetoric 193
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
Writing across cultures: contrastive rhetoric 195
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
8.1 Introduction
202
Towards a theory of writing 203
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?
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.
204 Theory and Practice of Writing
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
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
206 Theory and Practice of 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)
208 Theory and Practice of Writing
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,
212 Theory and Practice of Writing
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:
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.
(how)
A U TH O R ^ ► TEXT ► RECEPTOR
(who) (Text construction) (to whom)
(what)
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
Towards a theory of writing 219
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
220 Theory and Practice of Writing
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.
‘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
228 Theory and Practice of Writing
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.
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).
236 Theory and Practice of Writing
9.1 Introduction
237
238 Theory and Practice of Writing
■ 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.
■ 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:
From theory to practice 245
■ 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.
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
From theory to practice 251
clear that student considerations must enter into any planning for
a writing curriculum.
■ age
■ general educational and social background
■ cultural expectations
■ time since completion of teacher training
■ professional affiliations
■ years of teaching experience
■ academic subject knowledge/interests
254 Theory and Practice of Writing
9 .7 .1 .2 W hole-language a p p roach es
9 .7 .1 .3 G enre-centred a p p roach es
Note
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
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.
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.
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.
10.9 Conclusion
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
303
304 Theory and Practice of Writing
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.
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.
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.
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
NOTES
Teaching writing at
advanced levels
12.1 Introduction
341
342 Theory and Practice of Writing
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.
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.
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.
vides a useful check for the reader and re-establishes the major
issues and goals of the text.
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).
370 Theory and Practice of Writing
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
372 Theory and Practice of Writing
12.9 Conclusion
13.1 Introduction
377
378 Theory and Practice of 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)
884 Theory and Practice of Writing
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:
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.
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.
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
404 Theory and Practice of Writing
1 3.3.2.3 P er fo r m a n ce assessm en t
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.
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:
418 Theory and Practice of Writing
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
420 Theory and Practice of Writing
422
Conclusions: Writing in English 423
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
426 Theory and Practice of Writing
This list is provided tofacilitate easy access to a major part of this book.
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
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
York: Cam bridge University Press. 57-68.
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:
O xford University Press.
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
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
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
Composition and Communication 46, 30-45.
W h i t e , R. and V . A r n d t . 1991. Process uniting . L ondon an d New York:
Longm an.
Downloaded by [University of California, San Diego] at 23:05 19 June 2017
477
478 Subject 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,
Downloaded by [University of California, San Diego] at 23:05 19 June 2017