llNlYERSITY QF READING
( - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - " \\
CONCEPTUALIZATION, PLANNING AND DECISION -
MAKING IN THE PROCESS OF COMPOSING WRITIEN
ENGLISH DISCOURSE: IVOIRIAN UNIVERSITY
STUDENTS AT WORK
Anna MANOUAN
Thesis submitted for the degree of Ph. D
Department of Linguistic Science
October 1991

---

THESES 1
"
1979
Doctorat de 3eme cycle: Le Pn:tcrit anglais: son emploi par des
ctudianls ivoiriens de premiere annce d'universitc, Paris VII, France,
1991
Doctor of Philosophy (PhD): Conceptualization, planning and decision-
making in the process of composing written English discourse: 1vorian
University students at work, Reading University, United Kingdom,
'Plusieurs exemplaires de la these de troisieme cycle sont
disponibles au CAMES a Ouagadougou, Burkina Faso.
I
,~,

---

To mum and dad
.'.

---

ABSTRACT
One major assumption underlying this study was that research into
the composing process of university students will shed light on those
students' processes of solving cognitive problems.
Another assumption
Was
that the process of composing written discourse
integrates
those
cognitive abilities that are required for tackling academic tasks.
Four
main
research
questions
examined
quantitatively
the
relationship between
(l)
the advance planning patterns
(2)
time taken
on advance planning and writing performance;
(3)
the duration of
the
composing process and
(4)
the quality of
the written product and the
composing process.
A fifth question explored whether
the quality of
conceptualization
could serve as
an
index of
predictability of
the
overall quality of written products.
Forty video-protocols and forty scripts elicited from ten first
year English majors
(Universite Nationale d~ Cote d'Ivoire)
served as
data..
Half
of
those
protocols
and
scripts
covered
t"eferential
and
argumentative
discourse,
each
one
involving
both
familiarity
and
unfamiliarity with tasks.
Three main findings
have been yielded by correlational,
simple
and
multiple
general
linear
regression
analyses.
These
are:
(1)
Subjects
perform better
on
familiar
tasks.
(2)
The
most
successful (/
subjects adjust their advance planning to task demands.
For the most
part,
all
the
hypotheses
are
validated
when
an
argumentative,
unfami liar
task
is
considered.
(3)
The
subj ect
and
task
factors
contribute most
towards
explaining variance in
the composing process
(as measur~d by the duration of the processes of advance planning and
translating) .
Qualitative
findings
relqte
to
{1l
the
cognitive
processes
involved
in
composition,
(2)
problem-solVing
strategies
among
successful subjects, and
(3)
individual differences.
The
educational
implication
calls
for
an
indirect
'interventionist'
approach
inducing
autonomy,
self-commitment
and
critical thinking in the process of solving cognitive problems.

---

Table of Contents
List of illustrations
VIII
List of tables
x
AcknowLedgements
XII
List of abbreviations
XIV
1.
statement of the problem
2
1.1. Need for the study
2
1.1.1.
First justification:
General
situation
3
1.1.2
Second justification:
Theoretical framework
7
1.2. Definition of Writing and related concepts
11
1.2.1.
Definition of writing
11
1.2.2.
Definition of related concepts
18
1.3. Delimitation of the specific problem
27
1.3.1.
Background to the study .
27
1.3.2.
Statement of research questions
28
2
Review of the literature .
.
31
2.1. First language research on composing
33
2.1.1.
stage model process of composing
33
2.1.2.
The cognitive process model of
composing
.
.
.
.
.
41
2.1.3.
Revision process model
55
2.2. Second/third language research on
composing
.
.
.
67
2.2.1.
Empirical studies comparing
composing processes in L1 and L2
69
2.2.2.
Empirical studies with focus on
factors affecting composing process
76
2.3
Concluding remarks and research questions
81
IV
".

---

3.
Methodology . . . .
87
3.1. Introduction
88
3.1.1.
General perspectives and
academic context of inquiry
88
3.1.2.
Methodological lines of inquiry:
Theoretical framework
91
3.1.3.
Re-statement of major research
questions and
hypotheses
105
3.2. Methodology.
108
3.2.1.
SUbjects:
Overall presentation
108
3.2.2.
Data collection:
Instruments
111
3.2.3.
Experimental procedure
134
3.2.4.
statistical procedure
157
4. Results:
Quantitative and group-level
qualitative analysis
188
4.1. statistical results
188
4.1.1. Descriptive presentation
190
4.1.2. Analytical presentation
204
4.1.3. Summary of statistical findings
221
4.2. Qualitative results.
223
4.2.1. Identification of cognitive
processes in sUbjects' protocols
224
4.2.2. Attempting a typology of writers
229
4.2.3. Episodic structure.
238
4.2.4. Summary of qualitative findings at
group-level
244
5.
Analysis of individual sUbjects .
249
5.1. Descriptive presentation of identified
strategies
.
.
.
254
5.1.1.
Subjects' responses to the
demands of tasks 3 and 4
254
5.1.2.
Solving task 3 and 4: Tabular
presentation of identified strategies
260
5.2. Focus on individual subjects' responses to
task demands
260
5.2.1.
Responding to the demands of an
argumentative, unfamiliar task
266
v

---

5.2.2.
Responding to the demands of an
argumentative,
familiar task .
270
5.2.3.
Responding to task demands:
Comparative observations
271
5.3. Focus on individual subjects' problem
representation, planning of solution and
evaluation of solution
273
5.3.1.
Solving an argumentative,
unfamiliar task
274
5.3.2.
Solving an argumentative,
familiar task
291
5.4. Summary of qualitative findings at micro-
level .
299
5.4.1.
Processes of solving an
/
;
argumentative unfamiliar task:
Comparative results
300
5.4.2.
Processes of solving an
argumentative,
familiar task:
Comparative observations
304
6.
Conclusions . .
313
6.1. Interpretation of the results
314
6.1.1. statistical findings as related to
hypotheses .
315
6.1.2.
Macro-level qualitative findings
324
6.1.3.
Micro-level qualitative findings
330
6.2.
Discussion of the results
335
6.2.1.
Task effects and subject-
response to task demands .
336
6.2.2.
Episodic structure of the
composing process
353
6.3. Limitations of study,
need for further
research
and educational implications
356
6.3.1.
Limitations and need for further
studies
356
6.3.2.
Educational implications
359
Appendices .
366
Headings.
.
367
Appendix I
368
Appendix II
369
Appendix III
370
VI
'.

---

Appendix IV
373
Appendix V
374
Appendix VI
376
Appendix VII
378
Appendix VIII
379
Appendix IX
380
Appendix X
382
Appendix XI
387
Appendix XII
389
Appendix XIII
390
Appendix XIV
420
Bibliography .
.
•
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
. .
427
VII
'..;_"
"
".-
••.••,
"
<;t.
' "

---

Illustrations
2.1
The Flower and Hayes (1980,
1981) cognitive
model of composing
41
4.1
95% confidence interval limits representation of
overall performance (Oscore)
on four tasks:
argumentative, familiar
(Argf)j argumentative,
unfamiliar {Argu)jreferential,
familiar
(Reif);
and referential,
unfamiliar (Refu)
194
4.2
95% confidence interval limits representation of
time on reviewing process
(E)
on four tasks:
argumentative,
familiar
(Argf);
argumentative,
unfamiliar (Argu);
referential,
familiar
(Reif);
and referential,
unfamiliar (Refu)
194
4.3
95% confidence interval limits representation of
overall quality of introductory paragraphs
(Paraqual)
on four tasks:
argumentative,
familiar
(Argf)j
argumentative, unfamiliar (Argu);
referential,
familiar
{Reff};
and referential,
unfamiliar
(Refu)
195
4.4
95% confidence interval limits representation of time on
preparatory process
(or advance planning: P)
on four tasks:
argumentative,
familiar
(Argf);
argumentative, unfamiliar
(Argu); referential,
familiar
(Reff);
and referential,
unfamiliar (Refu)
196
4.5
95% confidence interval limits representation of time on
planning prior to-the first "active state" element in
translating (Pplus)
on four tasks: argumentative,
familiar (Argf);
argumentative,
unfamiliar (Argu);
referential,
familiar
(Reif);
and referential,
unfamiliar (Refu)
196
4.6
95% confidence interval limits representation of time on
translating process (T) on four tasks: argumentative,
familiar {Argf};
argumentative,
unfamiliar (Argu);
referential,
familiar
(Retf);
and referential,
unfamiliar (Refu)
197
4.7
95% confidence interval limits representation of time on
major pausing in translating process (Tpause)
on four
tasks: argumentative,
familiar (Argt);
argumentative,
unfamiliar (Argu);
referential,
familiar
(Rett);
and
referential, unfamiliar (Refu)
197
4.8
Graphic representation of the mast successful SUbjects'
advance planning on four tasks:
referential,
familiar
(Rett);
referential,
unfamiliar (Refu); argumentative,
familiar
(Argt);
and argumentative, unfamiliar (Argu)
202
VIII
..

---

4.9
Graphic representation of the least successful subjects'
advance planning pattern on four tasks: referential,
familiar
(Reff);
referential, unfamiliar (Refu);
argumentative,
familiar {Argf)j
and argumentative,
unfamiliar (Argu)
203
4.10
Subject-wise break-down of time on advance planning (P:
preparatory process)
209
4.11
SUbject-wise break-down of time on translating (T)
process
210
4.12
Mean Qscore (overall quality) on four tasks:
argumentative,
familiar (Argf)j argumentative, unfamiliar
(Argu)j
referential,
familiar
(Reff)j
and referential,
unfamiliar (Refu)
212
4.13
Association between overall quality (Qscore)
and pausing
time in the translating process of four tasks (a-d:
Reff:
referential,
familiar; Refu: referential,
unfamiliarj
Argf:
argumentative,
familiar; and Argu:
argumentative,
unfamiliar)
215
4.14
Subject-wise breakdown of group scores on four tasks:
referential,
familiar
(Reff)j
referential,
unfamiliar
(Refu)j
argumentative,
familiar
(Argf);
argumentative,
unfamiliar (Argu)
218
4.15
Association between Qscore (overall quality) and
familiarity (F:
familar; U:
unfamiliar) with tasks
219
4.16
Mean translating pausing time on four tasks:
referential,
familiar {Reff}; referential, unfamiliar (Refu);
argumentative, familiar (Argf); argumentative, unfamiliar
(Argu)
244
6.1
A descriptive diagram of the composing process
361
IX

---

Tables
2.1 Linear stage models of composing
35
3.1 Analysis of task demands
121
3.2 Summary of selection procedure
137
3.3
Ideal sample size
139
3.4 Sample resulting from criteria 1-5
141
3.5 Ordering of tasks
148
3.6 Subjects ranked in descending order on the
the four tasks
153
3.7 SUbjects ranked on the basis of overall scores
on the four tasks
154
3.8 Discriminating between successful and
unsuccessful sUbjects:
summary,
including
rank ordering,
sum of Qscore and 17.5 scOre
155
3.9 Pearson correlation coefficients for the
reliability of the tools for measuring the
written·products
158
4.1 Group means for Qscore,
paraqual,
Time
(seconds)
on P,
T,
E, Tpause,
Pplus
192
4.2 Pattern of behaviour of successful and
unsuccessful subjects' time
(seconds)
on
advance planning
200
4.3 Correlation between Qscore and P on the four
tasks
204
4.4 Analysis of variance between P and four
other variables
207
4.5 Analysis of variance between T and four
other variables
207
4.6 Analysis of variance between Qscore and
five other variables
211
4.7 Analysis of variance between Qscore and P
and T on an argumentative unfamiliar task
213
4.8
Analysis of variance between Tpause and Qscore
on an argumentative,
unfamiliar task
214
4.9 Analysis of variance between Qscore and nine
other variables
217
x
.<

---

4.10 Analysis of variance between Qscore, Paraqua1
and Pplus on an argumentative,
unfamiliar task
220
4.11 Time (seconds) on the translating process
225
4.12 Time (seconds) on the reviewing process
226
4.13 Matrix for observed cognitive behaviours
227
4.14 categorization of composing styles based on
Wason (1978)
230
4.15 Matrix for characteristic· sequence of
processes based on an argumentative,
least
familiar task
235
4.16 Number of major composing episodes as
identified in the protocols
241
5.1 Subject 4'5 response to the demands of task 4
255
5.2 Subject 7'5 response to the demands of task 4
256
5.3 Subject 2'5 response to the demands of task 4
257
5.4 Subject 4's response to the demands of task 3
258
5.5 Subject 5's response to the demands of task 3
259
5.6 Solving an argumentative,
unfamiliar task:
cognitive strategies activated by subject 4
261
5.7 Solving an argumentative,
unfamiliar task:
cognitive strategies activated by subject 7
262
5.8 Solving an argumentative,
unfamiliar task:
cognitive strategies activated by subject 2
263
5.9 Solving an argumentative,
familiar task:
cognitive strategies activated by subject 4
264
5.10 Solving an argumentative,
familiar task:
cognitive strategies activated by subject 5
265
XI

---

Acknowledgements
This research work was financially supported by an ODA
funding,
administered by the British Council.
Complementary funding
was available from the African Educational Trust,
the Leche Trust and
the Gilchrist Educational Trust.
To these institutions,
I extend my
gratitude,
as well as to the National University of the Cote d'Ivoire
for granting me a three-year's leave of absence.
A doctoral research in a particular area cannot be carried cui:
unless there is at least one person who expresses interest in it.
In
this respect,
I am grateful to:
-
Dr P.
D.
Kokora for strongly encouraging me to pursue this
line of inquiry;
-
Professor D. Wilkins and Mr R.V. White,
for their joint
supervision at the earliest stage of my work,
as well as to Or M.
Garman, whose special interest and appreciation acted as a catalyst
at a
late but crucial phase of this study;
-
the Centre Universitaire de Formation Permanente and the
Societe des Transports Abidjanais,
for their logistic support.
StimUlating discussions through formal and informal meetings
with researchers from different backgrounds do influence the
orientation of a research activity.
I am indebted to the following~
-
Professor P~ Smith,
Or B Butterworth,
Or O~ Green, Dr C. Weir
and Mr A.
Hughes,
for consulting with them;
-
Messrs O. Porter and D.S. McGovern,
Or A. Williarns, Mesdames
c.
Furneaux and K.
Spellman-Miller, M.L.
Kone,
and Ms M.
L.
Haag,
for their contribution in the form of marking scripts;
Dr. C.R.
Palmer,
Messrs J.B.L. Levin and S. Gilmour,
for
patiently coaching me in statistics;
XII

---

Acknowledgements
This research work was financially supported by an ODA
funding,
administered by the British Council.
Complementary funding
wag available from the African Educational Trust,
the Leche Trust and
the Gilchrist Educational Trust.
To these institutions,
I extend my
gratitude,
as well as to the National University of the Cote d'Ivoire
for granting me a three-year's leave of absence.
A doctoral research in a particular area Cannot be carried out
unless there is at least one person who expresses interest in it.
In
this respect,
I
am grateful to:
-
Dr P.
D. Kokora for strongly encouraging me to pursue this
line of inquiry;
-
Professor D. Wilkins and Mr R.V. White,
for their joint
supervision at the earliest stage of my work,
as well as to Or M.
Garman, whose special interest and appreciation acted as a catalyst
at a
late but crucial phase of this study;
-
the Centre Universitaire de Formation Permanente and the
Societe des Transports Abidjanais,
for their logistic support.
Stimulating discussions through formal and informal meetings
with researchers from different backgrounds do influence the
orientation of a research activity.
I
am indebted to the following:
-
Professor P. Smith, Or B Butterworth,
Dr o. Green,
Dr C. Weir
and Hr A. Hughes,
for consulting with them;
-
Messrs D.
porter and O.S. McGovern,
Or A. Williams,
Mesdames
C.
Furneaux and K.
Spellman-Miller, M.L.
Kone,
and Ms M. L.
Haag,
for their contribution in the form of marking scripts;
Dr.
C.R.
Palmer,
Messrs J.B.L. Levin and S.
Gilmour,
for
patiently coaching me in statistics;
XII

---

Hr A. J. Smith,
for his invaluable assistance with the computer
treatment of my data;
- Mr. R. Peattie, for incorporating a timing system into my
data so as to facilitate their manual treatment;
- Ms K.F. Willetts,
for facilitating access to valuable
research materials;
- Hr I. Cortinovis,
for patiently overseeing the typographic
presentation of this work; and
Mr. Campbell, for seeing to the smooth running of
wordprocessors.
My deep appreciation goes to Professor J. Tano and the Ivorian
Psychology majors, who helped in administering the Raven Progressive
Matrices,
and to those of my research mates and MA students who
generously gave up their own research activity to code samples of my
data.
I wish also to thank my siblings, those special relatives and
friends,
for their affectionate concern and care and/or prayer.
Last but not least,
I am indebted to my students whose thirst
for learning constituted the driving force.
XIII
;.

---

Abbreviations
Argf
argumentative,
familiar
Argu
argumentative,
unfamiliar
Ascore 1 analytic scorer 1 on the 7 TEEP attribute writing scales
Ascore 2 analytic scorer 2 on the 7 TEEP attribute writing scales
Bscore 1 analytic scorer 1 on 5 of the TEEP attribute writing
scales (excluding mechaniial accuracy I
and II)
Bscore 2 analytic scorer 2 on 5 of the TEEP attribute writing
scales
(excluding mechanical accuracy I
and II)
CALS
Centre for Applied Language Studies,
Reading
CIERPA
Centre Ivoirlen d'Etudes et de Recherches en psychologie de
]'Universite d'Abidjan
Cngn
consolidating gains
Cop
carrying out the plan
CUFOP
Centre Universitaire de Formation Permanente
E
time on reviewing (editing)
process
EFL
English as a Foreign Language
ESL
English as a Second Language
Evsl
evaluating the solution
F
most familiar
FLASH
Faculte des Lettres, Arts et Sciences Humaines
Fpb
finding
the problem
Hscore 1 holistic scorer 1
Hscore 2 holistic scorer 2
Ll
English as a first language
L2
English as a second language
L3
English as a third language
LTM
long term memory
P
time on preparatory process (or advance planning)
XIV
.<

---

Paraqual
overall quality of i~troductory paragraphs
PM 38
Raven Stanaard Progressive Matrices
PpIus
time on planning prior to the first
active state element in
translating process
PsI
planning the solution
Q
question
Qscore
overall performance on written products
Reff
referential,
familiar
Refu
referential,
unfamiliar
Rpb
representing the problem
SOTRA
Societe des transports abidjanais
STM
short term memory
T
time on
translating process
TEEP
Test in English for Educational Purposes
Tpause
time on major pausing in
translating process
Tl (or P)
time on advance planning
T2
(or T) time on translating
T3 (or El
time on reviewing
U
least
familiar
UNCI
Universite Nationale de Cote d'Ivoire
xv

---

CHAPTER ONE
1
-,

---

1.
Statement of the problem
This chapter undertakes three tasks:
(1)
Providing justification for the present
study by relating i t to a general educational
problem;
(2)
Articulating the general theoretical
framework underlying this investigation;
(3)
Defining the specific problem being
addressed,
analyzing key concepts such as
problem solving,
conceptualization,
preparation
and planning.
Each one of these points defines the main axes of
the major sections below.
1.1. Need for the study
The need for the present study lS best understood if
put in the perspective of a broader educational problem
situation and analyzed within the framework of cognitive
psychology.
To this end,
the examination of the specific
problem being addressed here will be preceded by the
2

---

description of the general situation within which the
educational problem arises,
followed by a brief review of
selected major studies linking writing and learning.
1.1.1.
First justification:
General situation
The educational problem being examined originated
from this investigator's dissatisfaction with informal
accounts by Ivorian university academics of the high
failure rate of first year students at the National
University of the Cote d'Ivoire.
As a rule sixty percent
of first year students, at the Faculty of Letters, Arts
and Social Sciences (FLASH),
fail their end of year
examinations,
as evidenced by the 1985-1986 statistics
from that Faculty.
Among many competing explanations for
this phenomenon,
a first cause may be attributed to the
so-called 'open-door' policy characterizing the
Francophone educational system.
From an external
observer's standpoint,
such a system, unlike the British
one, practises selection after admission l •
An
alternative account of the above problem is given by
Ivorian lecturers at FLASH.
Those lecturers ascribe the
high failure rate to laziness,
lack of motivation and
I
In the Ivorian context selection takes place at entrance.
Individual Departments define their admission requirements.
Section
3.2.1 spells out those of the Department of English at FL~SH.
The
rate of success at University is not determined by external screening
factors;
rather,
students'
performance on examinations determine9 the
success/failure ratio.
3

---

sustained work and/or to inadequate study skills.
At
times,
insufficient cognitive abilities are put forth.
Seeking to go beyond this impressionistic analysis
of the general problem,
this investigator carried out a
survey in 1985-1987,
involving both lecturers and
students at FLASH.'
Among other research questions,
one
research question relevant to this study was to discover
whether lecturers at FLASH shared common educational
objectives.
One main result of the 1985-1987 exploratory survey
was that two objectives attract general consensus among
lecturers at FLASH.
These main objectives were defined
as follows:
(1)
the ability to learn how to learn,
and
(2)
the ability to achieve autonomy.
The first objective
emphasized the ability to think,
to reason,
to identify
problems and to solve them.
The second objective called
for students' ability to exhibit self-dependence by
taking initiatives, making decisions and choosing the
best solution among conflicting or complementary
alternatives.
While the first objective was considered
to be partially met at the end of the first academic
year,
the second objective was far from being partially
met.
Indeed,
there was,
as suggested by lecturers,
a
great discrepancy between expected behaviour and observed
behaviour.
First year university students were said to
2
Detailed discussion of this survey can be found in manuscript,
Manouan (1989).
4

---

lack autonomy In that they exhibited no incentive for
exploring and discovering new ideas through wide reading
and independent study.
To improve the overall teaching and learning
situation at the Faculty of Letters, Arts and Social
Sciences,
it has been suggested that existing pedagogical
structures be reconsidered.
This entails an increase in
members of staff and thereby a decrease in the ratio of
tutees per staff member 3 •
Even though the need for
ameliorating present working and learning conditions
cannot be overemphasized,
this researcher strongly feels
that structural changes must necessarily be coupled with
due emphasis on writing since it plays a central role in
the evaluation of subject discipline areas at University
level.
Furthermore, writing,
as will be subsequently
shown,
incorporates all the cognitive skills required to
achieve the learning objectives as defined within the
context of FLASH,
National University of the Cote
d'Ivoire.
In sum,
these objectives constitute the first
reason for undertaking the present investigation.
Before examining another reason for the need for
this study,
i t is essential,
at this point,
to clarify
one or two issues likely to be misleading.
A call for
special emphasis on writing In such an academic context
might imply the following:
(1)
That little attention is
3
Such issues were discussed at a seminar organised by FLASH in
January-February 1983.
5

---

glven to writing In an academic context,
hence the need
to do justice to this discipline.
(2 )
That university
students experience difficulty with writing at an
academic level, hence the need for a therapeutic
treatment.
Admitting these to be the case,
what forms would the
solutions to those problems take?
Would potential
solutions take the form of:
Re-allocating 'adequate' writing time slots?
Investigating why university students experience
difficulty with composlng written discourse?
Developing a writing programme likely to enhance
students' composing abilities and to ensure
development of cognitive skills required to
undertake academic studies?
The ultimate goal of this study,
it is to be stated
here,
is not concerned with launching a programme of
curriculum innovation at the Faculty of Letters, Arts and
Social Sciences,
National University of the Cote
d'Ivoire.
Neither is it concerned with advocating the
primacy of a given methodological approach over another
one' .
The goal is,
given the objectives of FLASH,
to
analyze the problem solving processes of university
students to find out whether or not they exhibit the
Manouan
(1979)
raised the issues of first year Ivorian
university students'
readiness for undertaking English studies at
University level and the adequacy of the curriculum developed by the
Department of English at FLASH to cater for those students'
needs.
6

---

cognitive skills needed at university level.
Writing,
at
academic level,
is the example par excellence.
To quote
Greeno,
'the most demanding intellectual problems are
problems of composition'
(Greeno,
1978:264).
This statement leads naturally to examining the
second reason for pursuing research In the area of
composing written discourse.
This second reason is
offered by studies in written composition and cognitive
psychology.
1.1.2
Second justification:
Theoretical
framework
Considering the specific problem from its underlying
theoretical framework,
the need for this study seems
self-evident.
Composition is a
subject worthy of research and
understanding .... [It is]
concerned with one of
the most essential and valuable resources of
anyone's education:
the ability to write
(Irmscher,
1979:240).
Furthermore, writing represents
'a unlque mode of
learning'
(Emig,
1977 :34):
In reiterating Irmscher's
declaration of integrity of written composition as a
discipline and Janet Emig's strong claim,
the present
investigator wishes not only to re-emphasize the value of
writing as an autonomous field of investigation,
but also
to contend that successful learning at University level
7
.'.

---

presupposes awareness of
'the role for writing as
heuristic'
(Emig,
1977:34); that is,
a process of
discovery and reformulation during which concepts are
being discovered
(Applebee,
1982:367)c.
Theoretical evidence for the close relationship
between writing and learning is abundant.
A summary of
primary sources of evidence is given by Emig
(1977:34):
Lev Vygotsky,
A. R. Luria and Jerome Bruner
have all pointed out that higher cognitive
functions,
such as analysis and synthesis,
seem
to develop most fully only with the support
system of verbal language--particularly,
it
seems,
of written language.
Another interesting piece of evidence to support a
rapproachment between learning and writing lS Bruner's
(1960,
1962) analysis of three processes that occur
simultaneously in the act of learning:
(1)
acquisition;
(2)
transformation; and
(3)
evaluation.
The concurrent
activation of these processes enables the learner to
gather new information, manipulate knowledge so as to
adjust to the demands of new tasks,
and to check whether
information has been adequately manipulated as required
by a given context.
As will be shown in the discussion
of composition models,
the same processes operate in the
5
Applebee
(1982), drawing on Polanyi
(1958),
makes a
distinction between routine writing tasks and heuristic writing
tasks.
As suggested by his survey of American schools,
school
setting promotes routine writing rather than heuristic writing.
8

---

act of writing.
Given this observation,
the emphasis on
writing as a learning act is more than justified.
Apart from psychologists,
composition specialists,
informed by educational psychology,
have claimed that
writing promotes cognitive growth.
The unique
relationship between learning and writing is skilfully
examined in Emig (1977) where the author stresses that
just as
'successful learning is connective and selective'
so is writing which 1S 'integrative in perhaps the most
basic possible sense:
the organic,
the functional'
(Emig,
1977:37).
Other composition studies depicting the
relationship between writing and learning and/or
describing writing as a thinking or discovery process are
Applebee
(1982),
Zame1
(1982),
Odell
(1980), M.J.
Smith
(1979),
and Arapoff
(1967).
These varied studies constitute unquestionably ample
theoretical evidence to support the claim that writing 1S
instrumental in mental development and is conducive to
intellectual discipline.
This is achieved when a
subject-writer consciously attempts to attend to 'the
greater demand for explicitness
[in written dicourse as
opposed to spoken discourse and for making]
all
assumptions and premises explicit'
(Olson,
1977:278)
The assertion that writing promotes intellectual growth
'seems well-founded.
As a matter of fact,
writing as a
learning activity helps the learner to organize his
9
.<

---

thought and shape his experience of the extra-linguistic
world (Bruner, 1967; vygotsky,
1962).
Emphasis on stating the value of undertaking
research into written composition articulates the second
reason for carrying out the present study·.
The first
reason,
i t is to be recalled,
is an educational one,
which attempts to explore potential efficient means for
achieving the learning objectives as defined by members
of staff at the Faculty of Letters, Arts,
and Social
Sciences, National university of the Cote d'Ivoire.
Having delineated the general educational problem,
we shall first define writing and related key concepts,
then we shall state the specific problem constituting the
object of this investigation.
6
Our suggestion that writing as a cognitive problem solving
activity may shed light on subjects'
cognitive processes of problem
solving might be sUbject to strong criticisms centred on both
cultural as well as cognitive differences,
issues that are r&ised by
the Sapir-Whorf hypothesis; Kaplan's (~966, 1967, 1983, 1988)
discussion of rhetorical differences;
and by Carlson's
(1988)
analysis of cultural differences
in writing and reasoning skills.
Admittedly,
given subjects
from diverse linguistic background,
and
from comparable educational experience,
with varied degrees of
proficiency in English,
rhetorical/pragmatic differences are likely
to be observed in those subjects'
written products.
Simple
observations suffice to suggest that such differences might be
determined by socio-cultural and/or academic norms and expectations.
However,
i t would seem unwise to argue that the cognitive operations
that are activated in the act of problem solving are culturally
determined.
Furthermore,
it is not our contention to use writing to
distinguish between literate/illiterate cultures,
as posited in Olson
(1977).
On no account do we share Goody's (1968, 1977, 1980) extreme
position,
as reported in Faigley,
Cherry,
Joliffe & Skinner
(1985:75)
that
'the development of writing results in the development of new
cognitive structure'.
We are also aware of the study of the Vai in
Liberia by Scribner & Cole (1978).
(See also Cole and Scribner,
1974).
The point we are making here is that,
given a University
context,
a study of subjects' processes of solving writing tasks-
might help one discover whether they behave as relatively autonomous
problem solvers,
capable of making decisions and choosing the best
solution among competing alternatives.
10

---

1.2. Definition of Writing and related concepts
The discussion is centred on
(1) writing as a goal-
directed cognitive problem solving activity;
(2)
conceptualization,
concepts of preparation and planning.
1.2.1.
Definition of writing
Four recurrent descriptors have been used in
research on written composition to depict writing.
This
activity lS referred to as a thinking activity
(Arapoff,
1967; Emig,
1971,
1977; Stoehr,
1967)
or as a discovery
process
(Elbow,
1973; Emig,
1977; Murray,
1978; Odell,
1980; Smith,
1979; zamel 1982)
or as a creative process
(Zamel,
1976).
writing is also portrayed as a goal-
directed cognitive problem solving activity
(Bereiter &
Scardamalia,
1981,
1987; Flower & Hayes,
1980;).
Each of
these terms are clarified below.
Writing as a cognitive problem solving activity
To begin with,
it may be put forth that of these
defining concepts,
the last one seems a productive
descriptor in that it subsumes the others.
This last
definition could have provided a more economic conceptual
representation of writing by simply defining writing as a
problem solving activity as the ensuing discussion of the
11
.

---

nature of problem solving will indicate.
In Anderson's
words,
Problem solving is defined as any goal-directed
sequence of cognitive operations.
Problem
solving that requires the development of new
procedures is called creative problem solving;
problem solving that uses existing procedures
is called routine problem solving
(Anderson,
1980: 257) .
Such a definition of problem solving encapsulates
the concepts of thinking,
discovery and creativity.
In
addition,
i t includes the following essential and
necessary characteristics of problem solving in that i t
is
(1) goal-directed,
(2)
a sequence of operations,
and
involves
(3)
cognitive operations.
In other words,
to
qualify as a problem solving,
an activity must contain an
explicit goal,
the execution of which should include a
relatively orderly set of steps that activate thinking.
As one can observe,
by virtue of its nature,
writing
can be depicted as a creative problem solving since it
exhibits the three criteria stated above.
The resolution
of a writing problem calls for new procedures.
However,
writing as a creative problem solving
activity is distinct from other creative problem solving
such as mathematics.
As shown in Mayer,
Carkin & Kadane
(1984),
successful resolution of mathematics problems,
requ1res four kinds of knowledge:
(1)
linguistic and
factual,
(2)
schematic,
(3)
strategic and
(4)
algorithmic.
12

---

Put differently,
successful mathematics problem
solvers draw on the ability to comprehend sentences
expressing relations among variables,
the knowledge of
problem types,
the knowledge of appropriate strategies
and on the skill at applying arithmetic and algebraic
algorithms
(Mayer et al.,
1984).
While it is obvious that a successful writing
problem solver needs to be competent in the first three
kinds of knowledge--linguistic and factual,
schematic,
and strategic--this is not the case when considering
algorithmic knowledge,
even though a misreading of
Hillock's discussion of the four kinds of knowledge
needed in writing might lead to the contraD,7.
The
absence of algorithms typifies the difference between
mathematics problems solving and writing problems
solving.
The difference between writing and mathematics
as problem solving activities becomes clearer when the
distinction between algorithms and heuristics comes into
play.
It is worth emphasizing that while mathematics
problem solvers have access to algorithms,
writing
7
The four types of knowledge defined by Mayer et al.
(1984) are
not
to be mistaken for Hillocks's
(1986b)
four kinds of knowledge,
which are described as content knowledge,
procedural knowledge
. leading to content manipulation,
knowledge of discourse structures,
and procedural knowledge.
Clearly,
this last set of four- typesot
knowledge does not include algorithmic knowledge.
However,
a
parallel could be drawn between Hillocks's
(1986b)
four-way
distinction of knowledge and what Applebee
(1982)
identifies as three
domains of knowledge.
Widdowson
(1983)
makes two broad
categorizations in distinguishing between linguistic knowledge and
world knowledge.
13

---

problem solvers can resort to heuristics.
The former
have available 'methods guaranteed to provide a
solution'; the latter have access to 'methods
[that
tend to lead to successful solution but
[they] are not
guaranteed to do so'
(Anderson,
1980:268).
This
definition implicitly points to the 'trial-and-error'
nature of heuristics.
Could i t be that the 'trial-and-error' nature of the
term,
coupled with its etymology,
may have led a number
of written composition researchers to claim that writing
is a discovery procedure?
If this is the case, what is
to be understood by discovery?
Implicit in their use of
discovery is the connotation of haphazard search,
'groping into the darkness'
(Murray,
1978).
In the maln,
discovery is associated with 'insight'
(Wason,
1980),
unexpected,
unplanned,
unanticipated outcome or finding.
Is one to equate discovery in the composing process with
lack of planning?
Bereiter & Scardamalia
(1987)
strongly
assert that composing cannot be synonymous
with
planlessness.
One may add that planlessness cannot be
reconciled with the previous definition of writing as a
'goal-directed' (Flower & Hayes, 1980b, 1981a) activity.
What is then meant by discovery?
Where and when does i t
take place in the process of written composition?
14

---

Writing as a process of discovery and creativity
A first answer to this query is in the
representation of the process of written composition as a
three phase process.
In considering this model here,
attention is on how such a representation deals with the
discovery dimension of written composition.
I-Jason
(1980)
provides an explicit discussion of these phases.
In his
own words,
[The]
first phase is creative .... The aim is
simply to exteriorize thought without regard to
its expression and in accordance with the
cliche that you don't know what you are trying
to say until you have said it.
The intention
of the second phase is critical.
The text is
reordered,
sentences are changed,
and a lot is
deleted .... The second phase aimed merely to
structure thought.
The third phase is a
complete redraft [which has]
a compelling
force . . . .
In the third phase the cool,
detached
process of further clarification frequently
generates new thought
(\\'iason,
1980:132-133).
In sum,
the first phase is creative while the second
is critical and the third lS redrafting.
Of these
adjectives depicting each of the three phases,
creative
might be the most appropriate synonym for discovery as
understood by composition researchers.
As evidenced by
Wason
(1980),
analytical thinking and evaluative
examination are suspended in the first phase.
During
. this discovery phase, writers freely give way to their
thoughts allowing them to be exteriorized haphazardly
15
.'.

---

without any concern for order or structure,
this being
taken care of during the other two phases.
Such a description seems to lend support to the
cliche 'you don't know what you are trying to say until
you have said i t ' .
Clearly, Wason's
(1980)
description
of the first phase seems to suggest that priority is
given to content rather than to form.
Elbow's
(1981)
suggestion that the two conflicting skills,
creating and
criticizin~ be separated provides additional illustration
for this view
(Elbow,
1981:7).
There is no indication,
however,
that this stage is characterized by a total
absence of conceptualization and/or planning,
although a
close reading of Murray
(1978)
and Elbov:
(1973,
1981) and
others might point to the contrary.
A question that may be posed here is:
Does delayed
systematic structuring preclude concern for
conceptualization or concern for planning at the earliest
phase of the composing process?
Rohman & Wlecke
(1964)
have an explicit answer in this respect.
In their study, both stress that
'thinking-as-
discovery [which is different from thinking-as-
remembering]
is the crucial and initial stage of the
total thinking-writing process'
(Rohman and Wlecke,
1964:16).
Both refer to the first phase in the composing
process as the 'pre-writing'.
'the stage of discovery in
the writing process when a person transforms a
'subject'
into his own categories'
(see also Rohman,
1965).
This
16

---

is the stage at which conceptualization takes place in
the act of writing.
As a consequence,
Rohman and Wlecke
observe,
'hasty conceptualization by [student-writers]
obscured rather than illuminated the nature of the events
for which the concepts were supposed to provide insight'
(Rohman and Wlecke,
1964: 6) .
The position expressed by Rohman and Wlecke is
diametrically opposed to the prevailing consensus,
subsumed in the concept of discovery as understood by
Murray
(1978),
Elbow (1973,
1981)
and others.
Conceptualization and planning occur at the very
beginning of the composlng process.
Speedy attention to
these results in confusion in writing
(Rohman & Wlecke,
1964).
Chosson
(1975 ) echoes the same point.
Although
Chosson's study does not address the process of composing
written discourse,
his position is worth mentioning as it
stresses the importance of conceptualization and planning
prior to any course of action.
Arguing in favour of
activating these processes early in advance,
he asserts
that when one does not know what one is seeking,
one does
not know what one finds.
Indeed,
having a general and
vague idea about a course of action to be taken is not
conducive to operationalizing this idea for observation
(Chosson,
1975:92).
In similar vein,
Young,
Becker and Pike
(1970)
remark that unless the writer is aware of the problem (a
conflict between the writer's view and the worldview
17
'.

---

(Young et al.,
1970) or a gap to be crossed
(Hayes,
1987»
to be solved and unless he understands its nature
and knows its origin at the earliest stage in the
composing process,
there are limited chances for such a
problem solver to arrive at a satisfactory result.
Thus
Young and his colleagues emphasize the importance of
preparation time in the composing process
(Young et al.,
1970:89-118) .
To sum up,
the definition of writing as a cognitive
problem solving activity is well established in the field
of written composition.
Researchers in this area adhere
to the idea that such an activity involves mental
operations, creativity and discovery.
What has yet to
reach general consensus among those involved ln the field
of written composition is the role and place of planning
and conceptualization in the process of generating
written discourse.
Although conceptualization has been
used ln the preceding discussion,
the meaning of such a
term needs to be clarified.
1.2.2.
Definition of related concepts
Given the potential importance of conceptualization
and of preparation or planning during the composing
process,
special attention is now given to examining each
one of these concepts.
18

---

conceptualization:
Its nature
The term conceptualization occurs in composition
studies by Rohman and Wlecke
(1964)
and Stallard (1974).
Conceptual planning is coined by Bereiter & Scardamalia
(1987). What is to be understood by this concept?
Donaldson (1976) offers one way of understanding the
concept.
Conceptualization, she states, appears as a
multi-faceted concept which depends on the theoretical
position that is adopted.
Stressing the potentiality for
disagreement about how 'conceptualization' is to be
understood, Donaldson (1976)
examines two different
interpretations of the concept before offering her own.
She contends that an extreme behaviourist position would
consider that a subject is able to conceptualise if that
subject can demonstrate that he has
abstracted some common feature from [a set of
non-identical stimuli) and [that) he is
responding to this feature,
whether or not he
has any awareness of what the feature may be,
and indeed whether or not he has any awareness
of anything whatever
(Donaldson,
1976: 277).
Another tendency,
Donaldson remarks,
is interpreting
conceptualization within the framework of studies on
'concept learning' and 'concept attainment'
(see Bruner,
Goodnow and Austin, 1956; Bruner and Anglin,
1974).
Taking an independent line from the first two
interpretations,
Donaldson expresses the following view:
the process of "conceptualizing" is the process
of constructing complex representations for the
purpose of directing behaviour.
And I shall
consider i t appropriate to be concerned with
19
..

---

the degree to which consciousness accompanies
this process--consciousness of the world thus
represented and ultimately of the represent-
ations themselves (Donaldson, 1976: 277-278).
What is to be retained from Donaldson's view of
conceptualization is that this concept can be defined as
a process in which the subject consciously undertakes to
construct 'complex representations'
likely to control his
behaviour.
what is to be stressed is that Donaldson's
view reflects Johnson's analysis of the same issue as
seen below:
[The] equation of conceptualization and
abstract thinking ... is [the] ability to
abstract any feature of a problem situation
from its surrounding and manipulate it in any
way.
[This] requires ability to maintain a set
for that feature and shift the set if necessary
(Johnson, 1955:269, 272).
For a given subject to abstract, discriminate,
and select
from a complex problem solving situation only those
elements that. are essential for the solution of that
problem, adequate preparation is needed.
Preparation:
Its nature
Preparation,
as understood in its dynamic sense,
is
a
'process of getting ready, of adopting a preparatory
set,
based on present conditions and past learning'
(Johnson, 1955:146). The question is this: What form does
preparation for complex problem solving take?
Another
question of importance previously posed by Johnson is the
20
:t: .~.' ~..•

---

following: How can the preparation process be identified
in the solution of complex problems?
Comparing preparation for solving simple and complex
problems, Johnson has arrived at three main points:
[lJ ... ln mUlti-cycle problem solving the
preparation '"
cannot always be temporally
localized in one initial period.
[2J ... Time
span [for solving complex problemsJ
is longer.
Thus it becomes necessary to consider temporal
aspects of the set,
its function in steering
intellectual activity over a period of an hour
or so, resisting or succumbing to distraction
from within and without.
[3J ... Preparation
[for solving complex problemsJ
is less
explicit.
The thinker contributes to the
preparatory set (Johnson,
1955:157,158).
What is clear is the following:
Preparation may no'c
be restricted to the initial period of the problem
solving process.
Preparation occurs throughout the
process.
Given the cyclical nature of problem solving,
preparation and production alternate.
Finally, personal
involvement is necessary.
To arrive at a good understanding of the preparatory
process,
it is worth considering two complementary
aspects of preparation for complex problem solving.
These are surveying the problem and setting up the
requirements for the solution of the problem (Johnson,
1955:158).
In surveying the problem, Young et al.
(1970)
emphasize the need to identify and state the problem.
They go on to suggest that successful identification of
the problem requires understanding the nature and origin
of the problem.
A problem solver will have no incentive
to eliminate a problem unless he develops awareness of an
21
. '.

---

inconsistency in his beliefs and the external world.
Adequate statement of the problem is crucial during the
preparatory process. This implies explicit description of
the problematic situation and identification of the
unknown, which is in fact a partial description of the
solution.
The observations by Young and his colleagues that a
clearly stated unknown facilitates inquiry echoes
Chosson's
(1975)
comment cited earlier.
Young and his
associates go as far as to assert that the 'unknown is
not totally unknown;
i t is partially known,
and the [the
research question] will include this partial knowledge'
(Young et al., 1970:92). One major point that results
from the discussion on conceptualization and preparation
is that these components are inherent to the complex
problem solving process, and thereby to the process of
written composition. Although Johnson (1955)
remarks that
i t is difficult to unravel the preparatory process,
i t is
contended here that light can be shed on the preparatory
process if this is approached within the perspective of
decision-making or planning as discussed in cognitive
psychology. Making such a contention leads to examining
the concept of planning in the process of composing
written discourse.
22
; ;, ", ..
.'" ~ , .-

---

Planning in the process of composing written
discourse
Planning refers to the 'predetermination of a course
of action aimed at achieving a goal'
(Hayes & Roth,
1980:275-276).
Flower & Hayes (1981b)
distinguish
between two kinds of planning: rhetorical planning and
sentence-to-sentence level of planning.
Both types of
planning result from two different theoretical
perspectives:
rhetorical and linguistic.
The linguistic
hypothesis is that
writers pause in order to generate or plan what
they are going to say next.
[The rhetorical
hypothesis posits that] when people pause for
significant length of time, they pause in order
to carry out more global rhetorical planning or
problem-solving which is not necessarily
connected to any immediate utterance or piece
of text (Flower & Hayes,
1981b:230).
While these hypotheses,
as Flower and Hayes
acknowledge,
are not mutually exclusive, the former is a
'text-based perspective of the writing process',
the
latter is process driven.
As such,
they represent two quite distinct lines of
inquiry in current research--the bottom-up work
of linguistics vs the top-down approach of
cognitive psychology and rhetoric.
They
represent two significantly different models
and sets of assumptions about what is crucial
in the writing process (Flower & Hayes,
1981b:203).
This dichotomous view of planning nicely fits in the
Flower & Hayes (l980b,
1981b) cognitive model of the
writing process. This model gives a hierarchical
23

---

representation of the writing process.
In this
perspective, rhetorical planning,
including awareness of
the audience, genre, purpose, goal-setting,
is concerned
with the whole-text.
Sentence-to-sentence level
planning, on the other hand,
addresses local issues such
as immediate topic content.
Bereiter & Scardamalia (1987)
offer a similar
dichotomy.
They perceive two modes of planning in the
composing process:
conceptual planning and content-
generation planning.
conceptual planning comprises
setting goals, problems
(or task difficulties),
organization of text,
strategies (or procedures)
and
audience.
Content generation is the 'mental generation
of text' or 'the generation of material intended for
actual use in the text'
(Bereiter & Scardamalia, 1987:
203). Such preformation of language 'planning'
is closely
tied with immediate text production.
Both modes,
conceptual planning and content-
generation planning, are used by Bereiter & Scardamalia
(1987)
and others to discriminate among expert or skilled
writers and novice or unskilled writers.
Children are
said to be predominantly sentence-to-sentence level
planners.
Having internalized the schema of speech
production, children, as reported by Bereiter &
Scardamalia (1981,
1987), tend to transfer the 'forward-
looking thought' pattern of conversation to the written
composition situation.
Concern for establishing a
24

---

developmental index, added to the top-down hierarchical
view of composing, might have blurred the issue.
The question that comes to mind is the following:
Do these pairs (which are conceptual planning/rhetorical
planning,
on the one hand,
and content-generation
planning/sentence-to-sentence planning on the other hand)
apparently similar, describe the same reality?
Clearly,
conceptual planning and rhetorical planning denote
identical planning subprocesses such as goal setting and
audience identification.
However, the second elements of
both dichotomy might encompass different realities.
It
might be the case that Flower & Hayes (1981b),
and
Bereiter & Scardamalia (1987)
use content-generation
planning and sentence-to-sentence-Ievel planning to
describe a bottom-up type of planning.
If this is the
case, there is still another side of the question that is
left unanswered.
In other words, can one justifiably reduce content-
generation planning to sentence-to-sentence level
planning?
Admitting this to be the case, would a
sentence-to-sentence level planning be unequivocally
restricted to content-generation?
In any case, what is
to be understood by sentence-to-sentence level planning?
Some specification is needed here.
It seems that for a clear understanding of the
planning process in written composition to be achieved,
at least a tripartite view of planning is required.
Such
2S

---

a perspective will have to include conceptual planning,
semantic planning, and syntactic planning.
Combined
evidence from research on speech production and writing
production seems to provide strong ground for the claim
that planning in the composing process of written
discourse integrates the three kinds of decision-making
processes that underline writing production.
According
to Daiute (1981),
sentences are planned via set syntactic frames.
Clauses are planned and lexical items fit into
the frames . . . . Writers of grammatical multi-
clause sentences use automatic syntactic frames
for larger-than-clause structures (Daiute,
1981: 9).
Butterworth (1975),
rejecting Boomer (1965),
concludes that the major unit of speech production is
semantic planning rather than clause unit.
Another study
supporting semantic planning is Taylor (1969), who after
examining whether latencies would be influenced more by
content difficulty, or by structural complexity,
concludes that the main variable involved in sentence
production is content rather than structure.
Beattie
(1983),
reconciling both views,
strongly asserts that
speech production includes both semantic and syntactic
planning.
Both this view and the view that conceptual
planning is an obligatory component of planning in the
production of written discourse seem reasonable.
In the light of the discussion on writing as a
cognitive problem solving, and given the educational
26

---

objectives as defined by the members of the Faculty of
Letters and Social sciences, National University of the
Cote d'Ivoire,
this study has thought i t appropriate to
examine the phenomenon of conceptualization in the
process of written composition.
1.3.
Delimitation of the specific problem
An examination of the background of the specific
problem is dealt with,
followed by the statement of the
major and specific research questions of the present
study.
1.3.1.
Background to the study
In addressing the issue of conceptualization in
written discourse, this investigator is by no means the
first one to investigate this area.
Rohman & Wlecke
(1964)
are the pioneers in this respect.
Stallard
(1974)
stresses the need for examining conceptualization in
written discourse, making no explicit suggestion as to
how this could be implemented.
Bereiter & Scardamalia
(1981, 1987) have examined conceptual planning by
children.
This study differs from Rohman & Wlecke (1964) with
respect to their goals and methodological perspectives.
The latter has sought to:
27

---

1.
Isolate and describe the principle of
conceptual transformation and
2.
develop a course that would allow learners to
emulate the dynamics of the process of
conceptualization (see Rohman & Wlecke, 1964:13)
It is worth restating here that i t is not the
concern of this research study to devise a programme for
the teaching of writing but to discover students'
processes of solving a complex cognitive problem which is
exemplified by composing written discourse.
Although Stallard (1974)
suggested ten years after
the Rohman & Wlecke (1964)
study that attention be given
to conceptualization, except for the two studies by
Bereiter & Scardamalia (1981,
1987), research in this
area is scarce.
Given the paucity of research bearing on
conceptualization, i t seems justified to pursue this line
of investigation, raising the research questions below.
1.3.2.
statement of research questions
This study asks whether student-writers resort to
conceptualization and if they do whether this is
restricted to a specific point in the process of
composing written discourse.
If not,
is such a cognitive
activity evenly distributed throughout the composing
process?
Interest is alsb in the kind of decisions that
28

---

are made during the process of composing written
discourse.
Specifically,
i t seeks to discover the relationship
between the composing process and the quality of the
written product with respect to discourse types and topic
familiarity.
Another question being addressed is to
determine to what extent the quality of conceptualization
prior to the process of relatively continuous text
generation can help to predict the quality of the final
product.
Ultimately,
interest is in discovering the
cognitive processes that are activated to meet the
demands of solving writing tasks and in characterizing
the composing processes of L3 students.
These research questions, which underpin the overall
organization of subsequent chapters in this study, will
be redefined as a result of the review of the literature
to be undertaken in Chapter two.
Chapter three examines
the methodological perspectives,
including the
theoretical background, the research instruments and
methods of data analysis.
Chapter four presents and
analyzes the quantitative results as well as the
qualitative ones at group-level.
Chapter five deals with
individual sUbjects.
Chapter six covers the
interpretation and discussion of the results both at
~roup- and individual- level, providing general
conclusions, stating the need for further research and
delineating the educational implications of this study.
29

---

CHAPTER TWO
30

---

2
Review of the literature
A review of the literature on composing' may take
two alternative forms:
(i)
as a broad theoretical
approach or
(ii)
as a study of a paradigm shift,
The
broad theoretical form defines three theoretical
perspectives represented by the expressive and cognitive
views
(suggested by Berlin,
1984,
1988)
and by the social
view (suggested by Faigley
(1986), who provides an
interesting reVlew of the issues;
see also Faigley,
()..986) ,
The expresslve v~ew
perspective includes proponents
of the 'authentic voice' movement
(Coles
(1969); Elbow
(1973,
1981); Macrorie (1968,
1976); and Stewart
(1969),
among many others).
The cognitive view perspective
includes researchers emphasizing the cognitive dimension
of composing
(Bereiter & Scadarrnalia
(1981,
1987); Flower
& Hayes
(1977,
1980,
1981); Kroll
(1978);
Lunsford
(1980),
among many others).
The social view emerges from
the following:
Bizzell
(1982); Bruffee
(1984); Cooper
1
For a global review of the state of the art,
see Braddock,
Lloyd-Jones and Schoer (1963),
Hillock
(1986a)
and Krashen
(1984).
31
"

---

(1986); Heath (1983)
and Reither(1985)
to name only
these.
The paradigm shift,
suggested by Young (1978)
and
Hairstorn (1982)
inspired by Kuhn (1970),
articulates
research on composing around the product vs the process
dichotomy2.
While process-oriented studies seek to
unravel the mental operations that underlie the act of
composing, product-oriented research emphasizes the end-
result of the composing process.
The review of the literature in this study
arbitrarily focuses on the cognitive view,
integrating
the paradigm shift axis.
The decision does not preclude
interest in the expressive,
social,
and the affective
(suggested by this investigator)
views of composing.
This decision, though arbitrarily determined,
is
necessarily motivated by the major concern for
considering composing as a cognitive problem solving
activity, as seen in Chapter one.
In the following,
the review of studies bearing on
research on composing in English as a first-language will
precede the examination of studies on composing in
English as a second-,
a third- or even a fourth-language,
the first ones having determined the orientation taken by
the second set.
Note that throughout this stUdy, to
2
Although 'the claim of a paradigm shift is now viewed by some
as an overstatement'
(Faigley,
1986:527), this dichotomy remains a
useful one.
32

---

facilitate referencing L1 will designate first-language
and L2/L3 the other instances.
2.1. First language research on composing
This section reviews only studies that have
attempted to capture the nature of the composing process.
This emphasis seems justified given two recurrent
research needs expressed in the field of written
composition over the last ten years:
(1)
the need to
redefine the composing process;
(2)
the need for a
powerful framework that can define,
explain and predict
efficient performance (Beach & Briddwell
(1984);
Sommers
(1978,
1979); Witte
(1985,
1987).
In order to capture the characteristic features that
best depict the composing process,
a chronological review
will proceed from the stage model of composing to the
cognitive process model and to the revision process
model.
Attempts to buildup an extended cognitive
revision process model of writing will be made.
2.1.1.
Stage model process of' composing
This widely held view of the composing process as
stages or steps went unchallenged from the mid 1960s to
the early 1980s.
Its main characteristic feature could
33
,'.

---

be seen in its division of the composing process into
three discrete and sequentially arranged stages.
Different terminological sets have
been coined to
describe each constituent step.
The first set of
concepts was offered by Rohman & Wlecke (1964) and Rohman
(1965)
who visualized written composition as prewriting,
writing and rewriting.
Britton, Burgess, Martin, McLeod
& Rosen (1975)
constructed the second set--conception,
incubation and production.
Murray (1978a)
considered a
new set as pre-vision, vision and re-vision.
Other
formulations sharing the above writers' views of
composition are those of Collins & Gentner
(1980),
and
King
(1978). A slightly different formulation of the
stage model is the four-step variant as exemplified
respectively by Young, Becker & Pike
(1970)
and by
winterowd (1983). While Young et al.
(1970)
propose
preparation, incubation, illumination and verification,
winterowd
(1983)
refers to the steps as prewriting,
composing, rewriting and editing.
(Table 2.1 provides a
diagrammatic representation of the linear stage model of
compos ing . )
Questions that come to mind after examining the
different sets of labels are the following:
Is this
terminological diversity really justified?
Alternatively, do these different terminological sets
describe different realities?
Or are they simply new
labels to depict the same reality?
34
~-,:.... ....
.,.

---

TABLE 2.1
Linear stage models of composing.
a)
Thee-division variant
I Stage I
:;tage II
Stage III
Rohman & Wlecke
(1964)
Prewrit.ing
Writing
Rewriting
Britton et al.
(1975)
Conception
Incubation
(?)
Product. ion
Murray
(1978)
Pre-vision
Vision
Re-vision
b)
Four-division variant
I Stage I
Stage II
Stage III
Stage IV
Young et .1.
Preparation
Incubation
Illumination
Veri ficat.ion
(1970)
Winterowd (1983 )
Prewriting
Composing
Rewri ting
Editing
A first answer to the above queries is provided by
Murray's following assertion:
'I would like to propose
new terms for
consideration,
terms which may emphasize
the essential process of discovery through writing'
(Murray,
1978a:86).
Implicitly,
then,
previous labels
have failed to capture the central and intrinsic feature
of writing;
that is,
'writing as a discovery procedure'
(Murray,
1978a:86).
Indeed,
as Murray himself explicitly
states it,
'this process of discovery has not generally
been explored or understood'
(Murray,
1978a:8l).
While it may be the case that contrary to Murray,
other proponents of the stage model have not sought to
depict writing as a process of discovery,
this researcher
would argue that the use of such a metaphorical term as
illumination reflects the tacit acceptance of discovery
as a writing process even in models which do not adopt a
process viewpoint.
A contradictory Vlew of Murray's
position is voiced by Witte
(1985),
according to whom,
'from a theoretical and historical point of view,
35
.'.

---

Murray's view of revising is consonant with the
traditional linear-stage of composing'
(Witte,
1985:257).
It may even be added that the different linear-model
variants definitely share
two common components:
the
first and last stages.
The first stage,
referred to as
prewriting, conception, preparation or pre-vision, could
best be defined as the stage at which ideas are
conceived,
formulated,
reformulated and shaped.
Of all the terms offered, preparation,
which echoes
Jobnson's
(1955)
contention,
captures best the dynamism
inherent in this exploratory period.
Exploration into
the unknown is initiated when a writer, developing
initial awareness of a difficulty,
formulates the
difficulty as a problem.
Indeed,
the emphasis of Young
and his colleagues on the creation of a rhetorical
problem,
i.e awareness of a difficulty,
a conflict
between one's view and the extra-linguistic world,
is
significant, and is further developed in the Flower and
Hayes model to be discussed later,
in which the
development of the
rhetorical problem is a crucial
parameter.
Apart from the first stage,
consensus has also been
reached on the third stage which lS concerned with
delayed revision
(see Wason,
1980).
The words of Britton
and his colleagues nicely sum up the definition generally
accepted:
36

---

Revision is not just a question of correcting
and improving.
It is also the final stage of
the process by which a writer presents himself:
every piece of writing can be,
to some extent,
a declaration,
a tacit agreement with the
reader that the writer accepts responsibility
for his own creation
(Britton et al.,
1975:
47) .
An answer to the question raised above is that the
terminological diversity is not justified when the first
and last stages are examined.
Indeed,
even though
different labels have been coined to depict the first and
third stages of the linear model of composing,
these
stages appear to represent the same reality.
In other
words,
the terminological diversity In both instances
does not reflect different realities.
However,
a similar
claim cannot be made when the second stage is examined.
Indeed, while it seems reasonable to subsume the
notions of writing and composing and even vision as
describing the stage at which ideas generated during the
first stage are committed onto paper,
it would seem
wrong to assume that there is strict correspondence
between incubation and the other three terms mentioned
above, viz. writing, composing and vision.
The notion of
incubation involves rest periods and maturation.
On no
account does incubation refer to the physical activity
of transcribing.
The contention is that incubation,
illumination and verification should not be examined as
static stages,
rigidly
arranged and mutually exclusive.
Neither should they be perceived as fixed steps with no
37
.'.

---

possible overlap, but rather as mental operations that
intervene throughout the composing process and are not
therefore limited to anyone stage.
Strong criticisms,
in similar vein, have been
levelled against the linear-stage model of composing.
Written composition specialists have highlighted three
sorts of limitation.
Firstly,
they have questioned the
linearity of the model.
This conception,
they claim,
may be seen to arise from the influence of classical
rhetoric which,
in its attempt to apply the
'irreversibility of speech to writing',
has emphasized
the linear sequence of invention,
arrangement and style
(see Sommers,
1979,
1980; Nystrand,
1982).
Biological,
linguistic and mathematical theoretical models are other
influential sources that have contributed to the
misrepresentation of the composing process.
Rohman and
Wlecke's model exemplifies the influence of Abram'S
(1953) plant analogy, used to explain 'organic unity'
in
poetry
(Spivey, 1983 cited by Witte,
1985:256).
Britton
et al.
(1975)
and Winterowd
(1983)
have drawn their
inspiration from Jakobson (1960).
Henri Poincar~'s
(1913,
1946) model of creation has been instrumental in
producing the four-step variant
(see Crow1ey,
1977).
The second limitation of the linear-stage model can
be seen in its focus.
The model,
as it is, best
describes the unfolding pattern of the product rather
than the composing process itself
(Sommers,
1979, 1980);
38

---

Flower & Hayes
(1980).
The last shortcoming of this
model is its excessive emphasis on methodological or
pedagogical concerns at the expense of theoretical
considerations
(Sommers,
1979:46; Murray,
1978).
Winterowd is a case in point.
He acknowledges the
arbitrariness and artificiality of his division of the
composing process into discrete and uni-directional
stages.
However, he readily sticks to it,
asserting that
it is pedagogically strategic,
since he believes that
composition teachers need the
'know-how' to intervene in
every stage of the process.
The tension between theoretical and practical
considerations is also reflected in Britton et al. (1975)
The authors,
after insisting on the difficulty of
separating the three activities--conception,
incubation
and production --because they run concurrently, move on
to justify the need for prewriting to take precedence
over actual writing,
alleging once again pedagogical
reasons
(see Britton et al.,
1975:26).
Given the above limitations,
one may forcefully
claim that the linear-stage model misrepresents and
distorts our understanding of the composing process.
It
fails to capture the recursive and hierarchical nature of
composing
(Sommers,
1979; Perl,
1980; Flower & Hayes,
1981; Witte,
1985).
Along with Nancy Sommers,
this
researcher feels that such a model offers an a-priori
distinction of the composing components without providing
39
'.

---

evidence of their existence.
This researcher contends
that it lacks a clearly articulated theoretical base,
and fails to produce a theory of the composing process.
Given that those researchers were primarily concerned
with setting up a pedagogical tool,
rather than offering
a theoretical model of the composing process,
attacking
them on the basis of objectives that did not motivate
their study does not seem to do them justice.
Dissatisfaction with the linear-stage model has led,
as one could have expected,
researchers In the field of
written composition to seek alternative accounts of the
process.
Although there is a multiplicity of models
available
(Augustine,
1981; de Beaugrande,
1984)
the
present investigator proposes to give special attention
to the Hayes & Flower (Flower & Hayes, 1981; Hayes &
Flower,
1980) cognitive process model of writing because
this model lS the 'most widely cited model,
and has
tended
to fix the vocabulary people use in talking about
the composing process'
(Scardamalia & Bereiter,
1985b:781) .
A precursor to studies by Flower and Hayes is Emig
(1971) who may have paved the way to the process
movement.
A lesser known work is Zoel1ner (1969)
The orientation of investigations by Flower and Hayes lS
an improvement on Emig
(1964,
1971)
in that it fills a
theoretical gap by offering a cognitive theoretical
framework.
40

---

2.1.2.
The cognitive process model of composing
The graphic representation of the Flower & Hayes
(1980,
1981) model
(see Fig.
2.1)
includes three major
components:
(1)
the task environment,
(2)
the writer's
long term memory and
(3)
the writing processes.
The
constituent elements of the task environment are the
rhetorical problem and the text produced so far.
Chronologically,
the rhetorical problem,
as previously
alluded to in our examination of the preparation stage
and as emphasized by Flower and Hayes,
plays a crucial
role at the beginning of the act of writing.
The
rhetorical problem may either result from initial
TASK ENVIRONMENT
TIlE RHETORICAl.
TEXT PROIJIXE)
PROBLEM
Topic-
SO FAR
Audience
Exigency
'------'I
..WRITING ,.
PROCESSES
!'L'.NNlMJ
REVIEWING
~
mE WRITER'S LONG-
r--
1RANSUllNG
TERM MEMORY
~ 1 ORGANrm;GI
IEVAWA'llNG I
Knowledge of Topic.
Audienc.e.
i
and Writing
~I
R£VTSlNG
0
L - -
Pl...
~
I
I
I
I
I
MONOOR
I
Figure 2.1
The Flower and Hayes
(1980,
1981) cognitive model of
composing.
41
"

---

awareness of a difficulty by a potential writer or may be
initiated by an instructor who sets a writing task to a
student-writer.
The development of the rhetorical
problem is shaped not only by the task initiator,
but
also by the topic,
the audience being addressed,
the
writer's goals in writing or his motivating cues.
Assuming that the topic initiator is instrumental In
the formulation of the rhetorical problem,
one may
hypothesize that such an influential factor is likely to
determine the strategies that are activated In
retrieving needed information from long-term memory.
The
topic-initiator may affect the writing processes as a
whole by placing different exigencies and constraints on
the monitor, which is the executive system that controls
the three main processes of the total writing processes.
The planning,
translating and reviewing functions might
be activated,
regulated and weighted differently to
respond to the expectations of the task initiator.
A predictable difference between the traditional
linear-stage model and the cognitive process model is
illustrated by their units of analysis.
While in the
linear-stage model the 'major units of analysis are
stages of completion',
in the cognitive process model,
'the major units of analysis are elementary mental
processes,
such as the process of generating ideas'
(Flower & Hayes, 1981:367).
Another central difference
between the cognitive process model and the linear-stage
42

---

model is that the former emphasizes the hierarchical and
recursive dimension of composing.
Indeed,
unlike those in a linear organization,
the
events in a hierarchical process are not fixed
in a rigid order.
A given process may be
called upon at any time and embedded within
another process or even within another instance
of itself,
in much the same way we embed a
subject clause within a larger clause or a
picture within a picture
(Flower & Hayes,
1981:375) .
The recursiveness 3 of the Flower and Hayes model is
manifested by potential interactions between its three
major components,
task-environment,
writing processes and
long-term memory.
The double sets of two-directional
arrows linking the task environment to the writing
processes and the long-term memory to the writing
processes clearly suggest dependency among these
components.
In constructing this model,
Flower and Hayes were
greatly influenced by Newell & Simon's
(1972) models of
problem-solving.
The four underlying hypotheses could be
stated as follows:
(1)
The process of writing can be
defined as a set of distinctive thinking processes;
(2)
These cognitive processes are hierarchically arranged;
(3) The act of composing is goal-directed;
(4) Writing lS
a constructive act In the process of which writers create
their own goals
(Flower & Hayes,
1980,
1981) °
J
Myers
(1986)
stresses the inadequacy of the term as used by
composition researchers.
43
o.

---

The Flower and Hayes model seeks to account for the
mental processes at work in the act of composing.
It
undertakes to capture both external and internal elements
of this process.
How successful has i t been?
According
to Scardamalia & Bereiter (1985a),
the cognitive process
model provides
the most explicit account of mental operations.
The strength of the model lies in its claim to
account for the amazing diversity of mental
events during composing on the basis of a
relatively small number of such sub-
processes .•. The general structure of the
model, thus,
appears to do what i t is supposed
to,
which is to serve as a frame for working
out more detailed and possibly more
controversial accounts of how the mind copes
with writing tasks
(Scardamalia & Bereiter,
1985a:781) .
Witte also favourably reacts to the cognitive model.
His following words are quite revealing:
The Flower and Hayes model of composing clearly
represents a more robust theory of composing
than its predecessor,
the linear- stage model,
and i t seems consistent in many ways with the
views of composing offered by Sommers (1980)
and Nold (1981).
(Witte,
1985:26]).
Although,
as suggested by Scardamalia and Bereiter
(1985a), the cognitive process model of composing is well
received in the field of written composition,
this model
has not gone unchallenged.
Cooper & Holzmann (198])
strongly question the theoretical basis and
44

---

methodological perspectives of this model, making no
suggestions for an alternative model.
Given that Cooper
& Holzman's (1983) criticisms are often referred to in
the field of written composition,
it is worthwhile giving
special attention to their argumentation.
Specifically, both criticize Flower and Hayes for
their 'casual reliance on unarticulated theories'
and for
their loose terminological use of the concepts of plans
and goals.
Additionally, Cooper & Holzmann
(1983)
stress
the inconsistent view of the cognitive model as concrete
representation of the mental processes at work In
composing.
Finally,
they attack the cognitive model
whose methodological perspective is essentially concerned
with introspection and verbal protocols.
Cooper and
Holzman are suspicious of introspection as a research
tool which they consider antiquated because it dates back
to Plato.
They also dismiss this research perspective on
the ground that it has been scientifically discredited.
They reject
'thinking aloud' protocols, asserting that
'protocols,
far from being "extraordinarily rich data",
are exceedingly impoverished sources of information on
what writers are thinking about'
(Cooper & Holzman,
1983 :286) .
Apparently,
as suggested by Cooper and Holzman
(1983),
Flower and Hayes
'see no need for a distinction between
plans and goals',
even though by defining writing as a
goal-directed activity,
they tacitly consider cognitive
45
.'.

---

processes within the framework of purposive behaviour.
If
Cooper & Holzman (1983) are right,
the Flower and Hayes
model rests on theoretical shaky ground as a result of
terminological inconsistencies.
Central concepts--such
as goals and plans--would need to be precisely defined.
Furthermore,
given that Flower and
Hayes's perception of
writing as goal-directed clearly sets their work within a
purposive psychological framework,
and assuming that they
are well aware of the model of Schank & Abelson
(1977),
one would have expected them to justify their decision
for not referrlng to such a model and to specify how
their understanding of goals and plans differs from that
particular model.
A further cause for concern arises from the failure
of Flower and Hayes to take into account the repeated
warnings of psychologists as well as linguists that
models are not literal descriptions but abstract
representations of what they are modelling.
Indeed,
they
contend that their model is a concrete representation of
the mental processes at work in composing.
The model of the writing process
[they claim]
attempts to account for the major thinking
processes
and constraints we saw at work in
these protocols
(Flower & Hayes, 1980;368).
Such a claim,
as forcefully expressed by Cooper and
Holzman,
is misleading and ill-founded.
Others too have
warned that on no account should verbal protocols be
46

---

presumed to
give a direct record of the
mental processes of composing.
[ ... ] Thinking
aloud protocols are better thought of simply as
data to be used in conjunction with other data
to make inferences about the composing process
(Ericsson and Simon,
1980,
cited by Scardamalia
& Bereiter, 1985b:781).
Finally,
the need for corroborative data,
as
suggested by Ericsson and Simon above,
has simply been
ignored by Flower and Hayes.
Indeed,
they make no
specific attempts to supplement the incompleteness of
verbal protocols'.
A closer look at the criticisms made by Cooper &
Holzman
(1983)
reveals that both investigators are
suspicious of introspection and dismiss verbal protocols
as a variant of the discredited process of introspection.
It is also clear that their rejection of introspection
arises from their assumed 'so-called' triumph of
behaviourial psychology, an assertion which is somewhat an
alarming claim.
At this point i t lS worth noting that Flower and
Hayes's recourse to introspection5 --had this been the
case--is not methodologically unsound provided that such
an inquiry tool is demonstrated as being scientifically
rigorous.
It lS also important to distinguish between
diverse forms of verbalization, which, as Ericsson and
4
Flower & Hayes
(1983)
may be seen as an improvement as both
investigators have explicitly defined their procedure;
thereby
reducing potential sources of confusion in interpreting their
methodological perspective.
5
Sternglass
's
(1988)
discussion of the issue with respect to
research on reading and writing might be of interest.
47

---

Simon put it,
'are all jointly and loosely condemned as
"introspection"'
(Ericsson & Simon,
1980:216).
In this
respect,
it may be of interest to note how Ericsson and
Simon distinguish between concurrent verbalization,
retrospective verbalization and classical introspective
reports.
The term 'concurrent' verbalization refers to
the verbalization of ongoing thinking while a task is
being performed.
The term retrospective verbalization is
defined as a response to speclfic probes eliciting post
task account of mental processes.
Classical
introspective reports are produced by trained observers'.
Furthermore,
think-aloud protocol analysis,
used by
Flower and Hayes,
which has been independently initiated
by Duncker
(1926)
and Claparede
(1934)
to study problem
solving processes,
cannot be mistaken for introspective
analysis.
In their attempt to understand the mental
operations that are deployed in the process of solving a
problem,
protocol researchers are not primarily concerned
with eliciting the reflective thought of the problem
solver on his own mental operations;
rather their focus
is on the sequence of actions leading to the solution of
the problem and the mental processes underlying the
sequence itself.
The distinction between the nature of
the data elicited from a purely introspective perspective
and that resulting from think-aloud verbalization is
6
Introspective reports and verbalization of thought are used
interchangeably despite the above terminological distinction.
48

---

crucial in eliminating uninformed confusion between the
two forms of analyses and their variants
(see Steinberg,
1986 and Dobrin,1986).
To sum up,
to discount protocol analysis as a method is to
ignore the many contributions it has already
made to our understanding of written
communication, and to dismiss the knowledge it
is bound to uncover about composing in the
future
(Smagorinsky,
1989:476).
As to the incompleteness of verbal reports
emphasized by Cooper and Holzman
(1983),
such a
limitation should not be a matter of great concern if the
suggestions made by Ericsson and simon
(1980)
are taken
into account.
These two researchers argue that
verbal report data can be reliable provided
that experiments that record verbal responses
of any kind [have]
a rudimentary theory of how
subjects produce such responses--in what
memories the response information have been
stored, what demands the response makes on
short-term memory, whether responses can go on
in parallel with other behaviours
(Ericsson &
Simon,
1980:216).
In addition to the limitations of the cognitive
process model of composing highlighted by Cooper &
Holzman
(1983),
another important shortcoming of this
. model concerns its omission of short-term memory
(STM).
That such a model which seeks to capture both external
and internal elements of composing should ignore STM is
49
.,

---

rather surprising.
Indeed,
one would have assumed that
for this model to account fully for the mental operations
that are activated during the act of writing,
it must
necessarily include STM.
Could it be the case that
short-term memory is made redundant as a result of
the
potential direct retrieval capacity of long-term memory?
possible justification for ignoring STM may be found
In Flower and Hayes's contrasting definition of both
types of memories.
Opposing STM to LTM,
Flower and Hayes
emphasize the stability and systematicity of LTM.
'Unlike short-term,
[ ... ] memory long term memory lS a
relatively stable entity and has its own internal
organization'
(Flower & Hayes,
1981:371).
Three plausible inferences can be drawn from this
contrast:
(1)
STM is unstable;
(2)
It is a-systematic;
(3) As a consequence, it is not worth being considered.
These inferences,
far from being excessive,
are an
explicit account of Flower and Hayes's implicit view.
Hence,
the fact that STM is absent in their model does
not result from a casual omlSSlon.
This absence derives
from a deliberate theoretical option chosen by Flower and
Hayes.
Such a theoretical decision is somewhat analogous to
de Saussure's justification for attending only to
'langue' at the exclusion of
'parole'.
As one may
remember,
Saussure argues that since 'langue' is stable
50

---

and systematic it is worthy of interest,
whereas
'parole', because of its 'accidental' natur~ is to be
discarded.
This separation is no longer maintained,
as
testified by developments in the field of linguistics.
In
addition,
the study of performance has helped to re-
establish the balance between both aspects of
'langage'.
While the separation between these two facets of
language could be methodologically justified,
one would
be hard pressed to provide sound methodological reasons
for ignoring STM in a cognitive process model of
composing whose very purpose lS to account for mental
operations which are activated in language generation.
Another justification for Flower and Hayes's
omission of STM in their
(1980,
1981) model is offered by
studies involving impaired subjects which suggest a clear
STM/LTM dichotomy
(Baddeley,
1976).
Indeed,
Baddeley
goes on,
a great deal of the evidence available on human
memory is compatible with the view that human
memory comprises two major components: a short-
term store of limited capacity,
which relies on
phonemic coding, and a long-term store of
enormously greater capacity,
which depends
heavily on semantic coding
(Baddeley 1976:162)
Despite different kinds of evidence supporting the
separate memory system,
the review of studies by Baddeley
(1976)
attests to the complexity of the issue and invites
one to '!beware of oversimplification'.
The need for
being cautious is reiterated by Allport
(1985):
51
.,

---

The once-popular distinction between long-term
and short term stores ... as functionally
separable components has come under
increasingly severe criticism.
It is probably
fair to say that there is now no really
convincing evidence in favour of such a
distinction,
and much that is contrary
(Allport,
1985: 55).
Despite the complexity of the issue and the problems
raised by the discussion of LTM/STM, the fact that there
is 'some evidence that STS may function as a working
memory'
(Baddeley,
1976:161)
is rather appealing.
If one
agrees with the assumption that 'the short-term storage
system [may]
function as a working memory,
playing a
central role in reasoning,
comprehension of language, and
long-term learning'
(Baddeley,
1976:162), then there
might be a case for the inclusion of STM in a cognitive
process model of writing.
This suggestion seems to run along the same line as
Daiute (1981).
In her discussion of a psycholinguistic
model of writing,
she posits that
an explanation that accounts for the effects of
memory on sentence production is important to
the study of writing because writing involves
many activities that go on in short-term
storage.
During the composing stage,
the
writer is (1) generating ideas,
(2)
forming
propositions,
(3)
accessing lexical items,
(4)
planning clauses and sentences,
(5)
translating
from semantic and phonological representations
to orthographic ones,
and (6) planning
subsequent units (Daiute,
1981:8-9).
Given this, this researcher would suggest two
reasons for incorporating STM in the cognitive process
52
..'~
. -~.

---

model of writing.
Firstly,
during 'thinking aloud'
protocols,
subjects are likely to verbalize information
attended to in short-term memory. This may help to
explain incompleteness in verbal protocols.
Secondly,
inclusion of STM in a cognitive model of composing may
help one tap both the interaction between a STM loss and
writing rate
(Scadarmalia & Bereiter,
1982)
and the
interaction between STM loss and familiarity with tasks
(as observed in the data of this study).
The discussion of the Flower & Hayes
(1980,
1981)
cognitive model of the composing process has attempted to
show both the strengths and weaknesses of such a model.
Unquestionably,
the cognitive process model of composlng
represents a landmark in the
field of written
composition.
This model has been successful In
formalizing and systematizing the intuit ions of
practicians and theorists.
It has also managed to bring
together into a unifying perspective hunches and guesses.
However,
despite major advances initiated by the Flower
and Hayes model,
such a model as well as the other
'models of the composing process developed thus far are
what Kosslyn
(1980)
calls
'protomodels'.
They are
precursors of theory,
not instantiations of theories'
(Scardamalia,
1985b:783).
Pursuing Scardamalia's
'observation further,
this researcher would suggest that
potential weaknesses of the Flower and Hayes model arise
53
.,

---

not so much from its methodological approach as from its
theoretical considerations.
Reiterating previously stated criticisms,
this
writer would argue that for the cognitive process model
of composing to be fully efficient,
a number of
conditions are to be met.
Firstly,
a more explicit
theoretical framework.
coupled with rigorous definitions
of central concept~ is
to be made available.
SecondlY,
for the cognitive process model to provide an exhaustive
account of the mental processes activated during the act
of writing.
such a model must necessarily incorporate STM
into its system.
Additionally,
for this model to be
fUlly explanatory,
it must provide ways and means of
accounting for the developmental dimension of writing.
A final condition,
closely related to the third one,
lS
that only a wider view of revision is likely to
explain apparent contradictory results reported In
empirical investigations.
Such a composing
model, by
allowing an expanded built-in revision process in its
system,
should be explicit and powerful enough to explain
and predict successful as well as unsuccessful writing
performance.
This recommendation logically leads to
examining available theoretical and empirical studies on
revision so as to discover whether they provide powerful
alternatives to the linear-stage
and to the cognitive
models of composing.
54

---

2.1.3.
Revision process model
The discussion of the revision process model brings
together theoretical and empirical research on rev1s1on.
To claim that revision is an essential and necessary
process in
the act of writing is by now a truism.
The
importance of revision has reached general consensus
among professional writers,
teachers as well as
researchers
(Bartlett,
1981,
1982; Britton et al.
(1975).
Indeed,
'most writers accept rewriting as a condition of
their craft' (Murray,
1978:85).
If general consensus is reached as to the importance
of revision in the act of writing,
this 1S not generally
the case when one considers the place of occurrence of
the revision process.
Revision,
as very explicitly
stated in the linear-stage model,
is a separate activity
performed on a completed draft
(see Britton et al.
1975;
Murray,
1978; Winterowd,
1983).
A different view held by
Collins & Gentner (1980)
and Flower & Hayes
(1981)
is
that revision is a recursive activity, performed at any
point' .
Commenting on the place of occurrence and nature
of the revision process,
Bartlett
(1982)
has this to say:
Whatever its place in the writing process,
revision is invariably distinguished from text
generation by the fact that it involves some
fairly explicit processes of comparison,
generally between some segment of a text
(a
word,
phrase,
sentence, paragraph,
etc.)
and
7 Note
the definition of composing as recursive was first
suggested by Emig
(1964,
1969,
1971).
55
",

---

some representation of a writer's knowledge of
intention, which results in some attempt to
change existing text
(Bartlett,
1982:346).
Bartlett invariably limits the scope of the revision
rocess to textual
'touch-up'.
As such,
she
fails to
[take into account the need for expanding the concept of
'revision so as to produce an adequate theoretical model
I f ·
10
composlng.
Bartlett's restricted view of revision is
,not surpr1s1ng since,
although the value of revision 1S
IWidelY recognized,
'rewriting is one of the writing
skills least researched,
least examined,
least
understood, and normally --least taught'
(Murray,
1978:
85).
Sommers
(1980) makes a similar observation.
She
contends that
the absence of research on revision is a
function of a theory of writing which makes
revision both superfluous and redundant,
a
theory which does not distinguish between
writing and speech (Sommers,
1980: 379).
As one can infer from the preceding,
the lack of
research up to the early 1980s 1S ascribed to the linear
stage model of compos1ng.
Murray
(1978)
suggests another
reason for the scarcity of research on revision.
This
concerns
the negative perception of the term.
Rewriting
is stigmatized as failure,
negatively sanctioned.
Far
from being perceived by teachers,
students and even by
academics as 'the most exciting,
satisfying,
and
significant part of the writing process' or as an
56

---

'opportunity for discovery',
often,
'rewriting lS taught
as punishment'
(Murray,
1978:85).
Finally,
'most texts
[ . . . ] confuse rewriting with editing,
proofreading or
manuscript preparation'
(Murray,
1978:86).
At this
point
one may ask what is meant by reVlSlon and what
motivates its occurrence
(Hillocks,
1986a).
Definition of the concept of revision
Sommers
(1979)
has identified more than twenty-five
words that describe revision.
Among others,
terms such
as--revision,
rewriting,
tidying up,
clean-up,
rechecking,
proofreading--are used
ambiguously and interchangeably to depict
surface cosmetic changes and to equate revision
with an activity that is separate in quality
and isolated in time from writing
(Sommers,
1979:49) .
This conception of revision,
'which is reinforced by
the structure of most composition, textbooks',
is
consonant with Bartlett's
(1982)
view discussed above.
An interesting dichotomy which seems to offer a new
perspective is Murray's
(1978)
distinction between
internal revision and external revision.
One would have
hoped to consider internal revision as depicting the
evaluating and reformulating processes at work not only
during the generation of ideas but also during the
57
"

---

definition of the rhetorical problem.
However,
one
disappointingly notes that Murray's dichotomy is
restricted to textual retranscription
(Witte,
1985)
This limitation is best illustrated by possible
rapproachment between this set of terms and other terms
used in the linear-stage tradition.
The internal and
external revision
dichotomy reflects Winterowd's
(1983)
distinction between rewriting and editing or Flower &
Hayes's
(1981)
opposition between a
'writer-based' prose
and a
'reader-based' prose.
Witte
(1985)
would even see
similarity
with"
Perl & Egendorf's
(1978)
notion of
'retrospective restructuring'.
'Internal revision l
is
initiated by a
'completed first draft',
while 'external
revision'
lS initiated by the need to
'communicate
[ ... ]
with another audience'
(Murray,
1978a:85).
By restricting the scope of internal revision to
textual retranscription, Murray still remains entrenched
within the boundaries of the linear-stage model of
composing.
A possible way of breaking from the bonds of
the traditional paradigm of composing and thereby of
adopting a new theoretical framework is to redefine,
as
Sommers
(1979)
suggests,
the composing process not simply
'as a linear series of stages but rather as a
hierarchical set of subprocesses'.
One would galn insight into the composing process
not by looking at revision as rewriting,
rechecking,
proof-reading,
or editing, but as a recursive process, as
58

---

'an attempt to make writing consonant with changing
vision.'
Better, yet,
our understanding of writing would
be greatly enhanced if
'the entire process of revision' was
coupled with the examination of
'revision as a process of
making a work congruent with what a writer intends--a
process that occurs throughout the writing of a work'
(Sommers,
1979:48).
Such a definition of the act of
composing is more than appealing and should definitely
have paved the way to a number of empirical studies
within a new theoretical perspective.
Focus on empirical revision research
Although most researchers have wholeheartedly
adhered to the concepts of recursion and hierarchy,
they
have unfortunately been largely influenced by the
traditional stage model paradigm.
As Witte
(1985)
puts
i t ,
research on revision has largely limited the
phenomenon [ ... ] to the final stage in a
sequence of stages and has narrowly defined
revision as the manipulation and alteration of
features in written texts'
(Witte,
1985:271)
Clearly,
then,
revision investigations are
product-oriented (Bridwell, 1980) and,
in the main,
they
have sought to describe the effect of revision on the
8
Emphases added.
59
.',

---

completed draft.
Studies undertaken by Sommers (1978,
1980) make no exception.
As a rule,
empirical
investigations on revision have addressed three questions
which Humes
(1983)
and witte (1983c)
have summarized as
follows:
(1) At what point during the drafting of texts
do writers revise?
(2) What kinds of revisions do
writers make in their written texts?
(3)
What
differences in kinds of revisions of written texts occur
across writers of different abilities?
Answers to these questions have provided indices for
comparing skilled and unskilled writers.
Revision
studies have revealed that, unlike Skilled writers,
unskilled writers tend to revise prematurely, to attend
to local revisions and to be less inclined to revising.
As a rule, the number of revised drafts are taken as
indicators of proficient performance.
However, contradictory findings seem to invalidate
previous results.
Recent studies (Faigley & Witte, 1981,
among others)
indicate that,
although experienced adult
writers produce texts superior in quality to those of
unskilled adult writers, they produce less revising
within and between drafts (see Faigley & witte, 1981).
In sum, the contradiction is that proficient adult
writers do not necessarily resort to extensive textual
revision.
Hence, according to this view, extensive
textual revision cannot be considered as an index of
proficiency.
Neither can such a strategy be used as a
60

---

differential index scale to discriminate between skilled
and unskilled writers.
It is to be stressed,
however,
that no evidence is offered to account for the above
contradiction.
The orientation taken by most revision studies has
undoubtedly widened our understanding of what goes on
when a writer sets out to revise.
Developmental
psychological accounts have helped us understand that
revising involves specific psychological processes.
Yet,
by limiting their investigations to the final stage
of the linear conception of composition,
Sornrners
(1978)
as well as other researchers have failed to meet their
ultimate goal,
which lS to perceive
'the entire composing
process as a process of revision'
(Sornrners,
1979)
They
have also failed to fully exploit the implications of the
concepts of recursion and hierarchy inherent in the act
of composing.
Neither have they been successful in
explaining the contradiction discussed earlier.
As a
result,
they have been unable to account for
revision processes occurring prior to text generation
(pre-textual,
or during the processes of problem finding
and/or problem representation);
that is,
potential
revision processes that are activated prior to committing
words onto paper.
Whether considering revision as a post
task activity or as a
recursive activity,
most
composition specialists have failed to take into account
the need for expanding the concept of revision so as to
61

---

produce an adequate theoretical model of composlng.
This necessarily calls for an extended cognitive revision
process model ofcomposlng.
One might argue,
however,
that the Flower & Hayes
(1981)
and Hayes & Flower
(1980)
cognitive model of
composing is robust enough to account for pre-textual
revision processes.
Specifically,
one might argue that
the hierarchical structure of the cognitive processes for
writing, coupled with the recursive nature of the
monitor, which may at any time interrupt a glven process
or subprocess to allow another process or subprocess to
meet specific demands of,
say,
the task environment,
implicitly suggests that pre-textua1 revision might
precede textual retranscription.
It is true that the
Flower and Hayes model stands as an exception.
Indeed,
to use Witte's words,
i t is
[the]
cognitive process theory that
allows revision research to move beyond its
moorings in "traditional" conceptualization of
composing.
In the Flower and Hayes theory of
composing,
the concept of hierarchy as well as
the concept of recursion receive considerable
attention
(Witte,
1985:262).
It is also true that the central concepts of
hierarchy and recursion can best be accommodated by a
cognitive process model rather than by a linear-stage
model of composing.
Yet,
sharing, Witte's
(1985) view,
the present researcher would argue that
62

---

although revising a pre-text can be
accommodated in the Flower and Hayes model of
composing,
the model itself is not specific
enough to suggest the nature of that process or
the nature of the pre-text itself
(Witte,
1985:
271) .
An extended cognitive revision process model
of
composing
Clearly,
the above discussion highlights the need
for an adequate theory that would recognize that
writers are fully capable of completing writing
tasks,
some of them quite complex,
with most
substantive
'discovery' occurring well before
transcription
(Witte,
1985:271)_
Put differently,
along with Witte,
this researcher
contends that such a theory must rest on the assumptions
that
(1)
revising a pre-text --'the mental representation
[that]
can take multiple forms--from sensory images,
to
concepts,
to metaphors,
to feelings'
(Witte,
1985:264
informed by Flower & Hayes, 1984)--is as important as
revlslng a written text; and (2)
that as a result of the
hierarchical structure of composing,
the same
subprocesses operate on both pre-texts and texts.
It is
also claimed that such a model should be explicit enough
to account for not only the nature of pre-textua1 as well
as textual revising processes but also should be able to
explain contradictions or exceptions in the research on
63
-.

---

revision.
In other words,
this model should be powerful
enough to explain why,
as indicated by Faigley & Witte
(198l~ proficient adult writers do not necessarily resort
to extensive textual revision.
For such a model to achieve its purpose,
the
following adjustments are to be made:
(1)
Inclusion of the
central concepts of hierarchy
and recursion so as to accommodate the maJor
components of the act of writing as identified by
the cognitive process model;
(2)
Expansion of the concept of revising so as to
include
both retranscribing and pre-textual
revisions;
(3) Rethinking of the role of
'discovery'
not as
arising from
transcription,
or as what Flower and
Hayes term a
'generate and test strategy',
but
rather as substantive 'discovery'
resulting from
careful and deliberate planning occurring between
time of
'encountering rhetorical problem'
and time
of actual writing;
(4)
Supplementing 'retrospective structuring' or
recursive revising of written texts by a
complementary process referred to as
'projective
structuring' .
Assuming that these adjustments can adequately be
built into the proposed extended cognitive .revision model
of composing,
the present researcher contends that this
64

---

model should be powerful enough to account for the
developmental aspect of writing.
It is also claimed that
such a model should be able to account for the
contradictory findings previously discussed in section
2.1.3.
specifically, the extended cognitive revision
model should be able to explain differences between
writers as resulting from differences in abilities to
frame prior to transcription.
If this assertion is
accepted,
this researcher posits the following
hypotheses:
1)
Adequate framing prior to transcription results
in limited amount of textual revision.
2)
Inadequate pre-textual framing leads to
extensive textual revision.
The implication of the second assertion is that
skilled writers are likely to revise more at the
pre-textual level than at the textual level.
The first hypothesis rests on the assumption that
proficient writers call on pre-textual revision as early
as when ideas are being conceived,
formulated,
reformulated and shaped.
Put differently, pre-textual
revision processes are activated by the skilled writer
during exploration into the unknown.
This exploration is
initiated at the very moment when a writer develops
awareness of a difficulty,
and when he or she attempts to
formulate the perceived difficulty as a problem.
65

---

The second hypothesis assumes that skilled writers
have developed powerful medium- and long-term strategies
that allow them to create coherence In written discourse,
while unskilled writers are limited to short-term
strategies.
Indeed,
it is assumed that skilled writers
are provided with adequate
strategies for recognizing what information
might be relevant,
or assessing the extent to
which more information might be needed before
beginning to write
(Applebee,
1982:373).
Finally, glven that skilled writers naturally
activate,
during their production of written discourse,
certain key elements such as the writer's previously
acquired knowledge of a variety of schemata,
frames,
scripts,
and prose genres and different sets of
organizing principles,
such as syntactic,
semantic and
pragmatic,
this researcher posits that the activation of
these various elements by the writer involves deliberate
thinking and planning that occur during pre-textual
revision.
Even though at this stage
a fully-fledged extended
cognitive revision process model of writing cannot be
offered,
it is suggested that a promising research
perspective would be to consider written composition as
an ongoing revision process that is initiated when a
potential writer develops awareness of a problem.
66

---

This revising process might include evaluation and
reformulation subprocesses with potential checkpoints
identified as incubation,
illumination and verification.
The strength of the extended cognitive revision process
of composing definitely lies in its claim to define,
explain and predict successful and unsuccessful writing
performance.
Another strength is that,
by virtue of its
nature,
this model is likely to reconcile both the
process and product dimensions of written composition.
Note here that the hypotheses resulting from the
discussion of first language research on composing will
not be directly tested in this study.
They might serve
as explanatory hypotheses for the discussion of the
findings in this study.
It is worth recalling here that
Ll research on composing has determined lines of inquiry
In L2/L3 research on composing in ways to be examined
below.
2.2. Second/third language research on composing
In reviewing the literature on studies dealing with
L2/L3 composition,
a diachronic perspective could focus
on the two stages whose objects can be summarized In
Krapels's following words.
Early L2 researchers are apparently trying to
grasp whatever they can about the nature of L2
67

---

composing,
especially concerning which
behaviours seem to be successful or
unsuccessful in producing effective L2
compositions.
Later L2 researchers focus on
specific composing behaviours,
specific types
of L2 writers,
or features unique to L2
composing (Krapels,
1990:39).
Adopting such a diaChronic approach would lead one
to undertake a general-to specific description of changes
in the field of L2 composing.
Alternatively, one could
develop an approach based on three main issues inferred
from Krapels's
(1990)
review of the literature in this
field.
These issues are:
(1)
Empirical studies with
focus on identlfying similarities and differences between'
L1 and L2 composing processes;
(2)
Empirical studies with
emphasis on discovering factors affecting composlng
processes; and (3)
Pedagogically oriented studies
suggesting an application of process findings in the
classroom with a view to enhancing students' writing
performance.
This alternative organizing principle is
retained here,
excluding studies bearing directly on the
pedagogical exploitation of process models in the
classroom.
In the maln,
investigation in the field of second
language composing lS,
like research in L1 composing,
case study-based
(from 1 to 8 subjects,
4 on average,
from diverse linguistic backgrounds,
including Arab,
Burmese, Chinese,
Prench, German,
Spanish,
Turkish,
and
many others) using data which include think-aloud audio-
68

---

or video-protocols, written products,
questionnaires,
interviews and/or retrospective feedback.
Specific
studies making use of this methodology and illustrating
the two types retained in this study are now examined.
2.2.1.
Empirical studies comparing composing processes
in Ll and L2
Studies on second-language composing,
adopting a
comparative perspective,
have highlighted both
similarities and differences.
Findings bearing on the
similarities between Ll/L2 composing processes outweigh
those emphasizing differences between both composing
processes.
Among the few investigations singling out
differences between first- and second-language composing
features the oft-quoted pioneer study by Chelala
(1981)
This work is one of the first to have examined the
composing process of L2 writers.
After analyzing how two
'professional' Spanish-speaking subjects behave in the
process of composing written English discourse,
Chelala
arrives at a distinction between 'effective' and
'ineffective' writing behaviours, which includes the use
of Ll in the prewriting phase.
Having observed fewer
instances of reviewing and revising In the process of
composing in L2,
she concludes that L2 composing is
69

---

different from Ll composing.
Hall
(1990)
arrives at a
similar conclusion.
A stronger claim lS made by Raimes
(198Sb) who
expresses the view that
despite research findings,
all of us who have
tried to write something in a second
language ... sense that the process of writing
in an L2 is stardingly different from writing
in our Ll
(Raimes,
1985b:232).
This assertion lS clearly questionable when one
considers the case of those writers who produce written
discourse in three or more languages.
If such an
assertion were proven right,
the implication would be
that such writers would have to invoke composing
processes that are specific to each one of those
linguistic systems whenever such writers set out to
compose in any of those languages
(English,
French,
Italian and Spanish,
to name only these).
Findings invalidating Raimes's claim made in 1985b
are available in Raimes
(1987).
In this study as in most
studies on L2 composing,
differences and similarities
between Ll/L2 composing processes are observed, making i t
difficult to interpret and discuss the findings of such
studies.
One way of dealing with the findings of L2
composing research is to consider the principal criterion
for similarity and difference as suggested by Jones &
Tetroe
(1987).
In their own words:
70

---

It appears that studies that count quantity as
the principal criterion for similarity find
cross-language differences, whereas those who
look for patterns or quality find similariies
(Jones & Tetroe,
1987:39).
In their 1987 study on planning strategies in first-
and second-language, Jones and Tetroe observe that
'second-language composing is not a different animal from
first-language composing'
(Jones & Tetroe,
1987:55)
Further evidence of similarities between Ll and L2
composing strategies can be found in Edelsky
(1982),
Jacobs
(1982),
Lay (1982),
Zamel
(1982,
1983),
Gaskill
(1986),
Hall
(1987), Ardnt
(1987),
among many others.
There seems to be a general consensus among these
researchers that L2 subjects invoke the same 'cognitive
capacity'
(Jones & Tetroe, 1987)
required in performing
writing tasks in Ll.
Gaskill's
(1986)
comparative study
reports that identical revision strategies are deployed
both in English and in Spanish.
Hall
(1987)
and Arndt
(1987)
corroborate these findings.
The same processes underlying Ll writing are at work
In L2 writing
(Edelsky,
1982:227).
Zamel
(1982,
1983)
observes that advanced ESL writers,
like their native
counterparts, go through the same process of discovering
and creating meanlng.
She goes on to remark that these
'students
recognize the importance of being flexible,
starting anew when necessary,
and continuing to
71

---

rework their paper or time as they take into
account another reader's frame of reference
(1983:168) .
Two other issues related to that of the similarity
or identity of processes in the act of composing in L1/L2
are the transfer phenomenon and the interaction between
composing strategy and second language proficiency.
Jones
(1982)
addresses the issue of transfer of
composlng strategies from an L1 to an L2 composing
context.
This study of the composing behaviour of a
Freshman German speaker and a graduate Turkish student
shows that while the German-speaker subject is considered
as a
'good' writer, whose idea generation strategy leads
to text production,
the Turkish subject is labelled a
'poor' writer, whose writing strategy is text-bound.
The
findings are that writing strategies affect writers'
rhetorical structures.
The general conclusion is that
the 'poor' writer has never learnt to compose.
Implicit in this conclusion is that had the
'poor'
writer known what composing entails from his L1 composing
experience,
such knowledge would have been transferred in
the L2 composing experience.
Eldesky
(1982)
explicitly
expresses the view that
'knowledge about writing in first
language forms the basis of new hypothesis rather than
interfer6with writing in another language'
(Eldesky
1982:227).
In sum,
'writers will transfer writing
abilities and strategies, whether good or deficient,
from
72

---

their first language to their second language'
(Friedlander,
1990:109).
This transfer is independent of
other factors
(Jones & Tetroe, 1987; Mohan & Lo 1985,
referred to in Friedlander, 1990).
This last observation leads one to consider the
second issue which suggests a compensation hypothesis to
account for the interaction between strategy and
proficiency.
As Jones & Tetroe (1987) put it,
in the case of composing it might be that
second-language writers with effective
strategies are able to use those strategies to
compensate for the limitations imposed by their
imperfect knowledge of the language.
We would
expect to find little difference between the
composing process in the first and second
language in this case.
At the other extreme,
those with poor soategies may find that the
burden of second-language composing imposes
such additional constraints that their second-
language processes are markedly different from
those in the first
(Jones & Tetroe,
1987:36).
This point lS well taken and is worthwhile pursuing
if cross-language composing research is to develop
clearly articulated lines of inquiry.
One would wish to
take this issue a step further,
arguing that unless
researchers in the field of composing,
whether in a
first-,
second- or third-language,
adopt a broader
theoretical perspective that would construe composing In
relation to other cognitive problem solving,
investigations in composlng are bound to come up with
contradictory results likely to cause confusion rather
73

---

than shed light on the process of composlng written
discourse.
Commenting on differences among findings resulting
from L2 composition research,
Krapels asserts that
contradictions in second language writing
research ... may result from premature
generalizing on the part of researchers.
Case
study research usually yields data and findings
with limited generalizability, given the fact
such research
permits few subjects and often
precludes randomization of subject selection
IKrapels,
1990:50).
As stressed by Krapels
(1990),
contradictions among
L2 composition research findings clearly stem from
methodological perspectives.
Both the case-study design
of and the relatively small number of subjects involved
In those studies make it difficult to derive conclusive
generalizations
(Raimes,
1985b).
Another source of
contradictory results may lie in the underlying
assumptions of such investigations.
Indeed,
'ESL
students
[have been discussed] as if they were a
homogeneous group'
(Raimes,
1987:40).
Differences In
their academic background or educational experience are
disregarded,
as is the case in Jones
(1982).
A third source of contradiction stems from a lack of
shared understanding of the terminology
(viz unskilled vs
skilled: expert vs novice; basic writers and other
dichotomies)
used by most researchers.
Raimes
(1985)
lS
74

---

critical of the lack of rigour in the use of concepts
describing subject-writers.
To use her words,
there is no consensus on valid criteria for
measuring skill in writing and thus no clear
agreement on the meaning of unskilled.
Any
examination of unskilled writers must therefore
clearly address the question,
unskilled
relative to whom and according to what
criteria?
For generalizations from several
studies can be made only if the same criteria
are used
(Raimes,
1985b:231-232).
A fourth source of contradiction derives from the
level of interpretation of findings.
A distinction 1S to
be made between group-level and individual-level
findings.
Sources of contradiction in findings may be
ironed out if the sample is increased but above all if
individual differences are taken into account.
Such a
direction is worth considering as
'composing strategies
of each individual writer [are]
found to remain
consistent across languages'
(Arndt,
1987:257).
In addition to the need to attend to individual
differences so as to arrive at a set of composing styles
that can be generalized to a wider population,
the need
for considering factors likely to affect composing
processes cannot be overemphasized.
This issue is
examined in the next section.
75

---

2.2.2.
Empirical studies with focus on factors
affecting composing process
Three factors have been identified to have some
impact on the composing processes and performance of L2
composers.
These are:
(i)
subjects'
first language;
(ii)
familiarity with tasks; and
(iii)
task characteristics.
SUbjects have been observed to incorporate L1 into
the process of composing English written discourse.
This
phenomenon occurs irrespective of the sUbjects'
linguistic background (for instance,
chinese,
spanish,
Francophone Canadian).
Lay (1982), Martin-Betancourt
(1986),
Hall
(1987),
Jones & Tetroe
(1987)
and Friedland
(1990),
among others, have observed that L2 writers
resort to L1 to generate content.
Incorporating L1 into
L2 composing acts both as a facilitating and an
inhibiting factor.
It is to be recalled here that Chelala
(1981)
has
classified first-language use for prewriting in L2
composing into the 'ineffective strategies'.
Additional
evidence depicting the use of L1 as negative in the
composing process of L2 is offered by Zamel
(1982) who
reports that although her 'most proficient' subject-
writer activated translation as a composing strategy, her
other sUbjects in the study objected to such an approach,
one of them alleging that ' i t would be like being pulled
by two brains'
(p.
201).
Such an observation contradicts
findings supporting the view that resorting to L2 to plan
76

---

content leads to improved planning and thereby to better
quality product
(Friedland,
1990).
Though the studies above attest to the occurrence of
Ll in the composing process of L2 writers,
the role and
impact of Ll,
as a linguistic system,
is yet to be
proved.
Further studies are needed to ascertain whether
using Ll in the process of composing L2 is facilitating
or inhibiting.
What is also left to be determined is the
profile of those L2 writers who activate Ll in their
process of solving L2 writing tasks.
While one would
adhere wholeheartedly to the hypothesis that positive
transfer of efficient problem solving strategies from L1
into an L2 composition process is conducive to
enhancement of both L2 composing process and product,
one
is yet to be convinced that Ll content generating during
planning necessarily results in improved process and
product.
Familiarity 1S another factor that 1S related to the
composing process of written discourse.
Task familiarity
is considered to be the most critical factor in the act
of comprehending
(Tedick reporting Benhardt,
1986c).
A
satisfying way of examining the effect of familiarity
with tasks is achieved when this 1ssue is replaced,
as
did Tedick
(1988)
in the broader theoretical perspective
'based on the following axes:
(1)
cognitive psychology
(Bruner,
1956; Piaget,
1954; and Vygostsky,
1962);
(2)
constructivist theory (Rumelhart & Ortony,
1977;
77

---

Brandsford,
Barclay & Franks,
1972);
(3)
reconstructivist
theory
(Bartlett,
1932; Rumelhart,
1977; Spiro,
1977 and
(4)
schema theory (Anderson,
1977).
Such studies have enlightened research both in
reading and writing.
Although research in reading
investigating the effect of familiarity with task,
or
background or metacognitve knowledge,
or schemata on
comprehension abound
(see
Polanyi,
1958; Carell,
1983,
1984a,
1989a; Gass and Varonis,
1984),
writing research
examining the relationship between writing and
familiarity with tasks are scant.
Among the limited few
in Ll are Langer (1984),
Cheskey
(1984,
1987),
dealing
with secondary school subjects and McCuthen
(1986),
concerned with children.
Instances of L2 writing
research in this area are represented by Winfield and
Barnes-Felfeti
(1982),
Purves & Purves
(1986),
Tedick
(1988)
and Friedlander (1990).
The opposing pairs that are considered in these
investigations are high-/low- prior knowledge,
domain or
field-specific/general prompt,
and familiar/unfamiliar.
Other variables that are considered with respect to the
familiarity effect are rhetorical knowledge
(Bereiter &
Scardamalia,
1987;
(Jones,
1982) or discourse mode
(Matsuhashi 1979,
1981; Matsuhashi & cooper,
1978).
Background reading is a candidate variable suggested by
Campbell
(1990).
L2 studies investigating
78

---

the effect of task familiarity on composing have mainly
dealt with cultural knowledge.
Tedick
(1988), who is an exception,
is primarily
concerned with the effect of topic familiarity in
assessing written products.
Although the main concern of
Tedick
(1988)
differs from that of this investigator,
Tedick's study is worth mentioning as it is one of the
few L2 studies to have addressed the relationship between
familiarity with task and composing process and product.
Both L1/L2 studies that have examined the effect of
prior knowledge on the composing process/product have
arrived at parallel findings.
Subjects performed better
on writing tasks that are familiar.
Implicit is the
suggestion that familiarity with tasks affects the
composing process.
An interesting explanatory hypothesis
is suggested in McCuthen
(1986) who posits that in the
absence of domain-specific or appropriate general
knowledge needed to solve writing tasks, sUbjects will
invoke powerful composing strategies to compensate for
knowledge failure.
This hypothesis is disputed by Tedick
(1988) as seen in the discussion of the findings of this
study in Chapter six.
The discussion in this subsection has considered so
far two of the three variables
(L1 in L2 composing and
familiarity with the task) hypothesized to exert an
effect on the composing process, the third variable is
left to be examined.
79

---

The third factor influencing the composlng process
is task characteristics.
Studies that have investigated
directly or indirectly task-related influences
(Hamp-
Lyons,
1988) abound if all the variable elements making
up the complex factor of task
(Hamp-Lyons,
1990) are
taken into account.
The constituent elements of task
characteristics include the following:
- the writing tools
(use of paper and pen,
type-
writer,
or word processor,
as suggested by Hamp-
Lyons
(1990~;
- discourse purposes
(Matusuhashi,
1979,
1981)
- organization principles such as compare/contrast
(Langer,
1984);
- amount of time for writing
(Krashen,
1977,
1981;
Kroll,
1990); and
- essay question.
Investigating the effect on individual task
components on the composing process and product is
unquestionably informative.
For instance,
there is
evidence that the selection of the essay question is a
significant task variable
(Hamp-Lyon$, 1988).
Langer
(1984)
indicates that compare/contrast writing tasks are
likely to produce better structured written products.
The difficulty with investigations studying the
relationship between the composing process and task
characteristics is that, by adopting an atomistic
80

---

perspective, they fail to identify the effect of
confounding variables.
In the discussion of. the relationship between the
composing process and tasks,
it might be more productive
to examine the factor task as a bundle of components.
This could be achieved by defining a matrix for task
demands as discussed in Chapter 3,
section 3.2.2.
It may
be essential to consider task-related influences on
composing written discourse through a comprehensive
theoretical perspective, coupled with the appropriate
statistical techniques likely to identify, among the many
task characteristics, which ones contribute most to
explaining the interaction between tasks and composing
processes.
This suggestion is taken up and further
developed in the discussion of the findings in Chapter
six.
2.3
Concluding remarks
At the end of this selective review of the
literature on L1/L2 composing research, we shall now
pause and reflect on the general lines of inquiry that
can be derived from this survey. This review of studies
permits a number of general observations. To begin with,
what Johnson-Laird (1983)
said about the mind is perfect-
ly applicable here. Just as the 'mind is too complicated
81

---

to be seen clearly, or to be studied with advantage,
from
the perspective of a single discipline'
(Johnson-Laird,
1983:XI), so is the process of composing written
discourse.
Research questions raised by investigators in
written composition seem to call for the need for a
broader theoretical perspective that would consider the
process of composing written discourse with respect to
other cognitive problem solving activities.
Indeed,
implicit in the discussion of L1/L2
composing research is the same question that is raised
when the dichotomy LTM/STM is considered
(as seen in
Badde1y, 1976) or when the process of reading
comprehension in L1/L2 is examined (Freddle,
1985;
Kintsch & van Dijk,1978; Carson, 1990).
The question is
whether there are separate sets of cognitive processes
underlying L1 composing and L2/L3 composing or only one
set of cognitive processes functioning in two,
three or
more composing environments.
composing research findings seem to suggest both a
unitary and a separate view of the composing processes in
L1/L2.
The unitary view obeying the principles of
interference theory (McLaughin,
1987; McDonough,
1987;
Ellis,
1985; Gass, 1984) shows a range of similarities
between L1/L2 composing processes, while the separate
view highlights differences between these processes.
But
before any claim can be made, controversial findings
suggesting both similarities and differences in L1/L2
82

---

composing processes are to be re~examined in terms of the
unitary/separate perspective.
Additionally,
such a
perspective should be useful in accounting for both the
interaction between composing strategy and second
language proficiency and the compensation hypothesis that
effective problem solving strategies will compensate for
'imperfect'
linguistic knowledge or lack of readily
available field-specific knowledge.
Two other central issues in composing research
worth emphasizing are:
(1)
the relationship of reading
and writing and/or writing and learning;
and
(2)
the view
of the process of written composition that integrates
process and product.
First,
the relationship between
reading,
writing and learning cannot be overemphasized.
In these three instances,
the learner,
the reader and the
writer,
who may be the same subject,
concurrently
monitors the operations of acquisition,
transformation
and evaluation of knowledge by judiciously selecting and
establishing connections among the different components
of the material being manipulated.
Secondly,
it lS a
commonplace to state that focus on the composing does not
pre-empt incursions into the written product.
One way of
establishing a bridge between process and product lS
suggested by Witte & Cherry (1986)
who posit that the
complementary processes of framing and
topicalization
(can be viewed]
as exceedingly
rich constructs that help to establish a
83

---

connection between production processes and
written products
( Witte & Cherry,
1986:142)
To conclude this chapter whose main object has been
to review research studies on the process of composing
written discourse in L1/L2 so as to situate the present
study, attention is now given to the assumptions and
research questions that underpin this investigation.
The
assumptions can be formulated as follows:
Assumption I:
Composing written discourse involves
the same processes
(such as acquisition,
transformation and evaluation)
that are at work in
the act of learning.
Assumption 11:
Composing written discourse can be
seen as a cognitive problem solving activity.
Assumption Ill:
Composing written discourse entails
the same cognitive processes
(such as problem
finding,
problem representation, planning of the
solution,
evaluation of the solution and
consolidation of gains)
that are activated in the
process of solving cognitive problems.
Assumption IV:
The processes of revision and
evaluation are interdependent.
The evaluation
process which occurs thoughout the act of composing
involves reformulation subprocess with potential
checkpoints defined as incubation,
illumination and
verification.
84

---

The research questions that were stated in Chapter
one are re-examined here with respect to the need for
further research as arisen from this review of the
literature.
The first question to be put forth is:
Assuming that the processes of planning and evaluating
are instrumental in the act of composing written
discourse, what is the interaction between these
processes?
Another question is whether these processes
co-occur In the composing process irrespective of task
characteristics and subject characteristics.
A third
question is:
Assuming that subjects respond differently
to task demands,
what characterizes inter-subject
variation?
These research questions resulting from our reVleW
of the literature on composing are reformulated and
operationalized so as to be statistically tested.
The
reformulation of and the procedure for the quantitative
and qualitative examination of these questions constitute
the concern of Chapter three.
85

---

CHAPTER THREE
86

---

3.
Methodology
This chapter comprises three main sections:
(1)
introduction,
(2)
methodology,
and
(3)
qualitative
analysis.
The introductory section
(3.1)
covers three
points:
(i)
the general perspectives and academic context
of investigation
(3.1.1);
(ii)
the theoretical framework
underlying the methodological lines of inquiry
(3.1.2);
and
(iii)
the re~statement of the major research
questions and of the hypotheses
(3.1.3).
The
methodological section
(3.2)
is articulated so as to
include the following:
Subjects
(3.2.1); Data
collection:
Instruments
(3.2.2); Experimental procedure
(3.2.3); and
(3.2.4)
Statistical procedure.
The third
main section
(3.3), which deals with the qualitative
dimension of the study,
defines how the data in this
study has been segmented (3.3.1) and the attempt to
arrive at a typology of writers
(3.3.2).
Each of these
main points
lS
examined in turn.
87
-.

---

3.1.
Introduction
As indicated earlier,
the object of this section is
to define the general perspectives underpinning this
study,
the academic environment in which the experiment
has been carried out,
the theoretical framework of the
inquiry,
and the general and specific research questions
that are being addressed.
Each of these 1ssues 1S
attended to below.
3.1.1.
General perspectives and academic context of
inquiry
General perspectives
This study makes use of both a quantitative and a
qualitative perspective.
In adopting this dual
perspective,
it is hoped to provide a full account of the
process of composing written discourse,
including the
time when a subject develops awareness of a problem to be
solved to the solution of that problem.
Such a
perspective necessarily gives equal emphasis to the
composing process and to the written product.!
Only by
adopting this line of inquiry can one help subject
writers develop a'ltonomy under varied composing
IThe need to give equal emphasis to both the composing process
and the product is shared by most researchers calling for the need to
reconcile product and process
(see for instance Horrowitz,
1986; and
Cannar,
1987).
88

---

circumstances In general,
and under examination
conditions in particular.
In this respect,
there has
been an attempt,
in the context of gathering the data for
this study,
to reproduce a relatively stressful exam
condition familiar to those subjects.
It is to be
specified here that although focus IS on the composing
process In its entirety,
including both process and
product,
gIven the constraints of any research activity,
a discourse analysis of the product has not been
performed in this study.
Attention has been gIven to
identifying the cognitive processes that are activated
throughout the processes of solving four writing tasks
and to discovering the relationship between performance
on the writte~ products and such cognitive processes.
It is also to be stressed that this study is not an
experimental one in design.
No particular treatment has
been administered to subjects with a view to comparing
the
effect of that particular treatment on the behaviour
of an experimental group, using a control group as the
basis for comparison.
Instead,
it is empirical and more
of an observational type rather than experimental.
The
data are experimentally elicited,
as opposed to
naturalistically
(Garman,
1990).'
ZThe experimental technique of eliciting the data for this study
may have an effect on the s~bjects' composing process.
An objection
may be
levelled against the representativeness of such experimentally
elicited data on the account that
'writing under pressure is a very
unnatural situation'
(KraIl,
1990:141).
One might argue,
however,
that eVen under the mosL naLural writing conditions approximating
writing in real life external
factors such as deadlines
to meet do
influence the writing process to some extent.
89

---

To end the discussion of the general perspectives of
this investigation,
it is worth pointing out that this
study is,
in the main,
concerned with ten case studies
for the reasons provided in the discussion of the sample
size.
The discussion of the general perspectives leads
one to depict the academic context.
This is the concern
of the next sub-section.
Academic context of investigation
This study is concerned with Ivorian first year
University students majoring in English as a Foreign
language (EFL).
Instead of undertaking a study at the
University level or at Faculty level,
this research work
is deliberately limited to the Department of English for
two related reasons:
(1)
The researcher is a member of
Staff of that department;
(2)
Interest in such a study
arises from her experience with students majoring in
English as a Foreign language and from her interest in
discovering the nature of University students'
learning
difficulties.
Academically,
undertaking this study at this level
and with such a group of students seems more than
justified.
Indeed,
most work in the field of English
Composition has dealt either with native English speakers
or with students speaking and writing in English as a
second language,
and studying in the second language
90

---

environment.
Studies related to English Composition In
an EFL or an L3 environment are scarce.
Another reason for attempting to understand the
nature of the writing difficulties experienced by EFL
students is that one will not get to the heart of the
matter unless one strikes the problem at its roots.
It
seems appropriate to deal with students entering
University for the first time.
In so doing one lS likely
to arrive at an uncontaminated characterization of the
composing process of first year University students.
Having described the general perspectives and the
academic context of this research study,
the next step is
to set up the theoretical framework underpinning the
methodological lines of inquiry.
3.1.2.
Methodological lines of inquiry:
Theoretical
framework
In the following,
a reVlew of studies informing the
methodological perspective of the present investigation
is undertaken.
Such a review mainly deals with studies
that examined pausal phenomena in the production of both
spoken and written discourse.
91
'.

---

Pausal phenomena in spoken discourse
In discussing studies on pausal analysis,
an
historical perspective is adopted to identify the salient
changes that have occurred in this research area.
The review of the literature by Maclay and Os good
(1959)
and by Rochester
(19'13)
clearly suggest the
following:
(1)
Concern for pausal phenomena can be traced
back to the early 1930s;
(2)
Even though reference to
pauses and silence was frequent in linguistic
investigations, pausal pheno~ena did not constitute a
maJor area of research in linguistics and psychology
during the first half of the twentieth century;
(3)
Linguists were at variance as to the value of pausal
phenomena in linguistic analysis.
Pike
(1945) and Block
(1946)
stood out as exceptions
for having undertaken a systematic analysis of such an
aspect of speech.
Pike,
as suggested by Rochester,
was
the only linguist then to have asserted that
'pauses and
other prosodic features are not incidental.
These are
necessary to linguistic description')
(Rochester,
1973;52).
Pike,
undoubtedly,
struck a discordant note
with respect to both his predecessors and successors.
Block,
on the other hand,
was less assertive.
Despite
his systematic use of pauses in his linguistic analysis
of Japanese,
he maintained as the other linguists did
that
pauses had an optional status.
Yet, his dyadic ranking
JErnphases added.
92

---

of facilitative pauses into 'higher,
more constant'
and
'lower, more constant',
added to his
'minimal pause
group' as one criterion for
defining words ln Japanese,
allows one to remark that the position held by Block was
closer to that of Pike.
In the main,
pausal phenomena were considered,
up to
the early 1950s,
as
useful but not necessary . . . .
[They]
function
as non-significant events which may serve to
identify linguistically relevant units,
such as
junctures located at the boundaries of
phonemes,
morphemes,
words,
phrases,
and
sentences
(Maclay & Osgood,
1959: 20).'
Pauses were initially both negatively and positively
defined.
The terminology depicting pauses negatively
offers an interesting illustration of the situation.
Indeed, pauses were said to characterize the
'intermittent quality of speech'
(Rochester,
1973)
They
appear as
'disturbance in speech'
(Mahl,
1956)
or
'disjuncture'
(Brown and Miron,
1971) or as hesitations.
Hesitations interrupt the flow of speech while juncture-
pauses do not
(Lounsbury,
1954).
A positive definition of pausal phenomena is based
on criteria such as duration,
function,
and location.
A
definition which emphasizes duration is proposed by
Maclay & Os good
(1959).
Such a definition views
'This position is shared by B1oomfie1d
(1933) and Harris
(1951).
93

---

respectively hesitation as
'events
[that are]
relatively
gross and easily observable' and juncture-pauses as
'shorter in duration,
harder to observe and record'.
A
functional definition construes juncture-pauses as
listener-oriented while hesitation-pauses are speaker-
oriented
(Barik,
1968; Boomer & Ditmann, 1962; Garman,
1990).
Hesitation-pauses occur at points of lowest
transition probability (Maclay & Osgood,
1959).
They
mark off the beginning or end of speaker's units and
signal decision-making operations.
If up to the early 1950s, pausal phenomena were
considered with mixed feelings,
the significance of
pauses in the sLudy of speech became self-evident in the
second half of the twentieth century both in linguistics
and psychology.
The validity of pauses in speech as an
index of unit boundaries in language was established ln
the mid sixties.
Suci
(1967)
is a case in point.
The significance of pauses in psycholinguistics is
re-asserted by Rochester in his 1973 review of two
decades of research.
In that review,
Rochester views
pause analysis as an answer to a question Lashley posed
in 1951.
The question was concerned with how language
production is transformed into external temporal
sequence.
From Rochester's perspective, pause analysis
seeks to understand the decision-making or the cognitive
planning processes underlying speech production.
94

---

In their attempt to capture the cognitive planning
processes that are activated in the process of speech
production,
researchers have moved from a
'microstructure' perspective to a
'macrostructure'
perspective,
as indicated by Butterworth (1975).
'Microstructure'
studies,
essentially undertaken during
the first phase of pause analysi~ were mainly concerned
with defining and categorizing pauses with respect to
location,
duration,
and function.
'Microstructure'
studies have also been interested
In classifying and providing a physical description of
hesitations-pauses.
Hesitation-pauses are subdivided
into silence or unfilled pauses and filled pauses.
Filled pauses are further subcategorized into
repetitions--non-significant semantically: phoneme/word,
'I mean',
'sort of',
'well',
'you know'-- filled pauses--
e.g. Ah,
er,
um,--and false starts
(Lounsbury,
1954;
Maclay & Osgood,
1959; Tannenbaum, Williams & Hillier,
1965; Beattie, 1981), or 'mazes'
(Miller,
1987)
or 'non-
fluencies'
(Garman,
1990).
The primary concern of 'microstructure'
studies was
the 'search for units of encoding', which addressed the
general issue of the relationship between location and
function of pauses in speech
(Rochester,
1973).
Three
models were subsequently developed:
(1)
linear model:
lexical encoding hypothesis
(Goldman-Eisler,
1958a,
1958b);
(2) hierarchical model:
grammatical encoding
95
'.

---

hypothesis
(Boomer,
1965; Barik, 1968);
(3)
molar model:
semantic decision-making hypothesis. with respect to the
search for encoding units, studies related to the
semantic or 'content' hypothesis are worth mentioning.
These are Taylor (1969); Wilkes & Kennedy
(1970); and
O'Connell, Kowal & Hormann (1970).
The critical review
of the theoretical and empirical issues related to these
studies can be found in Boomer (1970)
and in Rochester
(1973) .
Unlike micro-level studies whose emphasis was on
'each individual pause,
its duration and location,'
macro-level studies have shifted their focus on 'overall
proportion of pausing in speech'.
Such research work has
shown, as reported by Butterworth (1975),
that several factors are associated with
variations in this proportion,
and these
reflect aspects of the cognitive process which
are concerned with content rather than
structure (Goldman-Eisler, 1961, 1968, Chap 4;
Lay and Paivio,
1969; Reynolds and paivio,
1968)
(Butterworth,
1975:76).
A finding worth highlighting is the evidence that
'in samples of spontaneous speech hesitant periods
predominantly of silence alternated with event periods
predominantly of phonation'
(Henderson, Goldman Eisler &
Skarbek (1966) reported by Butterworth, 1975).
At this point a question to be posed is the
following:
What can be retained from more than three
decades research activity on pausal phenomena in spoken
96

---

discourse?
Four points of maJor interest to this study
are worth being made.
Firstly,
there is hard evidence that pauses reflect
decision-making or cognitive processes underlying speech
production.
Secondly,
hesitation-pauses indicate initial
as well as final boundaries of speech units.
Thirdly, an
implication resulting from Rochester's
(1973)
review is
that planning in speech production is a
'multileveled
decision-making process'.
Fourthly,
as suggested by
Henderson et al.
(1966),
the episodic structure of speech
production exhibits a cyclic structure in that
hesitation-pauses and phonation alternate.
Such findings are crucial in determining the
orientation taken in this study.
Indeed,
in giving
special attention to pauses occurring in the composing
process,
it is hoped to capture most of the cognitive
processes that are activated throughout the process of
solving a writing task,
from the preparation process or
advance planning to the completion of the written
product.
Despite the common theoretical framework between
studies on pausal analysis in spoken discourse and the
present investigation,
one major difference is the
research data.
Pauses in conversational analyses consist
of filled and unfilled pauses in oral production.
By
virtue of its nature,
a study that is concerned with the
process of composing written discourse cannot solely
97

---

depend on pauses in oral production.
Research work on
composlng
has had to resort to overt signals of
cognitive planning.
This point is to be developed in the
following discussion of studies related to the composing
process of written discourse.
Pausal phenomena in written discourse
Since the studies examined in the previous section
are the very same that have informed research on
composing written discourse,
the question to be raised
here is, \\"hat differentiates the present investigation
from other research studies on composing?
To answer the above question, an examination of
three influential lines of inquiry will be undertaken,
highlighting differences and similarities between these
research studies and the present one.
The studies to be
considered are:
(1)
Bereiter & Scardamalia
(1981,
198'7),
(2) Matsuhashi
(1979,
1981); and
(3)
Flower & Hayes
(1981); and Swarts,
Flower & Hayes
(1984).
This study differs from Bereiter and Scardamalia
(1981,
1987).
In analyzing conceptualization in the
composlng process of children,
Bereiter and Scardamalia
focus exclusively on advance
planning.
This
investigator gives attention to conceptualization before
and during the translating process of the composing
process.
By their focus on advance planning and on
98

---

children,
the studies by Bereiter and Scardamalia
distinguish themselves from Matsuhashi's work.
Matsuhashi
(1979)
is the first study to have
systematically examined pausing in the process of written
composition of adult writers.
Matsuhashi,
adopting a
discoursal perspective, developed a seven level scheme to
code specific linguistic features.
These levels were
described as follows:
1. Abstract level--defined as possible hierarchical
relationships between T-units--based on a discourse
matrix devised by Nold & Davis
(1976);
2.
Sentence role--defined as semantic role--
developed by Cooper and Matsuhashi for the purposes
of Matsuhashi
(1979).
Five sentence roles were
identified:
generalizing,
rhetorical,
sequence,
relationship,
and development;
'3.
Paragraphs signalled by indentation or any
markers;
4.
Initial modifying structure;
5. Lexical cohesion--based on Halliday & Hassan
(1976) ;
6. Syntax; and
7. Content words.
This scheme which analyses pauses during the act of
writing,
by secondary school students,
is worth
mentioning because it is the first of its kind to have
been used to test empirically the effect of discourse
. purposes on productionS.
This scheme is not,
however,
5Note that in the present study discourse types
is used instead
of discourse purposes as understood in Matsuhashi
(1979,
1981).
99
'.

---

totally suitable for the purposes of the present study.
By concentrating only on pauses during continuous
writing,
such a scheme could be partially applicable to
this research.
That scheme is more applicable to a product model of
composlng rather than to a cognitive process model of
composing that seeks to capture planning and decision-
making throughout the composing process.
Specifically,
such a coding tool does not permit an adequate
representation of the preparation process,aprocess which
is mainly concerned with developing awareness of a
problem,
formulating the problem,
framing the task,
and
with generating ideas.
Neither does such a scheme allow
one to account for the organization process during which
ideas are formulated and reformulated and during which
the appropriate discourse frame is selected nor does i t
accommodate
the production of different drafts.
The methodological influence of Matsuhashi
(1979) on
the present investigation can be seen in its adoption of
video protocol as a technique of gathering data on the
composing process.
Another point of influence is the use
of pauses in scribal activity as overt indications of
cognitive planning and decision-making.
Matsuhashi
(1979,1981)
inspired the work of Flower &
Hayes
(1981) which has attempted to analyze the nature of
planning at the points of pauses In manual activity.
To
achieve their purpose,
Flower and Hayes have resorted to
100

---

a content analysis of 'think-aloud' verbal protocols.
They were thus able to establish the psychological
reality of composing episodes.
To sum up,
Flower & Hayes
(1981)
depend essentially
on pausing as revealed by verbal protocols.
Matsuhashi
(1979,
1981)
considers pauses In manual activity during
text generation.
What is being considered here lS
latency as revealed by a delay between time of
verbalization of thought or complete silence and the
beginning or resumption of manual activity or production
period.
A detail worth emphasizing is that in this
particular instance pauses in manual activity may be
matched with unfilled pauses
(complete silence)
or with
filled pauses such as hesitations, repetitions, false
starts in verbalized thought.
In the main,
pauses In
manual activity are instances of latency indicated by a
delay between verbalized thought and the beginning of or
resumption in manual activity.
Another discriminating point between Matsuhashi
(1979,
1981)
and this study is that while the former
considers pauses at seven different levels of discourse,
the latter examines major pauses during the composing
process.
Specifically,
this study has sought to examine
pauses occurring at initial boundaries of major composing
episodes.
An episode can be perceived as a unit of encoding in
written composition.
Smith (1982)
defines an episode as
101
.'.

---

•
the period of time during which an act takes
place,
a
"writing episode",
during
which a
writer's hand is on the
move,
when there is no
manifest pause for reflection,
and any
prewriting or rewriting that is being done
occurs simultaneously with the act of writing
itself ....
Writing episodes occur
intermittently, although thought is
continuous . . . .
Usually there are pauses of
varying lengths between episodes,
because the
writer's thought is reorganizing for
forthcoming episodes or because of extraneous
interruptions or distractions
(Smith,
1982:
105,
107).
Smith's v~ews of composing episodes is similar to
the definition given by Flower and Hayes in that they all
stress that composing episodes are 'units of
concentration' or 'periods of sustained focus'
in the
composing process.
Clearly, composing episodes are
defined with respect to the writing process and not In
terms of the product.
As evidenced by the data under
examination, major cognitive processes such as planning,
translating and reviewing,
as defined by Flower & Hayes
(1981), may and do co-occur within one single composing
episode.
What is of interest for the present study is the
dichotomy provided by Flower & Hayes
(1981).
The
distinction between major composing episodes and minor
composing episodes seemed productive for discriminating
high-levelled and low-levelled planning.
Major composing
episodes are 'clearly autonomous episodes with strong
boundaries', while minor composing episodes have 'weaker
boundaries
[,
with]
stronger connections to adjacent
102

---

episodes' (Flower & Hayes,
1981:234).
In sum, maJor
composlng episodes are relatively independent from other
composing episodes while minor composing episodes are
subordinate to others and function as sub-episodes within
the maJor ones.
Emphasis is here on pauses occurring at
the initial boundaries of major composing episodes.
Such a focus on initial boundaries pauses
necessarily excludes other kinds of pauses occurring
within composing episodes,
or between sentences,
or at
clause level as well as false starts occurring usually
after a major pause and overtly signalled in verbal
protocols by a shift or restructuring move.
Another type
of pauses that is not accounted for are long pauses
occurring after the first two,
three or four words
(e.g.
'It is ... '
or a noun phrase) marking off the beginning of
a major composing episode.
These usually occur after a
major pause.
One might conjecture that these pauses
indicate incompleteness of previous planning activity.
Although such phenomena might be interesting in
showing how writers move from planning to translating
and/or to reviewing,
these will not be discussed in this
study.
One may wonder why the present study is restricted
to pauses occurring at initial episode boundaries.
Such
'a decision is motivated by Flower & Hayes
(1981)
who make
the following assertions:
103
.<

---

1.-
boundaries are pivotal points at which
writers are breaking concentration and changing
focus,
they are likely to be decision points in
the writer's ongoing planning process.
2.-
[Initial episode boundaries]
appear to be the
source of many of the longer, more tantalizing
"pregnant" pauses others have observed.
3.-
Major episode boundaries have a high intuitive
discernabi1ity.
The selection of minor or sub-
episode boundaries will be more idiosyncratic since
readers respond to a wide range of events such as
shifts in topic, and the process shifts between
planning,
translating,
and editing,
intrusion of
metacomments
(Flower & Hayes,
1981:238).
If these observations are right,
it seems reasonable
to consider major episode boundaries to the exclusion of
minor episode boundaries
(see section 3.3.1 for the
identification of the episodic structure of the composing
process) .
The assumptions arising from this review of studies
could be summarized as follows:
Assumption I:
Conceptua1ization is inherent to
composing.
Pauses are overt manifestations of
conceptua1ization and planning and decision making
during latency.
The location and frequency of
pauses
(non scriba1 activity)
In the composlng
process is indicative of concern for planning in the
composing process.
Assumption 11:
At least,
three major levels of
planning and decision-making can be identified:
(1)
104

---

conceptual planning,
(2)
semantic planning and (3)
syntactic planning.
Assumption Ill:
Pausing and scribal activity
alternate during the process of composing
written discourse.
Assumption IV:
The quality of the introductory
paragraph of the written product reflects the
quality of conceptualization prior to text
generation.
These assertions have led to the formulation of the
major research questions and the specific hypotheses to
be stated below.
3.1.3.
Re-statement of major research questions and
hypotheses
Major research questions
This study seeks to examine the phenomena of
conceptualization,
planning,
and decision-making in the
process of written composition.
It seeks to examine
whether such
phenomena are localized to a specific point
in time during the composing process or whether this lS
evenly distributed throughout the composing process.
105
.'.

---

Specifically,
i t seeks
(1)
to discover the
relationship between the quality of the written ~roduct,
the composing process,
discourse type,
and familiarity
with tasks; and
(2)
to explore to what extent the
quality of conceptualization prlor to the translating
process can help predict the quality of the final
product.
Hypotheses
1.
The advance planning pattern of unsuccessful
subjects will be the same irrespective of discourse
type and degree of familiarity with the task,
whereas the planning pattern of successful subjects
will vary with discourse type and degree of
familiarity with the task.
2.
Assuming that advance planning plays a central
role in the composing process,
subjects who spend
more time On advance planning will score higher On
their written products than subjects who spend less
time.
3.
There lS a relationship between the duration of
the composing process,
as measured by time on the
preparation
(P)
and translating
(T)
processes and
familiarity with the task rather than with discourse
type.
4.
There lS a relationship between the quality of
the written product and the composing process,
as
106

---

measured by the duration of P and T, with respect to
discourse type and the degree of familiarity with
the task.
4.1. The quality of the written product is
associated with time spent on pausing during the
translating process.
4.2.
The quality of the written product is
associated with familiarity with the task rather
than vlith discourse type.
5.
The quality of conceptualization,
as measured
by an independent assessment of introductory
paragraphs,
rather than time spent on planning, will
constitute an index of predictability of the overall
quality of those products.
To sum up,
the previous main section has been
concerned with clarifying the general perspectives and
academic context of the investigation,
articulating the
theoretical perspective underlying the methodology in
this study,
and re-stating the general and specific
research hypotheses.
The next section presents the
specific methodological axes that have been used.
107

---

3.2. Methodology
Four points make up this section.
These are:
(1)
Overall presentation of subjects (3.2.1);
(2)
Data
collection: Instruments
(3.2.2);
(3)
Experimental
procedure
(3.2.3)
and (4)
statistical procedure (3.2.4).
3.2.1.
Subjects:
Overall presentation
The overall presentation of the sUbjects involved in
this study deals essentially with sUbject characteristics
and sample size and its justification.
Characteristics
As a rUle,
the bUlk of first year students at the
University of Abidjan is made up of students who are
Ivorian by birth or by adoption, whose age range is 18-24
years, with a mean age of 21 years upon entrance.
Academically,
all these students hold a secondary
school diploma such as the Baccalaureat6 or an
equivalent.
Most of the students admitted to the
Department of English have met the admissions
requirements as set by the Department7 •
strictly
~he Baccalaureat is the end of secondary school degree required
for admission to higher education.
1The
Department of English, National University of Cote d'Ivoire
offers the following courses: Linguistics,
Phonetics, Grammar,
Literature, Civilization,
Translation, English Composition, Reading
Comprehension,
Listening Comprehension,
Spoken English,
French
Literature and Grammar,
and a second foreign language
(German,
108

---

speaking, the Department admits only students who meet
the following requirements:
(1)
Secondary school graduates holding the
Baccalaureat, stream A2
(Philosophy,
French, two
foreign languages, History, Geography, Mathematics,
Physical sciences, and Physical education).
(2)
Such students should have scored at least the
following average grades in their final year in
Secondary School:
(i)
English:
a minimum average of
12 out of 20;
(ii)
French: a minimum average of lO
out of 20;
(iii)
Second language
(Spanish or
German):
10 out of 20.
Socio-economically, as revealed by questionnaires,
most students come from homes where the head of the
family is economically independent.
On the basis of
students' response to questionnaires, one can estimate
that three-quarters of the students are from an urban or
semi-urban home environment'.
To sum up,
one can remark that the population of
first year students can be considered as a relatively
homogeneous group with respect to age,
academic
achievement,
intellectual abilities and socio-economic
status (see Appendix I
on biographic and academic data).
Spanish or Port~guese}.
8T he
urban and semi-urban dichotomy ia suggested in Greenberg &
Bruner
(1966)
as a criterion for establishing a socio-economic
status.
109

---

The sample of subjects selected for this study was drawn
from that target popUlation.
Sample size and justification
Ten first year university students majoring in EFL
served as SUbjects.
These SUbjects constituted a
subgroup of a sample of 40 SUbjects who were involved in
a study that sought to examine the relationship between
writing competence and overall academic achievement
(manuscript, Manouan 1990).
The sampling procedure is
discussed in the section on the experimental procedure
(3.2.3).
Although previous studies such as Flower &
Hayes (1978,
1980, 1981)
and Matsuhashi
(1979,
1981)
that
have been concerned with analyzing the process of
composing written English discourse have dealt with only
four SUbjects, this study examined ten case studies. This
decision was statistically determined.
To obtain results
that would have some statistical significance,
i t is
advisable to work on a sample of at least 10 SUbjects.
Admittedly such a sample size constitutes a major
limitation of the quantitative dimension of this stUdy.
Clearly, practical issues,
as will be seen later,
overrode the principle of-generalization of the results.'
9This decision was made after consultation with the Department
of Applied Statistics,
University of Reading,
in July 1988.
110

---

Such a number can be easily handled when dealing
with a macro-level analysis that not only takes a
quantitative perspective but also considers the composing
process from a temporal dimension.
In selecting 10
subjects, this investigator intends to examine the
relationship between variables such as the quality of the
written product and the composing process,
defined in
terms of time on its different components--preparatory or
advance planning and translating processes--discourse
type and familiarity with the task.
The sample size will
be reduced,
however, to 4 sUbjects when undertaking the
qualitative analysis at a micro-level.
The selection
procedure is dealt with in section 3.2.3.
3.2.2.
Data collection:
Instruments
The instruments for collecting the data include the
following:
- Raven's
standard progressive matrices
(PM 38);
- Weir's test in English for educational purposes
(TEEP);
- Questionnaires and retrospective feedback;
- writing tasks; and
- setting and technical equipment .
. There was also a pilot study,
involving warm-up sessions.
Verbal protocols and written product resulting from the
protocols constitute the data for analysis.
I I I

---

Raven Standard Progressive Matrices
(PM 38)
Three versions of Raven's Standard Progressive
Matrices are available.
These are the advanced form,
PMAI-T, PMAII, the adult form,
PM 38, and the child form
PM 47.
In this study, the adult form of the Progressive
Matrices
(PM 38)
is used.
This test is a non-verbal
intelligence test, which seeks to measure the testee's
ability,
at the time of administration of the test, to
observe and reason upon non-specifically meaningful
geometrical figures. This test is organized on the basis
of relational reasoning.
Specifically, the testees' task
consists in perceiving the relationship among those
figures.
Those matrices illustrate sets of problems
deriving from general principles according to which non-
verbal tests that are likely to measure general
intelligence call on relations of eduction,
or
relationships among those figures
(see Appendix 11).
The test includes 60 problems, divided into 5 sets
of 12 problems each.
'Each problem consists of a design
or "matrix" from which part has been removed' 10.
Each
problem involves relating the 'geometrical figure which
completes a matrix of figures according to a logico-
mathematical rule which can be discovered by examining
the remaining figures'
(Evans & Waites,
1981:132). As a
rule, the testees observe the given figures,
taking into
account form and texture. They must capture spatial
JOIn Manual of instructions to the administration of the test.
112

---

relations related to the arrangement of the figures.
To
find the solution to a given problem, testees are to
identify the rule governing the set and they must use i t
as a reading frame
(see Appendix 11 for illustration).
Given that writing performance implies the ability to
establish relationships between spatial,
temporal and
linguistic elements, the selection of Raven's PM 38 seems
well-founded.
Raven's Progressive Matrices were selected because
they are the most well-known 'culture fair'
tests.
Additionally, they
provide a non-verbal series suitable for
measuring intelligence.
They can be given
successfully to almost any testee,
irrespective
of age,
linguistic ability, or physical defects
(Manual of Instructions,
Introduction).
Thus they may constitute a reasonable basis for
establishing the homogeneity of a group, despite Blum's
criticism that 'more than thirty years after the
pUblication of the test, unequivocal evidence of its
predictive validity is still lacking'
(quoted by Evans
and Waites, 1981:132).
Given the reasons stated above, Raven Standard
Progressive Matrices seem more than appropriate for this
study.
As a non-verbal test,
PM 38 allows the bypassing
of the linguistic problem that might hinder the
assessment of intellectual potentialities of individuals.
Besides, this test has been adapted by CIERPA (Centre
Ivoirien d'Etudes et de Recherches en Psychologie de
113
..

---

l'universite d'Abidjan) .
The three versions of the
Progressive Matrices discussed earlier, viz. PMAI-T,
PMAII-T, PM 38, and PM 47, have already been validated
against Ivorian populations.
specifically, PM 38 has
already been used with student populations similar to
those involved in the present study.
Finally,
PM 38
appeared suitable to help to ensure that those students
who had been selected for this study were comparable with
respect to general intelligence, and as such constituted
broadly speaking an intellectually homogeneous group.
In
other words, PM 38 could provide a reliable basis for
rUling out cognitive differences as a potential source of
variation among subjects in this investigation.
After the description of the Raven's Standard
Progressive Matrices, the next instrument to be described
is Weir's Test in English for Educational purposes(TEEP).
Weir's TEEP test: The Test in English for
Educational Purposes"
The TEEP test is meant to establish readiness of
non-native speakers of English to undertake academic
studies in British educational institutions,
at tertiary
level.
Specifically, this test provides 'a profile of
each candidate's proficiency in reading,
listening,
writing, and speaking'
(Weir,
1988:124).
"In weir (1983), this test is referred to as the Test in English
for Academic purposes
(TEAP).
In Weir (1988,
1990), the test is
known as the Test in English for Educational Purposes
(TEEP)~
114

---

This test is integrative in that it draws a close
relationship between three skills, reading,
listening,
and writing,
by setting writing tasks that compel testees
to exploit information drawn from reading and listening
tasks in producing a writing task. This test includes
three parts: two compulsory--papers I and II--and one
optional, paper Ill.
It includes a sUbject/specific
division by producing two versions of tests Paper 11 A
and Paper 11 B that are respectively suitable for non-
science and science students.
As reviewed by Pauline M. Rea
(1987), the TEEP is
comparable to the TO EFL (Test of English as a Foreign
Language) and the ELTS
(English Language Testing Service)
in its overall aim.
This test differs from the other two
in its overall design and. format.
In her own words,
~
'this test represents a nearly unique attempt to address
key issues of test validity and design from an empirical
perspective'
(Rea,
1987:78).
Another strength of the
TEEP lies in its having a positive backwash.
However,
the TEEP presents some limitations.
Three main ones have
been identified by Rea.
These concern:
(i)
the validity
of the sUbject/specific division;
(ii)
the length of the
examination;
(iii)
and the omission of an information
manual.
Rea is critical of the TEEP in that i t provides
·no evidence supporting the thematic distinction.
She
also stresses the lack of information needed to suggest
how to interpret and use the test results, how to
115
.,

---

integrate levels and to indicate how discipline areas may
influence the interpretation of results.
Rea laments the
omission of validation data for the TEEP in the test
manual.
Finally, she remarks that the length of the test
could be reduced to a realistic length.
Another criticism could be addressed to the
integrative nature of the test. How does one control the
unavoidably confounding nature of such a test?
Students'
low performance on the writing task may result from their
poor proficiency in reading and listening.
since the
writing test may not be measuring directly writing
performance,
a low score in writing may not necessarily
reflect inadequate performance in writing.
Despite the limitations of the TEEP enumerated
above,
in selecting this test,
the following were
considered:
(1) The objectives of the TEEP were in line
with our own concerns; that is, the assessment of
students' ability to:
A.- produce adequate written English for formal
academic writing tasks;
B.-
understand spoken English for listening to
lectures and discussion;
C.-
understand written English for reading
textbooks and other sources of information both
intensively and extensively;
D.-
produce adequate spoken English for taking
part in academic discussions and presenting
papers
(Weir,
1988:124).
All the four objectives are obviously shared by the
Department of English, at the University of Abidjan. It
is to be stressed that objective A, as spelled out in the
116

---

TEEP,
is directly related to our present concern.
(2) The
other reason for selecting the TEEP is that the absence
of overall results allows individual researchers to adapt
the interpretation of students' performance to their
areas of interest.
This flexibility has proven to be
useful in determining the sample size. This discussion
leads to examining how the specific tasks used in
eliciting the data for this study have been constructed.
writing tasks
The construction of the writing tasks 12 was guided
by the investigator's view of writing as a mode of
learning.
Four main points were taken into account when
developing the writing tasks. These were discourse type,
task-initiator,
sources of knowledge, and degree of
familiarity with the task.
Firstly,
instructions to perform the tasks were
formulated so as to cover two discourse types:
referential and argumentative as understood by Kinneavy
(1971).
The decision to focus on these two discourse
types to the exclusion of expressive discourse
seemed
justified given the academic context of the experiment.
Indeed, at the University of Abidjan,
English majors are
not expected to work on writing tasks requiring
. expressive discourse type per se, even though they are to
121n this study writing tasks or writing problems are used
interchangeably to designate what writing researchers refer to as
topics,
prompts or probes.
117

---

take positions and to express their personal point of
view on specific issues.
Excluding expressive discourse
was also motivated by practical considerations as seen
later.
Apart from Kinneavy
(1971),
other classifications of
discourse have been offered by Moffett
(1968), and
Britton et al
(1975)
and Britton (1978). Britton et al
(1975) review of different definitions of discourse types
clearly shows that consensus is far from being reached
among discourse theorists.
Kinneavy's four-way
distinction of discourse into expressive,
literary,
referential,
and persuasive was retained for the purposes
of this study.
within the perspective of such a
categorization, the aim of expressive discourse is self-
presentation.
In Odell, Cooper & Courts's (1978) words
'expressive discourse aims simply to articulate the
writer's personality or point of view.'
Unlike
expressive discourse, persuasive discourse is geared
towards change in the audience attitude or behaviour.
Referential discourse differs from the other two by
focusing on content, sUbject and reality (see van Dijk,
1985).
The main concern of referential discourse is to
display factual information, whereas the purpose of
literary discourse is to 'create a language structure
"worthy of appreciation in its own right'"
(Kinneavy,
1971: 39)
118

---

After specifying the types of discourse being
considered in this study, the next point regarding the
elaboration of the writing tasks was the task-initiator.
The tasks were only researcher-initiated.
These
restrictions seemed justified given the constraints on
students at University level.
As a rule,
undergraduates
in the Department of English,
at the University of
Abidjan,
are not requested to perform on self-initiated
tasks.
A third point of concern in developing the writing
tasks was source of knOWledge.
The construction of the
writing tasks drew on three potential sources of
knowledge: £§rsonal experience,
academic experience,
and
world experience. Tasks related to personal experience
can be defined as topics that call on subjects'
immediate
environment and informal experience without requiring
specialist knowledge.
Tasks drawing on subjects'
academic experience can be defined as tasks that call on
subjects' specialized knowledge that is expected to have
been acquired in a formal educational environment. Tasks
drawing on students' world knOWledge can be defined as
tasks that call on students' world knOWledge not necess-
arily available through their educational experience.
Such tasks may require informed or specialized knowledge
(see Appendix III for illustration).
The last concern when producing the tasks was the
degree of familiarity with the task.
Familiarity as
119
..

---

and not the discourse type in which given tasks should be
writ ten.
To clarify the term familiarity,
three
dictionaries were consulted. 13
The examination of these dictionaries revealed a
recurrent association made between substantives such as
familiarity,
acquaintance and habitude or between verbs
such as to familiarize,
to habituate,
and phrases such as
to make well acquainted or intimate.
Acquaintance is
defined in the Encyclopaedic Dictionary as
the
'act of
galnlng a greater or less amount of knowledge of any
person or thing.'
On the basis of this definition,
the following
understanding of familiarity was offered for the purpose
of the present study.
It was assumed that a topic is
familiar to students if they could talk or write about it
with adequate and sufficient knowledge or information.
In this respect,
only two degrees of familiarity were
identified:
most familiar
(F)
and least familiar
(U).
The clarification of the term familiarity is
essential since its relationship with discourse types,
the quality of the written product and the composing
process will be ultimately examined.
In section 3.2.3
due attention will be given to how subjects appraised
familiarity with tasks.
Table 3.1 provides a summary
representation of the task demands.
lJThese were:
New English Dictionary,
Shorter Oxford English
Dictionary and Encyclopaedic Dictiona~.
120

---

TABLE 3.1
Analysis of task demands.
[TaSk
1
2
3
4
I
Goal
explanatory text
explanatory text
persuasive text
persuasive text
focus on
focus on
focus on convincing
focus on convincing
presentation of
presentation of
the reader
the reader
information
information
not defined
explicit:
subjects
explicit:
subjects
implicit:
subjects
Ini tial state
to start with both
to start with
to start with
varieties of English
Ivorian parents'
excerpt of
."
position
Declaration of
Independence
Operators
referential
referential
argumentative
argumentat i ve
discourse type
not defined
compare/contrast
pros/cons
pros/cons
organization
Restrictions
on operators
duration
one hour
one hour
one hour
one hour
length
two pages
two pages
two pages
two pages
direction
not applicable
not applicable
not binding
not binding
audience
defined
defined
defined
not defined
Familiarity
familiar
unfamiliar
familiar
unfamiliar
Experience
personal
academic
personal
academic

---

A point worth making here is that the writing tasks
have been deliberately left unstructured.
Whether a
writing task should be structured or not is a valid
question to ask.
Ruth & Murphy
(1988)
provide an
extensive discussion of this issue.
Both structured and unstructured writing tasks can
be justified by the task's initiator as indicated by
Hamp-Lyons & Heasley
(1987):
There are arguments in favour of a clearly
laid-out topic being set,
providing students
with a strong structure to support the creative
writing they will do,
preventing them going
astray because of irrelevance,
poor
organization of arguments,
and so on.
But
there are also arguments in favour of leaving
the task very vague,
allowing students the
space in which to think and create,
to follow
their own interests and ideas,
to use their own
knowledge and test out their developing skills
without restrictions
(Hamp-Lyons & Heasley,
1987:135) .
The decision for setting unstructured writing tasks
seemed justified for the purposes of this study.
The
first
justification is that,
as noted by Hamp-Lyons and
Heasley,
such a procedure gives enough leeway to students
and allows them to focus and refocus the topics as they
wish.
Secondly,
the researcher is likely to obtain
detailed information on students' strategies while
attempting to formulate and reformulate,
plan and
organize during the act of composing, which structured
tasks would not only predetermine but also inhibit.
122

---

To sum up,
it is worth repeating tha~ even though
it would have been interesting to consider other
discourse types such as expressive and literary and to
allow for students'self-initiated topics,
taking such
considerations into account would not only have been
impractical but also would have yielded overwhelming
data,
which would have been unmanageable,
even under
ideal research conditions.
Thus for practical purposes,
this researcher's attempt at characterizing the composing
process of Ivorian EFL University students has
necessarily been constrained by practical issues.
The
limitation of the scope has resulted in abandoning the
examination of the interaction between the task initiator
and other variables such as the composing process, ~
familiarity,
and discourse types.
To recapitulate,
the elaboration of the writing
tasks took the following into account:
definition of writing as a mode of learning
discourse type
task-initiator
sources of knowledge
degree of familiarity with the task and
structured vs unstructured tasks.
The tasks developed along those lines were used to elicit
, the protocol data to be defined below.
123
'.

---

Verbal protocols:
Definition
A protoco],
as defined by Hayes
(1987:51~ 'is a
description
of the activities, ordered in time,
in which
a subject engages while performing a task'.
It is to be
emphasized that not every description of a task
performance can qualify as a verbal protocol; neither can
a description that includes only outcomes and goals
be considered a protocol.
A protocol is necessarily a
sequential and chronological ordered description of the
different stages the problem solver goes through,
from
the very moment the problem is found to the
very end of
the solution of the problem.
Three kinds of protoco1s are available:
motor
protocols,
related to physical activities such as
walking; eye-movement protocol,
'a record of the places
in a scene where the subjects fix their gaze as they
perform a task'; verbal protocol or 'thinking aloud',
which requires the problem solver to say aloud everything
that comes to his/her mind,
no matter how trivial or
irrelevant such information could appear
(Hayes,
1987:53).
It is worth reiterating that, despite
reservations expressed by Perl
(1980)
and Cooper &
Holzmann (1983), protocol analysis is an extremely
powerful tool for describing psychological processes.
As
such i t is the appropriate means of investigating the
composing process of unskilled writers.
There is general
agreement that verbal protocols provide reliable research
124

---

data
(Bakan
(1954); Radford (1974); and Ericsson & Simon
(1980,
1984);
Smagorinsky
(1989)).
It is to be emphasized that the present investigator
does not intend to undertake an exhaustive description of
both verbal protocols and video-recording.
However,
it
is believed that by combining video-recording and verbal
protocols,
one would hope to collect complementary sets
of data that would compensate for the incompleteness of
each technique of eliciting data,
and thereby overcome
some difficulties experienced by previous works.
For
example,
these complementary sets of data will allow this
researcher to assert that a given subject is silent
because he lS writing or scanning rather than make an
inference on the basis of the noise of pencils as was the
case in Arndt
(1987).
While video-recording focuses on observable
behaviours,
protocol analysis, as a supplementary tool,
enables one to draw plausible inferences on the content
of the writer's thoughts during the act of composing.
Finally,
the combined study of pauses and protocol
analysis makes it possible
to observe writers throughout
the composing process and to capture the mental activity
going on from the formulation of a given difficulty as a
problem to the production of the finished product.
The
gathering of the protocol data involved specific
experimental settingsand technical equipment.
These are
described in the following.
125

---

Setting and technical equipment
This section undertakes the description of the two
centres that were used for the pilot study and the actual
experiment respectively.
These were Centre Universitaire
de Formation Permanente (CUFOP)
and Centre Professionnel
de Formation de la Societe de Transport d'Abidjan
(SOTRA) .
Each of these centres will be described in the
following sections.
Centre Universitaire de Formation Permanente~ This
centre was opened in 1979.
Among other activities,
the
centre has been organizing pedagogical seminars for newly
appointed members of staff at the Universite Nationale de
Cote d'Ivoire.
It also provides assistance to
medical candidates preparing for the aggregation l4 in
medicine.
Finally,
it offers its expertise to private
companies seeking to improve their system of
productivity.
Situated on the main campus of the National
University of the Cote d'Ivoire,
CUFOP seemed the most
convenient place to start with.
The centre provided this
investigator with the following facilities: a
demonstration room inclUding a THOMPSON video recorder
14The Aggregation is a competition for the upgrading of both
secondary school teachers and University lecturers in the French
academic system.
Unavoidable differences in requirements are
observed.
126

---

VK306 with a remote control, a THOMPSON video camera TC
101 S,
a 51 cm THOMPSON monitor TV,
and a battery.
Such facilities were adequate enough to familiarize
subjects with the 'think-aloud'
technique of eliciting
protocol data but proved to be unsuitable to capture both
student writers and the writing pad.
Given the limited
research facilities available at CUFOP,
i t was felt
necessary to seek the logistical assistance of a better
equipped experimental environment.
SOTRA professional
training centre proved to be the most suitable in that i t
offered the audio-visual facilities required.
Centre de Formation Professionnel,
SOTRA:
SOTRA lS
the main semi private transport company catering for
public transport in Abidjan and its
neighbouring
suburbs.
Sotra professional training centre is situated
in an industrial suburb,
at approximately 15 kms from
Abidjan, and 20 kms from the University main campus.
The
centre provides for the training,
retraining and up-
grading of the company bus drivers,
technicians,
and
maintenance staff.
Among other facilities,
the centre lS
equipped with an audio-visual room geared toward
training in human relationships for the supervising staff.
127
".

---

Description of the audio-visual room and technical
equipment:
The main room,
usually used as a
demonstration
and teaching room,
included desks and
tables,
one SONY trinitron colour video monitor PVM-
1850PS,
two SONY movable tripod colour video cameras
DX1800P,
NO. 11451, with a flexible wire unit LO-22,
permitting remote control of the VCL-1106YB standard zoom
lens,
and with DXF-40 CE NO. 11368 viewfinder; a SONY
microphone,
and an RDSMN 8EC lamp.
Adjoining that main room,
a small room, which
contained the technical equipment,
served as a recording
room (See Appendix IV for the description of the
technical equipment).
The main room served as the
experimental room where subjects composed individually.
Both rooms were air-conditioned.
Note also that focusing and zooming could not be
adjusted at the pan bars of the tripod.
Apart from a
few
technical problems such as power cuts,
failure of the
microphone and lamp bulbs,
the pilot-study and the
collection of the verbal protoco1s took place in quite
reasonable conditions.
Attention is now given to
describing how the pilot study was conducted.
128

---

pilot study;
Warm-up sessions"
Three warm-up sessions were organized with a Vlew to
familiarizing subjects with the experimental procedure.
The first two sessions took place at CUFOP.
After
individual contacts with subjects,
the first session
consisted in defining and explaining the objectives,
the
nature,
and procedure of the investigator's research
study.
Subjects were reassured that this study was
apolitical,
solely academic,
and specifically In
fulfilment
of
a doctoral programme.
Stressing the apolitical nature of the work lS
essential in African developing countries as was revealed
by the reaction of an MA student,
at the University of
Reading,
during the pre-pilot study.
That subject
refused point blank to pursue the video-recorded think-
aloud experiment,
alleging failure to understand the
task.
The real reason,
as it turned out later,
was that
the content of the writing tasks had led the subject to
believe that the investigator was indirectly trying to
elicit information that might politically commit him
sooner or later.
Given this unexpected feedback,
this investigator
thought it worthwhile to reassure the subjects at the
National University of the C6te d'Ivoire.
After
reassuring these subjects and informing them of the
15The need for warm-up sessions was suggested by Cohen
(1987).
Gradual exposure of sUbjects to the experimental procedure arose from
this investigator's experience with first-year students at FLASH.
129
,',

---

purpose and nature of this research study,
the next step
of the first session centred on the viewing and
discussion of a pre-pilot study undertaken at the Centre
for Applied Language Studies
(CALSj,
Reading University,
during the summer of 1988.
This first session was extremely important in that
it allowed the subjects in this experiment to voice
different kinds of apprehension.
Besides the political
apprehension,
these subjects questioned the feasibility
of the 'think-aloud'
technique of eliciting data even
though some of them
acknowledged
'doing the same thing
as that subject'
observed in the pre-pilot study.
Others
even added 'Yes,
but we don't express i t aloud.'
In sum,
the real question was the following;
'How could it be
possible to think,
to talk,
and to write at the same
time?
Such a technique is too frustrating.'
The second warm-up session dealt with the 10
subjects' actual exposure to the 'think aloud'
technique
of eliciting data as independently initiated by Duncker
(1926), Claparede (1934)
and as practised by Flower and
Hayes
(1977,
1980a, 1980b,
1981a,
1981c) .'6
Students
were audio- and video- recorded individually.
The
gathering of data lasted 15 minutes for each subject.
The ten subjects were to write about one of two
researcher-initiated tasks bearing respectively on
16Appendix IV gives an account of the procedure suggested by
Swarts.
Flower & Hayes
(1984).
130

---

polygamy and the wearlng of uniform at school.
The
overall impression at the end of the second warm-up
seSSlon was that these subjects appeared reassured and
confident.
Neither the camera nor the tape recorder
seemed to have interfered with their work.
The
combination of thinking,
talking,
and writing no longer
appeared overwhelming as most of them admitted after that
session.
Only one subject reacted negatively after the
first exposure to the experiment.
She seemed depressed
after that session.
In fact,
she spoke so softly that
one could hardly hear her verbal protocol.
It later
emerged that this reaction resulted from her difficulty
In understanding the writing task.
The third and last warm-up session took place at
SOTRA professional training centre.
This session allowed
this researcher to expose these subjects to the very
context of the experiment and to potential difficulties
arising from technical equipment and assistance.
The
audio- and video- recording session lasted 30 minutes per
subject.
At the end of this session,
these subjects were
considered ready to start on the actual protocol
gathering that lasted one hour.
Indeed, most subjects
observed that 30 minutes were insufficient.
They even
suggested that for the forthcoming sessions the allotted
·time be increased to one hour.
Thus these subjects were
gradually introduced to the overall experimental
environment and procedure.
Such a gradual exposure
131

---

proved to be effective in that students were
progressively made aware of what verbal protocols
entailed, keeping the time pressure to a minimum.
The end result of protocols on the process of
solving writing tasks is a set of written products which
may be sUbject to different kinds of exploitation.
In
the following,
concern is with the written products
gathered for this study.
written products
Forty essays resulted from the 'think-aloud'
sessions in which the ten sUbjects participated.
Four
essays were produced by each subject.
Two sets of twenty
scripts dealt
respectively with argumentative and
referential discourse types.
Each set comprised two
subsets illustrating the two degrees of familiarity (most
familiar
(F),
least familiar
(U».
The evaluation
procedure of these finished products is described in
section 3.2.3.
Attention is now given to two other
instruments that served in eliciting information from
subjects.
Questionnaires and retrospective feedback
Questionnaire:
Two kinds of questionnaires were
used (see Appendices V and VI for illustration).
One was
132

---

concerned with eliciting biographical,
academic,
linguistic and socio-economic data from subjects.
The
other one consisted of closed and open questions.
This
instrument of data collection was articulated around
three axes:
Perception of writing,
method of enhancing
writing skill,
and composing process.
Information
covering these three axes may be useful in providing a
global description of subjects'
uncontaminated perception
of themselves as writers,
of their audience and of the
writing context.
Ultimately,
one would want to discover
whether or not EFL students entering university
understand what the process of composing written
discourse entails and the extent to which concerns with
accuracy affect this process".
The first draft of this
questionnaire sprung from an exploratory survey launched
at the end of the academic year 1986-1987.
This version
was subsequently revised and piloted in July 1988.
A
group of Pre-sessional students at the University of
Reading served as subjects for piloting the
questionnaire.
The final version of the questionnaire
was administered to all first year English majors at the
National university of Cote d'Ivoire in November 1988 '8 •
l1A paper will ultimately evolve from this questionnaire.
18A
modified version of this questionnaire has been published in
White & Mc Govern,
1991.
133
'.

---

Retrospective feedback:
The aim of the
retrospective feedback was to find out whether subjects'
attitude toward the 'think-aloud'
technique of eliciting
data had positively changed at the end of the experiment.
It also
sought to discover whether the experimental
environment interfered with subjects'
composing
processes.
Information secured from the questionnaires and the
retrospective feedback will be exploited with a view to
interpreting the statistical and/or the qualitative
results, as required.
Such a treatment presupposes a
rigorous experimental procedure.
This procedure is
discussed below.
3.2.3.
Experimental procedure
In this section,
six points are dealt with.
These
are:
- sampling,
- familiarity with tasks:
assessment,
- protocols:
writing sessions,
- evaluation of written products,
- discriminating among successful and unsuccessful
subjects,
and
- computer treatment of protocols.
134

---

Sampling
It is to be re-stated here that,
as mentioned in
section 3.2.2, the 10 subjects who contributed towards
the collection of the protocol data formed a subset of a
sample of 40 sUbjects involved in Manouan
(1990).
Five criteria guided the selection of those 40
subjects.
These were:
(1)
Nationality:
Ivorians
(vs non Ivorians);
(2)
First exposure to University:
Non-repeaters
(vs
repeaters) ;
(3)
Participation in the
Test of English for
Educational Purposes (TEEP), which seeks to measure
readiness for academic studies;
(4)
Level of performance on the TEEP test;
(5)
Parity of number of subjects; and
(6)
First composition scores and instructors'
recommendations.
The first criterion was used to control for
differences of educational background among sUbjects.
This brought the number of first year students who took
the TEEP test at the beginning of the academic year,
1988-1989, to 242 instead of 252".
The second
criterion, which took students' academic progression into
account, was concerned with retaining only first year
Ivorian university students who had registered for the
19From
the records of the National University of the Cote
d'!voire,
360 students were officially registered in 1988-1989.
252
took the test.
Of those 252,
10 were non Ivorians.
135
.'.

---

first time in 1988-1989.
Thus Ivorian first year
students who had been admitted to the University in 1987-
1988,
and who had been allowed to repeat their year in
1988-1989, were not considered.
This brought the number
of subjects down to 169.
The third criterion was to exclude,
among newly
admitted first year Ivorian students,
those who had not
produced a written passage as instructed by the TEEP
test.
The fourth criterion dealt with students'
performance on the TEEP tests.
A passing benchmark was
determined that allowed one to select low achievers and
high achievers.
(Detailed description of the procedure
will be given in due course).
After the application of
the fourth criterion,
the number of subjects came to 131.
The fifth criterion was concerned with equalizing
the number of subjects with respect to the level of
performance and sex.
In order to equalize the number of
low achievers with high achievers with respect to sex the
investigator had to reduce the number of high male
achievers.
This led to a reduction of the number of
subjects that were taken into account for the statistical
exploitation of the results.
Only 40 students could be
retained.
Of these 40,
10 were involved in gathering the
protocol data.
The selection of these 10 will be
ultimately described.
Table 3.2 provides a summarized
136

---

account of the procedure that helped determine the sample
size.
TABLE 3.2
summary of selection procedure.
criteria
Male
Female
Total
Nationali ty
182
60
242
Academic Progression
129
40
169
Participation in TEEP test
126
37
163
Performance on TEEP test
103
28
131
Parity
20
20
40
composition scores
+ Instructor's recommenda-
5
5
10
tion
The investigator is aware of the potential
limitations of this sampling procedure.
The limited
number of subjects may lead to results that may not be
statistically significant.
Neither will such a selection
procedure allow the investigator to generalize the
results to a wider population.
Such a decision,
however,
was motivated insofar as the concern for adopting a
rigorous sampling technique overrode the concern for
representativeness.
In establishing the sample size, performance on the
intelligence test, age and socio-economic status were not
taken into account.
The students' ages vary from 18 to
24 years,
which constitutes a six year interval.
Admittedly,
such an interval is rather wide.
It was
137
.•.

---

assumed that
age was unlikely to be a source of
variation likely to warrant any significant behavioural
differences among sUbjects.
This assumption seemed
reasonable given that the sUbjects involved in the
research study were no longer in an acknowledged
developmental period,
coupled with the fact that most
intelligence tests such as Wechsler-Bellevue and
stanford-Binet do not seek to measure the development of
intelligence beyond 16.
Such a consideration seemed to
provide a good ground for not retaining age as an
independent var iable2".
The socio-economic variable was
not considered as a criterion of selection because the
impact of home environment at the mean age examined can
be said to be negligible.
Besides, three-quarters of the
students considered spent most of their secondary school
training in a boarding school or with a guardian.
Hence,
were any environmental variable to be taken into account,
one would be well advised to consider school environment
since,
as indicated by
Greenfield and Bruner (1966),
instruction seems to be a determinant factor in
explaining differences among SUbjects.
While the application of the first three criteria
involved no partiCUlar diffiCUlties, this was not
~ It is to be mentioned that age intervals are rather narrow,
that is,
0-3 months when dealing with 1-3 years old children.
During
the second childhood
(or pre-pubertal period)
up to
12 years,
a one-
year interval is the maximum age range considered.
During the
developmental period
(12-16/18)
years old)
age variation is examined
over a wider interval.
Around 20 years,
one may consider a five-year
interval to be reasonable.
138

---

necessarily so when it came to applying the fourth and
fifth criteria.
Indeed,
in constituting the sample,
this
investigator had to come to grips with two related
constraints:
(1)
establishing the readiness benchmark,
and
(2)
determining the overall nu~~er of subjects to be
considered.
To test the hypothesis that there is a relationship
between writing performance and overall academic
achievement,
ideally at least 60 SUbjects would have been
required so as to contrast subJects on the basis of their
level of achievement and their sex."
An analysis of
such a relationship "'ould have implied that the sample
include the top 30 male and female and the bottom 30 male
and female,
distributed as shown in Table 3.3.
T.I'IBLE 3.3 Ideal
sample size
M
F
11
I
Top
15
15
Bottom
15
15
11
To determine subjects'
level of writing proficiency at
entry,
their performance on the writing component of the
TEEP test was used
as a reference
point.
Initially,
it was assumed
that
those
subJects who scored
2 on
the
TEEP
scale
could
have
been
considered
marginally ready to undertake academic studies
21The examination of
this hypothesis constitutes
the object of
Manouan
(1990)_
139
".

---

in English as a foreign language,
at the University of
Abidjan.
Indeed, an impressionistic
conversion of a
score of 2 on the TEEP scale into UNCI
(Universite
Nationale de Cote d'Ivoire)
scale,
coupled with
familiarity with the educational context, resulted in
scores of approximately 11-12.5 out of 20. n
since
strictly speaking, only students who score 12 out of 20
in English in their final year in Secondary School are
allowed to major in English at the National University of
the Cote d'Ivoire, to set the readiness benchmark at 12
out of 20 seemed
justified.
Doing this, however, would
have been unrealistic.
only 5 students,
including male
and female,
scored between 12.5 (2 on the TEEP scale)
and
14.5 (3 on the TEEP scale)--2 females scored respectively
12.5 and 14.5 while 3 males scored 12.5.
Given this number constraint,
the passing benchmark
was lowered to 10 out of 20
(6-10.5 = 1 on the TEEP
scale) .
This decision was somewhat in line with the
three-step screening procedure of the national screening
board, known as the Commission Nationale d'orientation
(see Appendix I
for sUbjects' background data).
By setting the readiness benchmark at 10 out of 20,
only 10 top female SUbjects, who scored between 10 and
14.5, could be selected from the overall popUlation.
~his impressionistic conversion was produced by a KELT member
of staff who had close~y worked with the Department of English,
at
FLASH,
for eight years.
This member of staff may be said to have
internalized the assessment system of that institution.
In addition,
he was familiar with the ELTS marking scheme and took time to study
the TEEP system.
140

---

This restricted number of top female subjects necessarily
brought the sample size to 40 instead of 60.
giving the
new distribution summed up In Table 3.4.
TABLE 3.4 Sample resulting from criteria
1-5.
I M
F
Top
10
10
Bottom
10
10
Since the number of male subjects who scored between
10 and 12.5 was higher than 10.
the top 10 male subjects
were randomly selected from the top male sub-set of
higher achievers.
The bottom 10 male and female subjects
were randomly selected from each sub-set of lower
achievers. 2l
Two main criteria guided the selection of the
subjects who took part in the collection of the protocol
data.
These were first composition scores and
instructors'
recommendations.
As said earlier,
ten
subjects were retained for this study.
These were
randomly selected among those subjects
(1)
who had scored at least 12' out of 20 on their
first
two class assignments that were holistically
23It is to be recalled that these 40 subjects constituted the
sample for examining the relationship between writing performance and
overall academic success.
The issue is addressed in Manouan
(1990).
Only ten of these served as subjects for this study.
141
"

---

marked,
with emphasis on organization and levels of
formality;
(2)
and whose third draft essays were judged by
their composition instructors as being the best
organized and audience-oriented.
The two class assignments that served as bases for
selection were of an expressive discourse type.
As a
first writing assignment,
all first year English majors
who had enro~d during the academic year 1988-1989 were
requested to write a letter to their best friend telling
him or her about how they felt after their first contact
with the National University of the Cote d'Ivoire.
The
second assignment consisted in writing the same letter to
a former secondary school teacher
(See Appendix VII for
details on assignments) .
To recapitulate,
the subset of 10 subjects
contributed to the data that served for the quantitative
and
qualitative analysis at group-level.
This sample
Slze was necessarily kept small given the amount of data
generated by video-protoco1s.
The qualitative analysis
at individual-level considered 4 subjects
(three most
successful and one least successful).
For this purpose,
the distinction between successful and unsuccessful,
as
established in Chapter 3,
section 3.2.~ proved useful.
In addition,
the advance planning pattern helped in
selecting the unsuccessful subject.
It was thought
interesting to give close attention to the top least
142

---

successful subject to discover what distinguished him
from the most successful subjects at a finer level of
analysis.
Having described the sampling procedure,
we
shall now consider how the degree of familiarity with
tasks has been determined by the ten subjects in this
study.
Familiarity with tasks:
Assessment
Prior to the actual writing sessions
(at the end of
the first warm-up session,
during the pilot study)
students were asked to assess the degree of familiarity
of two sets of 5 tasks,
developed as described in section
3.2.2 and covering the two identified discourse types--
referential and arqumentative--and the two degrees of
familiarity:
(1)
most familiar
(F);
(2)
least familiar
(U).
The following instruction was glven to the
students:
'A topic is familiar to you if you know about
it;
that is,
if you can talk or write about it with
adequate and sufficient knowledge or information'
(See
Appendix VIII) .
As a result of students' appraisal of the degree of
familiarity of each set of writing tasks,
two sets of two
writing tasks covering each discourse type were retained
for the experiment.
To ensure homogeneity,
individual
differences were disregarded.
All the ten subjects
involved in the case study were asked to write about the
143
".

---

referential and argumentative discourse writing tasks
that were considered most familiar and least familiar to
the majority of the group (see Appendix VIII for details
on assessing the degree of familiarity with tasks).
The
writing tasks that were selected for this study are shown
in Appendix Ill.
These writing tasks constituted the
bases for gathering 40 session protocol data to be
discussed next.
Protocols: writing sessions
The preparation sessions In the form of the pilot
study and the actual protocol experiment were extended
over a period of 7 weeks,
runnlng from late January to
early March 1989.
Timing for the gathering of protocols
was crucial.
The period selected was most suitable for
two main reasons.
First,
experience has shown that, by
mid January, most first year university students at the
National University of the Cote d'Ivoire have gone
through the first stage of adjusting to university life,
and that they may be considered more or less ready to be
exposed to other kinds of experiences.
Secondly,
students are most receptive and co-operative before the
first mid-term examinations usually programmed for mid
February.
Note that during the academic year 1988-1989,
the first mid-term examinations were held by mid March,
which was unusually late.
144

---

The verbal protocol data were gathered through a
synchronized audio- and video- recording.
A SONY tape
recorder,
placed on a writing desk,
was used for the
audio recording.
The video- recording was done on the
PAL system,
a system which is in used in the Cote
d'Ivoire.
To facilitate the exploitation of the video
data,
recorded from a U-matic recording system attached
to a special effects generator,
a VHS video-recorder was
directly connected to the u-matic recording system.
In
such an installation.
the U-matic recorder acted as a
relay circuit and enabled direct transfer of data from a
u-matic video recording system to a VHS system.
This
transfer system resulted in impoverishment of the quality
of the VHS recordings, yet relatively satisfactory for
the purposes of this study. 24
The video recording sessions lasted one hour per
subject and per session.
Each subject was allocated 4
sesslons.
Subjects worked individually in the
experimental room,
sitting at a desk,
writing on a lined
yellow pad, with a Felt pen.
Note that during the pre-
pilot study in 1988,
it was observed that the colour of
the writing pad and of the ink being used was essential
to obtain a relatively good quality picture on the video
24As the special effects generator operated with a U-matic video
recorder,
it was essential to transfer the video-recording onto a V~
recorder so as
to facilitate the exploitation of the data wherever
this investigator might be.
This relay system permitted a direct
transfer of the video-recording from a U-matic recorder onto a VHS
recorder while subjects were composing.
Although this arrangement
may seem difficult,
connecting a U-matic to a VHS video-recorder can
be easily handled when the appropriate cable is available.
145
'.

---

screen.
These details were taken into account during the
actual collection of the data.
The student-writer was recorded by one of the
movable colour tripod video camera~ adjusted at desk
level and focusing on the writer.
The other tripod
movable colour video-camera was mounted on top of a desk,
at the right-hand side of the writer,
adjusted so as to
focus on the writing pad.
Both images coming
respectively from both cameras,
and being viewed from the
two separate monochrome monitors,
installed in the
adjoining room, were channelled through the special-
effects generator.
Both synthesized images appeared on
the two trinitron colour video monitors,
one in the
experimental room,
and the other in the recording room.
The image of the writing pad appeared in full while the
image of the writer appeared as an insert on the screen.
Unlike Matsuhashi who used a split screen effect 1n her
1979 study,
this investigator opted for the insert
effect, which was more satisfying.
As stated earlier,
to ensure homogeneity, all the
ten subjects involved in the case study were asked to
write about the referential and argumentative discourse
writing tasks that were considered most familiar and
least familiar to the majority of the group.
It is to be
mentioned that unlike other studies on composing,
subjects were not encouraged to select a topic,
rehearse
and plan tasks before hand.
This was to avoid not only a
146

---

potential memorization effect,
but also to enable this
investigator to capture the composing process of those
subjects in its entirety,
under conditions approximating
stressful examination conditions.
To counterbalance the
'order effect',
writing tasks
were alternated both with respect to discourse type
and
with respect to degree of familiarity with tasks.
Sessions I
& III were devoted to referential discourse
type
; sessions 11 & IV were devoted to argumentative
discourse.
Subjects were divided into two groups of 5
each,
and performed alternatively on each task.
Both
groups performed on referential discourse in the first
and
third sessions.
Both groups performed on
argumentative discourse on the second and fourth
sessions.
While,
on session one,
group I was glven the
most familiar,
referential discourse task
(Reff)
to write about,
group 11 had to write about the least
familiar,
referential discourse task
(Refu).
Alternatively,
during the second writing session,
group
11 subjects were given the most familiar,
argumentative
task
(Argf)
while group I subjects were assigned the
least familiar.
argumentative task
(Argu).
The order was
reversed on session III and IV with respect to degree of
familiarity as shown in
Table 3.5.
These writing sessions generated forty essays,
the
evaluation of which is the concern of the following two
sections.
147

---

TABLE 3.5. ordering of tasks· .
I
Group I
Group I I
Session I
Reff
Refu
Session II
Argf
Argf
Session III
Refu
Reff
Session IV
Argf
Argu
·Reff:
Referential discourse most familiar
Refu:
Referential discourse least familiar
Argf:
Argumentative discourse most familiar
Argu:
Argumentative discourse least familiar
Evaluation of written products
The forty written products resulting from the
'think-aloud' video sessions were evaluated by native
speakers of English.
Seven composition assessors,
two
American at the National University of the Cote d'Ivoire,
four British and one Anglo-American at the University of
Reading, assessed these essays.
Four of these raters
randomly marked those essays, two analytically, using C.
Weir's 1988 TEEP attribute writing scales
(a four-point
scale 0-3,
involving content, organization,
cohesion,
grammar, vocabulary, mechanical accuracy I
and mechanical
accuracy 11) and two holistically~, having Hamp-Lyons's
1986 Formative Feedback Profile as a reference document.
~ A variety of quality rating schemes is available.
Among
others are: Jacobs,
Zikgraf, Wormuth,
Harfield & Hughey (1983)
analytic marking scheme;
Lloyd-Jones
(1977)
Primary Trait Scoring in
Cooper and Odell
(1977); and Diederich (1974) Analytic Quality Scale.
Homburg
(1984)
discusses the validation of holistic evaluation.
148

---

(See Appendix IX for details.)
A fifth marked selected
samples both holistically and analytically.
To control for sources of variation in marking,
four
maln precautions were taken.
First,
non-native speakers
were excluded to avoid dealing with variation between the
marking system of native-speakers and non-native speakers
(see Hughes & Lascaratou,
1982).
Secondly,
only expert composition teachers or
evaluators were used to exclude potential variation
between expert vs non-expert markers.
Thirdly,
all the
scripts were typed to eliminate the effect of handwriting
on the evaluation of the written products
(as seen in
Chase,
1968 and Markham,
1976).
Finally,
to
counterbalance the 'order effect',
the order of the
scripts was randomized before marking.
Two dimensions of the written products were
considered.
The first dimension was the overall quality
of the written product.
The second was the quality of
introductory paragraphs as reflecting the quality of
planning during the preparatory process.
The overall
quality of the written product
(referred to as Qscore)
was assessed holistically and analytically.
The
quality
of introductory paragraphs
(referred to as Paraqual) was
evaluated on a five-point impressionistic scale developed
for the purposes of this study
(Appendix
for details)
In each instance,
two raters marked four sets of
scripts.
The actual assessment of the scripts was
149

---

preceded by standardization sessions consisting in
practice marking,
using samples of writing by the
subjects involved in this study.
Marking sessions took
place individually.
Harmonization of marking between
analytic raters occurred before and at the end of the
evaluation of the forty scripts.
This researcher held
retrospective interviews with individual holistic and
analytic markers with a view to explaining differences
arlslng from evaluating the written products.
To check the reliability of the holistic and
analytic measuring tools used In this study,
a
correlation analysis was performed.
This point is
detailed In section 3.2.4.
To check the validity of the
analytic marking scheme,
a sample of scripts,
showing
variation among holistic and analytic raters,
were marked
both holistically and analytically by an independent
composition expert on the Pre-sessional course at the
University of Reading.
This seventh rater marked the
scripts in two stages.
The holistic marking preceded the
analytic marking.
This second type occurred a month
later so as to nullify the order effect.
The result of
this validity check was secured for interpreting the
statistical and/or qualitative results,
if needed.
After discussing the precautionary steps taken
towards a systematic evaluation of the written products,
attention is now given to describing the evaluation
150

---

procedure
(see Appendix IX for the instructions glven to
the assessors) .
The scores obtained from evaluating the written
products will constitute the basis for the statistical
and/or the analytic examination of the hypotheses to be
tested in this study.
Given the high correlation between
the analytic raters,
as shown in section 3.2.3,
the
analytic scores were used to discriminate among
successful and unsuccessful subjects as described below.
Discriminating among successful and unsuccessful
subjects
In the absence of pre-established criteria for
discriminating among successful and unsuccessful subject
writers in a given research context,
this study adopts a
de facto discriminating procedure by envisaging three
ways of distinguishing both types of subjects.
These were
as follows
:
1.
Ranking subjects in descending order on the
basis of their performance on individual tasks,
as
measured by the TEEP attribute writing scales,
described in the preceding section;
2.
Classifying subjects on the basis of their
overall total scores obtained across tasks;
151

---

3.
Classifying subjects on the basis of the target
score of 17.5. 26
The discussion that follows lS concerned with a
combined application of the three-way discriminating
procedure.
First way of discriminating among subjects:
Ranking
subjects in descending order on the 4 tasks yields
respectively 4 top clusters and 4 bottom clusters
observed in Table 3.6.
A general observation is the consistent occurrence
of the same set of subjects in the top clusters and of
another set of subjects in the bottom clusters.
Specifically,
it is observed that subjects 2,
4 and 7
appear systematically in the top clusters,
whereas
subjects 1,
9
and 8 consistently fall within the bottom
clusters.
Subject 5 occupies a
'middle' position.
Although the pattern of behaviour of subjects 3 and 10 "cS
less consistent,
these subjects are likely to be grouped
into the top clusters.
In the main,
the observed consistency In clustering
allows a crude discrimination into successful and
unsuccessful categories of subjects.
Subjects 2,
3,
4, 7
26The target score of 17.50 was established after consultation
with a chief examiner of the AEB.
This board commissioned Weir
(1983).
Note that the readiness benchmark discussed earlier differs
from the success benchmark to be considered.
While the former has
been determined for the constitution of the sample size,
the latter
is used to delineate profiles of composers as they behave in this
study.
152

---

and 10 can be tentatively considered successful and
subjects 1,
6,
8 and 9 unsuccessful.
To refine this
classification,
the second way of discriminating among
subjects will be considered.
TABLE 3.6 SUbjects ranked in descending order on four tasks.
TASK
Discourse type
Familiarity
Argumentative
Retp.rencial
Qscore
Rank
SUbject
QSCOT>2
Rank
Subject
19.0
1
2
19.0
1
7
,
19 0
2
3'
18.0
2
18.5
3
10'
16 .5
]
5 .
,
,
,
Familiar
18.0
16.0
2
,
,
18.0
7
16.0
3
16.5
6
6
H. 5
6
8
1 ~ . 5
7
5
14.0
7
10'
,
15.0
1
1l. 5
8
1
,
15.0
8
9
11.0
6
14.5
10
8
08.5
0
9
,
17 .5
1
17.5
1
10'
,
16 .0
2
2
17 .0
2
16 .0
2
7
16.5
3
2
,
15.5
5
16.5
3
7
14 .5
5
3
15.5
5
5
,
Unfamiliar
14.5
5
10'
14.5
3
13.0
7
8
13.5
7
1
12.0
8
1
13.5
7
8
10.5
9
6
13.0
9
6
I
07.0
10
9
12.5
10
9
• Starred numbers indicate performance of subjects who break the observed pattern
Second way of discriminating among subjects:
The
above tentative classification is refined when the
overall total score across tasks
(sum of Qscore obtained
153
'.

---

on the TEEP Attribute writing Scales)
lS
examined.
Table
3.7 illustrates the point.
TABLE 3.7 Subjects ranked on basis of overall scores on
the four tasks.
S~ of Q Scores
Rank
Subjects
70.5
1
4
69.5
2
7
67.5
3
2
64.5
4
10
64.0
5
3
63.0
6
5
56.0
7
8
52.0
8
1
49.0
9
6
43.0
10
9
Taking the s~ of Qscore is definitely an
improvement in that it permits a distinction between the
4 top subjects and the 4 bottom ones.
However,
such a
basis of discrimination provides no clear-cut dividing
line between successful and unsuccessful subjects.
Indeed,
the question is whether subjects 3 and 5 can be
considered successful or not,
given the s~ of their
Qscores.
Could the third method of discriminating among
subjects be helpful in makingadecision?
Third way of discriminating among subjects:
To answer
the above question,
subjects were examined with respect
to their ability to ' h i t ' the target score of 17.5 at
154

---

least once.
This third way of distinguishing between
successful and unsuccessful sUbjects provides evidence
for excluding sUbject 5 from the successful class.
Table
3.8 provides the final classification as a result of the
combined application of the three ways of discriminating
among subjects.
TABLE 3.8
Discriminating between successful
and unsuccessful
subjects:
summary presentation,
including rank ordering,
sum of Qscore and 17.5 target score.
Sum or
Targ.:t
SllCC~SS
Suh'eels
Rank
Osro~~
score'"
4
1
705
3
7
2
695
1
Succcs~ful
2
3
67.5
1
10
4
6..1.5
2
3
5
64.0
1
5
6
63.0
0
8
7
56.0
0
Unsuccessful
1
8
52.0
0
6
9
49.0
0
9
10
43.0
0
~
... Numher
of
times
th~
tllrgcl
score
has
been
'hit'
Combining the evidence from the three-way
discriminating
procedure permits the classification of
sUbjects into successful and unsuccessful groups as
illustrated by the table above.
Computer treatment of protocols
For the computer treatment of the data,
the 40
'think-aloud' video protocols, segmented as discussed
155
.'.

---

above,
were organized around eight variables:
sUbject,
discourse,
familiarity,
phase,
episode,
state,
endtime,
and duration.
These were defined as follows:
Subject
Number of student-writers involved In
the experiment
Discourse
Argumentative,
and referential
discourse types
Familiarity
Familiarity with writing task:
F
(familiar);
U (familiar)
Phase
Subdivision of
'think-aloud' video
protocols into P (advance planning),
T (translating process),
and E
(editing,
sub-process of reviewing
process)
Episode
Component of P,
T,
and E
State
Descriptor of components of an
episode:
'active' (scribal activity),
'passive'
(pause in manual activity)
Endtime
Time at which an episode ends
Duration
Total time spent on one episode
The following schematic representation will help In
understanding how a phase is articulated:
Phase
1-------------1---------------------1------1
P
T
E
Episode
1--------------1-------1-----1------1--1---1
1 2 3
4
5
6
State
I------p-------I-p--a--I-p--al-p--a-Ipalp-al
Appendix X (related to the timing of composing episodes)
provides detailed illustration of the procedure.
156

---

3.2.4.
Statistical procedure
Two points are examined:
(1)
the methodological
questions and
(2)
the statistical techniques used to test
these questions.
Methodological questions
In investigating the relationship between the
quality of the written product and the composing process
and in comparing successful and unsuccessful writers with
respect to their pauslng pattern,
this researcher was
faced with the following methodological questions:
- Which measure of the quality of the written
product lS to be selected?
- How does one discriminate among successful and
unsuccessful subject writers?
- What are the most important temporal variables
associated with quality?
- Which pausing time elements are best associated
with successful writing?
These methodological questions are examined In the
sections below.
Selecting a tool for measuring written discourse:
An analysis of the correlation within and between
analytic and holistic scorers has been undertaken ,~ith a
157
,

---

vlew to discovering which of these two ways of assesslng
the quality of written products lS more reliable.
Computing the correlation involved two stages.
The
first was to examine how systematic the scorers were in
distributing subjects with respect to the seven TEEP
criteria writing attribute
scales,
as described in
section 3.2.2.
The second stage excludes the two
mechanics related criteria
(e.g.
spelling,
punctuation)
to discover whether excluding these criteria increases
the strength of the relationship among scorers.
A general conclusion that can be drawn from the
correlation analysis is that,
as expected,
analytic
marking is more reliable than holistic marking.
In other
words,
the overall agreement between analytic scorers is
higher than the overall agreement between holistic
markers
(r: 84 % as opposed to r:
74 %;
see Table 3.9)
TABLE 3.9
Pearson correlation coefficients
for the reliability of
the tools
for measuring the wri t ten products.
.
Ascore 1
I Ascore 2
I Hscore 1
I Hscore 2
I Bscore 1
I Bscore 2
..... score 1
1.0000
0.8381
o. 67 64
0.7381
0.9600
o.8207
Ascore 2
0.8381
1.0000
0.6434
0.7071
0.8251
0.9735
Hscore 1
0.6764
0.6436
1. 0000
0.7368
0.6748
0.6255
Hscore 2
o.7381
0.7071
0.7368
1.0000
0.7421
0.6564
Bscore 1
0.9600
0.8252
0.6748
0.7421
1. 0000
0.8409
Bscore 2.
0.8207
0.9735
0.6255
0.6564
0.8409
1.0000
Ascore 1
analytic scorer 1 on the 7 TEEP attribute writing scales
Ascore 2
analityc scorer 2 on the 7 TEEP attribute writing scales
Hscore 1
holistic scorer 1
Hscore 2
holistic scorer 2
Bscore 1
analytic scorer 1 on 5 of the TEEP attribute writing scales
(excluding mechanical accuracy I
and III
Bscore 2
analyt ic scorer 2 on 5 of the TEEP attribute writing scales
(excluding mechanical accuracy I
and III
158

---

Another point worth making is that although
agreement between analytic markers is stronger
(r:89 %)
when mechanics are excluded from the overall quality
score,
excluding or including spelling and punctuation
hardly affects how analytic markers distribute subjects.
Given these observations,
the statistical testing of the
hypotheses involves only overall analytic scores,
i.e the
sum of scores obtained on the 7 criteria as defined by
the TEEP writing attribute
scales.
The next question to
be addressed is whether subjects can be separated into
successful and unsuccessful on the basis of their scores
on the written products.
Discriminating among successful and unsuccessful
subject writers:
Discrimination among successful and
unsuccessful subjects in this study was arrived at after
the three-way discriminating procedure discussed in
section 3.2.3.
The target score of 17.5 was definitely
the dividing line which separated the most successful
from the least successful ones.
To be qualified as
successful,
a subject must have 'hit'
the target
benchmark at least once.
As a consequence, only those
subjects who had 'hit'
the 17.5 target score at least
once were considered successful. After discussing this
point,
the next question concerns the selection of the
major temporal elements associated with quality.
159

---

Important variables associated with quality?
The
third question leads to examining whether the observed
three major temporal components of the composing process,
identified as advance planning time
(T1:P),
translating
time
(T2:T),
and editing time
(T3:E),
should be equally
considered when examining the relationship between the
quality of the written product and the composing process.
Note in passlng that total composing time was fixed to be
one hour
(60').
The specific question being addressed
here is whether E should be retained since total
composing time equals P+T+E=60.
Given that these three
variables
(P,
T and E)
are interdependent,
once P and T
are included in any model,
the inclusion of E in any
model is redundant as E cannot yield any further
information.
This phenomenon is known as
multicollinearity.
To avoid this effect,
it was decided
to exclude E from any model that studies the relationship
between any given variable and the composing process from
a temporal dimension.
Which pausing elements to select?
The last
methodological question dealt with the type of pausing
elements to retain among the different pauslng elements
observed in the data.
Only three of these pausing
elements were considered on the basis of evidence
available in the literature or of empirical observation
to be discussed in the following sections.
160

---

As discussed in Chapter one,
section 1.2.2,
the
preparatory process or advance planning is theoretically
instrumental in creating a satisfactory written product.
Strong claims made by Rohman & Wlecke
(1964),
Young et
al.(1975),
Bereiter & Scardamalia
(1981,
1987) might be
sufficient to retain advance planning as the sole pausing
time element in discussing the association between the
quality of the written product and pausing.
However,
as
Johnson
(1955:57)
rightly pointed out,
preparation or
planning cannot be limited to one point in time during
the process of complex problem sOlving.
The process of
complex problem solving displays a cyclic pattern in
which pausing and event alternate.
This claim is confirmed by the episodic structure of
the composing processes examined in this study.
This
being the case,
in examining pausing in composing written
discourse,
every pausing element is to be retained.
This
claim and observation set the ground for including
pausing during the translating process in the present
analysis.
Another empirical observation,
which is not
theoretically based but worth exploring,
is that advance
planning,
as evidenced by the data under examination,
is
always followed by a pause which initiates the
translating process and which constitutes the
'passive
state'
component of the first major episode in the
translating process.
As discussed in section 3.2.3,
under the heading of the computer treatment of the data,
161
.".

---

the
'passive' state describes the pausing element
preceding the scribal element of a composing episode
occurring within the translating process.
If the preceding analysis is well-founded,
then it
seems reasonable,
in examining the association between
pausing and successful writing,
to glve due attention to
each of the three pausing time elements discussed above,
VlZ.
(1)
pausing time as represented by advance planning
time
(P),
(2)
pauSlng during the translating process
(Tpause) , and (3)
sum of advance pausing time and first
pause in T (Pplus).
Having dealt with the methodological questions
raised in the previous section,
one may reasonably
envisage how the different hypotheses in this study can
be statistically examined.
Such a concern constitutes
the object of the next section.
Testing of Hypotheses:
statistical techniques
After examining the four methodological questions
raised in the introductory section of the statistical
procedure,
the next major point to be dealt with was the
statistical testing of the specific hypotheses formulated
in this study.2?
27For the statistical treatment of the hypotheses,
this study
used the SAS programme available at Reading University.
The tabular
presentation of the results was partially informed by F. Woodward and
Frances 1988.
162

---

Testing hypothesis 1:
The statistical question was the
following:
Could the range of time, minimum and maximum,
that subjects spend on advance planning on the four tasks
in this study,
reveal the planning pattern of successful
and unsuccessful subjects?
To answer this question,
a
tabular representation (involving 3 tables)
seemed most
adequate for a descriptive treatment of the hypothesis.
Table 4.2
(a-c)
displays firstly,
subjects ranked In
descending order as a result of the discriminating
procedures discussed earlier and secondly,
the time those
subjects spent on advance planning on each of the four
tasks
(4.2
(a)).
Then time spent on advance planning of
familiar tasks is compared with time taken on advance
planning of unfamiliar tasks
(4.2
(b)).
Thirdly,
an
examination of time taken on advance planning with
respect to discourse type
(argumentative vs referential)
is considered
(4.2
(c))
Testing hypothesis 2:
The hypothesis is
that
the
amount of time spent on advance planning lS a
discriminating variable among successful and unsucceSSful
subjects.
To test the second hypothesis,
a correlation
analysis has been undertaken with a view to discovering
whether there is an association between time spent on
advance planning and overall quality across the four
tasks and separately on individual tasks.
163

---

Testing hypothesis 3,
4, and 5:
Linear regression
models,
of the form Y = ~ + ax for a simple model and Y =
~ + aXl + aX2 + .
aXn for a multiple model,
were used
to predict the response variables
(y variables)
from the
covariate variables
(x variables)
involved in these three
hypotheses.
Model I: seeks to answer the following statistical
question:
Which of the variables--subject,
discourse,
familiarity,
and the interaction between
familiarity and discourse
(covariates)-- contribute
most towards explaining and predicting respectively
P and T (response variables)?
Model 11:
This 'Full Model'
seeks to find out,
globally,
whether Qscore is associated with P and T
or whether Qscore is associated with task. The
response variable is Q and the covariates are P,
T,
task,
and the interaction
among these
variables.
Model Ill:
This
'Reduced Model'
examines whether
Qscore is associated with P and T on individual
tasks.
Qscore is the response variable,
P and T the
covariates.
Model IV:
This simple linear regression model
investigates whether Qscore is associated with
pausing during the translating process
(Tpause)
on
individual tasks.
(Response variable:
Qscore;
Covariate variable:
Tpause.)
164

---

Model V:
This multiple linear regresslon model
seeks to discover whether subject,
discourse,
familiarity,
P, and T (covariates) equally
contribute towards explaining and predicting Qscore
(response variable) .
Model VI:
This last multiple linear regression
model examines whether Qscore is associated with the
quality of introductory paragraphs
(Paraqual)
or
whether Qscore is associated with time on planning
prior to translating
(Pplus).
(Response variable:
Qscore;
Covariates:
Paraqual,
Pplus.)
3.3. Qualitative analysis"
In segmenting the 40 protoco1s under examination,
i t
was felt necessary to consider the composing process as
having both a temporal and an episodic dimension.
In
examining the temporal dimension,
the question was
whether the protoco1s of the subjects involved in this
study could be segmented within the Flower and Hayes
2BThe qualitative analysis was underpinned by the
following
concerns:
(1)
integrating multiple perspectives
(Atkinson,
1979);
(2)
considering multiple strategies of triangulation
(Denzin,
1978); and
(3)
combining
'theories and methods carefully and purposefully with
the intention of adding breadth or depth to our analysis,
but not
for
the purpose of pursuing
·objective"
truth'
(Fielding &. Fielding,
1986:33.
See also Hammersley &. Atkinson,
1983).
Hence,
in adopting
such a perspective,
we thought
it essential to seek how to
integrate
different sources of
information,
such as pause analysis,
content
analysis of the verbal protocols and the quality of the written texts
to ensure the validity of the analysis.
165

---

cognitive model of writing.
Assuming that the major
cognitive processes are observed,
what would the duration
of processes such as preparation,
translating and
reviewing reveal about the process of composing of those
subjects?
These questions were quantitatively and
qualitatively examined.
The specific hypotheses and their statistical
treatment appear in sections 3.1.1.,
3.2.3 and 4.2.4.
The qualitative treatment of the data is the concern of
sections 3.3.1 and 3.3.2. and 3.3.3.
Note that,
given
the inherent subjectivity in identifying the episodic
structure of the composing processes,
the episodic
dimension has been dealt with only qualitatively.
Sections 3.3.1. and 3.3.2. provide respectively an
account of the procedure for identifying the major
cognitive processes and the episodic structure of the
protocols and a
frequency count of these episodes.
The examination of the temporal and the episodic
structure of the verbal protocols was
facilitated by a
timing system that was incorporated into the data by the
technician of the Centre for Academic Staff Development
and Training,
Reading University.
This built-in timing
system greatly facilitated manual exploitation of the
protocols. 29
29Timing was generated by copying the film through a panasonic M
7 camera and by overwriting.
using a panasonic titler
(or character
generator) .
166

---

The procedure used is discussed in the following.
Emphasis is first on the identification of the cognitive
structures of subjects' composing processes
(3.3.1.).
Then the
identification of the episodic structure of
those composing processes 1S examined
(3.3.2.).
Finally,
an attempt is made to set up a typology of writers
(3.3.3).
3.3.1.
Identifying the cognitive structures of
composing processes
For the sake of the segmentation,
i t was assumed
that the protocols would represent three time elements:
time one
(T1),
time two,
(T2),
and time
(T3).
These
three time elements would represent the duration of the
preparatory process,
or advance planning process
(T1:P),
of the translating process
(T2:T),
and of the reviewing
process
(T3: E) .
Note
that such a segmentation procedure might lead
one to believe that this study views the composing
processes as linear processes.
Although it is to be
stressed that there 1S some linearity in the composing
process,
given that it unfolds along a time axis,
with a
beginning and an end,
the position taken here is not a
uni-linear one. Writing as a composing process is
unequivocally recursive,
as understood in composing
research.
In addition,
even though
T1 necessarily
167
'.

---

precedes T2,
T3 does not necessarily follow T2.
In other
words,
T3 representing the duration of the reviewing
process does not always occur at the very end of the
translating process.
The reviewing process
(E)
may occur
at different points in the translating process, as
illustrated by the data under examination
(Appendix X
provides instances of such a pattern) .
This note of caution having been made,
the next
concern is to establish the boundaries of the three major
cognitive processes that are activated In the process of
solving the four writing tasks retained for this study.
Identifying the three maJor components of the
composing process:
P,
Tt
E:
To establish the boundaries
of the major cognitive processes,
shift in focus was used
as the sole criterion.
This shift in focus,
signalled by
pause in manual activit~ is necessarily global in scope
and signals overtly the beginning and the end of the
major cognitive processes,
as defined by Flower & Hayes
(1981)
and Hayes & Flower (1980).
The initial boundary of P was set when subjects
started reading the writing tasks silently or aloud.
The
final boundary was signalled by a shift in focus overtly
expressed by the subjects when they uttered 'Now I am
going to
...
the introduction'
or indicated by a pause
which usually followed the reading of the task and led to
the actual writing of the introductory paragraphs of
168

---

essays.
T was initiated by the first shift in focus
leading to the translating process or by overt cues In
the 'think-aloud' protocols when subjects expressed 'Now
I'm going to start the body of my essay'.
The end of T
was overtly marked off by the shift in focus indicating
that subjects were about to focus on reviewing.
The
same criterion shift in focus was used in determining the
boundaries of E.
As shown in the examination of the
three types of re-reading
(section 3.3.2,
Stage Ill:
Focus on the reviewing process),
E may be an instance of
type III re-reading activity.
The discussion of the
duration of P,
T and E is the concern of section 3.2.3,
Computer treatment of protocols.
The question that arises when attempting to study
the composing process of non-native composers of English
written discourse within the framework of the Flower and
Hayes cognitive model of writing is the following:
To
what extent can such a model that is designed to account
for skilled native writers of written English discourse
be suitable for analyzing the composing processes of
unskilled non-native writers?
The first answer is provided by Flower and Hayes.
Such a model,
they contend,
can serve as a guide to the
diagnosis of writing difficulties.
Secondly,
if the goal
of such an analysis is to account for the cognitive
processes in composing written discourse,
then it seems
reasonable to posit that the Flower and Hayes cognitive
169

---

model should be applicable to any population of composers
in any linguistic situation.
Having clarified the position taken here,
one may
want to express three research questions related to the
application of the Flower and Hayes cognitive model.
These are:
(I)
Assuming that the Flower and Hayes cognitive
model of writing may serve as a diagnostic tool,
what would such a tool reveal about the composing
process of non-native first year university
composers of English discourse?
(2)
In using the Flower and Hayes cognitive model as
a framework,
what type of differences,
if any, would
one observe in the process of solving the four tasks
examined in this study?
(3)
Among the variables that are considered to be
good descriptors of the behaviour of skilled native
composers of written English discourse.
which ones
can adequately help describe the composing behaviour
of the subjects under examination?
3.3.2.
Identifying the episodic structure of composing
processes
The. task was to segment the composing processes of
subjects according to common informalized intuitions
about episodic structure.
It is worth stressing that
170

---

there is no correct way of coding protocols,
as
emphasized by Swarts et al.
(1984).
These researchers
insist on the fact that there is no existing grammar for
analyzing such data.
Knowledge of the data to be coded
is of utmost importance.
As such familiarity with the
data helps develop awareness of the internal logic of the
composing processes of sUbject-writers.
They also stress
that only a de facto decision between two knowledgeable
judges helps discriminate between major composing
episodes and minor composing episodes.
Even though great familiarity with protocols is
helpful in discovering the internal logic of the
protocols to be analyzed,
it was felt in this study that
to reduce the degree of sUbjectivity in segmenting
protocols, and to ensure systematic segmentation, an
observation-segmentation scheme was to be devised.
A
tentative version of that scheme was tried out during the
first preliminary segmentation rounds; that first version
of the segmentation scheme was revised and finalized two
months later.
The revised version was used by this
investigator during the third systematic segmentation of
the data and the 12 independent jUdges who segmented
samples of the protocols so as to establish the
reliability of the segmentation scheme (For illustration
see Appendix XI).
At this point it is to be specified that the
segmentation of the protocols focused only on the major
171
.<

---

composing episodes occurring within the translating
process.
The reasons for attending solely to major
composlng episodes were given in section 3.1.2.
The
decision to examine only composing episodes occurring
within the translating process was determined by
practical issues.
Undertaking to record all the pauses,
as revealed by the verbalization of thought,
would have
been impractical,
unreliable and self-defeating.
This
investigator had no access to a pauseometer nor to any
other sophisticated measuring instrument.'o
As a result,
it was crucial for her to resort to an objective planning
cue,
such as pauses In manual activity.
Only the
translating process could display a systematic pausal
pattern in manual activity.
Given those constraints,
it
seemed reasonable to undertake a detailed segmentation of
the translating process into major composing episodes.
Thus pauses that did occur during the preparatory and the
reviewing
(editing) processes were not attended to.
In determining the episodic structure of the
composing processes,
to ensure rigour and objectivity,
three criteria suggested by Swarts et al.
(1984) were
adopted in this study.
Paus~in manual activity,
coupled
with these three criteria, guided the identification of
the episodic structure of the protocols.
These three
criteria were defined as follows:
JOEven though minimum cut-off points have been defined by speech
analysts
(Goldman-Eisler 1958,
1961,
1968; Boorner,
1965; and Wilkes &
Kennedy,
1970),
this study does not consider these.
172

---

(1)
Shift in focus
(e.g. on audience,
organization,
content,
setting up goals or
plans,
strategies,
task exigencies or
difficulties) .
(2)
Shift in setting up new objectives
(e.g.
influence reader's attitude,
improve writer's
status,
succeed to produce a good quality
product) .
(3)
Shift in train of thought.
Note that this second shift In focus
(type 11)
differs from the first one (type I)
considered in the
identification of the major cognitive processes.
Whereas
the scope of shift in focus type I is global,
the scope
of shift in focus type 11 may be considered local in that
this type of shift in focus operates within a maJor
cognitive process.
Its function is to set off the
initial and final boundaries of composing episodes
within,
for example,
the translating process.
Another
point of interest is that these three criteria
(shift In
focus type 11, shift in setting up new objectives and
shift in train of thought)
illustrate two modes of
planning.
Criteria
(1)
and
(2)
exemplify what Bereiter
& Scardamalia (1981,
1987) refer to as conceptual
planning.
Criterion
(3)
is an instance of semantic
planning as suggested by Butterworth (1975)
and Beattie
(1983).
Both kinds of planning were addressed in Chapter
one,
section 1.2.2.
The question that can be posed is
173
.'.

---

this:
How effective can one be in applying those
criteria to segmenting protocols?
What follows should
provide an answer to the question.
The examination of the episodic structure of the
composing processes was undertaken in keeping with the
three previously discussed criteria, using the
observation/segmentation scheme previously discussed.
As
said earlier, this observation/segmentation scheme was
used by this researcher and the 12 independent judges who
segmented samples of the forty video protocols. Appendix
XII illustrates agreement among jUdges and shows the
result of the application of this scheme by four
independent jUdges,
including this researcher.
To facilitate the segmentation task,
i t appeared
necessary to proceed in stages:
stage I: Focus on scribal activity
(referred to,
in
the computer treatment of the data,
section 3.2.3,
as the 'active state' of a composing episode) within
composing episode.
stage 11: Focus on pauses occurring between
composing episodes boundaries.
stage Ill: Focus on the reviewing process:
(i) Treatment of re-reading activity.
(ii) Treatment of revision.
174

---

Stage I:
Segmenting translating process into
composing episodes
The investigator coded all the 40 hours of
'think-
aloud ' video-protocols,
using the segmentation scheme
discussed above.
After familiarizing herself with all
the composing processes constituting the data under
examination,
she undertook preliminary segmentation of
the data by segmenting the four sets of ten composlng
processes into composing episodes, by first recording the
time of the scribal components
(or 'active state'
elements or event units)
of the composing episodes.
The
segmentation of one hour-long video 'think-aloud'
protocol required one and a half to two hours,
depending
on the style of subject-writers.
A
second round of systematic re-segmentation of
the data to check the accuracy of the recorded time of
the 'event' periods previously identified in the first
segmentation was deemed necessary,
given the limited
sophistication of the video player being used.
The
Sansun video player,
available at CALS and to which this
investigator had access,
did not allow for the slowing
down motion used by other researchers such as Matsuhashi.
The check for accuracy in reaction time
<time for
pressing the pause button) was arbitrarily performed 48
hours after a preliminary segmentation.
In the main,
the
reaction time was accurate; differences In timing,
if
any,
averaged one frame.
Such differences proved to be
175

---

quite negligible, given human limitations.
Note that the
sensitivity of the equipment being used,
a Panasonic
video-recorder player for example, might reveal one
second difference in reaction time.
A third and final
segmentation was undertaken to
ensure intra-coder rellability.
This occurred three
months after completing the check for accuracy ln
reaction time.
The third systematic segmentation was
performed with a Vlew to
checking the investigator's
intuitions about the episodic structure of the data.
FinallY,
a selective comparison between this third
segmentation and the first rounds of segmentation was
carried out:
(1)
to check systematicity in segmenting;
and
(2)
to verify whether doubts or hesitations about
considering specific composing episodes as sub-episodes
of major composing ones had been dissipated.
It appeared that,
after the first two runs of
systematic segmentation of the protocols into episodes,
this investigator had developed sufficient familiarity
with the video 'think-aloud' protocols.
This familiarity
allowed her to overcome potential linguistic,
syntactic
and discursive interferences,
such as anaphora,
deixis,
connectives, which she had to come to terrn$with ln the
first trials and which can heighten segmentation
difficulties that novice judges might experience.
176

---

Stage 11:
Treatment of between boundary pauses:
All the pauses occurring at initial and between
boundaries of the major composing episodes were tallied.
A close examination of pauses occurring between
boundaries revealed that such pauses did not constitute
single units as Flower & Hayes
(1981) might have
suggested.
In some instances,
such as between boundaries,
pauses were observed to be of two types:
those marking
off the end of a previous composing episode and those
signalling initial boundaries of a subsequent composing
episode.
A final composing episode boundary pause
occurred when the subject, after prolonged peLiods of
hesitation-pause,
clearly marked off the end of a given
composing episode by simultaneously saying
'full stop'
and/or by reinforcing the dot on the paper.
As a rule,
final boundary pauses were immediately followed by
initial boundary pauses indicating the beginning of the
following composing episode.
These observations somewhat parallel those made In
research on speech production.
Keeping in mind the
distinction that is made between juncture-pauses and
hesitation-pauses
(Boomer
(1965);
Barik (1968); Hawkins
(1971)),
this investigator felt
it important to
discriminate between final and initial pauses occurring
between composing episodes.
177

---

stage Ill:
Focus on. the reviewing process:
In
focusing on the revision process, special attention was
given to re-reading activity and revision.
The following
will be concerned with how re-reading and revision were
treated.
(i)
Treatment of re-reading activity:
An additional run was undertaken for the treatment
of re-reading occurring within the translating process.
Three types of re-reading were observed.
Type I:
Immediate re-reading of the text just
generated.
Type 11:
Delayed re-reading of the text
generated in one or two previous composing
episodes.
Type Ill:
Re-reading of the text generated in
more than two composing episodes,
including
introductory paragraphs or parts of the body of
essays.
Type I re-reading activity
had an evaluative
function of the text immediately preceding it.
As such,
type 11 re-reading activity sUbsequently closed off the
composing episode that included that text.
Such re-
reading was thus considered as the final boundary pause
occurring at the end of the composing episode directly
related to it.
Type I re-reading was usually followed by
another pause indicating the beginning of the succeeding
composing episode.
178

---

Type 11 re-reading activity was concerned with
scanning.
It essentially consisted in structuring and
planning moves leading to text generation.
Type 11 re-
reading was treated as an initial boundary pause.
Type III re-reading activity consisted of all other
kinds of re-reading that dealt with editing.
Type III
re-reading activity was treated as an editing operation.
Apart from types of reading activity,
revision was
the other area of concern.
(ii)
Treatment of revision:
Since substantial revislon
(beyond editing)
was not
a sub-process common to all the video protocols examined,
reVlSlon was arbitrarily
subsumed under the translating
process.
As a consequence, all those composing episodes
that were clearly revised version
(s)
of previously
generated composing episodes were simply integrated into
those composing episodes.
It was then the case that two
or three composing episodes occurring at different points
in time during a given composing process were combined
into one single composing episode
(Appendix XI
illustrates the procedure) .
It is to be stressed here that the 12 independent
judges who segmented samples of the 40 video protocols
were not required to go through these three stages.
However,
their attention was drawn to the phenomenon of
revision as this could have influenced their segmentation
179

---

of the data.
Given that the judges were essentially
concerned with identifying major composing episodes,
they
did not have to attend to the different types of re-
reading activity observed in the protocols.
Attention
was on familiarizing them with the
observation!
segmentation scheme discussed earlier.
The procedure In
their training sessions is described in the following
section.
Training of independent judges:
As mentioned
earlier,
to ensure inter-observers'
reliability,
12 In-
dependent judges segmented ten to thirty minutes samples
of the four sets of the ten video
'think-aloud' protocols.
Three judges segmented each set.
The samples size
was limited to ten to thirty minutes for practical
reasons.
Those judges were trained by this investigator.
Judges were trained individually or in palrs,
depending
on their availability.
The training sessions lasted two
hours or more,
allowing for personality factors.
Those
training sessions were organized as follows:
i)
Discussion of the observation-
segmentation scheme,
including clarification of
concepts such as
'units of concentration',
minor and major composing episodes.
Demonstrations consisted in viewing samples of
the video protocols and examining samples of
180

---

previously segmented video protocols by this
investigator so as to establish common ground.
ii)
-
Practice segmentation of samples by
independent judges followed.
iii)
-
Feedback from investigator:
Final
clarification of concepts and discussion of
practical issues related to handling of the
observation sheet.
3.3.3.
Composing styles: attempting a typology of
writers
To depict the composing styles of the subjects
involved in this study,
three alternative ways of
analyzing the data were envisaged.
Three questions were
posed.
The first question addressed the extent to which
the two contradictory pedagogical models suggested by
Wason
(1978),
and mentioned in Anderson
(1980),
could
provide a good description of the composing processes
studied here.
The next question concerned the type of
composing styles that would emerge from the present data.
Specifically,
it was thought to be of interest to find
out whether composing styles that would emerge from such
an analysis would present the same configurations as
those discussed In Hayes & Flower (1980).
Finally,
it
was of interest to discover to what extent the
characteristic sequence of problem solving available In
181
.

---

Hayes
(1987)
could be used as a coding scheme to yield a
satisfactory account of the process of composing of the
10 subjects in this study.
Each of the three
alternative ways of setting up a typology of writers are
dealt with in the sections that follow.
Two contradictory pedagogical models
In this particular instance the procedure consists
in classifying the ten sUbjects on the basis of the two
contradictory pedagogical models of composition.
A
reductionist view of these models would separate writers
into planners and non-planners.
The planner-type (or
outliner, as described in Anderson,
1980), starts by
generating ideas.
These ideas are transformed into an
outline which is then translated into a first draft.
The
non-planner or the non-outliner, on the other hand,
generates ideas while composing the first draft.
Note that under an extreme model I,
no additional idea
generation is expected after creating the outline.
Under
an extreme model II, no outline is made.
Two remarks are worth being made with respect to
these models.
Firstly,
implicit in both models is the
provision for producing more than one draft.
Secondly,
the model does not accommodate compromisers.
As such
neither of these models account for the composing style
of skilled writers as,
to use Anderson's words,
'most
182

---

good writers represent a compromise between the two
extremes'
(Anderson,
1980:459).
Given the built-in
limitations of these models,
it is expected that neither
of these contradictory pedagogical model will provide a
good account of the composing styles of subjects.
An
alternative way is to identify idiosyncratic composing
features that are likely to emerge from the protocols.
This alternative perspective is the concern of the
section to follow.
'Configurations' of writers emerging from the data
To arrive at a typology of writers emerging from the
data,
a close examination of the preparatory process is
deemed important.
An attempt is made to identify
distinctive patterns of behaviour that are observed from
the time sUbjects develop awareness of the problem, after
reading the question (0),
to the first composing episode
in the translating process as defined by Flower and
Hayes.
The results of this analysis are shown under
point 4.2.3.
Although the typology of writers developed
on the basis of the actual protocols may give some
indication about the composing styles of the sUbjects
under examination,
such a typology may fail to shed light
·on what the sUbjects do in the process of solving writing
183

---

problems.
Is one likely to observe what Hayes
(1987)
considers to be the characteristic sequence 3l ?
Establishing characteristic sequence of compos~ng
processes
The above question can be g~ven an answer after
applying what can be considered a coding scheme.
This
scheme suggested by Hayes
(1987)
provides an economic
representation of the process of solving a problem.
This
scheme includes SlX points listed below.
1. Finding the Problem:
recognlzlng that there lS a
problem to be solved.
2. Representing the Problem: understanding the
nature of the gap to be crossed.
3. Planning the Solution: choosing a method for
crossing the gap.
4. Carrying Out the Plan
5. Evaluating the Solution: asking 'How good lS the
result?' once the plan is carried out
6. Consolidating Gains: learning from the experlence
of solving (Hayes,
1987:1).
This scheme seems to offer a linear description of
the process of solving problems.
A linear ordering of
these courses of action may be observed when solving
simple problems.
This is not, however,
the case when
dealing with complex problems.
Indeed,
to quote Hayes
31Emphases added.
184

---

In hard problems, we may have to do a great
deal of backtracking. For example, whenwe
evaluate what we have done, we may decide that
our solution is terrible,
and go back to
planning. Or while trying to execute a
solution,
we may discover something about the
problem which will lead us to represent it in
an entirely new way. Retracing of this sort is
characteristic of problems that are called
"ill-defined"
(Hayes,
1987:2).
Clearly,
the process of solving complex problems lS
far from being linear.
The amount of backtracking,
coupled with the forward and backward movement that can
be observed in the problem solving process,
points to the
contrary.
It is therefore hoped that by applying the
above coding scheme to analyzing the process of solving
writing problems, which are instances of complex
problems,
one might get insight into the composing
processes of the subjects being examined.
In identifying the characteristic sequence of the
process of solving complex cognitive problemsby the 10
subjects involved in this study,
the data for examination
are the 10 video protocols on an argumentative unfamiliar
task.
This decision is motivated by the statistical
results discussed in Chapter 4,
section 4.1.
This
characterization of the process of solving a writing task
operates at a group-level of analysis.
An individual-
level analysis will focus on the content of the processes
of representing the problem and of evaluating the
solution.
To this end,
only four subjects are considered
for the reasons indicated in Chapter 3,
section 3.2.3.
185
'.

---

This chapter has undertaken to define the general
perspectives and academic context of this study and the
theoretical framework underpinning the methodological
perspective.
It has also spelled out the underlying
assumptions and reformulateathe major research questions
and hypotheses whose quantitative and qualitative
examinations constitute the foci of Chapters four and
five.
186

---

CHAPTER FOUR
187
'.

---

4. Results:
Quantitative and group-level qualitative
analysis
The results presented here derive from a macro-
level,
or group,
analysis.
It has been felt convenient
to bring together the findings bearing on both the
statistical and qualitative testing of the hypotheses
which involve the sample as a group.
The implication of
such a decision is that individual patterns of response
are not the primary concern of this chapter.
The quantitative and qualitative results at group
level define the two main sections of this chapter.
The
first section (4.1)
is subdivided into three
(4.1.1 -
4.1.3).
The second section
(4.2)
is articulated around
four subdivisions
(4.2.1 - 4.2.4).
Each of these
sections and their subsections are examined,
in turn,
below.
4.1. Statistical results
The statistical analysis is based on the data
displayed in section 4.1.1.
These include subjects'
188

---

performance scores
(Qscore)
on the end products of the
protocols,
the quality of introductory paragraphs
(Paraqual) ,
time spent on the major composing processes
(advance planning (P),
translating (T)
and reviewing
(E) ),
time on planning during the translating process
(Tpause)
and time on overall planning prior to the first
active component of the translating process
(Pplus).
Detailed discussion of how these variables have been
determined and measured has been undertaken in Chapter 3,
section 3.2.3.
However,
three points are worth
summarlzlng here.
First,
Qscore represents the mean sum of scores on
the written products as assessed by two analytic
markers, using the TEEP seven-point attribute
writing scales.
Paraqual stands for the mean overall quality of
introductory paragraphs as independently assessed on
a five-point impressionistic scale by two raters.
Secondly, major composing processes
(advance
planning (P),
translating
(T)
and reviewing
(E))
have been determined within the framework of the
Flower & Hayes
(1981)
cognitive model of the writing
process.
The boundaries of these components,
P,
T,
and E have been determined,
as discussed in section
3.2.3, by pause in manual activity,
coupled with
shift in focus.
Duration of P,
T and E has been
measured by the timing system incorporated into the
189
.,

---

protocols
(see section 3.2.3 on computer treatment)
Pplus
(pausing time prior to the first
'active'
element of the translating process)
equals time on P
+
time on first
'passive'
element in T.
Thirdly,
note that pausing time in the translating
process
(Tpause)
includes only total pausing
signalling the beginning of major composlng
episodes,
as described in section
3.3.2.
As a
consequence,
total pausing initiating minor
composing episodes has been excluded.
Let us recall
that,
as seen in section 3.3.2,
the analysis of
minor composing episodes has not been included in
this study for practical purposes.
Section 4.1 covers three main points.
These are:
- descriptive presentation 4.1.1,
- analytical presentation 4.1.2, and
- summary of the statistical findings 4.1.3.
4.1.1. Descriptive presentation
This presentation essentially deals with the summary
of the data examined quantitatively and with the testing
of the first hypothesis.
Given the procedure--tabular
presentation--used to test hypothesis 1,
i t has seemed
appropriate to include this hypothesis in this section
which deals with the descriptive examination of the data.
190

---

Summary presentation of data
The data that have been quantitatively exploited
are displayed In Table 4.1. and graphically represented
In figures 4.1 - 4.9.
Table 4.1. sets out group mean data for Qscore,
Para~Jal, and for time on P,
T,
E,
Tpause and on Pplus
for the four tasks.
Figures 4.1-4.7 depict the data summarized in Table
4.1.
The purpose of describing these data by way of
charts is to look at task differences as they appear on
each of the seven measures.
To this end,
the 95 percent
confidence interval limits have been used.
Each 95
percent confidence interval is computed by taking the
sample mean,
and adding or subtracting 2.262
(from t-
distribution with 9 df)
times Standard error of the
mean 1 •
Looking at Table 4.1,
row by column, within the
framework of the 95 percent confidence interval limits,
one arrives at the descriptive results presented below.
On the basis of the 95 percent confidence interval limits
representation,
differences between 2 variables are
considered not important if the graphic representation
shows total overlap.
When partial overlap is identified
such a pattern is said to suggest some difference.
1
This formula was used for the 95% confidence interval limits
representation:
x t t df • SE (x)
191
"

---

TABLE 4.1
Group means
tor Qscore,
Paraqual,
Time
(seconds)
on Pi
T,
El
Tpause,
Pplus·.
Tasks
N
Qscore
Pa raqual
P
T
E
'!'pause
pplus
Medn
(SE)
Mean
(SE)
Meo'Jn
(SE)
M€dfl
(SE)
Mean
(SEl
Melln
(SEl
",elln
(SEl
Ar\\l r
10
16.90
(0.5667)
2.70
[O.290'J)
4) I .4
( 12 9.02)
~7)S.7
( 1 6'i
'lO)
198. B
(5 B 92)
~ 64 . B
(123.21 )
4 B 2
D
( 1 46 .07)
Argu
10
1') • 6':>
(0.9889)
1 . 90
(0.]]17)
549 . 1
(B) . 09)
2(,38.9
(189.65)
269.8
(108.961
{,()S.ll
(1l3.1'J)
605.7
(852.]])
Rerf
10
14.50
(1.O461)
2.35
(0.26921
') 7 ') .9
(112.111
2948.4
1170.62)
124.6
(50. J2)
633. '}
010.20)
407.8
(118
491
Reru
10
15.00
(0.5773)
2 . 60
(0.2))))
427 . (,
( 131.9))
2783.1
(138.]6)
) ) J . tl
(93. SS)
654 . 2
(B II .22 )
510.7
(140.57)
Oscore:
Overall
performano:e on ..... rltto=n prodW':t.s;
Paril'ltldl:
Overc.'lll q\\Mlity (,[
int.t'JdlKlory ~)drdgraphs; p, Time on pr-=paralory process
(or
ddvdn,:~
planning); 1':
1'lme on t.ranslat.ing process;
E,
"!me on revleo.ling
(edillnql
p,·ocess;
'T'paus .. · 1'lme on major pausing
in t.rolln9Lallng process;
Pplus:
1'Jme
on plaJ1Jllng prior to the
tirst olIct.ive stat.e element
in t-r,1[lsldting process,
Alg!.
,'rgtJmentollt.!Ve,
famlliollr llls)o;;
Arqu:
Argumentdlive,
un!~ml11~r t~s)o;.
Ref!
Referentl~l, [olImll iar ta9k; Refu: Referential. ullfamlJ lar tas~

---

Though statistical significance cannot be automatically
claimed when no such overlap can be clearly observed,
such non-overlapping is considered to constitute an
important difference.
The first set of descriptive results to be reported
highlights
(given the pattern)
some difference observed
on the four tasks relative to overall performance on
written products
lQscore),
and to time on reviewing
(E).
These results are summarized as follows:
(1)
When comparing the four tasks on the basis
of Qscore,
Fig.
4.1 suggests some difference,
as indicated by the 95 percent confidence
interval limits,
between the overall quality of
an argumentative,
familiar task
(mean Qscore =
16.90)
and the overall quality of an
argumentative unfamiliar task
(mean Qscore =
13.65) .
(2)
Examining the four tasks in terms of time on E,
Fig. 4.2 suggests some difference,
as signalled by
the 95 percent confidence limits,
between times on
the editing subprocess on a referential,
unfamiliar
task
(333.0 sec)
and on a referential,
familiar task
(126.6 sec).
193
'.

---

OVERAlL PERFORMANCE ON WRITTEN PRODUCTS
20
18
16
a
SCORE
14
I
12
10
ARGF
ARGU
REFF
REFU
TASKS
Figure 4.1
95% confidence interval limits representation of overall
performance
(Qscore)
on four tasks:
argumentative.
familiar
(Argf);
argumentative.
unfamiliar
(Argu);
referential,
familiar
(Reff);
and
referential,
unfamiliar (Refu).
TIME ON REVIEWING (editing) PROCESS
Seconds
600
500
400
E
300
100
I
200
O . L . . - - - . . . . : t . . . . . . . . -
ARGF
ARGU
REFF
REFU
TASKS
Figure 4.2
95% confidence interval limits representation of time on
reviewing process
(E)
on four tasks: argumentative.
familiar
(Argf);
argumentative,
unfamiliar (Argu);
referential,
familiar
(Reff); and
referential,
unfamiliar (Refu) "
194

---

The second set of descriptive results to be
discussed points to not important differences among the
four tasks with respect to:
overall quality performance
(Qseare)
on
referential tasks
(Fig.
4.1);
overall quality of introductory paragraphs
(Paraqual)
(Fig.
4.3);
time on advance planning
(P)
(Fig.
4.4);
overall pauslng time prior to the first
'active'
element in the translating process
(Pplus)
(Fig.
4.5) ;
time on translating processes
(Fig.
4.6);
major pausing time on translating
(Tpause)
(Fig.
4 .7) .
OVERAlL QUALITY OF INTRODUCTORY PARAGRAPHS
score
4
3
...J
<I:
::J
f I
o
2
<I:
a:
<I:
c..
0 - ' - - - - - - - - - - - - - - - - -
ARGF
ARGU
REFF
REFU
TASKS
Figure 4.3
95% confidence interval
limits representation of overall
quality of
introductory paragraphs
(Paraqual)
on four tasks:
argumentative,
familiar
(Argf);
argumentative,
unfamiliar
(Argu);
referential,
familiar
(Refl);
and referential,
unfamiliar
(Refu).
195
'.

---

TIME ON PREPARATORY PROCESS
Seconds
(or advance planning)
800
600
400
P
200
0 - ' - - - - - - - - - - - - - -
ARGF
ARGU
REFF
REFU
TASKS
Figure 4.4
95% confidence interval limits representation of time on
preparatory process
(or advance planning:
P)
on four tasks~
argumentative,
faml1iar
(Argf);
argumentative,
unfamiliar
(Argu);
referential,
familiar
(Reff); and referential,
unfamiliar
(Retu)_
TIME ON PLANNING PRIOR TO THE FIRST
"ACTIVE STATE" ELEMENT IN TRANSLATING
Seconds
1000
800
600
P
PLUS
400
I
j
20:
ARGF
ARGU
REFF
REFU
TASKS
Figure 4.5
95% confidence
interval
limits representation of time on
planning prior to the first
'active state'
element in translating
(~lus) on four tasks: argumentative,
familiar
(Argf); argumentative,
unfamiliar
(Argu);
referential,
familiar
(ReEf);
and referential,
unfamiliar
(Refu).
196

---

TIME ON TRANSLATING PROCESS
Seconds
3200
3000
I
2800
T
2600
2400
2200 ..1-
"'---
_
ARGF
ARGU
REFF
REFU
TASKS
Figure 4.6
95% confidence interval
limits representation of
time on
translating process
(T)
on four tasks:
argumentative,
familiar
(Argt);
argumentative,
unfamiliar
(Argu);
referential,
famiJiar
(Retl);
and referentlal,
unfamiliar
(Refu).
TIME ON MAJOR PAUSING IN TRANSLATING
Seconds
PROCESS
900
800
700
600
T
PAUSE
500
400
300
200
ARGF
ARGF
REFF
REFU
TASKS
Figure 4.7
95% confidence interval
limits representation of time on
major pausing
in translating process
(!pause)
on four tasks:
argumentative,
familiar
(Argt);
argumentative,
unfamiliar
(Argu);
referential,
familiar
(Reil};
and referential,
unfamiliar
(Retu).
197
'.

---

Two general observations are to be made at the end
of this descriptive presentation of the data.
Firstly,
as suggested by Table 4.1,
a description of the four
tasks relative to the seven measures
(Qscore,
Paraqua1,
P,
T,
E,
Tpause and Pp1us)
cannot always adequately be
undertaken by considering discourse
(argumentative vs
referential)
and familiarity
(familiar vs unfamiliar)
In
isolation.
Secondly,
the description of the data from
the 95 percent confidence interval limits does not allow
one to conclude that a discourse effect,
a familiarity
effect, or the interaction between discourse and
familiarity account for the suggested differences.
Neither does such a descriptive framework help to
determine the level of significance of such potential
differences.
Such concerns are beyond the scope of this
section and will be dealt with in section 4.1.2.
Having looked at the descriptive pattern of the
data, we shall now examlne the findings in relation to
the first hypothesis.
Testing of hypothesis 1:
The advance planning
pattern of unsuccessful subjects will be the same
irrespective of discourse type and degree of familiarity
with
tasks, whereas the planning pattern of
successful subjects will vary with discourse type and
degree of familiarity with tasks.
198

---

The purpose of the tabular presentation below is to
discover whether the range of time taken on advance
planning of the four tasks would display a planning
pattern that would help to differentiate successful
subjects from unsuccessful ones.
Note that in Chapter 3,
section 3.2.3, subjects have
been grouped respectively into the most successful and
the least unsuccessful sets on the basis of evidence
resulting from the three-way discriminating procedure
(detailed in section Chapter 3,
3.2.3).
This involves:
(1)
ranking subjects in descending order on the basis of
their performance on the four tasks;
(2)
ranking subjects
with respect to the sum of their overall scores on the
four tasks;
(3)
separating subjects who 'hit' at least
once the 17.5 target benchmark score from those who
'missed'
the expected target on the four tasks.
Table 4.2
(a-c)
summarizes the advance planning
behaviour of these two sets of subjects.
Considering the tabular presentation above,
the
following observations can be made.
Table 4.2
(a)
shows
that the five successful subjects
(4,
7,
2,
10,
3)
spent
more time on advance planning (P)
on an argumentative,
unfamiliar task, but that only four of these five
subjects
(4,
7,
2,
10)
spent less time on P on an
argumentative,
familiar task.
As far as the unsuccessful
subjects
( 5,
8,
I,
6,
9)
are concerned,
no consistent
advance planning pattern can be observed.
199

---

TABLE
4.2
Pattern of behaviou.r of successful
and unsuccessful
subjects time
(seconds)
on advance planning.
a)
Time taken on advance planning of individual
tasks.
r",,;,.s
DiscourSe
I ;o-f2! ~nt 1,,:
argumentdt J.ve
Rank
Suhje':t
F
U
F
U
,
•
.c,:>,:-
·;.H.
~
."
2
7
~'H
'08 ~
157
683
--
3
2
~ '1'::
4')0
30~
540
4
10
1,,10
~,O ')
2Q
639
5
·
ill
l~' ~
1 t 7
281
·
,
,
1 1 ) ..
""
'"""
o.
,
7
"
55
89
\\09
8
1
~,I; :;
4S4
608
5' 5
--
9
6
') &:::
49
72'
4. 7 ~
I
\\0
9
CS
'! :'7
24 <;1
460
Bold numbers represe,.t mJ:-.:iroclm time 0n a(1'Jar!ce p!anLl:,r:.
Underlined l:umbe,s
~-e~)~'eSe;;L n,j:<irnwn L:r.,c- 0n aOV,'i'.<:e pl';";lr.!~,,;:.
-
']':'::::',-
1
?"ml \\ (an VI
gunI'.
Sub1eCt
I
F
U
U - F
\\8,
Successful
\\
4
~~. i
1050
2
7
:3 ;~
'1.187
61&
3
2
7 ~6
1030
234
4
\\ "
? ~?
'H4.
1 '11
5
3
:: 0::
4. J::
131
,
Unsuccessful
5
::: Do:
:~ 53:::'
- 52
7
,
1 ~:
164
43
8
1
670
10&9
199
I
,
6
1::91
521
--no
I,
10
9
274
587
313
I,
cl Time on advance plarllling of argum~~ncatlV'" and re{.:'r-2ncJal
casks.
TASKS
Discou:-se
EEF
ARG
REF-ARG
Rank
.sub,€ct
I
Successful
1
4
661
884
177
2
1
,
'"
840
122
3
992
84<
-138
4
\\ 0
~ 1 1 ~
982
-133
5
3
"2 £7
448
16\\
Unsuccessful
C
5
200
2546
-124
,
7
Sf
198
111
8
\\
7<:<:,
1193
447
9
;
611
1 2D 1
590
\\0
.~
1 S~
109
551
!

---

In Table 4.2
(b),
it lS observed that the most
successful set displays positive and fairly large
differences between time on advance planning of familiar
and unfamiliar tasks.
By contrast,
the least successful
set exhibits,
as a rule,
negative and relatively smaller
differences.
Table 4.2
(c)
reveals that,
given discourse
type,
none of the se~point
to consistent advance
planning.
In sum,
these three tables taken together
indicate that the most successful subjects in this study
exhibit a relatively consistent advance planning pattern
of behaviour while their least successful counterparts
fail to do so.
Figures 4.8 and 4.9 clearly illustrate
the difference in behaviour between the successful and
the unsuccessful subjects.
After the examination of the statistical results at
a descriptive level,
the concern in the sections below is
to report results deriving from an analytic perspective.
201
,

---

700
600
soo
·
· 400
•
~
~
~
300
200
lBT
Sl&ECT
#t-IIr/It 2
e-.e--e 3
Q-Q.-Q "
e-&-& 7
E-E-£ 10
Figure 4.8
Graphic representation of the most successful
subjects' advance planning on four tasks:
referential,
familiar (Reff);
referential.
unfamiliar (Refu);
argumentative.
familiar (Argf);
and argumentative.
unfamiliar
(Argu) .
202

---

1500
"00
1300
1200
1100
1000
900
·
·
· aoo
·~
700
L
·~ 600
L
SOO
400
300
200
100
TEST
&--&-it 5
Q.--6-G 6
Figure 4.9
Graphic
representation of
the
least
successful
subjects·
advance
planning pattern
on
four
tasks:
referential.
familiar (Reff);
referential.
unfamiliar (Refu);
argumentative,
familiar (Argt);
and argumentative,
unfamiliar
(Argu) .
203
".

---

4.1.2. Analytical presentation
Two sets of statistical results are presented in
this section:
those yielded by a correlational analysis,
which examines hypothesis 2 and those resulting from
regression analyses that seek to test hypotheses 3-5
quantitatively.
Testing of hypothesis 2:
Assuming that advance
planning plays a central role in the composing process,
subjects who spend more time on advance planning will
score higher on their written products than subjects who
spend less time.
A correlational analysis has been undertaken with a
view to discovering whether there lS a positive
relationship between the overall quality of the written
product
(Qscore)
and time on advance planning
{Pi.
The
results,
that are displayed below,
have been obtained
from such an analysis.
TABLE 4.3
Correlation between Qscore and P on the
four tasks·.
Together
Separately
,
Argf
Arqu
Reff
Refu
Correlation
coefficient
0.1017
-0.3325
0.3369
0.1799
0.4462
Signi !icance
level"
0.5325
0.3478
0.3411
0.6190
0.1961
Argf:
argumentative,
familiar task
Argu:
argumentative,
familiar task
Re:ff:
referential,
familiar task
Re[u,
r(?ferential,
unfamiliar: task
t<
0.05
204

---

Table 4.3 permits two kinds of observation.
Firstly, when the relationship between Qscore and P is
examined, with focus on the four tasks taken together,
the correlation coefficient lS greater than zero
(r:
0.1017) _
However,
the test of the null hypothesis
(HO:
r=O)
gives a significance level 0.5325,
i.e. greater than
O. OS.
Although the data reveal that there is a slight
positive correlation between both variables
(Qscore and
P),
the relationship between Qscore and P is not
statistically significant.
Secondly,
when the relationship between Qscore and P
lS examined with respect to individual tasks,
none of
them point to a significant relationship between Qscore
and P.
In three of the tasks
(Argu,
Reff,
Refu) , Qscore
and P are slightly positively correlated
(0.1799~0.4462)
In one of the tasks
(Argf)
Qscore and P are slightly
negatively correlated (-0.3325).'
After considering the correlational findings
relative to the second hypothesis,
attention is now glven
to the statistical results yielded by the different
linear regressions models used in the quantitative
analysis of the data.
Note that in the presentation of the results,as
yielded by different regression models,
all the models
2SAS computes the significance level using t-distribution
involving r,
with 8 df
(n-2);
that is,
t= r
,,(
(n-ll/(1-r 2 )
) .
The significance level is
independent of whether the correlation
between variables is positive or negative.
205
'.

---

that have been used for the analysis of the data arc
considered useful for explaining the data under
examination if their overall P-values are less than 0.05
(P<0.05).
Similarly,
any covariate is said to contribute
in any way towards explaining the variability in the
response variable if P< 0.05.
Testing of hypothesis 3:
There is a relationship
between the duration of the composing process,
as
measured by time on the preparation (P)
and translating
(T)
processes and familiarity with the task rather than
with discourse type.
A multiple linear regression model
(Model I,
see
section 3.2.4) has been used to find out which of these
covariates
(subject, discourse,
familiarity and the
interaction between familjarity and discourse)
would best
account for the variability in the response variables P
and T.
'rwo analysis of variance tables
(4.4 and 4.5)
display a summary of the results.
Tables 4.4 and 4.5 indicate that the multiple
regresslon model I provides a good description
(R2 =
0.8473)
of the P (advance planning)
element in the
composing processes, whereas it gives a reasonably
accurate description
(R' = 0.6202) of the T (translating)
element in the composing processes.
This model accounts
206

---

TABLE 4.4
Analysis of variance between P and four other
variables' .
Source of
Sum of
Mean
F
variation
OF
Squares
Square
Value
PR>F
R'
Regression
12
4217978 .9
352498 .2417
12.48
0.0001
0 .8473
Residual
27
760383 .1
28162 .3370
Corrected
Total
39
4978362.0
Source
OF
s~ of
F
Squares
Value
PR>F
Subject
9
4057025. 5
16.01
0 .0001'"
Familiarity
1
71740. 9
2 . 55
0 .1221
Discourse
1
78322 .5
2 . 78
0 .1069
Fam"'Disc .
1
10890. 0
0 .39
0 . 5393
• p,
Time on advance pla~,ing (0, preparatory processl
Four other variables: Subject,
filffilllarity,
discourse,
Fam"'Di'icourse
P<O.OOl, very highly significant
TABLE 4.5
Analysis of variance between T and four other
variables· .
Source of
s~
of
Mean
F
variation
OF
Squares
Square
Value
PR>F
R'
Regression
12
6548927.4
545743.950
] .67
0.0024
0.6202
Residual
27
4010240.6
148527.429
Corrected
total
39
10559168.0
Source
OF
Sum of Squares
F Value
PR>F
Subject
9
6047011.7
4.52
0.0011""
Familiarity
1
171741.0
1.16
0 .2918
Discourse
1
318444.0
2.14
0 .1547
Farn*Disc
1
11730.6
0.08
0.7808
T:
Time on translating process
Four other variables: Subject,
familiarity,
discourse,
Fam*Disc
P<0.05,
highly significant
207

---

respectively for 85 % of the variability 1n P and 62 % of
the variability in T.
Another detail of importance 1S
that,
in such a model,
subject accounts for most of the
variability in both P,
with P<O.OOOl
(i.e. very highly
significant)
and T,
with P<O.OOll
(i.e. highly
significant)
Figures 4.10 and 4.11 provide a subject-
wise break-down of time on the advance planning processes
and on the translating processes of the four tasks
discussed in this study.
These figures illustrate inter-
subject variation.
Testing of hypothesis 4·
There is a
relationship
between the quality of the written product and the
compos1ng process,
as measured by the duration of P and
T,
with respect to discourse type and the degree of
familiarity with the task.
Two multiple linear regression models
(Model II:
the
'Full Model'
and Model III the 'Reduced Model')
were used
for the statistical examination of the general hypothesis
4 .
The 'Full Model' which statistically examined
whether,
given the four tasks taken together,
the overall
quality of written products
(Qscore)
is associated with
tasks yields the results as summarized in the analysis of
variance shown in Table 4.6.
208

---

Seconds
,"
g
.~
5
6
1 10 2
9
"
3
7
B
5
7
10 4
1
2
6
9
3
B
5
10 6
2
I
4
1
,
So..bJ .at
I
ARG/F---
r
ARGlUr---
I
RUfF
I
I
REFIU
I
T_'
Figure 4.10
Subject-wise break-down of time on advance
planning (P:
preparatory process).

---

Seconds
T-phase time
~.&
e 3 ;( ... 1 10 5 1 11 9
9'
e I 4 11 3 2 5 7 10
B 4
6
2
3
10 9
I
7
5
9U:lJ..ct
1
ARG/U
I
I
Rff/f
I
I
REf/u
I
T_k
~t~tA
. . . . . P
~ A
Figure 4.11
Subject-wise break-down of time on translating
(T)
process.

---

TABLE 4.6
Analysis of variance between Qscore and five
other variables".
source of
s=
of
Me=
F
variation
DF
Squares
Square
Value
PR>F
R'
Regression
11
157.38057049
15 21541550
3.15
0.0058
0.55]8
Residual
28
134.86]17951
4.81554213
Corrected
total
]9
]02 .24375000
source
DF
Sum of Squares
F Value
PR>F
p
1
2.7580
0.57
0.4547
T
1
5.5396
1.38
0.2502
Task
]
59 .QQQ4.
1o.0S
0.015""
P"'Task
3
18 .0024
1. 25
0.3119
1'"'1'a5).:
]
80.9800
5.50
0.0039"
Qscor"€':
Overall quality ot written products; Five other variables:
P
(time on
advance planning),
T {time on translating processl: Task
(interaction between
discourse and familiarity),
P*Task,
and
T*Task.
P"O.OS,
significant
P<O 01.
highly significanr_
The analysis of variance Table 4.6 indicates that
the full linear regression model provides a relatively
accurate description of the variability in the quality of
the written products
(Qscore).
In this model,
55 % of
the variability in Qscore is accounted for'.
Of interest
are the observations that task and the interaction
between T and task contribute towards explaining the
variability in Qscore.
Figure 4.12 illustrates the
first
point.
The 'Reduced Model'
that examInes the association
between Qscore and P and T,
taking the four tasks
Other sources of variation are factors such as SUbject,
task-
raters and the rating instrument.
This point is developed in a papGr
presented at IATE~L 1991,
EXETER
(see Manouan,
1991).
211

---

Mean Q scores
Oscore
"I.
16
.5
..
13
12
If
::
o
:o
5
•
2
o
F
U
F
U
I-- ~---i
I-- REF---i
Figure 4.12
Hean Qscore Coverall quality) on four tasks:
argumentative. familiar CArgf);
argumentative, unfamilIar
CArgu);
referential, falDitiar (Reff);
and referential.
unfamiliar (Refu).
212

---

separately,
permits the results summarized In the
analysis of variance shown In Table 4.7.
TABLE 4.7
Analysis of variance between Qscore and P and T on
an argumentative unfamiliar task+.
Source of
s~
of
Me=
F
variation
DF
Squares
Square
Value
PR>F
R'
,
Regression
74 7240
37.3621
19 .66
0.0013
0.8489
Res idual
7
13 .3010
1.9001
Corrected
total
9
88.0250
Parameter
Estimate
T
for HO:
Parameter ~ D
PR>T
Standard error of estimate
-,
Intercept
.9246
-J .D9
0.3111
2.6788
P
0.0075
4.05
0.0049"
0.0018
T
0.0047
5.54
0.0006'"
0.0008
P
(Time on advance planning or preparatory proc€ss)
T (Time on translating process)
P<O.OS,
highly signific<U1t
P<O.Ol.
very highly significant
Table 4.7 shows that the reduced linear regression
model provides a good description of the variability In Q
when an unfamiliar argumentative task lS considered.
In
such an instance,
85 % of the variability In Q lS
explained.
Note that In this particular case,
the major
elements In the composing process P (with P<O.0049,i.e.
highly significant)
and T (with P<0006,
l.e.
very highly
. significant)
are both significant predictors of Qscore.
213
'.

---

Testing of hypothesis 4.1.
The quality of the
written product is associated with time spent on pausing
during the translating process
(Tpause).
A simple
linear regression model
(Model V,
see section 3.2.4)
allows the results as displayed in the analysis of
variance shown in Table 4.8.
TABLE 4.8
Analysis of variance between Tpause and Qscore On
an argumentative,
unfamiliar task.
Source of
s=
of
Mean
F
variation
OF
Squares
S~lare
Value
?R>F
~
Re~ression
1
34.8951
34.8951
5.25
0.0511
0.3964
Residual
8
53.1299
6.~412
Table 4.8 reveals that,
when an argumentative,
unfamiliar task is taken into account,
the major pauses
in the T element of the composing process account for 40
% of the variability in the quality of the written
product
(Qscore) ,
(Fig. 4.13 gives a visual
representation of the results) .
Testing hypothesis 4.2.
The quality of the written
product
(Qscore)
lS
associated with familiarity with task
rather than with discourse type.
The results yielded by the multiple linear
regression model
(Model V,
see Chapter 3,
section 3.2.4)
are displayed in the analysis of variance Table 4.9.
Table 4.9 illustrates that the second 'Full Model'
used in the analysis of the data provides a good
description of the data.
This
'Full Model' accounts for
214

---

GHorr
,., GseOr!
,.,
.w
(al Rrll r: 0.2321
.w
(b) Rrfu r:-O.OM2
'~
11
r l
~
•
•
i5-
17
~i5- "l17
•
C
••
•
•
•
0
I ~
g 11
•
'~
'~
.w
•
.w
•
'~
Il
• ..-1
lJ
~
~
...
0
o
•
0.
11
~
0
•
•
@11
U
U
1
r l
r l
r l
•
r l
7
~
~
"
<I!
:>
"
<I!
:>
\\;
r
,
1
. - - - - - - - - . -
,-------,-
0
100
JOO
500
700
90Q
1100
1300
1500
o
100
300
500
700
900
BOO
lYIO
150(J
PholU 1 P4US! I src l
Phut T pausp Isrc)
Dscore
,.,
>-.Dscorr
le) Argf r~ 0.2219
.w
.w
Id) Arqu r: O.b2~~
'~
'~
11
r l
11
r l
~
••
•
~
••
&17
i5- 17
C
s::
II
0
o 11
•
'~
•
••
'~
•
.w
•
.w
•
•
'~
I)
'~ IJ
~
0
o.
•
0.
11
•
~
~ 11
0
U
•
U.
r l
r l
r l
r l
~
~
"<I!
"<I!
:>
J '
0
lOO
3(10
100
700
900
1100
DOO
I~OO
6
10~
300
100
700
900
1100
doe
150Q
Phis! 1 polu5e l SI!C I
Phut 1 pause Isrc)
Figure 4.13
Association between overall quality
(Qscore)
and pausing time
in the translating process of four tasks
(a-d:
Refl: referential,
familiar;
Refu:
referential,
unfamiliar; Argf: argumentative,
familiar;
and Argu:
argumentative, unfamiliar)_

---

77 % of the vu~iubility i~ QscorP.
In this case,
subject
and familiarity
contribute towards explaining the
variability in Qscore.
As such,
subject and familiarity
are the best predicto~s of Qscore.
Figures 4.14 and 4.15
provide a visual representation between,
on the one hand,
Qscore and the subject factor and on the other hand,
Qscore and farriliarity with tasks.
216

---

TABLE 4.9
Analysis of variance between Qscore and nine other variables·
Source o[
Sum
o[
Mean
v21riation
CF
Squares
Squares
F Value
PR>r
R'
Regression
17
232. 0441
13.6496
4.28
0.0009
0.7677
~_e5idual
22
70. 199&
3.1908
Corrected
1;ota1
39
302.207
Source
DF
Sum of
Sauares
F Value
PR>F
Subject
9
189.0562
6 .58
0 .0002···
Discourse
1
2.7562
0 .86
0 .] 62 8
Fami 1 iari ty
1
18.9062
5 .'?3
0 .0235'
P
1
2.1748
0 .68
0 . ~ 17 9
T
1
0 .1690
0 · D~
0
8200
P<-Di5cQurse
1
1 .0085
0 .32
0 .5797
,
To-Oiscourse
1
.0184
1 .89
0
1835
p'Tami 1 iad ty
1
0 .OOOJ
0 .00
0 .9921
""Familiaritv
1
11 .9540
3 · 7 5
0 . 0 659
T
for HO:
Standard
parameter
Estimate
Parameter
PR>T
Er(or- of
-0
Estimate
Intercept.
15.0540
2 . 92
0 . Q079
5 . 1534
,
subJects
1
-1 _9773
-1 .
0 .1724
1 .4018
2
1 .33t-
0.: i
0 .3707
1. ~&22
J
-0 .104.:;
-0.07
0 .9468
1 .5431
,
2 .0340
1 .36
0 .1864
1 .4912
5
1 .4247
0 .68
0 .5019
2 . OB 65
,
-1 .9261
-1 · J 1
0 .2042
1 .4716
7
1 .0009
0 .76
0 .4~&0
1 .3190
8
-1 .9564
-1 .05
0.3064
1 .6704
9
-, . 4844
-2 .93
0.0077
1.5285
10
0.0000
Discou["se
Arg
-, .90B7
-1 .01
0.3221
4.6458
Ref
O. 0000
Familiarity
F
10 .3216
2.16
0.0420
4.779~
U
0 .0000
P
-0.0007
-0.30
0.7694
0 .0026
T
-0.0000
-0.01
0.9946
0 .0016
p"Discourse
Acg
0.0006
0.27
O. 790~
0.0021
Ref
0.0000
T Discourse
Arg
0.0019
1 .22
0.2342
0.0015
Re[
0.0000
P·Familiar
F
-0.0024
~ 1 .11
0.2766
0.0022
: u~core, uveral~ quallty o~ wrltt~n proaucts
Nlne other vanables:
SubJect,
dIscourse,
randl iarity,
P,
'I',
P·Oiscourse,
1'·Disccurse, P"Oiscourse,
P"Famillarity and T"Famil iarity
p",O. 001, very highlY signi ficant::
p<0.05,
significant
217
'.

---

QSCOfe
Qua LIt Y scores
t
7
"
5
2
:5
8
10 1
6
9
10 "
2
7
5
:5
1
8
6
9
2
:5
10 "
7
6
5
I
9
8
"
2
7
5
10 3
8
1
6
9
~Jec:I't
t
l.RFfIF
I
I
2.RF'1"1U
I
I
3.ARGIf'---
,
4.ARGIU
I
TFor
Figure 4.14
Subject-wise breakdown of group scores on four
tasks:
referentiaL
familiar
(Reff);
referential,
unfamiliar
(Refu);
argumentative,
falDiliar (Argf);
argumentative,
unfamiliar (Argu),

---

OSCOfe
16
11
B
o
! 7
6
5
4
3
2
u
Figure 4.15
Association between Qscore (overall quality) and
familiarity
(F: familiar;
U:
unfamiliar) with tasks.
Testing of hypothesis 5:
The quality of
conceptualization,
as measured by independent assessment
of introductory paragraphs
(Paraqual),
rather than time
spent on advance planning (Pplus), will constitute an
index of predictability of the overall quality of those
products.
A simple linear regression model
(Model VI,
see
section 3.2.4) has been used to arrive at the results
displayed in the analysis of variance shown in Table 4.10
219
",

---

TABLE 4.10
Analysis of variance between Qscore and Paraqual and
Pplus on argumentative unfamiliar task.+
Source of
s=
of
Meaan
F
vaariation
Df
Squaares
Squ3.re
ValuE
PR>F
R'
Regression
2
42 4468
21.2234
3.26
0.0999
0.4822
Residual
]
<5
5782
6.~1l2
Corrected
Total
9
88.02<'0
s=
of
Source
DF
Squares
F VcJ1 uc
PR>F
Paraqual
1
]0.0396
3 . 61
0.0688'
pplus
1
0.4105
0.06
0.8090
Qsco,e:
OVErall
quality of written products Paraqual:
Quality of
introductory
f'ar'lgraphs
P<O.lO,
significant
Table 4.10 shows that the simple linear regression
model used accounts for 48 % of the variability in
Qscore.
Given an argumentative,
unfamiliar task,
the
quality of introductory paragraphs
(Paraqua1)
constitutes
a reasonable predictor of the overall quality of written
products when P<0.10).
Note that a significance level
set at P<O.OS will not give significant results; however,
there might be some evidence that paraqual may help
predict Qscore if P<0.10 is considered.
After the examination of the statistical results
with respect to individual hypotheses,
one may want to
pause here and consider what kind of general findings
220

---

have been obtained.
This point is the concern of the
next section.
4.1.3. Summary of statistical findings
The major statistical findings in this study are
dealt with below.
1.- When familiarity with tasks is considered,
successful subjects display a consistent advance planning
pattern while unsuccessful subjects exhibit inconsistent
advance planning pattern behaviour.
2.- Even though for the most part the data reveal a
slight positive relation
(r= 0.1017)
between the overall
quality of the written product
(Qscore)
and the duration
of advance planning
(Pl,
such a relationship is not
statistically significant
(significance level:
0.5325,
i.e. greater than 0.05).
When an argumentative,
familiar
task is considered,
there lS a slightly negative
correlation (r = -0.3325)
between overall quality and
time on advance planning.
3.- There is strong evidence that the factor of
subjects accounts for the variance in the duration of
advance planning process
(P)
and in the duration of
translating process
(Tj.
In this particular instance,
p-
values for the variable subjects are respectively
P<O.OOOl
(i.e. very highly significant) and P<0.0015
(i.e. highly significant).
There is no evidence that the
221
,

---

duration of advance planning is associated with discourse
type.
Neither is there any evidence that time on the
translating process is associated with discourse types or
familiarity with tasks.
In sum,
subject is the only
variable that contributes towards predicting the duration
of both the advance planning process and the translating
process.
4.- The
overall quality of the written product is
associated with tasks
(P<0.0159,
i.e. significant))
and
the interaction between tasks and familiarity with tasks
(P<0.0039,
i.e. highly significant).
Only in the case of
an argumentative,
unfamiliar task lS the
overall quality
of the written product associated with time spent on P
and T.
5.- There is
some evidence that time spent on
planning In the translating process is associated with
the overall quality of the written product only when an
argumentative,unfami1iar task is examined.
6.- The overall quality of the written product is
associated with familiarity with tasks
(P<0.0245,i.e.
highly significant), with tasks and with subjects.
7.- The quality of introductory paragraphs may
constitute a reasonably good predictor of the overall
quality of the written product only in the case of an
argumentative,
unfamiliar task and at a significance
level less than 0.10.
222

---

The summary of the statistical findings ends the
quantitative analysis,
one of the major points in this
chapter.
The next major point to be examined concerns
the qualitative analysis at group-level.
The sections
below report the findings resulting from such a
perspective.
4.2. Qualitative results
The qualitative analysis has been undertaken so as
to provide complementary information and to shed light on
other phenomena that were not or could not have been
captured by the purely quantitative analysis of the data.
To this end hypotheses guiding and arising from the
qualitative perspective,
at group-level,
are examined.
Four main subdivisions have been retained here.
These are:
- Identification of the cognitive processes In
subjects' protocols
(4.2.1);
- Attempting a typology of subject-writers
(4.2.2);
- Episodic structure of the subjects' protoco1s
(4.2.3);
- Summary of qualitative findings at group-level
(4.2.4) .
223
'.

---

4.2.1.
Identification of cognitive processes in subjects'
protocols
It is worth reiterating
the three research
questions underlying the identification of the cognitive
processes:
(1) Assuming that the Flower and Hayes cognitive
model of writing may serve as a diagnostic tool,
what would such a tool reveal about the composing
processes of non-native first year University
composers of English discourse?
(2)
In using the Flower and Hayes cognitive model as
a framework,
what type of differences,
if any,
would
one observe in the processes of solving the four
tasks examined in this study?
(3) Among the variables that are considered to be
good descriptors of the behaviour of skilled native
composers of written English discourse,
which ones
can adequately help describe the composing behaviour
of the subjects under examination?
The examination of the first two research questions
has led to the following two tabular representations
(4.11 and 4.12).
These respectively display time on the
translating and reviewing processes in the course of
solving the four tasks.
The advance planning or
preparatory process (P)
having already been dealt with in
section 4.1.1,
reference to time on P will be made with
respect to Table 4.2.
224

---

TABLE 4.11
Time
(seconds)
on the translating process.
Discourse
}\\.rgum..nt"tlv,;,
Referentllll
FlUI\\ili"rity
I
F
U
F
U
Rank
Subject
successful
1
4
3271
3035
3199
3082
2
7
2420
2676
2235
2227
3
2
3224
3108
2966
3035
4
10
2101
2579
2018
3649
5
3
2952
3236
2994
2792
Unsuccessful
6
5
2170
2494
2485
2124
7
8
3511
3493
3568
3614
8
1
2766
2294
3362
2627
9
6
2284
1770
3082
3049
10
9
2058
1704
3575
2632
Final ranking retained as a result of the three
discriminating procedures discussed in section
3.2.3. Underlined numbers denote minimum time on
translating. Bold numbers denote maximum time on
translating
This table shows that,
except for subject 5,
unsuccessful subjects
(8,
1,
6 and 9)
spend less time on
the translating process when solving a persuasive,
unfamiliar task.
Three of these subjects display longer
time on the translating process when solving a
referential,
familiar task.
No general pattern con®on to
most successful SUbjects can be observed.
225
'.

---

TABLE 4.12
Time
(seconds)
on the reviewing process_
Discourse
A.rqUll\\entative
YlllIlilierity
F
U
F
u
Renlo::
SUbject
Successful
1
4
101
0
168
84
2
7
331
241
267
237
3
2
72
0
67
75
4
10
437
382
337
446
5
3
453
83
407
656
Unsuccessful
6
5
0
0
0
0
7
8
0
0
0
0
8
1
226
517
0
489
9
6
0
1077
0
502
10
9
368
398
0
841
Final ranking retained as a result of the three
discriminating procedures discussed in section 3.2.3.
As indicated in Table 4.12,
the reviewing process
was not activated in 14 protocols
(viz.
35%)
out of a
total of 40;
12 (viz.86%)
out of these 14 concerned least
successful subjects
(rank 6th-10th).
None of this subset
resorted to the reviewing process when solving the
referential,
familiar task.
Two subjects
(5 and 8)
from
that subgroup never activated this cognitive process
during their processes of solving the four tasks included
in this study.
As a rule,
the least successful group
spend more time on the reviewing processes of unfamiliar
tasks.
226

---

These two tables
(4.11 and 4.12)
as well as the one
displaying time on advance planning
(4.2)
indicate that,
generally speaking,
these protocols display the three
major cognitive processes
(planning,
translating and
reviewing) .
These tables as well as Fig.
4.8-4.9 depict
variability among subjects with respect to the three
major processes.
The next question related to the identification of
the major cognitive processes is:
In addition to these
three major ones, which of the cognitive behaviours
characterizing good native writers can be observed prior
to and during the act of writing of the subjects under
examination.
Using Stallard (1974)
and other studies
(Pianko
(1979);
Perl
(1978); Raimes
(1985bV as a
descriptive framework,
the following matrix has been
drawn.
TABLE 4.13
Matrix for observed cognitive behaviours.
Subjects
1,10,6
4
7
2
5
3
,8,9
.
Planning behaviour
-
+
+
+
-
+/-
..
Revision Behaviour
-
+
-
-
+
-
Audience awareness
+
-
+
-
-
-
Consideration of
purpose
+
-
-
-
-
-
stylistic concerns
+
+
-
+
-
-
stopping to read at
intervals
+
+
+
-
-
-
+
, p resence at: DehaVlour
absence of behaviour
refers to some form of advance outlining
*
**
refers to some form of revision beyond cosmetic editing
227
.'.

---

The matrix for observed cognitive behaviours permits
a first classification of subject-writers into six
categories.
As shown in Table 4.13 only sUbjects 4 and 7
exhibit more than half of these features.
Both sUbjects'
protocols show stylistic concerns and stopping to read at
intervals (concern for style is shared by subject 5, and
re-reading is also exercised by subject 2).
While
subject 7 displays some form of advance outlining (just
as do subjects 1,
2, 5, 6, and 10)
and revision beyond
cosmetic editing (shared with subject 3),
subject 4
exhibits no such patterns.
Instead,
audience awareness
(also explicit in subject 2's protocol)
and consideration
for purpose are evident in his protocol.
The application of the Flower and Hayes cognitive
model, coupled with the cognitive behaviour matrix to
analyzing the protocols, enables one to depict
differences among sUbjects and provide objective bases
for comparing writing styles.
Such a descriptive
analysis, however,
is bound to be limited to a surface
level of inquiry.
To gain greater insight into subjects' composing
processes,
three other descriptive attempts are ventured.
The first attempt is a straight application of the two
contradictory pedagogical models
(based on Wason,
1978 as
discussed in Chapter 3, section 3.3.3).
The second
descriptive procedure seeks to identify composing styles
as they emerge from the data being considered.
Finally,
228

---

an exposition of the classification resulting from
Hayes's coding grid is available.
In applying the coding
scheme based on Hayes
(198~ as seen in Chapter 3, section
3.3.3
,
the underlying research question is whether one
is likely to observe what is claimed to be the problem
solving characteristic sequence in those subjects'
composing processes.
4.2.2. Attempting a typology of writers
The object of this section is to arrive at a
typology of those subjects'
composlng styles after
succeSSlve consideration of the three descriptive
frameworks
(the two contradictory pedagogical models,
composing styles emerging from the data and the coding
scheme based on Hayes
(1987» discussed above .
. Specifically,
the aim here is to seek answers to three
questions:
(1)
Could the two contradictory pedagogical
models provide a good description of the composing
processes under examination?
(2) would
the composing
styles emerging from the data approximate the same
configurations as the ones described in Hayes and Flower
(1980)?
(3)
To what extent can the coding scheme derived
from Hayes
(1987)
offer an adequate characterization of
the composing processes as revealed in the protocols?
229
•

---

Application of the two contradictory pedagogical
models
The answer to the question seeking to find out
whether these models are adequate enough to account for
the data in this study is summarized in Table 4.14.
TABLE 4.14
Categorization of composing styles based on Wason
(1978).
Model I
Model IT
Outliner
Non-outliner
Subjects
Subjects
1,2
3
Compromisers or Mitigators
Subjects
6,7
I
Extreme
Extreme
Outliners
Non-outliners
Subjects
Subjects
5,10
4,
8,
9
Two remarks arlse from this table.
Firstly,
the
subjects under examination do not seem to lend themselves
to being classified as either outliners or non-outliners
(or extensive revisers vs non-revisers
(Beach,
1976), or
radical outliners vs radical brainstormer (Reid,
1984a).
Despite their potential extension to include extreme
cases of outliners and of non-outliners,
both models
fail,
as expected,
to provide comprehensive and accurate
characterizations of the composing processes.
Variety in
composing styles emerging from the data is not captured.
For instance,
by labelling subjects 5 and 10 as extreme
230

---

outliners,
and subjects 4,
8,
and 9 as extreme non-
outliners,
one unavoidably overlooks important details
that differentiate these subjects.
Table 4.13
clearly
instantiates these differences.
Secondly,
these models rest on two assumptions:
(1)
Planning 1S necessarily externally represented;
(2)
Planning 1S restricted to the early phase in the
compos1ng process.
As will be shown later,
these
assumptions are questionable.
In short,
these two
contradictory pedagogical models do not provide a valid
representation of the composing processes that emerge
from the present data.
A close examination of the 40
protocols suggests four different composing styles
emerg1ng from the data,
as presented in the following
diagram:
, - - - - - - - - - - - - - - A - - - - - - - - - - - - - - ,
f - - - - - - - B - - - - - - - - - - - ,..
DETAILED
c
Q
- - - - - - >
------>1 FORMULATION I
Style A:
Q ---------> Exposition
Style B:
Q ----->Plan-----> Exposition
Style c:
Q---->Formulation--->Plan---->Exposition
Style D:
Q---->Forrnulation----->Exposition
As a rule,
style A composers
(subjects 8,
9)
start
generating text immediately after reading the question.
231
.

---

Subject 8 is inclined to rehearsing composing episodes
before committing them onto paper,
whereas subject 9
tends to fall silent.
Only once,
in the process of
solving the persuasive,
unfamiliar task,
did subject 9
spend time on articulating the problem that was set.
Style B composers
(subject 10)
read question,
write
out a detailed outline,
then develop the plan into a
text.
Subject 10 refers to his outline after completing
individual major composlng episodes.
Such a subject
systematically spends time on editing at the end of the
translating process.
Style D composers
(subjects 3(4)
read question,
overtly attempt to capture the gist of the
problem,
then proceed to generate text.
Subject 3 lS
singled out by her use of
'quick-writing'
technique and
by her systematic production of several drafts,
two to
three,
depending on tasks.
Subject 4 exhibits neither of
these characteristics nor does he produce an outline.
However,
subject 4 plans as he goes along and makes a
seemingly powerful use of the evaluation process.
Subject 4 fails twice,
on a referential familiar and on
an argumentative unfamiliar tasks,
to complete the tasks
given at the end of the allocated time.
Note that,
despite this failure,
subject 4 succeeds in 'hitting'
the
target score of 17.5.
While composing styles A,
Band D include one or two
subjects,
composing style C consists of half of the
subjects
(1,2,5,6,7)
in this study.
variation in style
232

---

surfaces,
as might be expected.
In addition to
formulating, i.e. overt attempt to clarify the problem at
hand,
the subjects in this category may either generate a
detailed outline or seriate points to be developed during
exposition.
Note that subject 5 is the only composer In
this subgroup who systematically produces a detailed
outline.
Unlike subject 10, who is a style B composer,
subject 5's plan is written out in full sentences.
His
advance planning includes conceptual planning,
semantic
planning,
syntactic and lexical planning.
Another point
worth making is that subjects 5 and 8 never succeed in
solving the assigned tasks in their entirety at the end
of the allotted time.
To conclude,
a general observation is that composing
styles A and B provide a relatively accurate description
of the composing processes of the subjects classified as
such. The protocols of those subjects reveal that those
subjects'
composing processes are linear irrespective of
discourse type or degree of familiarity with tasks.
Such
composing processes unfold along fixed and rigid patterns
that can be easily predicted.
Composing style C does not
permit a distinction between successful and unsuccessful
subject writers.
A confrontation of composing style C and the matrix
. for observed cognitive behaviours enables one to refine
the classification of subjects into composing styles.
As
indicated by the cognitive behaviour matrix (see Table
233

---

4.13),
subject 7 exhibits cognitive strategies that
distinguish her from the other subjects in composing
style C.
Similarly,
subject 2 cannot be lumped into the
same composing style as subjects 1 and 6.
Subjects 1 and
6 could adequately be classified into the same set as
subjects 8 and 9,i.e. composing style A.
Such regrouping
seems justified if the first classification based on the
matrix of cognitive behaviours is taken into account.
Furthermore,
a close look at reformulation by subjects 1
and 6 suggests that these subjects limit their overt
formulation of the problem to paraphrasing.
No instances
of powerful and efficient activation of cognitive
strategies are available.
This point is further
developed in the analysis of subjects' processes of
problem finding
(See the discussion on the characteristic
sequence of composing processes below) .
One may want to pause here and put the following
question:
How do these composing styles emerging from
the protocols compare with those identified in Hayes and
Flower
(1980)?
An answer to this question is provided in
chapter 6.
Another observation related to the composing styles
identified in this study is that none of these composing
styles allow one to capture the characteristic sequence
observed in the process of solving cognitive problems.
The question is whether such students exhibit such a
sequence. The next sub-section addresses this issue.
234

---

Characteristic sequence of composlng processes:
Application of Hayes's 1987 scheme
To establish the characteristic sequence of the
process of solving cognitive problems by the 10 subjects
involved in this study,
the protocols on the
argumentative,
unfamiliar task serve as data.
The
question being investigated lS whether a coding scheme
based on Hayes
(1987)
could satisfactorily account for
the composing processes of those subjects.
The characteristic sequence matrix set out in Table
4.15 permits the following
observations:
TABLE 4.15
Matrix tor characteristic sequence of processes based on
an argumentative,
least
familiar task .
.
Processes
Subjects
Fpb
Rpb
PsI
Cop
Evsl
Cngn
I
+
+
+
+
2
+
+
+
+
3
+
+
+
+
4
+
+
+
+
+
5
+
+
+
+
+
6
+
+
+
+
7
+
+
+
+
+
8
+
+
+
9
+
+
+
+
10
+
+
+
Fpb
Finding the problem
Rpb
Representing the problem
PsI
Planning the solution
Cop
carrying out the plan
Evsl
Evalu&ting the solution
Cngn
Consolidating gains
235

---

1.- No instances of consolidation of gains
(Cngn)
are observed.
2.- The process of evaluating solution
(Evsl)
is scarcely activated.
Only three subjects
make use of it,
though at different levels.
3.- The other processes are observed in most of the
protocols.
Except for two subjects whose verbal
protocols do not overtly present any form of problem
representation,
for the most part subjects resort to
internal and/or external problem representation.
4.- There is evidence that these processes do
not occur linearly.
The process of finding the
problem in the form of reformulation,
for
example,
occurs after problem representation.
Fine tuning of problem representation occurs
when and where necessary.
5.- The application of the coding scheme derived
from Hayes
(1987)
provides evidence that the
internal structure of composing episodes,
as
identified in the translating process,
lS cyclic.
The alternative occurrence of planning of solution
and carrying out of plan is a case In point.
In
other words,
there is strong evidence that composlng
episodes are always initiated by planning,
signalled
by filled and unfilled pauses In scribal activity,
236

---

hesitation or structuring moves.'
Such planning
moves are immediately followed by the execution of
the plan.
In the main,
problem finding
(Fpb)
lS reduced to
reading the task.
Very rarely do subjects attempt to
offer a personal reformulation of the problem to be
solved.
No differences are apparent in the process of
carrying out the plan
(Cop).
Variation,
however,
can be
observed in the processes of problem representation
(Rpb) , of solution planning
(Psl),
and of solution
evaluation
(Evsl).
To capture these differences,
a
content analysis of these processes as they are revealed
in five protocols has been undertaken.
Looking at 5
protocols from this perspective is the object of Chapter
5 .
After identifying the major cognitive processes and
the compoSlng styles of the ten subjects under study,
another issue to be addressed is the episodic structure
of these protocols.
The main concern is whether an
examination of the data from such a perspective would
yield complementary information about composing processes
in general and about individual differences among
subject-writers.
4
Filled pauses here refer to scribal inactivity coupled with
verbalization of thought.
Unfilled pauses indicate both scribal
inactivity and absence of verbalization of thought.
237

---

4.2.3.
Episodic structure
It is to be recalled that the analysis of the
episodic structure considers subjects' protocols in their
entirety.
In other words, the focus is not on the
product as such.
An attempt has been made to segment the
translating process as defined by Flower & Hayes
(1981)
and Hayes & Flower (1980).
In doing so,
the guiding
principle has been to capture the mental activity behind
the verbalized words, attending to three criteria defined
in terms of shift in focus,
shift in setting up new
objectives and shift in train of thought (as discussed in
Chapter 3, section 3.3.4).
Determining the episodic structure is far from being
an easy and neat undertaking.
Concern for achieving
reliability in this respect has resulted in the
painstaking procedure described in Chapter 3,
section
3.2.3.
The segmentation of the protocols into composing
episodes has led to two findings worth reporting.
The
first major finding yielded by the segmentation of the
translating process into composing episodes is the cyclic
nature of composing episodes.
As evidenced by the
presence,
in composing episodes, of an 'active' state
element and a
'passive' state element (see Chapter 3,
section 3.2.3 dealing with the computer treatment of
protocols),
composing episodes always include two main
elements:
(i)
pausing
(in manual activity)
and
(ii)
238

---

event
(scribal or manual activity).
It lS to be recalled
that in the computer treatment of the episodic structure
of the protocols,
the pausing element is referred to as
the passive state of an episode and the event element is
referred to as the active state In an episode.
Both
constituents alternate throughout the translating
process.
The segmentation of the protocols into composlng
episodes provides an alternative source of evidence that
the internal structure of composing episodes is cyclic,
in addition to the application of Hayes
(198?)
to the
data.
Indeed,
it has been observed that planning of
solution and execution of the plan alternate within
composing episodes.
In other words,
identifying the
characteristic sequence of composing processes and
identifying the episodic structure of protocols have
revealed that composing episodes are systematically
initiated by planning leading to scribal activity.
Such
a finding is extremely important,
as will be shown in the
discussion of the results.
The second finding to be highlighted concerns the
nature of between-boundary pauses.
Close focus on the
translating process has shown that between-boundary
pauses are not all alike.
Two types of such pauses can
be identified (see Chapter 3,
section 3.2.3).
These
pauses serve two different functions:
(i)
episode
closure,
and
(ii)
episode initializer.
Pauses with a
239
"

---

closure function signal the end of previously generated
composing episodes.
Pauses with an initializer function,
on the other hand,
indicate the beginning of a subsequent
episode.
In the main,
a between-boundary pause with a
closure function is immediately followed by another
between-boundary pause with an initializer function.
Another observation of some interest resulcr from
examining the reviewing process occurring within the
translating process.
Such a
focus indicates that re-
reading as an activity is a recurrent strategy used by
most subjects.
Three types of re-reading have been
recorded:
Type I
re-reading:
This first type concerns
immediate re-reading of the text just generated and
has an evaluative function of that text.
Type I
re-
reading closes off the composing episode which
includes that text.
It is usually followed by
another pause In scribal activity,
which indicates
the beginning of the succeeding composing episode.
Type II re-reading is a delayed re-reading of the
text generated in one or two previous composing
episodes.
This second kind of re-reading deals with
scanning and consists essentially in structuring and
planning moves leading to execution of solution.
Type III re-reading covers more than two composing
episode.
It includes all the other kinds of re-
reading related to editing.
240

---

Among the three types of re-reading examined in Chapter
3,
section 3.2.3,
type Ill,
editing,
is the preferred one
1n that it occurs in the protocols of most subjects.
Type I,
as an evaluative strategy,
is limited in use.
A fourth finding 1S that substantial revision beyond
editing is not common practice among subjects.
Finally,
the segmentation of the protocols into composing episodes
displays variability in pausing during the translating
process among subjects.
Note, however,
that Table 4.16,
which displays the frequency of composing episodes by
subjects on the four tasks,
does not reveal such
variation.
TABLE 4.16
Number of major composlng episodes as identified in
the
protocols.
Discourse
Arg
Ref
Familiarity
Familiarity
Familiarity
Falllil ier
lJnt!l/l'llliar
FaJrIi liar
Unfllll\\iliar
Subject
1
19
21
22
27
2
20
22
20
13
3
10
14
14
20
4
26
15
20
22
5
19
15
17
18
6
18
16
25
26
7
22
20
16
15
8
26
31
22
19
9
18
10
22
15
10
19
25
20
19
All
197
189
198
194
Mean
20
19
20
19
Min
10
10
14
13
Max
26
31
25
27
241
.'.

---

The mean number of composlng episodes lS basically
the same across the four tasks.
The question is,
What
does the apparent variability In pausing styles, as shown
in Appendix X, which displays subjects' pausing time
during the translating process,
reveal?
The statistical examination of this question' has
been articulated around two questions:
1.
Is the duration of pausing during the
translating processes of individual subjects the
same?
Conversely, are there differences in pause
duration among subjects?
2.
Is the duration of the active element In
composing episodes related to the duration of the
immediately preceding passive element?
To answer the first question,
an examination of the
mean duration of pausing time during the translating
phase has been undertaken.
Fig. 4.16 illustrates the
point.
Fig. 4.16 summarlzes the results concerning the
first question.
Here again,
variation is observed among
subjects.
Examination of Fig. 4.16 reveals that this
variation is to be attributed to subjects 2,
4 and 9.
5
This statistical question is post-hoc in nature in that
it
results
from the qualitative analysis.
242

---

(Seconds)
Mean T-phase pause duration
<
.
~
~
.3
1
5
10 9
6
~
2
9
7
5
8
.3
1 10 6
2
~
7
9
B
,
5
.3
10 6
2
9
7
B
~
5
1
6
.3
10
So..bJ-.ot
AAGIT------;
IIRG/U
I
~Fff------
REFIU
r
T_'
Figure 4.16
Hean
translating pausing time on
four
tasks:
rererential.
ramiliar
(Reff);
re&rential,
unfamiliar
(Refu);
argumentative.
f~miliar (Argf);
argumentative.
unfamiliar
(Argu).

---

The correlational analysis performed to investigate
the strength of the association between the duration of
the active element in a composlng episode and that of the
passive element in that composlng episode shows no
relationship between the duration of both variables.
4.2.4.
Summary of qualitative findings at group-level
The purpose of this section is to attempt to
integrate the different findings resulting from the
qualitative examination of the data at macro-level.
In
this respect,
attention lS on showing how each of the
three descriptive techniques
(the identification of the
cognitive processes,
the categorization of writers'
composlng styles and the identification of the episodic
structure of the protocols)
provide complementary sources
of evidence.
A first finding is that the three major cognitive
processes as understood by Flower and Hayes
(planning,
translating and reviewing)
are observed in most of the
composlng processes analyzed in this study.
The
processes of problem finding,
problem representation,
planning of the solution, as defined in Hayes
(1987),
are
also apparent in the protocols.
There is evidence that
in both cases these processes do not necessarily occur
sequentially.
The reviewing process may be activated in
the translating process.
Problem finding may take place
244

---

after problem representation.
When needed,
problem
representation may be invoked during the solution of the
problem.
Another set of findings,
which evolves around
comparing the most successful subjects with the least
successful ones,
is that,
as observed in section 4.1.1,
time on advance planning constitutes a distinctive
feature between successful and unsuccessful subjects,
as
determined in Chapter 3,
section 3.2.3.
The most
successful subjects,
as opposed to their least successful
counterparts,
demonstrate consistent advance planning
behaviour.
A second discriminating point between both
subgroups is that the least successful subjects exhibit a
consistent pattern of translating.
Four of these
subjects spend maximum time on the translating process of
a referential, familiar task and minimum time on the
translating process of an argumentative unfamiliar task.
By contrast,
the most successful subjects display no such
consistency in their translating process.
The translating pattern of the least successful
sUbjects seem to indicate that this subgroup is likely to
spend little time on trying to solve demanding tasks.
In such a situation,
this group seems to compensate for
the limited time on the translating processes of
·unfamiliar tasks by spending longer time on reviewing.
A third difference that sets both subsets apart is the
activation or non-activation of the reviewing process.
245

---

As a rule, when the Flower and Hayes cognitive model of
the writing process is considered,
successful subjects
can be said to be systematic reviewers,
whereas
unsuccessful subjects are not.
Additionally,
two of the
three subjects who activate the process of evaluation of
the solution, as understood in Hayes
(1987),
belong to
the successful subset.
A third set of findings bears on the episodic
structure of the protocols.
Two major findings are worth
emphasizing.
Firstly,
composing episodes present an
inherently cyclic nature.
Pausing indicating planning
and scribal activity alternate within composing episodes.
Secondly,
two types of pauses in scribal activity occur
between boundaries,
those functioning as initial
boundary markers and those functioning as final boundary
markers.
Fourthly,
there is no relationship between the
duration of the activity and that of the immediately
preceding pausing element within the composing.
As a general conclusion to this chapter,
a number of
points are worth emphasizing.
The composing process
reflects a cyclic structure In which pausing and event
occur alternately.
Such an observation points to the
need to reconsider the importance given to advance
planning in the process of composing written discourse.
Both the quantitative and qualitative treatment of the
data at macro-level provide evidence for individual
246

---

differences ln the processes of solving and in the
performance of the four tasks examined in this study.
As
a rule,
subjects respond differently to task demands and
specificity,
including familiarity and the interaction
between familiarity and discourse type.
There is clear
indication that the most successful subjects adjust their
cognitive strategies to familiarity with tasks.
Discourse type as a maln effect does not seem to elicit
characteristic cognitive patterns of behaviour
irrespective of the degree of success in solving the
writing problems in this study.
Although pausing patterns during the translating
processes of the four tasks reflect inter-subject
variability, varlance in pausing patterns can be ascribed
to three subjects.
Finally,
there are indications that
successful problem solving results from optimizing
efficiency in activating all sorts of cognitive
strategies in the processes of problem representation,
planning of solution and evaluation of solution.
Such a
conjecture constitutes the working hypothesis underlying
the close examination of individual protocols to be
carried in Chapter 5.
247

---

CHAPTER FIVE
24B

---

5.
Analysis of individual subjects
It was suggested in Chapter 4 that a content
analysis of individual protocols should help pin down
differences in subjects' processes of problem solving.
This involves problem representation (RpbJ,
planning of
solution (Psl)
and evaluation of solution
(EvslJ.
This
chapter undertakes to identify these differences by
examlnlng filled pauses in subjects' verbalization of
thought.
It is assumed that such pauses reveal
different strategies that are activated by individual
subjects to meet the demands of processes such as
understanding of the problem to be solved,
planning of
solution and evaluation of solution.
In this respect,
hesitation moves such as halting hands as well as fillers
such as
'all right' and 'okay' are excluded Slnce they do
not lend themselves to a productive content analysis.
Interest is also in discovering how individual subjects
respond to specific task demands.
The data for the content analysis might have
included samples of protocols produced by both successful
or unsuccessful subjects,
as defined in this study.
Such a perspective would have permitted a systematic
249

---

comparison of cognitive strategies activated by
relatively expert and novice sUbject-writers.
Another
perspective would have involved a comparison of the five
most successful subjects constituting the upper group so
as to arrive at differential profiles among members of
this subset.
A third way of undertaking the content
analysis of individual protocols is to consider three
experts and one novice, as is the case in most protocol
analysis in writing research (see studies by Flower and
Hayes) .
This perspective is adopted in this chapter.
Focus is on sUbjects 4,
7 and 2,
the three top sUbjects
in the most successful group,
and on subject 5,
the top
sUbject in the least successful group.
It might be interesting,
for example,
to discover
whether different cognitive processes and strategies are
invoked by sUbject 4, the most successful in the sample,
and sUbject 5, a borderline case, who never 'hits' the
17.5 target score and whose overall total score differs
from that of sUbject 3, ranked fifth in the upper group,
by one point (Qscore 64 vs 63; see Table 3.7, Chapter 3,
section 3.2.3, under the heading of Discriminating among
successful and unsuccessful sUbjects).
Such differences
may prove useful in accounting for qualitative variation.
As to the selection of protocols resulting from
solving an argumentative. unfamiliar task,
this decision
is statistically motivated.
Given that for the most part
the hypotheses in this study are validated when an
250

---

argumentative task is considered,
it seems reasonable to
give special attention to the processes of solving that
task.
Additionally,
given that familiarity is important
in accounting for differences in overall performance,
it
seems justified to examine the process of solving an
argumentative,
familiar task so as to find out:
(1)
whether or not a successful subject resorts to the same
cognitive strategies when solving a familiar and an
unfamiliar task within the same discourse type;
(2)
how
two subjects,
one successful and one unsuccessful,
behave
in their processes of solving a familiar task,
when
discourse is kept constant.
As regard to the conventions when making reference
to extracts from individual protocols,
the following are
to be specified:
Capital letters and numbers in brackets
have been used to facilitate reference to individual
protocols.
A reference such as
(XIII S4 T3,
21-32)
should be interpreted as follows:
XIII: Appendix XIII.
S4: Subject 4.
T3: Task 3.
21-32: Lines 21-32.
Note also that in extracts from individual protocols
a period after a
filler,
such as So.,
has the same
. function as an ordinary period, whereas a period within
251
•

---

brackets
I.) indicates a pause In the external
verbalization of thought.'
Two concepts that need to be clarified are process
and strategies.
In Chapter two,
process was opposed to
product;
the former signals the mental operations that
underlie the act of composing while the latter represents
the end-result of such an activity.
The question here is
whether process and strategies are two distinct
constructs.
Bialystock
(1990)
raises and thoroughly
examines this question, without offering, however,
a
hard-and-fast definition of these terms,
which probably
attests to the difficulty of the issue.
For the purposes of this study,
the definition of
process suggested in Chapter two is retained.
Strategies
are defined as the relatively observable manifestations
of a given process.
In other words,
strategies are the
mental devices that are activated towards a process;
these exhibit three features
(suggested in Bialystock,
1990)
(i)
effective
(or productive),
(ii)
systematic and
(iii)
finite
(not an idiosyncratic creation of the
problem solver) .
Problem representation,
planning of
solution and evaluation of the solution are instances of
the processes that underpin problem solving.
Re-reading,
reformulating and questioning exemplify the strategies
that are invoked by the subjects in this study.
1
Distinctions between different kinds of pauses as suggested by
Crystal
(1969),
and referred to in Garman
(1990),
are not considered
here.
252

---

This chapter covers four ma~n points:
- descriptive presentation of data;
- analysis of responses to the demands of tasks 3
and 4;
- analysis of three cognitive processes accessed by
four sUbjects;
- summary of findings as revealed by a content
analysis of the data.
The first main section (5.1)
g~ves a presentation of
subjects'
responses to task demands and of cognitive
strategies that are called on in the processes of solving
tasks 3 and 4.
The second main section (5.2)
~s
primarily concerned with examining how subjects respond
to the constraints built into an argumentative,
unfamiliar task and an argumentative,
familiar task.
The third main section
(5.3)
is devoted to a content
analysis of five individual protocols.
Such an analysis
is undertaken with a view to providing additional
information on the strategies that are activated towards
three main cognitive processes:
problem representation,
planning of solution and evaluation of solution.
The
fourth main section gives a comparative account of the
findings obtained from undertaking a content analysis of
individual protocols.
253

---

5.1. Descriptive presentation of identified strategies
This section is structured around two sets of
tables.
The first set
(5.1-5.5)
presents individual
subjects'
responses to the demands of tasks 3 and 4,
the
argumentative,
familiar task and the argumentative,
unfamiliar task.
The second set
(5.6-5.10)
gives a
summary of the strategies deployed by individual subjects
towards the processes of problem representation,
planning
of solution and evaluation of sOlution.
5.1.1.
Subjects' responses to the demands of tasks 3
and 4
Tables 5.1 -
5.5 have been constructed within the
framework of the matrix for task demands as outlined in
Table 3.1
(in Chapter 3,
section 3.2.2).
These demands
are defined in terms of goal
(what is to be achieved),
initial state (from where to start),
operators
(the means
by which the problem is to be solved),
restrictions on
the operators
(specific discourse type,
time allocated,
number of pages allowed),
familiarity with tasks,
kind of
experience needed and challenge as perceived by subjects.
The purpose of this set of tables is to aid the search
for explicit cues which are present in individual
protocols and which refer to each one of these demands.
254

---

TABLE 5.1
Subject 4's Lesponse to the demands of task 4.
Explicit
Illustrations
cues
Goal
+
Then I would have written a winning composition
(XIII 54 T4,
302 )
Initial
+
This demonstration is to be laid on the basis of this extLact here
( , )
extract written
state
and revised by JeffeLson
(50-52)
"
operators
general
articulation
+
So we are now compelled to analyze the cons of our topic
(XIII 54 T4,
280-281)
Restrictions
time
+
Hey what time is it
(XIII 54 T4,
91-92)? I
only have nineteen minutes left.
Time is
running by.
What a pity (240-241) ! We'll
try to make a short conclusion 'cause we have
only ten minutes left.
Oh dear!
I've once mOLe been taken aback.
Okay.
It doesn't
matter (274-277)
Familiarity
+
Do I know anything about
the American Democracy?
I've never been interested in
politics
( . )
It serves me right today.
What am I
going to say? (XIII 54 N,
53-56)
Experience
personal
+
Of course we've neveL been to
America but
(171-175)
academic
+
When studying in the course of AmeLican political histoLY can I
notice here some
moments when America fail
to preserve human rights human freedom?
(150-153)
world
+
.. . thLough the information we've received we have heard on newspapeL or why not by
hearsay (172-175)
Challenge
+
This is quite a tricky subject
(XIII 54 T4,
53)
+
To indicate presence of explicit cues referring to specific demands of task 4.

---

TABLE 5.2
Subject 7's response to the demands of task 4.
Explicit
Illustrations
cues
Goal
-
Initial
+(7)
No
(.)
d
myth because of what
is happening in certain countries like South
state
Americd
I.) which I.) even in here in Africa I.) soutllAfrica or I.) no not
Sout"Africd in Africd in generd1
IXIII S7 T4,
152-156)
Operators
general
+
First of ,,11 I
SdY
I.) Yes Yes. No I.) umm " myth because (152-153)
articulation
Restrictions
direction
+
And now do I
think that
the American democracy is a myth
1142-143)?
So Il m supposed to say
(.)
to give my opinion about
r.)
what Jefferson said
length
+
1145-147).
[two-page essay)
tsheah
(18)
Familiarity
-
Experience
-
Challenge
+
Gosh
11augh)
red11y though
(4); mon dieu,
phuuh
(13);
[myth}
I.)
Prrr
(20);
oh
boy this is tough job (68).
+(?)
: To indicate that
the explicit cue referring to the
initial
state points to the deletion
by subject 7 of the initial state as defined in the problem statement.
+
:
To indicate presence of explicit cues referring to specific demands of
task 4.
To indicate absence of explicit cues reterring to specific demands of task 4.

---

TABLE 5.3 Subject 2's response to the demands of task 4.
r
Explicit
Illustrations
cues
Goal
+
... And to see to convince my reader that this kind of policy in America is
not a myth but i t ' s very close to reality (XIII 52 T4,
238-240)
Initial
+
I
must try to know how the American society is governed and see i f ( . ) what's
state
being said in that topic is
( . )
I
mean respected
( . )
or i f what is being said
is reality or
( . )
a myth
(75-80)
,"
Operators
general
+
I
have to argue for or against the claim that the concept of American
articulation
democracy is a myth
( . )
(41-42)
others
+
I
need a real analysis of the American society
(75-76)
But
to build my introduction I have to define the word democracy
(89-91)
Restrictions
direction
+
Personally,
I don't think i t ' s a myth
(134-135) ;
I have to argue against that
claim
( . )
which says that the American democracy is a ...
for i f this kind of
democracy in America i t ' s I'S a myth
( . )
it means that i t ' s
far from
audience
+
a
certain reality and I don't
think so
(139-143)
To convince my reader (8,238); I have to prove it
(172)
Familiarity
-
Experience
+ (7)
What
I know also is that this kind of pOlicy tends to bring a certain
stability in a society (34-35)
Challenge
+
First of all I must say it's a very nice topic for it's dealing with the
I
concept of democracy
(26-28) .
My problem is that I don't understand the real meaning of myth
(149-151)
I
+ (?)
To indicate difficulty in determining the nature of the experience

---

TABLE 5.4 Subject 4's response to
the demands of task 3.
Explicit
Illustrations
cues
Goal
-
Ini tial state
+
All right.
So I must pay attention I must not
forget
that my study is to be
laid with reference to the Ivorian realities.
The Ivorian realities must lead
me in my composition
(35-38)
Operators
general
+
All right.
So we've made a kind of analysis of the use of a higher of analysis
articulation
of the use ot a higher education
(1l9-122) .
All right.
Thus we've analyzed
which use it is for someone to get a higher education.
And now we're trying
to analyze the use i t is
for someone
to get a
job 1138-140)
All right.
So now we';y.e compelled to make a comparison between all
these
values . .. We are compelled to make a comparison between these kinds of benefits
and to express our point of view (184-185,
190-192) .
Restrictions
-
Familiarity
-
Experience
-
Challenge
-
I
+
To indicate presence of explicit cues referring to specific demands of task 3.
To indicate absence of explicit cues referring to specific
demands of task 3.

---

TABLE 5.5 Subject 5's response to the demands of task 3.
Explicit
Illustrations
cues
Goal
-
.-
Initial
•
Here the caSe lies on Ivory Coast Cote d'Ivoire.
Let '5
say i t can be extended
state
to the rest of Africa.
It is the same gap to be filled
(84-89)
Operators
general
•
I
think I ' l l deal
wi th my topi c
in two parts.
First
of all I ' l l
try to show
articulation
what can be good tor such idea.
I ' l l
try to then to argue for
them.
Argument
for the parents .. . then on the second part I ' l l
try ...
to argue against
those
parents
(32-41)
Restrictions
direction
•
AmI to choose or to deal with both sides
(125-126)
Familiarity
-
Experience
-
Challenge
+
This is quite intellectual
(126)
+
:
To indicate presence of explicit cues referring to specific demands of task 3.
To indicate absence of explicit cues referring to specific
demands of task 3.

---

5.1.2.
Solving task 3 and 4: Tabular presentation of
identified strategies
Tables 5.6 - 5.10 have been laid out so as to
capture the following:
- cognitive processes such as problem
representation, planning of solution and evaluation
of solution; and
-
strategies as defined In terms of their levels or
form,
their structure/type and their frequency.
Problem representation may take two forms:
internal
(mental)
and external
(diagrams,
notes,
outlining).
Planning and evaluation of solution may occur at two-
four levels
(conceptual,
semantic,
syntactic and
lexical),
depending on subjects.
The strategies invoked
via these forms or at these levels are of different
types.
These are re-reading, questioning,
reformulating, paraphrasing and structuring.
These
concepts are clarified in section 5.2 which provides a
content analysis of these strategies in their context of
occurrence.
5.2. Focus on individual subjects' responses to task
demands
Subjects' responses to task demands may be analyzed
in terms of compliance or non-compliance with the
260

---

TABLE 5.6 Solving an argumentative,
unfamiliar task: cognitive strategies activated by subject 4
Strategies
Number of
Cognitive Processes
Form/Level
Type/Structure
occurrence
Problem
Re-reading
(task)
4
.
Internal
Questioning
4
Representation
reformulating
(task)
4
Conceptual
Structuring
13
Questioning
7
Planning of Solution
Semantic
Questioning
6
Structuring
3
Re-reading
2
Recapitulating
1
Conceptual
Questioning
7
Assessing
(+ )
3
Assessing
(- )
1
Evaluation of solution
Semantic
Re-reading
2
Syntactic
Assessing
1
Lexical
Assessing mechanical accuracy
2

---

TABLE 5.7 Solving an argumentative,
unfamiliar task: cognitive strategies activated by subject 7.
Strategies
Number of
Cognitive Processes
Form/Level
Type/Structure
occurrence
Problem
Re-reading
(task)
15
Representation
Internal
Questioning
7
+
Reformulating
(task)
3
external
Rephrasing
8
Recapitulating
2
Conceptual
Structuring
6
Questioning
4
Recapitulating
1
Planning of Solution
semantic
Questioning
3
Structuring
3
Re-reading
3
Syntactic
Rehearsing
3
Conceptual
Assessing
1
Evaluation of Solution
Syntactic
Re-reading
1
Reformulating
2

---

TABLE 5.8 Solving an argumentative
unfamiliar task:
cognitive strategies activated by subject 2.
l
Strategies
Number of
."
Cognitive Processes
Form/Level
Type/structure
occurrence
Problem
Re-reading
(task)
4
Representation
Internal
Questioning
2
Reformulating
(task)
4
Conceptual
Structuring
23
Questioning
1
Re-reading
2
Planning of Solution
Semantic
Questioning
3
Rehearsing
2
Re-reading
6
Evaluation of Solution
Conceptual
No overt cue
(hesitation-false start)
0

---

TABLE 5.9 Solving an argumentative
familiar task:
cognltive strategies activated by subject 4.
i
Strategies
Number of
Cognitive Processes
Form/Level
Type/Structure
occurrence
Problem
Re-reading
(task)
1
Represe.ntation
Internal
Questioning
3
Reformulating
(task)
1
Conceptual
Structuring
16
Questioning
3
Re-reading
9
Planning of Solution
Recapitulating
5
Semantic
Questioning
2
Structuring
5
Re-reading
4
Rehearsing
1
Evaluation of Solution
Syntactic
Assessing
(- )
1
Lexical
Questioning
2
Assessing
(- )
3

---

TABLE 5.10 Solving an argumentative
familiar task: cognitive strategies activated by subject 5.
l
Strategies
Number of
Cognitive Processes
Form/Level
Type/Structure
occurrence
."
Problem
Re-reading
5
Representation
Internal
Questioning
4
+
Reformulating
3
externa.l
Tra.nslating
(L2 --> L31
1
Conceptua.l
Structuring
6
Semantic
Questioning
3
Planning of Solution
Structuring
11
Re-reading
1
Rehearsing
3
Translating
(L2 --> L3 I
7
Conceptual
Assessing
4
Evaluation of Solution
Lexical
Searching for synonym
5
Assessing mechanical accuracy
2

---

constraints built into the formulation of the problem and
into the task instructions.
In so doing,
one may want to
discover whether the presence or absence in subjects'
protocols of explicit cues referring to the seven-point
demands
(summarized in Table 3.1 in Chapter 3,
section
and presented in 5.1.1)
indicates awareness or a lack of
awareness of what successful solutions to those writing
problems entail.
This section examlnes how individual subjects
respond to tasks 3 and 4 by comparing the two subsets'
responses respectively to an argumentative,
unfamiliar
task
(task 4)
and to an argumentative,
familiar task
(task 3).
5.2.1.
Responding to the demands of an argumentative,
unfamiliar task
Tables 5.1-5.3
(in section 5.1.1)
show that,
In the
main,
subjects' protocols contain overt responses
(as
indicated by the presence of explicit cues)
to each one
of the seven-point task demands.
variation among
subjects is observed.
These subjects differ in their
ways of making explicit their awareness of the
constraints built into the writing problems at hand.
Subject 4 responds explicitly to all the seven
demands,
subject 2 to six of these demands,
and subject 7
to four.
Subject 4 is the only one who attempts to
266

---

access overtly the different kinds of knowledge
(defined
as personal experience,
academic experience,
and world
experience)
required for successful performance of the
task.
In this respect,
he lS the only one to assess the
state of his knowledge.
Thirdly,
subject 4 is the only
one to show explicit awareness of time restriction.
The
way he copes with anxiety resulting from his insufficient
knowledge of the problem situation and from the limited
time allocated will be shortly dealt wj.th.
The three subjects clearly find task four
challenging.
Illustrations of their feelings are
provided by the following extracts from the protocols:
'This is quite a tricky subject' (XIII S4 T4,
53) .
'My problem is that I don't understand the real
meaning of myth'
(XIII S2 T4,
149-151).
'Gosh (laugh)
really tough . . . .
Mon dieu~
Phuuh!
Myth!
Prr!
Oh boy,
this isatough
job!'
(XIII S7 T4,
13; 20; 67-68).
The number of exclamation marks and qualifiers depict the
difficulty experienced by these subjects in their
processes of solving this argumentative,
unfamiliar task.
It is worth mentioning that in this particular instance,
subject 7 resorts to French, which is de facto her mother
tongue,
though officially her second language.
In responding to goal,
subject 2's protocol contains
explicit cues pointing to awareness of audience,
as
illustrated by the following:
267
'.

---

to convince my reader that this kind of
policy in America is not a myth but i t ' s very
close to reality . . . . I have to prove it
(XIII 52
T4,
238; 172).
By contrast,
subject 4 demonstrates implicit
awareness of audience.
His words permit this inference:
It's rather look like a mere statement all my
composition.
If I could get information from
newspapers or from lectures
(.)
to constitute a
very sound basis to my composition then I would
have written a winning composition
(XIII 54 T4,
298-302) .
Finally, as said earlier,
subject 4 is the only one
to overtly monitor his anxiety in the process of solving
an argumentative,
unfamiliar task.
Three sources of
anxiety are observed.
These are:
(i)
insufficient
knowledge,
(ii)
limited time allocated for solving the
problem, and (iii)
short-term memory (STM)
failure.
To
alleviate the anxiety caused by these three constraints,
subject 4 resorts to three kinds of 'pacifier'
techniques
such as,
comment on the process of composing,
activating
self-control strategy, and cursing.
Cursing occurs after unsuccessful attempts to
retrieve information from short-term memory:
All right.
What was I going to say about those
organizations? ... But anyway I can still
remember an American
(.)
delivered a very
bright a very interesting lecture last time
about the American foreign policy.
What's his
name this guy?
(.)
Uuh the bloody guy what's
his name?
Okay.
I
forgot his name.
What is
his name this guy?
Okay.
It doesn't matter
XIII 54 T4,
292-293; 302-307).
268

---

Comment on the process of composing takes the
following form:
We'll try to write something.
Sometimes it is
that you don't have any precise idea about what
you're going to write at first.
But once
you've begun writing some sentences you
discover that some ideas are crossing your mind
and you then realize that you knew something
about the topic....
I
can feel the ideas
crossing my mind now.
How strange it is when
I
realize that at the beginning I
didn't have
no idea about the American democracy!'
(XIII
S4
T4,
155-161; 184-187).
This comment on the composlng process illustrates
two aspects of subject 4's strategic way of coping with a
challenging unfamiliar task.
The first part of the
comment indicates that brainstorming is appropriate when
in despair.
Activating brainstorming here illustrates
subject 4's flexibility In problem solving in that
alternative ways are to be called on when the
'preferred'
strategy turns out be unproductive.
The second aspect
derived from the comment concerns gap-filling decision
(Hayes,
1987)
as an efficient way of coping with ill-
defined and challenging problems.
As for the 'pacifier'
techniques deployed so as to
reduce anxiety the second half of the comment on the
composing process lS a good example.
Additional
illustrations can be seen in phrases such as,
'Okay' and
'All right' or sentences such as,
269
".

---

I must cheer up.
I must never feel desperated
even if sometime at first I don't feel
(.)
very
learned about the topic' .... '50 cheer up cheer
up cheer up cheer up.
Never let yourself go
to despair even if you don't have any idea try
to cheer up. The ideas will come.
(XIII 54 T4,
161-163;187-189) .
Repetitions of these phrases 'cheer up'
and 'no
matter' and other alike are overt indications of subject
4's attempt at monitoring and regulating the anxiety due
to the constraints from inadequate knowledge,
limited
time for composing and 5TM failure.
The analysis of how subjects 4,
7 and 2 respond to
the demands of an argumentative,
unfamiliar task has
revealed that subject 4 is by far the most sensitive to
the demands of such a task,
assuming that the presence of
explicit cues referring to the specific demands allows
one to make such a conjecture.
Is one to observe
similar exhibition of awareness of task demands in his
process of solving an argumentative,
familiar task?
The
answer to this question is the concern of the next
subsection.
5.2.2.
Responding to tbe demands of an argumentative,
familiar task
Tables 5.4-5.5
(in section 5.1.1
) exhibit the
responses by subjects 4 and 5 to the demands of an
argumentative,
familiar writing task.
Both subjects 4
and 5 overtly respond to the initial state and to the
270

---

operators of this specific problem.
Only subject 5 ~s
sensitive to the lack of specificity as to how the
solution to this argumentative problem should be
directed.
The real difficulty is that the task
instructions do not explicitly state whether there ~s a
binding restriction as to how the solution to the problem
should be articulated.
In other words,
as formulated,
the instructions do
not explicitly state whether subject-writers are to argue
for or against the proposition or whether they can adopt
a middle-position that would allow them to consider both
sides of the problem.
Unlike subject 5,
subject 4 ~s
not hampered by this lack of restriction on direction,
given that from the start his decision is to argue
against the position expressed in the problem.
Although
subject 5 seems to find task 3 somewhat challenging
('This is quite intellectual', XIII SS T3,
126),
subject
4 squarely deals with this task.
5.2.3.
Responding to task demands: Comparative
observations
Responses to task demands may vary with respect to
individual subjects and individual tasks.
Differences
in how subjects respond to the constraints of specific
tasks are indicated by the presence
(+)
Or the absence
271
..

---

(-)
of explicit cues,
which may take the form of overt or
covert signals in those subjects' protocols.
Variability among subjects may result from how anxiety is
dealt with by individual subjects.
Lastly,
the need to
adjust one's response to task demands is exemplified by
subject 4's different cognitive and affective behaviours
to cope with specific task demands.
Comparing the responses by subject 4 on respectively
task 3
(the argumentative,
familiar one)
and task 4
(the
argumentative,
unfamiliar one),
it is interesting to note
that his protocol on solving an argumentative,
familiar
task does not contain explicit cues referring to goal,
restrictions on the operators or to other kinds of
constraints that could hinder his process of solving task
3,
The absence of cues pointing overtly or covertly to
the other five demands of the task may suggest tacit
recognition by subject 4 of the adequacy of his different
kinds of knowledge required for successful performance of
task 3.
Clearly,
the absence of
'pacifier'
tips in
subject 4's protocol on an argumentative,
familiar task
depicts his ease at accomplishing this task.
One may
wonder here again whether the differences in how subject
4 responds to tasks 3 and 4 provide evidence for his
ability to consciously regulate his cognitive strategies
so as to meet the constraints of different writing
problems.
272

---

Would such an ability attest to efficiency in
problem solving strategies?
This lS another point to be
considered in due course.
Attention is now given to
individual subjects' processes of problem solving.
These
are,
as mentioned earlier,
problem representation
(Rpb) ,
planning of solution (Psl)
and evaluation of solution
(Evsl).
In this analysis,
the strategies activated
towards these three cognitive processes are also
discussed.
5.3. Focus on individual subjects' problem
representation, planning of solution and evaluation of
solution
This section undertakes a detailed content analysis
of the recurrent types of strategies that are activated
by individual subjects In their processes of solving
tasks 3 (argumentative,
familiar)
and 4 (argumentative,
unfamiliar).
Tables 5.6-5.10
(in section 5.1.2)
show
that the identified strategies are activated at different
levels and with varied frequencies by different subjects.
It is to be specified that the frequency of occurrence
represents the number of times these strategies are
actually invoked in individual subjects' protocols.
Such frequency is by no means related to the number of
pauses or planning moves made.
273

---

The maln focus of this analytical section lS
individual subjects' processes of problem representation,
planning of solution and evaluation of solution while
these subjects perform tasks 3 and 4.
This analysis is
based on five protocols generated by subjects 4
(2
protocols),
7,
2 and 5.
The limitation to those
protocols is statistically and qualitatively motivated,
as indicated In Chapter 3,
section 3.2.3 and in the
preview to this chapter.
5.3.1.
Solving an argumentative,
unfamiliar task
Subject 4
Problem representation:
What strategies are
activated by subject 4 towards this process?
In order to
understand the nature of the problem represented by task
4,
sUbject 4 resorts to re-reading the task,
questioning,
reformulating.
Focus is also on selecting relevant
information,
and on assessing his understanding of the
problem statement.
Re-reading,
as a strategy, lS activated,
three times
(see Table 5.6 in section 5.1.2), and concerns the
problem statement and/or instructions.
The first re-
reading of the task for understanding,
which is limited
to the problem statement,
precedes immediately the
reading of the task for discovery of the problem to be
274

---

solved
(XIII S4 T4,
21-22).
The second re-reading,
with
emphasis on the gap to be crossed,
deals with the
instructions
(XIII S4 T4,
41-43).
The third and final
re-reading,
which includes both the problem statement and
instructions
(XIII S4 T4,
46-51),
leads to reformulating
the problem.
Reformulating,
used four times,
seems important in
fine tuning of the understanding of the problem (XIII S4
T4,
46-51;
97-106).
In reformulating,
subject 4 selects
relevant information that enables him both to identify
the initial state of the problem (e.g.
'this
demonstration is to be based on this extract here ... · (50-
51) and to establish a relationship between the external
world or the problem situation and the problem statement
In other words,
I am asked to wonder or to
judge whether this declaration of independence
by Jefferson is really respected in the course
of the American policy
(XIII S4 T4,
77-80).
The reformulating strategy involves an implicit
definition and clarification of key concepts such as ~
and democracy.
Questioning lS the third strategy that subject 4
resorts to in his attempt at understanding the problem.
This strategy is activated four times.
In the first
instance,
it is meant to facilitate the subject's
discovery of the problem,
as illustrated by the following
words:
'So.
What am I wanted to do really in this
topic?
What am I wanted to do?
(XIII S4 T4,
42-43).
In
275
.-.

---

the second case,
this strategy seeks to assess the
subject's knowledge of the problem situation:
Do I
know anything about the American
democracy?
Oh do I
know anything about it?
(.)
I've never been interested in politics.
(.)
It serves me right today.
What am I going to
say
(.)
(XIII S4 T4,
53-56?)
In the main,
re-reading,
questioning,
and
reformulating are the three strategies that are activated
by subject 4 towards problem representation.
These
strategies do not occur sequentially.
They are invoked
when deemed appropriate.
Planning of solution:
What kinds of planning moves
are made by subject 4?
Three types of strategies are
deployed.
These are questioning
(13 times),
recapitulating (3 times)
and structuring
(6 times).
Structuring is essentially concerned with the
articulation or organisation of the active state
(scribal
activity)
of an episode.
These strategies are
subsequently discussed with respect to the internal
structure of filled planning moves.
The first planning move,
initiated by a question,
is
global in scope and addresses the overall articulation of
the solution
(e.g.
'So.
What am I going to say?
Which
aspects are to be considered in dealing with such a
question?' (XIII 54 T4,
85-88)
This question is followed
by a specific strategy leading to problem solution:
What
276

---

I must first be concerned with is to try to write an
introduction on the basis of the faint
idea I already
have upon the subject'
(XIII 54 T4,
89-911.
This planning move is an illustration of conceptual
planning.
Its main focus is the preparatory process
(advance planning),
which ends with a final attempt at
problem representation.
The second planning move,
though conceptual,
lS
local in scope as it occurs after a pause pointing to a
shift in focus,
from the preparatory process to the
translating process.
This move aims to articulate the
structure of the introduction around a general to
specific organizational pattern.
As an illustration,
the
following can be quoted:
Okay.
Sure.
Okay.
What am I going to begin
by in n¥ introduction?
(.)
I ' l l begin by a
general observation(.)
a fact which everybody
knows and thereof trying to get my subject an
interrogated
(XIII S4 T4,
107-111).
In the third planning move
(XIII 54 T4,
122-126),
which is also conceptual,
subject 4 chooses to raise a
number of questions that are themselves induced by the
initial problem.
These questions mark the end of the
introductory paragraph and suggest how the problem is to
be solved.
These are followed by a decision to re-read
the text produced so far for an evaluative look.
This
point will be examined under the process of evaluation.
The fourth planning move is an instance of long-term
memory
(LTM)
probing to compensate for limited adequate
277
.'.

---

knowledge about the problem situation.
The subject's own
words offer an interesting exemplification of retrieval
from LTM.
What follows helps to capture the process:
Since I've no precise information about
(.)
the
American politics what am I going to do?
(.)
Okay.
But at least I'm sure that I have some
information about it.
Some historical
information (.1 must prove very useful to me.
Some historical information
(.)
All right.
When studying the course on the American
political history, can I notice here some
moments when America fail to preserve human
rights human freedom?
(XIII S4 T4,
146-152) >
This fourth move and the fifth one are instances of
conceptual planning.
The fifth differs from the fourth
in that i t envlsages alternative sources of knowledge
such as newspapers and hearsay.
The sixth move illustrates semantic planning.
First,
one notes recapitulation of preceding points that
have been discussed.
Then questions are raised to
discover specific content areas such as instances of
freedom of speech and freedom of the press in America
(XIII S4 T4,
200-205).
The seventh-tenth planning moves represent
conceptual planning.
In making the seventh move,
subject
4 defines the operators by which the solution to the
problem is to be pursued.
Structuring,
as a strategy,
lS
invoked when subject 4 decides to focus on the analysis
of the 'economic level' and justifies such a decision.
There is also indication that this aspect should
278

---

constitute the last point to be discussed
(see XIII S4
T4,
253-256).
In the eighth planning move
(XIII S4 T4,
274-275),
a decision is made to conclude the essay given
time constraints.
The ninth planning move
(structuring at conceptual
level)
signals an explicit intention to pursue the
process of solving the problem.
Subject 4 feels
'compelled to analyze the cons
... the elements that
America is not really a country where democracy prevails'
(280-287).
It is worth noting that this move seeks to
question the position taken earlier.
Such a decision was
not initially considered.
There are no overt cues nor
implicit ones to account for the sudden change in the
planning of the solution.
The tenth planning move,
which
is a conceptual one,
is characterized by a search for
supporting evidence to strengthen the refutation of the
previous position.
To end the description of the filled planning moves
made by subject 4,
it is worth pointing out that nearly
all these moves are instances of conceptual planning.
Only the sixth planning move deals with semantic
planning.
Evaluation of solution:
What moves are considered
'by subject 4?
Four levels of evaluative moves are
observed:
(i)
conceptual,
(ii)
semantic,
(iii)
syntactic
and
(iv)
lexical.
Three types of strategies are observed
279
'.

---

In this particular instance.
Subject 4 uses re-reading
of the task and text produced so far,
questioning and
recapitulating as strategies in his process of evaluation
of the solution.
Re-reading is undertaken so as to check the solution
against the statement of the problem (XIII S4 T4,
131-
143).
In addition,
re-reading as well as the other two
evaluative moves are concerned with assessing the
understanding of the nature of the problem,
and with the
relevance and adequacy of the solution
(e.g.
XIII S4 T4,
260-272).
Most conceptual and semantic evaluative moves
have led to positive assessment of the solution by
subject 4
(143-144;
235-239;
272-274).
Decisions for
thorough revision of planning are not observed.
Could it
be that subject 4 does not feel the need to revise
solutions or to consider alternative solutions?
When
negative conceptual evaluation occurs,
failure to rev)se
does not result from unawareness of the need for
revision; rather,
this stems from restricted knowledge of
the problem situation (see 296-301) .
The syntactic evaluative moves,
however,
have
resulted in negative assessment.
Although subject 4
considers a given construction 'clumsy', he has no
incentive for syntactic revision,
as exemplified by the
following protocol extract:
Oh this is a very clumsy construction.
I
should have written
and also in view of the
280

---

great range of private TV and radio channels.
Okay.
It doesn't matter.
It's only a mistake
of style (.)
a few points off ...
it doesn't
matter'
(XIII S4 T4,
247-251).
Can this absence of reVlSlon be ascribed to a lack
of commitment or to the potential artificiality of the
experiment?
Subject 7
Problem representation:
What strategies are
activated by subject 7?
Re-reading,
questioning and
paraphrasing are the strategies used by subject 7 in her
attempt to understand the problem being examined.
Re-reading for understanding is initiated by the
following strategic move:
'So.
Let's see what he lS
talking about then'
(XIII S7 T4,
21).
This move is
subsequent to reading for problem finding.
As shown In
Table 5.7
(in section 5.1.2),
ten observations of re-
reading for understanding have been recorded.
Re-reading
could best be described here as chunk re-reading.
The
problem statement and instructions have been broken down
into smaller 'manageable' units.
The reading of such
units is followed by paraphrasing, which is signalled by
statements such as:
'Okay.
Here
(.)
the text says ...
'(XIII S7 T4,
26),
so the truths are ... (27); so the aim of
the government is.,. (35)
[or]
that means that ... (41)'.
281
'.

---

Subject 7 resorts to this technique to facilitate her
understanding of the problem.
paraphrasing, as a strategy to activate problem
representation, differs from the strategy of reformulating
used by subject 4.
paraphrasing, which is always
preceded by the re-reading of chunks,
is to a large
extent text-bound.
Subject 7 takes no liberty in
departing from the text.
Unlike reformulating,
paraphrasing does not reflect the subject's previous
knowledge.
One may venture that reformulating is
conceptually oriented while paraphrasing is semantically
guided.
Questioning here has a semantic function.
This is
of the form:
'What does endowed mean?'
(23)
or
'What do
they mean by a myth?'
(105).
No instances of
questioning,
with a conceptual orientation,
have been
observed.
What does the absence or presence of these two
types of questioning signal?
Planning of solution:
What kinds of moves are made
by subject 7?
The moves made by this subject operate at
two levels of planning: conceptual and semantic.
The first planning move,
conceptual in function,
lS
made during the preparatory process.
This move which is
meant to set the overall orientation of how subject 7
intends to solve the problem does not fulfil the initial
purpose that is built into the task instruction.
The
282

---

difficulty seems to arlse from interpreting the ultimate
goal of this first move,
as reflected by subject 7's own
words:
'I think I am for it because it is true that
everyone is created equal
(.)
and everybody has
unalienable rights.
But now how am I going to start?'
(XIII S7 T4 77-79).
Should such a move be subdivided
into two different planning moves:
conceptual and
semantic?
If one takes into account the concern for
understanding
(e.g. 80-83)
that occurs within thi.s
planning move,
it seems reasonable to assume that this
first move is basically conceptual In nature.
The focus
of this move provides additional evidence for the
prevlous interpretation.
Can the second planning move be labelled conceptual,
semantic or both?
The nature of this move is equivocal.
A neat classification of such a move might be misleading.
This move is articulated as follows:
Reading of a
segment of the problem statement followed by a question
How am I going'
[followed by the reading of
another segment of the problem statement,
coupled with hesitation pauses]
' I ' l l start
with this.
Humm.
No.
I
think I ' l l start by'
[leading immediately to the execution of the
plan]
(88-93).
The structure of this planning move depicts the
'difficulty in establishing a typology.
What is clear lS
that this move initiates the translating process.
283
.'.

---

The third planning move includes both concern for
conceptual as well as semantic planning (142-169).
It
essentially deals with the overall organization of the
remaining elements of the solution,
ie. with structuring
the paragraphs making up the body of the essay.
This
move clearly indicates that in solving this problem
sUbject 7 intends to examine both sides of the problem.
Questions such as 'And now do I think that the American
Democracy is a myth? .. Why isn't i t a myth?' (142-143;
162-163)
suggest conceptual planning. The semantic one
involves details that seek to answer both questions
raised at the conceptual level.
A statement such as ' I t
depends on the degree of emancipation of a country of the
development of a country ... for example in the united
states everyone is . . . What else?'
(163-168)
illustrates
the point.
The examination of this third planning move helps
one predict why sUbject 7 is likely to be unsuccessful in
her attempt to solve this argumentative. unfamiliar task.
In considering 'certain countries like South America,
South Africa ... Africa in general' from the start, subject
7 has clearly failed to identify the initial state of the
problem.
The problem situation, as suggested by the
problem statement,
is the United States of America within
the context of the declaration of independence.
Failure
to recognize this built-in constraint has led to an
extension of the initial state.
The deletion of the
284

---

constraint inherent in the initial state by subject 7 has
heightened her difficulty in solving task 4.
The fourth planning move,
which is semantically
oriented,
is signalled by a question 'What next?'
(219-
220).
To answer this question,
subject 7 refers to her
notes.
Only four filled planning moves have been recorded.
The first one is clearly conceptual while the fourth
fulfil a semantic function.
The second and third moves
include a dual function of conceptual and semantic
planning procedures.
Evaluation of solution:
What moves are considered
by subject 7?
In evaluating the solution to the problem
that is to be solved,
subject 7 makes no conceptual and
semantic evaluation.
Syntactic and lexical evaluative
moves are observed.
Three moves of this kind are
recorded.
The first linguistic evaluative move is
signalled by lexical repetition and rehearsal of the
first
thought to be committed to paper.
This evaluative
move occurs within the first planning move
(117-141)
and
is completed when subject 7 makes a final decision to
start the scribal activity (e.g.
'I think I ' l l start by
Jefferson's ... '
(93)).
The second evaluative move signals a false start.
This move lS preceded by a re-reading of part of the
instructions to perform the task,
which leads to the
285

---

deletion of lexical items followed by a syntactic
revision of the text produced so far
(111-117).
A re-
reading of the introduction, which signals the third
evaluative move,
results in its thorough syntactic
reV1Slon.
Irrelevant items are deleted
(e.g.
'I don't
think all this is necessary . . . . I
think I
am going to
write this all over'
(124-126).
The implementation of
this decision is preceded by a recapitulation of the gist
of the problem.
This recapitulation seems to suggest the
need for subject 7 to check her understanding of the
problem statement prior to the actual revision that is to
follow.
Once again revision takes the form of deletion
and reformulation
(134-142).
As seen above,
the activation of the evaluative
process by subject 7 operates at syntactic and lexical
levels.
Failure,
by subject 7,
to invoke the evaluative
process at a conceptual and/or semantic level may have
resulted in her inability to resolve the conceptual
problem created by this subject's expansion of the
initial state of task 4.
The question is,
Had subject 7
made a conceptual evaluative move,
would she have been
able to detect the error introduced by her deleting the
constraint implicit in the initial state?
286

---

Subject 2
Problem representation:
Whac strategies are
activated by subject 2°
In his process of understanding
the problem under examination,
subjecc 2 invokes
questioning,
re-reading,
and reformulating as strategies.
Reformulating towards knowledge representation
occurs once.
Questioning starts off problem
representation.
In the main,
the questioning strategy
consists in selecting and defining key words such as
democra<::Y. and myth
(XIII 52 T4,
SO-54).
Re-reading for
understanding the nature of the problem enables subject 2
to sift out the gist of the problem,
as shown in the
following:
Well if I understand this passage is
(.)
laying
the stress on the equallcy among the people
(.)
equality in everything.
In conclusIon,
they're
created equal.
So ic's dealing with equality'
(XIII 52 T4,
50-54).
Planning of solution:
What kinds of planning moves
are made by subject 2?
Conceptual and semantic planning
are deployed.
The first planning move indicates that the
solution to the problem can be worked out by confronting
the initial state and the American society.
To quote
subject 2:
I think I need a real analysis of the American
society.
In other words,
I must try to know
how the American society is governed and see if
(.)
what is being said in that tOplC is
(.)
I
mean
(.1
respected or ...
a myth
(XIII 52 T4,
75-80) .
287
.'.

---

As it can be seen this first planning move,
global in
scope,
is conceptual.
It is made during the preparatory
process.
The second,
third and fourth moves are concerned
with conceptual planning.
The second indicates both the
beginning of the translating process and preliminary
articulation of the introductory paragraph
(XIII,
52 T4,
76-80; 104-108).
The third planning move exemplifies
digression
(80-104; 109-119)
that reiterates the extended
definition of democracy previously examined.
This
planning move ends with re-reading the text produced so
far when subject 2 directs himself to go back to the
introduction (119-120).
The fourth planning move, which
is conceptual,
is initiated by a question 'Now what can I
say?'
(XIII 52 T4,
131).
Although this question may
suggest semantic planning,
the answer to that question
specifies options that are available.
This move is
clearly procedural.
Subject 2 chooses to reject the
claim that the American democracy is a myth.
He goes on
to account for his position.
uneasiness about his lack
of understanding of the concept ~ is expressed.
Here
again digression lS noted
(XIII S2 T4,
129-153).
This
planning move,
instead of leading to the execution of the
plan,
is followed by a false start which culminates in
syntactic revision
(154-162).
One may venture here that
revision results from a change in planning.
288

---

Two questions are raised in the fifth planning move.
These questions are meant to generate a search for ways
of concluding the introductory paragraph.
Given the
nature of these questions,
this fifth move deals with
conceptual and semantic planning
(e.g.
'How can I end?
Does an analysis of this kind of policy ... to claim that
the American policy lS a myth or not?'
(162-165)).
The sixth move is marked off by the re-reading of
the task instructions
(168-170).
Concern is with finding
ways of demonstrating that the 'American democracy is
close to reality'.
To achieve this,
a brief analysis of
the American society is suggested.
Re-reading of the
task and recapitulating and digression take place.
Subject 2 ends this planning move (173-179; 198-199),
which is conceptual,
by choosing to emphasize the social
and economic
'fields' in solving task 4.
The seventh planning move starts with a semantic
orientation followed by a conceptual guideline,
as
illustrated by the following words:
'What else can I
say
apart from religion and freedom of undertaking?
I
think
I can move on the political field ... '
(218-220).
The eighth planning move
(237-240)
is predominantly
conceptual in nature.
Subject 2 decides to end the
process of solving the problem by summarizing so as to
persuade the reader.
Thus concern for audience is
manifest in this move.
Note that partial preliminary
rehearsal precedes the execution of the solution.
Such
289
'.

---

rehearsal may point to semantic and even linguistic
planning.
The ninth move illustrates semantic planning,
as
shown by this question 'What else can I
add?'
(255-256)
followed by re-reading of text immediately preceding the
question.
To sum up,
five of the nine moves made by subject 2
In his process of solving an argumentative,
unfamiliar
task are clearly conceptual in function.
One is
unequivocally semantic and three operate both at a
conceptual and at a semantic level.
Evaluation of the solution:
What moves are
considered by subject 2?
No explicit evaluative moves
are suggested by the analysis of his filled pauses
(see
Table 5.8 in section 5.1.2).
Despite this,
assuming that
re-reading and revising are overt instances of evaluative
moves,
then one might assert that some evaluation takes
place in subject 2's process of solving task 3.
As
observed earlier,
the fourth planning move, which,
given
its initial focus,
might have signalled the beginning of
a new composing episode,
ends in a syntactic revision of
the previous episode
(XIII 52 T4,
154-162).
The
hesitation and the false start which led to the syntactic
revision are evidence of evaluation at the level of
planning.
However,
no explicit cue in subject 2's
290

---

protocol allows one to pursue such a conjecture any
further.
5.3.2.
Solving an argumentative,
familiar task
Subject 4
Problem representation:
What strategies are
activated by subject 4?
Re-reading, questionilig and
reformulating are the three strategies used by subject 4
to activate the process of problem representation.
Re-reading for understanding the problem occurs
after the process of problem finding.
This strategy
generates questioning to identify the very nature of the
gap to be crossed.
Identification of this gap leads to
reformulating the problem as a question
(e.g.
'Which use
is i t to get a higher education' XIII S4 T3,
30-32).
Subject 4 perceives such a question as being 'the main
question that underlies all the topic'
(XIII S4 T3,
31-
32) .
Selection of relevant information characterizes
the reformulating strategy.
Of importance is the
identification of the initial state by subject 4.
In
his own words,
I must pay attention I must not forget that my
study is to be laid with reference to the
Ivorian realities.
The Ivorian realities must
lead me in my topic in my composition (XIII S4
T3,
35-38).
291
..

---

Planning of solution:
What kinds of planning moves
are made by subject 4?
Conceptual and semantic planning
are observed in subject 4's process of planning the
solution to an argumentative familiar problem.
The first planning move, global,
deals with the
general articulation of the solution
(e.g.
'So.
I begin
this topic with a question of the following one.
What
use is it to get a higher education?
What use is it?'
(XIII 54 T3,
33-35)).
This planning move which is made
during the preparatory process is conceptual.
The second planning move is also an instance of
conceptual planning.
It functions as a shifter in focus
by setting the beginning of the translating process.
unlike the first planning move,
the second one is
localized in scope.
It is mainly concerned with
structuring the introduction (e.g.
'All right.
We'll
begin by an introduction a very short one'
(39-40)).
The third planning move, a semantic one,
consists '-11
reformulating so as to clarify the preceding thought
(e.g.
'All right.
That is to say ... '
(51)).
The fourth planning move,
a conceptual one,
involves
re-reading the text produced so far,
questioning as
search for organizational procedure (e.g.
'How shall I
keep on? How shall I continue the introduction?
... 50 the
question I must never forget it:
What use is i t to get a
higher education?'
(60-63)).
292

---

The fifth,
sixth,
seventh,
eighth,
ninth,
and tenth
planning moves may be classified as semantic ones.
The
fifth
and ninth take the form of re-reading texts
produced so far.
The sixth,
seventh,
eighth and tenth
planning moves end with linguistic cues.
These overtly
signal concern for semantic planning.
The following will
nicely illustrate the point:
'Someone who has a
knowledge what kind of power is that he's got?'
(101-
102),
and statements such as
'All right.
We'll also be
about to say that ... We can say that ... These are . . . '
(109-
110).
Recapitulating is observed in the seventh and
eighth planning moves.
The eleventh and twelfth moves are conceptually
oriented.
These moves define
the operators used to
organize the execution of the plan.
Specifically,
in
making such moves,
subject 4 expresses his intention to
compare and contrast
(eleventh planning move)
and to
conclude by synthesizing
(twelfth planning move).
Both
planning moves end with semantic planning.
The thirteenth planning move illustrates a
shift In
focus,
from the translating process to the editing
process.
As such i t is conceptually oriented.
A general observation to be made is that,
in his
process of solving an argumentative,
familiar task,
subject 4 activates equally conceptual and semantic
planning strategies
(respectively 6 and 7 moves).
By
contrast,
in his process of solving an argumentative,
293
'.

---

unfamiliar task,
subject 4 is planning moves that operate
predominantly at a conceptual level
(nine out of ten
moves).
One may wonder what such a difference reveals.
Evaluation of the solution:
What moves are
considered by subject 4?
Linguistic evaluative moves are
the only ones made by subject 4 in his process of solving
an argumentative,
familiar task.
Such moves are made on the basis of normative
considerations and of level of register.
Appreciations
such as
' ... motto is better'
(48-49),
'contribute is
better'
(53)
are illustrations of normative evaluation.
Instances of phrases rejected to meet the demands of
formality required by the task are the following:
Thus someone who jobs someone who works not
jobs that's quite slangy
1165-167) . . . . The best
sight would be that of a man who may have as
much stuffing in his belly.
Well this doesn't
sound very good English'
(XIII 54 T3,
(198-
199) .
This last negative assessment results neither In
syntactic revision nor in lexical revision.
Note here again that,
as opposed to his evaluative
moves in the process of solving an argumentative,
unfamiliar task,
the evaluative moves made by subject 4
when solving an argumentative,
familiar task are
linguistic ones,
specifically lexical in nature.
For
instance,
'who jobs'
is replaced by 'who works',
which,
294

---

in subject 4's own words,
'sound better';
'to contribute
to development'
is selected instead of
'to partake'
Is it the case that an argumentative,
familiar task
readily calls for linguistic evaluation while an
argumentative, unfamiliar task requires conceptual level
evaluative moves?
Subject 5
Problem representation:
What strategies are
deployed
by subject 5?
Re-reading and questioning are
the two strategies used by subject 5 in his attempt to
understand the problem represented by a familiar
argumentative task.
Problem representation as a process lS initiated by
a question raised immediately after reading for problem
discovery.
This question is geared towards the
identification of
'the key words or the broad idea of the
topic'
(XIII SS T3,
15).
Questioning leads to re-reading
for understanding,
which in turn results in reiterating
part of the question previously posed.
Semantic
clarification of the phrase 'on-the-field' is offered'.
Problem representation by subject 5 includes
consclous identification of the initial state
(82-89)
and
2
In Manouan
(1991),
it has been pointed out that differences in
understanding the phrase
Non-the-field N are potential sources of
variation among the raters of the written products resulting from the
verbal protocols.
295
'.

---

the restrictions (124-126)
that a statement of a problem
may contain.
The following question by subject 5
'Am I
to choose or to deal with both sides'
clearly emphasizes
the lack of specificity of the restrictions of the
problem to be solved.
A personal decision is needed.
At
this point
subject 5 makes no such a decision
explicitly.
Planning of solution:
What planning moves are made
by subject S?
Four levels of planning are invoked.
These are:
conceptual,
semantic,
syntactic and lexical.
Subject 5 makes his first planning move during the
preparatory process.
This first move is interesting ln
that it fully integrates the other problem solving
processes such as problem representation,
planning of
solution and evaluation of solution.
It also involves
conceptual,
semantic and syntactic planning.
Conceptual
planning spells out the overall articulation of the
solution as indicated in the following:
Okay.
I think that I ' l l deal with my topic in
two parts.
First of all I ' l l try to show what
can be the good for such idea.
I
try then to
argue for them.
And on the second point of
view if they are grounds to say such a thing we
must also recognize that higher education is
also important.
Then on the second part I
try
to see I ' l l try to point out to argue then
against those parents
(XIII S5 T3,
33-43).
Semantic planning is written out ln full sentences
(e.g.
37-65)
or signalled by questions such as
'What else?
296

---

Means of transportation.
What else education can bring
as a good thing?'.
The result of semantic planning ln
this first move might constitute a pre-draft.
Another interesting feature relative to this first
planning move is that points to be developed at a later
stage are examined in this first planning move
(e.g.
' ... let's say health ground.
This will be developed
during on during the body of my work'
(74-75)).
The
first planning move ends when subject 5 decides to start
the introduction
(e.g.
'Okay.
I have 40 minutes left.
I
think I must begin with the introduction ... in order to
finish.'
(XIII S5 T3,
170-172).
The second planning move is concerned with both
conceptual and semantic planning:
'Introduction.
I must
go from a general idea to specific.
Okay I think that I
must first of all stress the fact that unemployment has
increased in these so-called developed countries.
Unemployment.
No unemployment.
Let's say.
Let's
begin'.
This last statement leads to the execution of
the solution.
This second planning move sets off the
beginning of the translating process.
The third planning move illustrates a decision to
include the initial state:
'I must try to insert the fact
that this is in Cote d'Ivoire ... '
(238-240).
This third
move seems to incorporate both conceptual planning and
semantic planning.
The fourth planning move depicts
conceptual planning.
By this move--'Let's come to the
297
'.

---

development now'--subject 5 decides to start examining
the body of the essay.
As observed,
the planning moves made by subject 5
indicate overlap of different planning levels.
Such
overlapping points to the need to explore all the levels
of planning that are likely to occur within any single
planning move made by subject 5.
Evaluation of solution:
What evaluative moves are
considered by subject 5?
Four types of evaluative moves:
(i)
conceptual,
(ii)
syntactic and lexical,
(iii)
register and
(iv)
mechanical accuracy.
A conceptual
evaluative move appraises the value of a point as
important,
irrelevant or redundant
(e.g.
'This lS quite
important . . . . This is not a point . . . . This is irrelevant
. . . . 1 have already mentioned i t '
XIII SS T3,
110-111;
117;
204).
A linguistic evaluative move involves lexical
selection to expand the range of lexical items
(e.g.'
not as a plague as one of the greatest
problem
Let's check a synonym of unemployment
Let's say men to avoid repetition' XIII SS T3,
210-211;
216-217 ).
An evaluative move related to formality
involves inserting lexical items between inverted commas
or brackets
(139).
Finally,
an evaluative move that
deals with mechanical accuracy seeks to check the
spelling of a lexical item:
'specialize with an 2 or a
~
? I
298

---

The four-level evaluative moves that are activated
by sUbject 5 lead to revision at each one of these
levels.
For example,
a point that is perceived as
redundant is deleted.
No further attention is given to
such an item.
Decisions to find alternative synonyms to
specific lexical items result in selecting alternative
words.
5.4.
Summary of qualitative findings at micro-level
At this point one may want to find out what the
content analysis of the three processes of problem
representation,
planning of solution and evaluation of
solution has added to the characterization of the writing
styles of subjects 4,
7,
2 and 5.
What kinds of
strategies are activated towards the three processes
being considered?
Are there any idiosyncratic features
that help differentiate these subjects?
It is to be recalled that section 4.2.2 in Chapter 4
undertook to set up a typology of the subject-writers
involved in this study.
This categorization produced a
diagram illustrating four composing styles that emerged
from the protocols.
The following grouping was obtained:
. (1)
Composing style C:
subjects 7,
2 and 5
(Q---------->
Formulation ---> Detailed outlining ---> Exposition); and
299

---

(2)
Composing style D:
subject 4
(Q---------> Formulation
------> Exposition).
In that section, variation in style has been
suggested.
Given this,
the question that one may wish
to pose here lS,
In what respect do subjects 7 and 2
vary? In the same section 4.2.2, when discussing the
characteristic sequence of composing processes,
i t was
reiterated that variation is observed in the processes of
problem solving,
namely in the processes of problem
representation (Rpb) , planning of solution
(Psl)
and of
evaluation of solution
(Evsl).
What has the content
analysis unravelled?
5.4.1.
Processes of solving an argumentative
unfamiliar task: Comparative results
First,
i t is to be observed that when subjects 4,
7
and 2 are engaged in solving an argumentative,
unfamiliar
task,
four main strategies are activated towards problem
representation.
These are:
(i)
re-reading, questioning,
reformulating, and paraphrasing.
The three subjects
make use of re-reading as a strategy.
Subjects 4 and 2
invoke reformulating while subject 7 calls on
paraphrasing.
Paraphrasing,
it has been observed,
lS
text-bound.
Note in passing that subject 7 paraphrases
only when solving task 4.
Reformulating aims to achieve
accurate understanding of the nature of the problem to be
300

---

solved.
This is achieved by making explicit connections
between the problem statement and problem situation (or
the external world),
by selecting relevant information
likely to ensure adequate identification of the initial
state of the problem,
and finally by clarifying key
concepts.
Subject 4 is a good user of such a kind of
reformulating.
The three subjects resort to questioning to
facilitate problem finding in the process of problem
representation.
Subject 4 uses questioning to assess
his knowledge of the problem situation.
This function
of questioning sets him apart from subjects 7 and 2.
Concern for the initial state is a source of
variation among subjects.
Subjects 4 and 2 attend
overtly to the demands of the built-in constraint of the
initial state of task 4.
Subject 7, on the other hand,
deletes such a constraint, which may constitute a reason
for her failure to produce a successful solution.
The
three subjects consider it essential to clarify the key
concepts in the problem statement.
In clarifying those
concepts, subjects 4 and 2 draw on their previous
knowledge.
Surprisingly,
subject 7 fails to do so.
While subject 7 makes use of external representation to
facilitate her understanding of the problem,
subjects 4
and 2 resort only to internal representation.
Variability among subjects also stems from the type
of planning moves that are made.
The most recurrent
301
..

---

planning strategies are conceptually and semantically
oriented.
Syntactically and lexically oriented planning
is observed.
The three levels of planning are observed
in the planning process of sUbject 7.
subject 2 makes a
relatively balanced use of conceptual and semantic
planning strategies.
Note that digression is an
idiosyncratic feature of sUbject 2.
SUbject 4 can be
depicted as a conceptual planner.
Nine out of the ten
filled planning moves made by subject 4 are geared
towards conceptual articulation of the problem.
Only one
instance of semantically oriented planning move has been
recorded.
Can one rightly assume that the use of
fillers such as 'all right' and 'okay' by subject 4
indicates semantic planning moves?
Another interesting
question to be raised is whether major planning
necessarily calls for conceptually oriented moves?
Alternatively, does focus on conceptual planning
contribute to efficient and successful planning of
solution?
Emphasis on the evaluation of the solution is
explicit in the process of sUbjects 4 and 7.
SUbject 2
seems to exhibit implicit concern for evaluation,
providing that re-reading and false starts leading to
revision constitute reasonable ground for making such an
inference.
Linguistic evaluative moves best characterize
sUbject 7's evaluation process.
Such moves were observed
to lead to syntactic revision.
Three different levels of
302

---

evaluative moves occur In the evaluation process of
subject 4.
Only linguistic evaluative strategies result
in lexical revision.
In that particular case,
revision
seems to be dictated by normative and aesthetic
considerations,
such as
' i t sounds better'
or
' i t doesn't
sound good English'.
Register is also taken into
account,
as expressed implicitly in the following words:
'not jobs that's quite slangy'.
Several general observations can be made:
In the
process of solving an argumentative,
unfamiliar task,
subjects 4,
7 and 2 systematically call on re-reading and
questioning as strategies for problem representation,
planning of solution and evaluation of the solution.
Recapitulating is more systematically activated during
the planning of the solution and the evaluation of the
solution.
Reformulating seem to be restricted to the
representation of the problem being examined.
Structuring or procedural moves are only invoked in the
process of planning the solution.
What evidence accounts for differences In
performance?
Although subjects 4,
7 and 2 are
categorized as successful,
subject 4 is the only one,
out
of the ten subjects in this study,
to have been
successful in solving task 4
(the argumentative,
. unfamiliar) ,
i.e.
to have reached the 17.5 target score.
Attempting to answer this question will constitute one of
the concerns of chapter 6.
303

---

5.4.2.
Processes of solving an argumentative,
familiar
task: Comparative observations
The next question to be asked In seeking a form of
generalization is this:
Are the strategies activated
towards problem representation,
problem solution and
evaluation of solution when an argumentative,
familiar
problem is being solved any different from those that are
invoked when an argumentative,
unfamiliar problem is
being dealt with?
The content analysis of the two video protocols by
subjects 4 and 5 reveals that,
basically,
the same
strategies are called on in the processes of solving task
3
(the argumentative,
familiar one).
For instance,
when
considering problem representation,
both subjects 4 and 5
make use of re-reading,
questioning and reformulating to
facilitate their understanding of the problem.
In
addition to these two,
subject 4 reformulates the
problem.
The main difference relates to the frequency of
occurrence of such strategies.
Indeed,
comparing
subjects 4 and 5 on the number of occurrence of each one
of these strategies towards problem representation,
it is
observed that both subjects activate reformulating three
times.
Re-reading and questioning are invoked once by
subject 4 and three times by subject 5
(see Tables 5.9
and 5.10 in section 5.1.2).
An observation worth making
lS
that subject 4 behaves differently when he is engaged
In solving an
argumentative,
unfamiliar task.
In such
304

---

an instance,
re-reading of the topic is deployed three
times, while questioning and reformulating are called on
four times (for comparison,
see Table 5.6 and Table 5.9)
Does this evidence suggest that fine tuning for
understanding the argumentative.
familiar problem is
achieved immediately, making further trials unnecessary,
whereas fine tuning of understanding the argumentative,
unfamiliar problem takes longer?
Clearly,
this
difference in cognitive strategic behaviour suggests that
sUbject 4 monitors and regulates the frequency of
occurrence of individual strategies to meet the demands
of specific tasks.
This point is worthy of close
examination and is to be pursued further in the
discussion of the results.
Apart from the difference in the number of
occurrences of specific strategies deployed by sUbjects 4
and 5,
translating
(from subjects' second language
(L2)'
French,
into their target language (L3):English,
and
vice-versa),
as a strategy,
is invoked only by sUbject 5.
Just as sUbjects 4 and 5 find i t important to attend
to the initial state of a problem when examining an
argumentative, unfamiliar task so do they when an
argumentative,
familiar task is being dealt with.
Indeed,
both sUbjects explicitly acknowledge the need to
focus on the initial state of the problem at hand.
Subject 5 is the only one,
however,
to suggest the need
for clearly stated and binding restrictions as to the
305

---

direction of the operators.
Concern for explicitness in
restrictions lS illustrated by subject 5's own words:
'Am I
to choose or to deal with both sides'
(125-126)?
Differences in the form taken by problem representation
are to be pointed out.
While subject 5 resorts to
internal and external representations of the problem,
subject 4 makes use of internal representation only.
As a general observation regarding the planning
processes,
as a result of the content analysis of the
video protocols of subjects 4 and 5,
Tables 5.9-5.10
show that the same levels of planning moves--conceptual,
semantic and syntactic
(except for the lexical ones
present in subject 5's protocolsJ
and nearly the same
strategies
(except for recapitulating and translating
from L2 to L3 and vice-versa)--are observed in the
processes of planning both an argumentative,
unfamiliar
task and an argumentative,
familiar one.
Of interest lS
the number of occurrencesof each category.
Subject 4
makes 12 filled planning moves.
These are instances of
both conceptual and semantic procedural cues.
As such,
these moves can easily be subdivided into two distinct
types--conceptual and semantic.
Five of these are
clearly conceptual in nature while seven are classified
as semantic.
It is worth reiterating here that, when one
compares the planning moves made by subject 4 in his
processes of solving an argumentative, unfamiliar task
(task 4)
and an argumentative,
familiar task
(task 3),
306

---

the planning moves made by subject 4 in his process of
solving task 4 are predominantly conceptual,
nine out of
10.
While the planning moves made by subject 4 when
solving task 3 lend themselves to two distinct subgroups
(conceptual and semantic),
this is not necessarily so
with subject 5.
The first planning move made by this
subject incorporates conceptual,
semantic,
syntactic and
lexical planning.
There is also evidence that some
initial planning of some composlng episodes occurs in
advance planning well ahead of the execution of the
solution.
Another observation concernlng the planning
moves made by subject 5 is that nearly all his filled
planning moves represent both conceptual and semantic
procedural cues.
As for the evaluative moves,
subject 4
only makes syntactic and lexical evaluation.
Subject 5,
on the other hand,
resorts to conceptual,
semantic and
lexical evaluation.
What are the main features that set these two
subjects apart?
It is to be recalled that subject 4 is
considered a successful subject.
He has hit the target
score of 17.50 three times and has scored the highest
overall quality total scores on the four tasks,
as
defined In section 3.6.5.
Subject 5 lS unsuccessful.
He has not succeeded in 'hitting' the expected target
score of 17.50 once.
307

---

One may pause here and ask whether these subjects'
processes of problem representation,
planning of the
solution and evaluation of the solution help to account
for this difference.
The question that may be raised at this point is:
To what extent can such a hypothesis help to predict and
to account for the cognitive behaviour and performance of
the subjects in this study?
Put differently,
to what
extent can such a hypothesis be generalized to the sample
being considered?
Assuming that adjusting advance planning to task
demands is an indication of efficient monitoring of the
cognitive load in the process of composing written
discourse,
then failure to exhibit such behaviour
testifies to inefficient control of the cognitive domain.
If this is the case,
one may venture that the least
successful subjects'
advance planning behaviour portrays
inefficient cognitive monitoring insofar as their
planning pattern is not articulated so as to meet the
constraints of specific tasks.
Their inconsistent
advance planning pattern,
the negative differences or
relatively small differences between time on planning
familiar and unfamiliar tasks illustrate the point.
Another instance of inadequate cognitive monitoring
is their relatively longer time on the translating
processes of familiar tasks as opposed to shorter time on
the translating processes of unfamiliar tasks.
A
308

---

compensation hypothesis has been suggested in Chapter
four to explain longer time taken on the reviewing
processes of unfamiliar tasks.
The hypothesis is that
the least successful sUbjects spend longer time on the
reviewing processes of unfamiliar tasks in order to
compensate for the shorter time taken on the translating
processes of such tasks.
Assuming this to be the case,
the least successful subjects would systematically be
engaged in deleting,
adding and reformulating previous
solutions arrived at.
A closer examination of those
subjects'
verbal protocols,
however,
reveals that,
in
such a situation,
time on reviewing consists in re-
reading without any ensuing 'fix-up' or 'repair'
(Levelt,
1983)
strategies for reconsidering and sUbstantially
revising their solutions to the problems at hand.
Re-
reading in this case may be considered a
'default'
(Bereiter & Scardamalia,
1987)
strategy activated towards
'time killing'.
If display of cognitive behaviours as represented in
Table 4.13
(Matrix for observed cognitive behaviours)
1S
an indication of efficient cognitive monitoring in the
process of written composition, then here again i t can be
said that the least successful group's behaviour and that
of some of the most successful set epitomize inefficient
-use of such an executive function.
Indeed,
one observes,
in their protocols,
a total absence of explicit cues
showing concern for personal reformulation of problem,
309

---

audience,
purpose, style and concern for re-reading for
evaluation. Absence of evaluative strategies produces
linear and/or fairly predictive composing patterns,
as
seen in section 4.2.2.
Inefficient and insufficient
activation of the evaluation process,
for instance,
leads
SUbject 2 not to consider alternative forms of problem
representation. Had such a subject combined both internal
and external problem representation when faced with an
ill-defined writing task,
he might have cut down the
amount of digression that characterizes his protocols.
Failure to activate evaluative strategies also
results in the production of poor quality products.
Comments by holistic raters on such products provide
insight into the difficulties experienced by unsuccessful
SUbjects.
These SUbjects 'tend to go off the topic' or
'not [to] address topic correctly.'
They produce 'too
many irrelevant points'.
In some cases,
their
'introduction[s]
[are] wholly irrelevant to topic
[and/or
the] development [is] partially irrelevant,.3
The lack
of concern for task demands by the unsuccessful group is
instantiated by the following words:
Not much argument; mostly returns to a
description of the situation,
rather than
sticking to an argument.
Organization too
loose altogether.
Too much running on about
different ideas.
Length of essay by no means a
JQuoted comments are extracted from marginal comments made by
the two holistic markers.
Appendix XIV provides further
illustration.
310

---

criterion of quality.
This person mistakes
quantity for quality.
Organization unclear,
content does not seem well throughout.
Just an
unorganized flow of thought.
Poor grammatical
usage and application of vocabulary
(Appendix
XIV) .
Finally,
if the hypothesis that successful problem
solving entails ability to control and regulate the
affective domain of the composlng holds true,
then
failure to monitor that dimension may result In
unsatisfactory solution to writing problems.
Subject 7
lS a case in point.
Though this subject belongs to the
most successful set in this study,
she fails to
'hit'
the
target score of 17.50 twice.
As shown in her protocols,
anxiety generated by a lack of familiarity with the tasks
is overwhelming and inhibiting.'
Unlike subject 4 who
consciously accesses
'pacifier tips'
to cope with anxiety
resulting from time constraints, memory lapses and a lack
of familiarity with the task,
subject 7,
in similar
situations,
exhibits none of those strategies likely to
alleviate the pressure caused by such factors.
The purpose of this chapter has been to account for
inter-subject differences in terms of the cognitive
processes of problem representation, planning of solution
and evaluation of the solution.
To this end,
a content
analysis of individual protocols on solving an
4For the
interaction between the cognitive and affective
domains,
see Lay & Paivio,
1969 and Reynolds & Paivio
(1968).
311

---

argumentative,
familiar and an argumentative,
unfamiliar
tasks has been undertaken.
An attempt has been made to
show to what extent the hypothesis that successful
problem solving results from efficient monitoring of the
cognitive and affective domains involved in the process
of composing written discourse could be generalized to
the sample in this study.
In the next chapter, whose object is to interpret
and discuss the findings in this study In the light of
similar findings in related studies,
an attempt will be
made to find out whether, despite the small sample size,
both quantitative and qualitative results could parallel
other findings.
312

---

Chapter six
313

---

6.
Conclusions
This chapter focuses on summarizing the major
findings in this research, discussing these findings with
respect to other studies considering similar variables,
highlighting the limitations of this study,
and
suggesting further lines Of inquiry and educational
implications.
Four main articulations are retained:
section 6.1. findings,
section 6.2. discussion of findings,
section 6.3.
limitations of study and suggestions
for further investigation,
and
educational implications.
6.1.
Interpretation of the results
To start, one would want to ask what this study has
aChieved.
An answer to this question is geared towards
summarizing the major findings as they relate to the
hypotheses which have been tested both quantitatively and
qualitatively.
In all cases,
the primary concern here is
314

---

to attempt to weave together findings and general as well
as specific research questions that have been addressed
In this study.
Three subsections constitute the general
articulation of section 6.1.
Quantitative findings
obtained from the statistical testing of hypotheses are
dealt with in section 6.1.1.
The qualitative testing of
hypotheses at group-level are exposed in section 6.1.2.
Cognitive behavioural patterns generated by the content
analysis of individual protocols are examined in section
6.1.3.
Each of these points are attended to,
in turn,
below.
6.1.1.
statistical findings as related to hypotheses
Hypothesis 1 is confirmed. As expected,
the advance
planning patterns of successful sUbjects vary with
discourse type and degree of familiarity with task.
By
contrast,
unsuccessful sUbjects adopt the same advance
planning pattern irrespective of discourse type and
degree of familiarity with task.
Supporting evidence is
provided by adaptability (i.e graphic inconsistency)
and
non-adaptability to tasks of respectively the successful
and unsuccessful sUbjects in their advance planning (as
observed in Table 4.2
(a-c)
and Fig.
4.8 and 4.9).
Two related questions are to be posed here. Does
adaptability in the advance planning of the most
315

---

successful sUbjects indicate that this group is more
inclined to adjust their planning to meet the demands of
different writing problems? Conversely,
Does the observed
inadaptability in the advance planning of the least
successful group illustrate failure by these subjects to
develop appropriate problem solving strategies, causing
them to adopt fixed planning strategies irrespective of
the specific demands of different problems?
Another
question that one might want to venture is the following:
Assuming that non-adaptability in advance planning
exemplifies failure to respond to different problem
situations, could this failure result from rigidity in
approach?
On the other hand, does the graphic
inconsistency, which illustrates the planning pattern of
successful sUbjects reveal flexibility in adjusting to
different problem situations?
Hypothesis 2 is invalidated.
There is no hard
evidence to support the conjecture that sUbjects who
spend more time on advance planning will score higher on
their written products than sUbjects who spend less time.
The findings indicate that the longer sUbjects spend on
advance planning the less likely are they to produce
quality writing,
as illustrated by the negative
correlation (r= -0.3325)
between Qscore and time on
advance planning (P)
(see Table 4.3 in Chapter 4,
section
4.1.2) .
316

---

Hypothesis 3 is partially confirmed.
As
expected there is no relationship between the duration of
the composing process, defined in terms of time spent on
advance planning (P)
and time taken on translating
(T),
and discourse type.
However, the hypothesis that the
duration of the major elements in the composing process
(P and T)
is associated with tamiliarity with tasks 1S
invalidated.
There is no evidence supporting this
hypothesis.
The multiple general linear regression model
used to test this hypothesis reveals that the sUbject
variable is strongly associated with P and T (P<O.0001
and P<O.0011,
respectively).
This result is interesting in itself and leads us to
put forth the following:
Does this evidence suggest that
the sUbject variable is a ma1n variable when one seeks to
predict the duration of the composing process?
In other
words,
Is it the case that when attempting to discover
which variables contribute best towards accounting for
variance in the response variables P and T,
given
familiarity with tasks and discourse type,
one may
reasonably consider only subjects as a covariate?
Put
differently,
it would seem justified to claim that the
SUbject factor accounts for most of the variance in the
composing process,
as measured by time taken on advance
. planning and on translating.
317

---

Hypothesis 4.1 is partially confirmed.
The quality
of the written product is associated with time spent on
pausing during the translating process only in the case
of an ?rgumentative, unfamiliar task.
In this instance,
an increase in the duration of pausing lS conducive to an
increase in the overall quality
(Qscore), of the written
product, as exemplified by the following
Qscore =
10.32 + 0.005503*Tpause.
This supporting evidence results from the simple linear
regression model
(see the description of Model IV in
Chapter 3,
section 3.2.4 and the results displayed in
Table 4.8 in Chapter 4, section 4.1.2).
Hypothesis 4.2.
As expected, the overall quality of
the written product
(Qscore)
is not associated with
discourse type.
Instead, Qscore is associated with
familiarity with tasks.
The supporting evidence is that
the scores obtained by an average sUbject are increased
by 2 marks on familiar tasks
(see Table 4.9 in Chapter 4,
section 4.1.2).
Such a finding resulting from examining
the relationship between Qscore and familiarity with
tasks, when the four tasks are considered globally,
remains valid when the mean average scores on the two
argumentative tasks are analyzed.
Although the reverse
phenomenon (see Fig.
12 in Chapter 4,
section 4.1.2),
which is observed when referential tasks are considered,
might lead one to qualify such a claim, the finding that
318

---

sUbjects' performance 1S higher on familiar tasks remains
true.
Indeed,
the differences between the mean average
scores
(15)
of the referential, unfamiliar task and the
mean average scores (14.50)
of the referential.
familiar
task are not important, as revealed by the 95% confidence
limits representation.
Any differences are only very
slight.
A large sample size may reveal greater
differences.
This observation may suggest that task
characteristics can be an explanatory factor to be taken
into account when examining the relationship between the
overall quality of written products and familiarity with
tasks.
Hypothesis 4 1S confirmed when an argumentative,
unfamiliar task is concerned.
When the relationship
between the overall quality of the written product and
the composing process (as measured by time taken on
advance planning (P)
and translating (T))
is examined
globally,
i.e. across the four tasks,
no association can
be observed between quality and the composing process.
Quality (Qseare)
is associated instead with task
(P<
0.0159)
(see Table 4.6 in Chapter 4, section 4.1.2).
When the relationship between Qseare and the composing
.process (P,
T)
is examined separately for the four tasks,
there is no evidence that the overall quality is related
to the composing process on respectively a referential,
J19

---

familiar task,
a referential, unfamiliar task,
or an
argumentative,
familiar task, with an extremely low R':
0.1132.
The quality of written products is,
however,
related
to the composing process, as represented by its major
temporal components P and T,
only in the case of an
argumentative, unfamiliar task, as illustrated by the
high R':0.8488 and the significance level of P<0.0013.
In
this particular instance, P (P<0.0049)
and T
(0.0006)
are
both very important predictors of quality
(see Table 4.7
in Chapter 4,
section 4.1.2).
The parameter estimates
show that a gain in Qscore is higher on a
'second of P'
than on a
'second of T',
e.g. Qscore = -2.92 + 0.0075*P +
0.0047*T,
where 0.0075 represents an increase in Qscore
for each additional second on P and 0.0047 an increase in
Qscore for each additional second on T.
Hypothesis 5 is confirmed when the case of an
argumentative, unfamiliar task is examined.
The
hypothesis that the quality of conceptualization, as
measured by an independent assessment of introductory
paragraphs,
constitute an index of predictability of the
overall quality of written products finds some supporting
evidence when this hypothesis is tested with respect to
an argumentative, unfamiliar task.
In such a case the
quality of introductory paragraphs is a better predictor
(P<0.0688,
slightly higher than P<0.05))
of the overall
320

---

quality of the written product than the amount of time
spent on planning the production of those paragraphs.
To sum up,
it is of interest to highlight three
major findings:
(1)
The overall quality of the written
product is associated with familiarity with tasks;
(2)
SUbject as a variable is strongly associated with the
duration of the composing process
(as defined by P and
T);
(3)
There is recurrent evidence that task 4
(i.e. the
argumentative,
unfamiliar one)
1S
a main variable to
consider when one seeks to predict the overall quality of
written products,
or the duration of the major elements
of the composing processes, or to use the quality of
introductory paragraphs as an index of predictability of
the written product.
Given this set of evidence,
i t
seems reasonable to give special attention to the
following three variables:
familiarity with tasks,
subjects and the argumentative,
unfamiliar task.
Such
emphasis on these three variables is the concern of the
qualitative analysis.
Before examining the findings
arrived at from a qualitative perspective,
one may want
to make a number of conjectural comments on the
statistical findings.
An interesting finding lies in the validation of
.hypothesis 4.2 which states a relationship between
overall quality and degree of familiarity with tasks.
In
this respect,
i t is believed that,
given the limited
321

---

sophistication of first year students,
such sUbjects
would be inclined to be more sensitive to the degree of
familiarity with tasks.
Indeed,
i t seems reasonable to
assert that such students can easily appraise whether or
not they have access to the appropriate knowledge that is
required to solve a familiar problem.
On the other hand,
i t is assumed that such sUbjects are less likely to be
aware of the demands of differences in discourse type. As
a result,
they may not exhibit,
consciously or
unconsciously, any concern for,
nor deploy adequate
strategies to satisfy,
the exigencies of different
discourse types.
Another important finding to be highlighted is that,
except for the above hypothesis 4.2.,
all the other
hypotheses tested by linear regression models are
confirmed only in the case of an argumentative,
unfamiliar task.
This finding is interesting for two
main reasons.
The first one is that i t suggests that any
study that seeks to predict the overall quality of
written products from a set of explanatory variables can
satisfactorily limit its scope to an argumentative,
unfamiliar task.
The second reason which emerges from
respectively Hayes-Roth (1979)
and Hayes-Roth & Hayes-
Roth
(1979)
is that task characteristics affect the
planning strategies that are invoked.
Could i t be that
the inherent structure of an argumentative,
unfamiliar
task is more suitable in examining conceptualization in
322

---

the process of written discourse?
Is i t the case that an
argumentative. unfamiliar task would call for a top-down
planning strategies?
In such a case,
is the structure of
such a task robust enough to compel sUbjects to resort to
a top down planning strategy,
irrespective of their
inclination?
A productive line of investigation in discussing the
findings in this study is the personality characteristics
hypothesis.
In their 1979 study, Hayes-Roth & Hayes-Roth
claim that personality characteristics may constitute
another variable that influences the planning pattern.
Thus it might seem of particular interest to take a close
look at sUbject characteristics and their interaction
with planning.
This aspect is worth delving into since,
as shown earlier in Table 4.2, Chapter 4, section 4.1.2,
as a rule,
successful sUbjects spend maximum time on
planning ahead an argumentative. unfamiliar task and
minimum time on advance planning an arqument?tive.
familiar task.
This evidence clearly indicates that
successful sUbjects, unlike their least successful
counterparts,
adapt their planning strategies to the
demands of different tasks.
Is this an indication that
successful subjects demonstrate flexibility in planning
while the least successful ones are inclined to exhibit
rigidity?
Does this suggest that adequate planning
strategies likely to enhance the quality of the solution
to a given problem stem from flexibility in personality,
323

---

whereas inadequate planning strategies which are bound to
negatively affect the end~result originates from rigidity
in personality?
6.1.2.
Macro-level qualitative findings
The qualitative analysis at group-level has been
guided by three main concerns:
identification of cognitive processes in sUbjects'
composing processes;
-
setting up a typology of composing styles;
and identification of the episodic structure in
the translating process.
To highlight the salient findings resulting from
such a perspective,
i t is worth reiterating the research
questions underpinning these concerns.
The formulation
of specific research questions will precede the
exposition of findings as they relate to individual
concerns.
Concerns relating to group-level qualitative
analysis
(1)
Identification of the cognitive processes. Three
questions guided the search.
These are:
324

---

What would the Hayes & Flower
(1980)
cognitive
model,
as a diagnostic tool, reveal about subjects'
composing processes?
Are there any differences in subjects' processes
of solving the four writing tasks considered in this
study?
Which of the cognitive behaviours characterizing
good native writers can be observed in subjects'
composing processes?
A first finding is that the Hayes and Flower
(1980)
model can adequately serve diagnostic purposes.
Such a
framework has revealed that, globally speaking, L2/L3
subjects resort to planning,
translating and reviewing in
their processes of composing written English discourse.
Individual differences are observed among these sUbjects.
using the scores on their written performance as a basis
of comparison,
one observes that,
unlike their least
successful counterparts, the most successful subjects
tend to adjust their advance planning time to the demands
of the problems to be solved.
The translating pattern of
the least successful subjects seem to indicate that this
subset is likely to spend little time on trying to solve
demanding tasks.
Finally, the most successful subgroup
can be portrayed as systematic reviewers, while the least
.successful one is not.
As to how well the set of cognitive behaviours
exhibited by skilled native writers depicts these
325

---

subjects,
i t can be said that none of the features
defining the matrix for observed cognitive behaviours
seem to be useful descriptors of the composing processes
of the subjects in this study.
The following cognitive
behaviours,
such as planning,
revision,
audience
awareness,
consideration of purpose,
stylistic concerns
and stopping to read at intervals,
are not shared by most
subjects.
Here again,
individual differences arise.
Such a
statement needs, however, to be qualified.
This
observation made on the basis of the matrix for cognitive
behaviour
(see Table 4.13 in Chapter 4,
section 4.2.2)
is
true only when an argumentative. unfamiliar task is dealt
with.
Had this matrix been applied to the processes of
solving the other three tasks
(argumentative.
familiar,
referential.
familiar and referential. unfamiliar),
i t
would have been apparent that re-reading,
as a cognitive
strategy,
is invoked more often than Table 4.13 would
lead one to infer.
The need for qualifying the previous
generalization urges one to emphasize that task effects
are an important variable in discussing the composing
processes of sUbjects.
Differences among subjects can best be explained by
redefining the concept of planning and its place of
occurrence in the process of composing written discourse.
An alternative account for variation among sUbjects
326

---

concerns individual sUbjects' responses to task demands,
as will be shown later.
(2)
setting up a typology of composing styles.
The
three underlying questions are formulated as follows:
To what extent could the two contradictory
pedagogical models
(suggested by Wason,
1978)
provide a good description of sUbjects' composing
processes?
To what extent could a coding scheme based on
Hayes (1987)
satisfactorily account for the
composing processes of sUbject-writers.
In other
words,
Is one likely to discover in those sUbjects'
protocols the characteristic sequence observed in
the process of solving cognitive problems?
How do composing styles emerging from the
protocols compare with those identified in Hayes and
Flower (1980)?
Attempt at grouping sUbject-writers into two
opposing sets--outliners vs non-outliners--(as suggested
by the two contradictory pedagogical models)
clearly
fails to provide full and accurate characterizations of
composing processes.
such a dichotomous view of the
composing processes does not (as already stated in
Chapter 4, section 4.2.2)
capture variety in composing
styles that emerge from the data.
The limitation of such
a model,
it has been suggested,
lies in two questionable
327
.,

---

underlying assumptions:
(1)
Planning is necessarily
externally represented.
(2)
Planning is restricted to
the early phase in the composing process.
Reasons for considering the coding scheme based on
Hayes
(1987)
as adequate are twofold.
Firstly,
i t
permits an objective categorization of the protocols.
SecondlY,
this scheme may provide a basis for inter-
sUbject comparison.
It is to be stressed that there
might be a shortcoming in limiting the application for
this scheme to solely identifying the presence or the
absence of given cognitive processes
(finding problem,
representing problem,
planning solution,
carrying out
plan,
evaluating solution and consolidating gains).
Such
a scheme may prove,
however, productive if supplemented
by a content analysis of a set of cognitive processes
whose activation seems to display differences among
sUbjects.
As for the third question underlying the second
concern of group-level qualitative analysis, given that
an answer to such a question calls for a comparison
between the four composing styles emerging from the data
and the four
'configurations' of writers suggested by
Hayes & Flower (1980), this third question will be
adequately dealt with in the section on the discussion of
the results.
328

---

(3)
Identification of the episodic structure in the
translating process.
Two independent sources of
evidence,
the application of the coding grid based on
Hayes
(1987)
to ten protocols on an argumentative,
unfamiliar task and the segmentation of the data into
composing episodes,
have shown that the internal
structure of composlng episodes is inherently cyclic.
Alternative occurrence of pausing for planning and
scribal event is observed.
This observation clearly
indicates that planning is not localized to a specific
point in time.
This finding lends support to the general
research question in this study.
The general research
question is concerned with
discovering whether the
phenomena of conceptualization,
planning and decision-
making are localized for the most part to a specific
point during the composing process or whether these are
evenly distributed throughout the composing process.
Implicit in such a research question is the claim that
planning is not restricted to any point in the composing
process.
The importance of such a claim, as will be
shown in the discussion,
cannot be overemphasized and
will be subsequently dealt with in relation with other
studies sharing similar views.
Another finding resulting from examining the
'episodic structure of the protocols concerns the nature
of between boundary pauses.
These are of two types:
initial boundary markers and final boundary markers.
329

---

Parallel findings are available In research on speech
production.
Finally, surprisingly,
there is no relationship
between the duration of scribal event element in a given
composing episode and the duration of the immediately
preceding pausing element in that same composing episode.
Three concerns relating to group-level qualitative
analysis have been dealt with in this section.
These
have been to:
- discover the cognitive process at work in the
protocols;
- establish a typology of composing styles; and
- depict the episodic structure in the translating
process.
Focus has been on interpreting the findings resulting
from each of these concerns at group-level qualitative
analysis.
We shall now examine the qualitative findings
at individual sUbjects'
level.
6.1.3.
Micro-level qualitative findings
A content analysis of individual protocols has been
performed so as to detect sourCes of variability among
sUbject-writers at a finer level.
What has such an
undertaking revealed?
The findings are organized around two main points:
330

---

-
(i)
individual sUbjects' responses to task
demands;
-
(ii)
strategies activated towards the processes of
problem representation, planning of solution and
evaluation of solution.
The analysis of subjects' responses to task demands
shows that, out of the seven-point response matrix (goal,
initial state,
operators, restrictions on the operators,
familiarity,
experience and challenge)
the four sUbjects
considered are likely to respond to the initial state and
operators.
In other words, when SUbjects respond overtly
to task demands,
they attend to where to start and how
the problem is to be solved.
By contrast, these SUbjects
are less likely to assess overtly whether they have
adequate knowledge for successful solution of the problem
at hand.
There 1S noticeable variation in how subjects'
respond to individual task demands.
In this respect,
SUbject 4's protocols on an argumentative,
familiar task
and on an argumentative, unfamiliar task offer
interesting explicit cues that illustrate his ability to
cope, both cognitively and affectively, with task
specificity.
As noted in Chapter 5, section 5.2.1,
he is
the only one to assess to what extent his knowledge of
the problem situation (When dealing with an
argumentative, unfamiliar task)
can help him arrive at a
successful solution of that particular problem.
When
331

---

different sources of anxiety (resulting from his
inadequate knowledge of the task, the limited time
allocated for solving the problem,
and from STM break-
down)
become overwhelming, subject 4 activates certain
'pacifier' techniques to control and regulate such an
overpowering and inhibiting feeling.
Given that he can be considered the best writing
problem solver in the sample, one may wish to suggest
that his ability to monitor both cognitive and affective
factors accounts for the overall higher quality of his
writing performance.
Put differently,
one may wonder whether differences
in how sUbjects respond to task demands provide
sufficient evidence for positing that conscious
adjustment of cognitive and affective strategies to
problem constraints is likely to result in successful
performance on different writing tasks.
Evidence for
pursuing such a question further may be given by focusing
on the major findings resulting from examining subjects'
strategies towards problem representation (Rpb) , planning
of solution (PsI)
and evaluation of solution (Evsl).
Successful solution to non self-initiated writing
problems requires efficient activation of three cognitive
processes:
Rpb,
PsI and Evsl.
Four recurrent strategies
are deployed towards these cognitive processes.
These
are: re-reading, questioning, recapitulating and
reformulating.
In the processes of planning and
332

---

evaluating the solution,
these strategies may operate at
four different levels: conceptual, semantic,
syntactic
and lexical.
Potential sources of variation are defined with
respect to the identification of the initial state and
restrictions on the operators,
form of problem
representation
(internal and external)
and types of
planning and evaluative moves.
As illustrated by the
case of sUbject 7,
adequate identification of initial
state and restrictions is crucial in problem
representation.
Another source of variation stems from the frequency
with and the level at which each one of the four
cognitive strategies are invoked.
In other words, the
number and level of occurrences of re-reading,
questioning,
recapitulating and reformulating is
determined by sUbject and by task.
Focus on subject 4 will illustrate the point.
This
sUbject seems to dispense with fine tuning for
understanding an argumentative,
familiar task, while the
need for fine tuning towards understanding an
argumentative, unfamiliar task is deemed necessary.
A brief way of depicting sUbject 4 as a problem
solver could be formulated as follows:
An effective
<controller and regulator of the cognitive processes of
problem representation, planning of solution and
evaluation of solution as well as the affective dimension
333
,

---

of composing.
Such a characterization may be useful in
accounting for differences in composing styles between
sUbject 4 and the other sUbjects (7,
2 and 5).
Although
just as sUbject 4,
sUbject 5 is basically sensitive to
task demands,
this sUbject's composing style is best
described by Anderson's
(1980)
coordinating hypothesis
and Collins & Gentner's (1980) hypothesis that confusion
between idea generation and text production results in
writing difficulty.
Other descriptions of subject 5 can
be found in Bereiter & Scardamalia's (1983)
discussion of
proper allocation of mental capacity and in Britton
(1983)
whose analysis of 'shaping at point of utterance'
stresses the need to associate 'spontaneity and moment by
moment interpretative process'.
Given these observations,
it seems reasonable to
posit that successful writing performance results from
efficient monitoring of the processes of problem
representation, planning of solution and evaluation of
solution as well as efficient controlling and regulating
the affective domain.
Conversely, unsuccessful writing
performance stems from inefficient monitoring of both the
cognitive and affective domains.
This point is to be
examined in the discussion of the results.
334

---

6.2.
Discussion of the results
In this section attention is given to how the
findings of related studies provide supporting evidence
to the findings yielded by this study or to how they
diverge.
The discussion of the findings is articulated around
two main issues:
(1)
task effects and sUbject-response to
task demands;
(2)
episodic structure in the composing
process.
The examination of the first major issue is
organized around three axes:
(i)
familiarity with tasks,
(ii)
task characteristics and
(iii)
sUbject
characteristics.
These three axes constitute the three
sUbsections of the second main section in Chapter 6 which
discusses findings relative to:
the effect of familiarity with tasks on the
composing process and on writing performance;
the effect of task characteristics on the
composing process and on writing performance;
the effect of sUbject characteristics on the
composing process and writing performance.
The second main issue is essentially concerned with
demonstrating the
cyclic nature of composing written
discourse.
These issues are respectively considered
'below.
335

---

6.2.1.
Task effects and sUbject-response to task
demands
The concern here is to discuss findings as they
relate to familiarity with tasks,
task characteristics
and the sUbject factor in written composition and on
performance scores.
The ultimate goal of such an
articulation is to demonstrate that task effects on
written discourse composing process and performance
cannot be itemized.
A definition of tasks involves a
combination of features
(familiarity,
discourse type
among others),
the isolation of which provides a partial
understanding of the task phenomenon.
Additionally,
one
may need to consider the effect of tasks on the composing
process and on performance in conjunction with subjects'
responses to the demands of specific tasks.
Having stating the need to take an integrated
approach to discussing the issue of task effects,
one may
want,
for the sake of clarity in the discussion of the
findings,
to consider in turn each of the three axes
delineated above.
Effect of familiarity with tasks on the composing
process and on writing performance
The finding that sUbjects' overall writing
performance is enhanced by familiarity with tasks is
supported by evidence in studies on reading and writing.
336

---

L1 and L2 reading investigations
(Carrell,
1983; Gass &
Varonis,
1984; among many others) that have examined the
impact of prior or background knowledge On reading
comprehension have discovered that familiarity with topic
is instrumental in facilitating reading comprehension.
Similar results have been arrived at in L1 and L2 writing
research.
The effect of topic familiarity is reflected
in increase in the length of texts produced by subjects
(Cheskey,
1984,
1987; McCutchen,
1986; Winfield and
Barnes-Felfeti,
1982; Tedick,
1988).
In accounting for subjects' good performance on
tasks that are familiar to them, Tedick (1988)
asserts
that '[subjects'] holistic scores increased 011 the field-
specific topic because their knowledge base was
organized,
contextualized, and more fluent'.
Their poor
performance on 'general prompt' tasks,
she contends,
is
ascribed to difficulty in articulating and organizing not
readily available information so as to meet the demands
of academic writing,
in a limited time.
When explaining
differences among sUbjects with different linguistic
proficiency, Tedick, arguing along the same line as
McCutchen (1986),
claims that beginning sUbjects'
inability to produce 'longer error-free T-units on the
field-specific topic than on the general topic',
as
opposed to their more advanced counterparts,
indicates
that 'prior knOWledge of the subject matter of the topic
337

---

does not compensate for immature writing strategies'
(Tedick,
1988:134).
Tedick readily accepts McCutchen's account of
children's failure to produce accurate long texts.
In
doing so, Tedick establishes an equation between
children's immature writing strategies and beginning ESL
writers' composing strategies.
However, Tedick rejects
the claim
of McCutchen
(1986)that mature writing
strategies could compensate for lack of prior knowledge,
alleging that her findings do not lend support to such a
hypothesis.
In other words,
from Tedick's point of view,
McCutchen's compensation hypothesis is contradicted
because the advanced subjects in her 1988 study could
exhibit linguistic superiority only when they could
access their subject matter knOWledge.
Implicit in Tedick's reasoning is the equation
between second language proficiency and cognitive
ability.
Put differently,
beginning ESL SUbjects resort
to immature cognitive strategies while advanced ESL
subjects invoke mature cognitive strategies.
Lack of
evidence in Tedick's
(1988)
study to support the
hypothesis that advanced SUbjects produce comparatively
better quality texts,
irrespective of familiarity with
topic,
leads that researcher to reject the claim that
mature strategies could compensate for unfamiliarity with
task.
338

---

Reasons for questioning such an argumentation are
two-fold.
First,
the assumption that all elementary or
intermediate ESL/EFL writers of English would necessarily
resort to immature strategies seems rather a strong
claim.
similarly,
the claim that all advanced ESL/EFL
writers would exhibit mature strategies needs to be
qualified.
Secondly, the assumption that linguistic
performance necessarily reflects cognitive ability seems
unjustified.
Indeed,
a number of L2 studies in written
composition, Raimes
(1985b,
1987)
and other more recent
studies
(see Kroll,
1990),
indicate that linguistic
proficiency does not necessarily result in efficient
composing strategies.
Besides, L2 and L3 sUbjects do not
constitute a homogeneous group
(Raimes,
1987).
Finally
background knowledge varies in degree and in nature.
without delving into these issues any further,
one
would want to pursue McCutchen's
(1986)
compensation
~ypothesis further.
This hypothesis can be reformulated
as follows:
mature strategies could compensate for lack
of familiarity with tasks, whereas 'extensive content
knowledge does not necessarily compensate,
however,
for
immature writing strategies'
(McCutchen,
1986:442).
Sharing this view,
this investigator would want to
suggest that solving a familiar task may not call for
sophisticated
cognitive
strategies
while
solving
an
unfamiliar
task
may
necessarily
require
deployment
339

---

of efficient problem solving
strategies.
This
contention finds supporting evidence in two observations
derived from this study:
(1)
Taken together,
all the
subjects perform better on their familiar writing tasks
irrespective of their degree of success or their
strategies activated in the solution of these tasks.
(2)
Taken individually,
subject
4 is a case in point.
As
seen,
an argumentative,
familiar
task causes him to
reduce the amount of fine tuning for understanding the
problem and the need for evaluating his planning moves.
other findings backing up the above conjecture are
offered by Bereiter & Bird (1985)
and Casanave
(1988).
Although both studies deal with reading,
they are worth
mentioning here as they provide a dichotomy likely to
explain the absence or presence of sophisticated problem
solving strategies.
Specifically,
both studies indicate
that routine strategies are invoked in problem-free
reading while repair strategies
(Bereiter & Bird,
1985)
or non-routine strategies
(Casanave,
1988)
are called on
when dealing with challenging reading.
In any case, successful and efficient activation of
problem solving strategies implies attending to the
specificity of the problems to be solved as shown below.
Put differently,
familiarity with tasks may not
adequately account for the degree of sophistication of
the cognitive strategies that are deployed in the process
of solving complex problems such as writing tasks.
As
340

---

indicated earlier,
task effects constitute a multi-
faceted factor,
among which is the degree of familiarity
(most familiar
(F)jleast familiar
(U)).
An atomistic
view of task characteristics is bound to offer an over-
simplistic interpretation of the effect of each
individual facet considered separately.
Having made this
point,
one would want to adopt a holistic approach to the
relationship between task characteristics and overall
performance as well as the composing process.
Effect of task characteristics on composing
processes and writing performance
The finding that task is related to sUbjects'
overall performance is born out by similar results in
other studies.
In this respect, Ruth and Murphy
(1988)
have this to say:
The interaction between students and essay test
questions has been recognized and demonstrated
by psychometric research (citing Coffman,
1971a, p.289) . . . . Despite the fact that
researchers in educational measurement have
acknowledged topic effects,
the larger
theoretical issues of essay task construction
and typology have seldom been addressed.
There
has been little concern about determining the
frequency,
magnitude,
or underlying nature of
essay task effects.
Topic effects have tended
to be treated anecdotally or incidentally in
the context of other research objectives
(Ruth
& Murphy,
1988:55).
341

---

Ruth & Murphy (1988)
refer to Godshalk, et al.
(1966)
to illustrate their point.
In that study,
topic effect was identified as a source of
variability in the reliability of the test
score,
but i t was not discussed as an
interesting phenomenon on its own,
worthy of
further investigation
(Ruth & Murphy,
1988:
56) .
Such a statement by these two investigators is
somewhat surprising since among other points covered in
their discussion of task instruction,
'Effects of mode or
type',
as a sUbsection,
is devoted to a review of product
studies
(Crowhurst,
1978; Crowhurst & Piche,
1979;
Perron,
1977; Whale & Robinson,
1978; San Jose
(1972)
Rosen,
1969)
whose findings are instances of the
relationship between discourse type and syntax.
Specifically,
these studies suggest that argumentative
written products are,
to quote Ruth and Murphy,
'more
syntactically complex than narrative papers at grade 6,
10, and 12'.
Matsuhashi
(1979,
1981,
1982)
on process
found that planning varied with discourse purposes of
reporting and generalizing.
The difficulty in dealing with task characteristics
may be two-fold:
(1)
the elusive nature of such a
variable;
(2)
the attempt of researchers to itemize the
different components.
One way around this difficulty (as
already indicated in Chapter 2),
it is suggested here,
is
to consider task effects as a bundle of features that can
342

---

be adequately accounted for within the framework of a
task demands matrix as described in Chapter 3,
section
3.2.2.
Such a matrix, which incorporates goal,
initial
state,
operators, restrictions on operators,
familiarity,
and kind of experience,
enables the present researcher to
explain why,
given two familiar tasks,
scores on an
argumentative,
familiar task are higher than scores on a
referential,
familiar or why scores on a referential,
familiar are surprisingly lower than scores on an
unfamiliar task.
This matrix,
for instance,
helps to
determine whether differences in performance can be
ascribed to whether a task is well-defined or ill-defined
and to the degree of explicitness or lack of explicitness
in task instructions.'
Specifically,
a comparison of the two referential
tasks
(familiar
(task 1)
and unfamiliar
(task 2)
within
the matrix for task demands
(see Table 3.1 in Chapter 3,
section 3.2.2)
enables one to observe that task 1, though
assessed as familiar by the ten subjects in this study,
is an ill-defined problem while task 2 is a relatively
well-defined problem.
Considered in terms of the
referential focus,
both tasks
(referential,
familiar and
referential, unfamiliar)
can be said to be equally i l l -
defined problems.
Indeed,
an operator such as the verb
explain,
by implicitly suggesting a referential discourse
I
A discussion of ill-defined and well-defined tasks can be
found in
Bourne,
Ekstrand & Dominowski
(1971)
who provide an
informative introduction to the understanding of problem solving.
343

---

~, eliminates the probability for a discussant text
but by no means makes provision for an organizational
pattern.
As a consequence,
subjects are left to infer
that explain may include either define, describe,
and
illustrate or the three.
Nothing in the task
instructions warns sUbjects against the negative
sanctions that might result from an exclusive emphasis on
one of these operators.'
Task 2,
however,
may be
perceived as a better structured task as i t contains a
built-in organisational pattern suggested by the terms
differences and similarities.
This built-in
organizational pattern proves to be a relatively robust
one and leads to better organized products
(Langer,
1984) .
Of interest then is the inherent structure of a
given task when examining task effects on the composing
process.
A problem which exhibits an inherent
hierarchical structure necessarily calls for a top-down
planning strategy,
as illustrated by Byrne
(1977),
(Hayes-Roth & Hayes-Roth 1979).
Similarly, severe time
constraints lead to a top-down strategy, whereas minimal
time constraints encourages bottom-up planning process
(Hayes-Roth,
1979).
Task structure,
amount of time for
solving the task,
and other features make up the complex
task factor,
which,
as discussed in Chapter 2,
section
~ As seen in Manouan (1991), one source of variation between the
two analytic assessors stemmed from differences in their
interpretation of explain.
344

---

2.2.2,
is associated with the composing process and
performance.
The interaction between the task factor,
as
a bundle of features,
and the sUbject factor constitutes
an attractive alternative to consider since it is
quantitatively demonstrated that
different topic tasks administered to writers
with different language and cultural
backgrounds elicit responses that are
linguistically measurable and measurably
different
(Reid,
1990:191).
The design and perspective of the study by Reid
(1990),
a discourse analysis approach,
does not permit a
comparison with the findings of the present study.
However, what is of interest is the findings that provide
supporting evidence of the impact of topic
(or task)
on
writing performance.
A comprehensive view of the issue
of the interaction between task effects and composing
process as well as performance is to be achieved only if
such a discussion include the subject factor and its
interaction with the demands of tasks.
Effect of sUbject characteristics on composing
processes and writing performance
A set of quantitative findings arrived at in this
study suggests that inter-subject variation is the main
effect in predicting or explaining response variables
such as
(1)
the duration of the main components of the
345

---

composlng process
(advance planning,
translating and
reviewing),
(2)
the overall performance scores on written
products and
(3)
the mean duration of pausing at initial
boundaries of major composing episodes.
Qualitative
findings,
both at group- and individual-level,
provide
confirming evidence of individual differences.
Basically
these differences illustrate variance in activating sets
of cognitive strategies towards the processes of problem
representation, planning of solution and evaluation of
sOlution.
Research evidence of variation among sUbjects is
available in other studies,
some of which bear directly
on composing written discourse,
others on learning
ability and general performance.
Accounting for
'individual differences in processes'
(Resnick and
Neches,
1984)
and in performance has been based on
various paradigms.
A first attempt to explain individual differences
among composers,
on the basis of pauses,
is illustrated
by van Brugger (1946).
Among other findings,
that study
found that the rate of word flow increases with academic
factors
(reading, vocabulary,
spelling), personal
features
(perseverance, extroversion vs introversion),
sex and socio-economic status.
A second explanatory
perspective is Rowher's
(1978)
'elaborative conception'.
In this study, Rowher suggests that such a conception
could adequately account for
'differences among persons
346

---

in their performance on learning and memory tasks'.
From
this perspective, age-related differences and individual
differences derive from the same sources,
i.e.
metamemorial knowledge,
semantic knowledge and strategies
or operators.
A third explanatory line is exemplified by the
flexibility vs inflexibility dichotomy (Hayes-Roth,
1979;
Hayes-Roth & Hayes-Roth,
1979; Rose,
1980).
Flexibility
in planning approach allows successful subjects to adapt
their planning strategies to task specificity and task
demands.
Inflexibility or rigidity in problem solving
strategies by unsuccessful subjects results in adopting
fixed and inappropriate planning moves irrespective of
task characteristics.
An alternative account for individual differences in
process and performance is considered in the present
study.
It is suggested that inter-subject variation can
be accounted for in terms of the ability to monitor and
regulate both the cognitive and affective load exhibited
by problem demands.
Such an explanatory perspective
integrates the following:
(1)
the flexibility hypothesis
(Hayes-Roth,
1979);
(2)
the expertise hypothesis
(Hayes-
Roth & Hayes-Roth,
1979; Hayes-Roth & Thorndike,
1979;
Bereiter & Scardamalia (1987); Hayes & Flower, 1980),
(3)
·the rhetorical strategies hypothesis
(Bereiter &
Scardamalia, 1987; Matsuhashi,
1981; Jones,
1982),
(4)
the familiarity hypothesis,
(5)
the broader schema
347

---

hypothesis,
and (6)
the affective dimension of written
composition (Green,
1982).
However,
it does not provide any means for measuring
personality features,
as shown by van Brugger
(1946),
neither does i t accommodate the academic variable
emphasized in that same study.
This fourth explanatory
perspective may prove useful,
although Resnick and Neches
(1984)
find it premature to address the issue of
individual differences in processes prior to an
understanding of the processes.
The alternative explanatory line suggested here is
useful in that it attempts to offer a comprehensive
examination of individual differences by considering
globally potential variables that may contribute to
inter-subject variation given specific writing-problem
demands.
Ability to monitor and regulate the cognitive load
consists in successful activation of the cognitive
processes of problem representation, planning of solution
and evaluation of solution.
Satisfactory problem
representation entails monitoring comprehension; that is,
'checking [one's] representation [of problem] very
carefully against the problem statement before [one]
launch[es]
into any massive solution attempt
(Hayes,
1987:18).
Successful comprehension monitoring involves
considering alternative problem representation and/or
independent detection of source and nature of difficulty.
348

---

Put differently,
adequate monitoring of understanding of
problems implies recognition,
by the problem solver,
of a
built-in and/or an induced difficulty.
For example,
appropriate problem representation
strategies are to lead the problem solver to acknowledge
that added difficulty in understanding a problem may
result from his/her approach to tackling the problem.
It
may be the case that the problem solver's failure to
understand the writing problem sterns from three potential
sources:
(i)
failure to select important information,
create objects,
and to establish a relationship between
the different components of the problem, taking domain-
specific or background knowledge into account;
(ii)
limitation of form of problem representation to internal
or mental representation;
and (iii)
absence of deploying
accurate LTM memory search for retrieval of knowledge
when domain-specific knowledge is not readily available
in STM.
In sum,
adequate comprehension monitoring
(as shown
by Thorndike,
1917; and Casanave,
1988) requires critical
assessment of one's previous understanding of the nature
of the problem at hand and readiness to adopt 'reflective
scepticism'
(Brookfield,
1987) or 'playing the doubting
game'
(Elbow,
1973).
Such an approach instantiates
flexibility in activating cognitive strategies towards
the process of problem representation.
349

---

Monitoring of planning process is another component
whose efficient activation is conducive to successful
problem solution.
Planning is viewed as a load reducing
strategy (Flower & Hayes,
1980a), which implies conscious
attention to the two problem spaces and their interaction
(Bereiter & Scardamalia, 1987)
or separating idea
generation and text generation (Collins and Gentner,
1980) .
Time on advance planning has been used to separate
expert from novice writers of English as a first language
(Stellard,
1974; Pianko,
1978).
The same criterion
serves as discriminating variable between skilled vs
unskilled (Flower & Hayes,
1980a)
or between native
writers
(L1)
and speakers of other languages
(L2/L3).
The commitment hypothesis has been advocated by pianko
(1979),
Sommers,
(1978,
1979) and Perl
(1979),
among many
others, to explain the non-existent or scant advance
planning of inexpert writers.
Raimes
(1987)
seems to
invalidate the commitment hypothesis.
Efficient planning, as seen in this study, requires
even redistribution of planning throughout the composing
process and appropriate adjustment of planning time taken
on advance planning (or during the preparatory process)
to task specificity.
For example, different allocation
of time on advance planning is necessary when a problem
solver is tackling
(1)
a familiar,
well-defined task,
(2)
350

---

a familiar,
ill-defined task,
(3)
an unfamiliar,
well-
defined task,
and
(4)
an unfamiliar,
ill-defined task.
Each of these tasks,
as we have seen,
calls for different
allocation of time on advance planning so as to meet the
demands of problem representation or the overall
organizational structure of composition.
Pursuing the analysis of sUbjects' responses to task
specificity further,
one may venture that the efficient
writing problem solver may be portrayed as one who
readily considers alternative courses of action when
faced,
for instance,
with ill-defined problems.
In such a
case,
he/she might resort to gap-filling decisions or
jump into the problem before he/she fully understands i t
(Hayes,
1987).
Or as the case may be,
he/she may
envisage hypothetical reasoning
which involves adding to the problem
representation by making hypothesis and working
out the consequences of these hypotheses to
learn more about the problem (Hayes,
1987:21).
Another point to be attended to in efficient
monitoring of the planning process is adjusting overall
planning strategy to the internal structure of tasks
(Hayes-Roth & Hayes-Roth,
1979) .
Given an inherent
hierarchical task structure,
as exemplified by Byrne's
(1978)
planning meals,
a top-down planning strategy is to
be invoked,
while other tasks might call for a bottom-up
or an 'opportunistic planning' as instantiated by Hayes-
Roth's
(1979)
errands.
351

---

To summarize the discussion on monitoring the
cognitive load of the composing process,
i t is worth
stating here that unless the process of evaluation is at
work throughout the composing process of written
discourse,
control and regulation of the processes of
problem representation and planning of solution cannot be
optimized.
Just as the cognitive domain is instrumental in the
writing process and on the writing performance so is the
affective domain.
Although research evidence supporting
this assertion is scarce,
as deplored in Green
(1982),
this point cannot be overemphasized.
The qualitative
analysis at micro-level has revealed that the most
successful sUbject in this study is the one who
consciously controls and regulates anxiety generated by
unfamiliarity with task, time constraints and memory
failure.
To end the discussion on task effects and inter-
subject response to task demands, two general
observations are worth making.
Firstly,
i t is to be
stated that one cannot offer a neat comparison between
the Hayes & Flower's (1980)
four 'configurations' of
writers and the four composing styles that have emerged
from this study.
Secondly,
a promising comparative
perspective may originate from adopting an integrated
approach to the interaction between task-related
differences and the sUbject factor.
352

---

The previous main subsection has dealt with the
issue of task effects and inter-subject response to task
demands.
After addressing the first main issue in the
discussion of the results,
focus is now on the second
main issue in this section.
6.2.2.
Episodic structure of the composing process
A question that can be raised is, What has the
episodic structure revealed about the nature and function
of pauses?
A first observation concerns the nature of
between boundary pauses.
Initial boundary- and final
boundary- pauses parallel the juncture- and hesitation-
pause dichotomy made in speech studies (Boomer & oitmann,
1962; Boomer,
1965; Barik,
1965).
Secondly, the episodic
analysis of the composing process has highlighted three
types
(type I-type Ill)
of re-reading,
examined in
Chapter 3,
section 3.3.2.
These three types of re-
reading exemplify three functions of pauses in the
process of composing:
(1)
evaluating,
(2)
scanning and
(3)
editing.
Additionally, pausing (in scribal activity)
may reflect concerns for revising beyond surface editing.
In discussing the finding related to the cyclic
nature of the composing episodes, it might be interesting
-to consider jointly this issue and that of planning and
its place of occurrence.
First, one may wish to
reiterate along with Zamel (1984)
that
353
..

---

writing cannot necessarily be broken down into
neat categories such as pre-writing, writing,
and rewriting,
for revision and invention have
been observed to co-occur throughout the act of
composing
(Zamel,
1984:986).
Another observation from our data, which supports
Rochester
(1973),
is that planning is a multi-level
decision-making process.
Thirdly,
the finding that
composing episodes exhibit an inherent cyclic structure
which allows pausing and event to alternate is borne out
by Johnson's
(1955)
assertion that preparation cannot be
limited to a given point in the process of solving a
complex problem.
A similar finding is available in Henderson et al
(1966).
As reported in Butterworth (1975),
'in samples
of spontaneous speech hesitant periods predominantly of
silence alternated with event periods predominantly of
phonation'.
One would tentatively consider as additional
sources of supporting evidence Goldman-Eisler (1968),
Butterworth
(1980), Beattie (1983).
In the discussion of the episodic structure of the
composing process, we have highlighted the nature of the
between-boundary pauses as indicating initial episode
markers and final episode markers.
Emphasis has also
been on the cyclic nature of individual episodes.
The
claim that planning and event alternate within a given
episode has been re-asserted.
It must be emphasized here
that there is no settled view on the issue regarding the
internal structure of episodes.
As observed in these
354

---

data,
conception,
incubation and production may run
concurrently, making i t difficult to 'be clear about the
demarcation of planning activity'
(Garman,
1990:132)
The observation by Garman
(1990),
coupled with his
balanced review of the issue,
suggests a moderate claim.
Thus,
given limited supporting evidence,
no strong
claim is made as to the inherently cyclic nature of
composing episodes.
The question that can be posed is:
Does this finding run counter the Flower & Hayes's
(1980,
1981)
cognitive model of writing?
Could i t not be argued
that the presence of the monitor accommodates to some
extent this characteristic of the composing process,
given that planning may be activated at any time?
An answer to both questions is that,
assuming that
the cognitive model allows for planning to occur
throughout the composing process,
such occurrence is only
virtual but not necessary while the finding in this study
suggests that planning and event necessarily alternate.
Furthermore,
for the cognitive model to offer a parallel
account,
this will have to be explicit.
As it is,
its
implicitness constitutes a limitation in itself.
The observed cyclic nature of composing episodes
allows one to restate that planning is not limited to the
initial phase of the composing process.
If i t is granted
-that planning occurs from the time a potential writer
develops awareness of a problem to the very end of the
solution of that problem,
i t can be argued that lack of
355

---

or limited advanced planning does not preclude planning
elsewhere.
Insufficient advanced planning may be
ultimately compensated for.
The implication of this
interpretation will be seen below.
6.3.
Limitations of study, need for further research
and
educational implications
At this stage one needs to pose and evaluate to what
extent the research questions raised in the statement of
the problem and subsequently reiterated elsewhere have
been answered.
To what extent have the methodological
perspectives been useful in the attempt to answer those
questions?
Finally,
in what respect can the findings
arrived at shed light on similar problem situations or be
generalized to the process of composing in general?
What
kind of educational implications can be derived from this
study?
These questions may not all be satisfactorily
answered.
However, starting with the limitations of the
present study and the need for further studies, a number
of points can be made.
6.3.1.
Limitations and need for further studies
The first limitation of the study can be seen in the
size of the sample considered.
Admittedly,
such a small
356

---

sample does not permit generalizability of the findings
to a wider population.
Furthermore, given the size of
the sample interesting statistical patterns have failed
to emerge.
It is be recalled that concern for rigorous
sampling,
coupled with practical issues,
overrode concern
for generalization of the findings.
This limitation can be tackled once the sample size
is increased to at least 25 for statistical significance.
The qualitative dimension could be manageable if the
number of tasks is reduced to one,
focusing on an
argumentative,
unfamiliar task as most of the hypotheses
examined in this study have been validated when that task
is considered.
Improvement of the qualitative perspective can be
achieved by developing an explicit grammar of protocol
analysis likely to facilitate the discrimination between
major and minor composing episodes.
Such a device could
be developed on the basis of a hierarchy of cognitive
operations occurring within specific composing episodes.
Further research is needed to investigate the
following:
(1)
Given a larger sample size, would one find
confirming evidence that all the hypotheses,
tested
quantitatively in this study,
are confirmed when sUbjects
,perform on the same argumentative, unfamiliar task?
357

---

(2)
Given a larger sample size, would subjects
perform better on the same referential,
unfamiliar task
than on the same referential,
familiar task?
(3)
Given a larger sample size, would there be
statistically significant evidence that time on advance
planning is related to the overall quality of written
performance?
(4)
Expert L3 student-writers will resort more to
higher level planning strategies than novice L3 student-
writers.
conversely,
novice L3 student-writers will
resort to sentence-to-sentence level planning.
(5)
Assuming that conceptual planning is a
distinctive feature of mature writing,
there is a
positive correlation between rhetorical planning and
syntactic maturity.
(6)
Assuming that idea-generation and text-
,
generation are two distinct processes,
simultaneous
occurrence of these two processes in the act of composing
results in greater cognitive loads.
(7)
Could sensitivity to discourse type and concern
for revision be developmental processes that are not
related to linguistic proficiency?
(8)
What factors cause individual subjects to
successfully monitor anxiety resulting from lack of
familiarity with tasks (and/or task difficulty as
perceived by subjects), amount of time for composing and
STM memory failure.
358

---

(9)
Assuming that the social context affects the
composing process, what social factors contribute the
most to variation in the process of composing written
discourse?
After considering the limitations and need for
further study,
the next point to be examined here is the
educational implications of the present investigation.
The question is what has one learnt from this study that
can be of any value to any educator?
6.3.2.
Educational implications
The fact that some hypotheses have not been
validated is an interesting finding in itself.
Though i t
is widely held that advanced planning is conducive to
better quality writing,
no empirical evidence is
available in this study to support such a claim.
In this
study,
i t has been observed that time on advance planning
does not constitute a discriminating feature among the
subjects examined.
A potential explanation is that such
a claim may find verification only when dealing with
expert writers while it may be invalidated when unskilled
writers are considered.
The ensuing implication is that it is not enough to
.prescribe a particular treatment and hope that it will
work at all times.
What is advocated here then?
Is one
to suggest a more individualized approach or a clinical
359

---

one to improving the process of composing written
discourse as a problem solving activity?
Would this
alternative approach be viable in the learning context of
'overpopulated' classes?
At this point it is worth stating the obvious; that
is,
initial contributions of students as problem solvers
are to be taken into account.
Problem solvers are not
tabulae rasae as often assumed.
To help sUbjects in
their process of solving a writing task,
for example, the
following open diagram (Fig.
6.1)
could be useful in
promoting a sense of commitment to the problem to be
solved,
by emphasising the uniqueness of any problem and
the ensuing personal answer to such a problem.
Before engaging in a description of Fig.
6.1, three
points are worth re-stating.
Basically,
i t has been
assumed throughout this study that central to any problem
solving situation is the definition of a problem as a
clash between the problem solver's worldview and the
problem statement.
It has also been stressed that
successful problem solving implies that the problem
solver recognizes the problem at hand as unique and that
its solution requires a personal answer.
Another claim
is that adequate monitoring of both the cognitive and
affective dimensions of composing entails constant
activation of the process of evaluation and optimization
of the proceSSes of problem representation and planning
of the solution.
360

---

WHIT HH'E
r DONE Hi
HHE I lODEU TO
'H1T OTHEII sOLUTlOHs
I
rmHG TD UHDERTIHD?
W BE EHVls.,GED'
THE REsTRll1IDNs
IN THE PRDBLEM
smE.IEH~
mBE I fIIU TIlE lIS!
f
HOi CA! I OPT/!llE!1
C1lCIilTIIE STlIHEGIES'
WE I !lUX ILL
iHIT CA!< I 00 TO RELP
BOmG BEClIlSE I oon
TIlE cOIISTRmrs
!E GET TD TIlE BEART
U!DEIISTlHD IT
IHTO _tCCOU~T?
Of TIlE IUITER ?
15 TIIERE OHLI ONE iH
I
I1I I.~
i'\\
00 I iAl!f
Of iESPOJNDING TO TIlE TlsK!
IHOWLEIIGE TD
E.lIsmG~~~~
BOW CA! I USE KY
INmlP1IET
CONVI/ICf
TIllS PRDBLE1l?
/fY READER?
BOli l!l I OOIHG ~
TI~lsE?
= WE I CONSIDERED
iHH mD OF
=
mBE
~ TO OPT/mE
IIIAT
VrEi THIT IUm:Rs?
KNOlII.EIIGE IXI
= THE AUDTED TIXE? TIlE WOI!:!
Wll\\T 15 THE ImT
I HIVE?
WANTS IS Nil!
EnICIM W.H Df
aWCIll
IlHAT DTllEII THIlIGs
USING THE mE
WBEii CIN I F111II I1IE
~ IIIIIT ISTHEImT
CAll I TlllNI: Of?
.u.wcITtD'
KNOliLEIIGE TllIT IS IiEEDED?
=SlflSFrlNG SOLIlTIDN 00 I HAlT
TO TIlE PIOBLE1l?
TD I!FOiR
XI WDEII?
HOII A! I COPING
TIlE WH I fEEL IBOUT
THE PII0BLE1l 15
illAT 00£5
ilTII THE ~OBLE1l?
WH.\\! 15 TRE PIOBLEIi
IHPOIIToo
TIlE
TO BE SOLlED'
IHSTRIJ<:i\\lII
"un?
OOES TIlE TASK nu. FOi
1
SPECIFIC STRATEGIES?
/5 THIS P1IOBLEII
OLD?
~ IS IT XI PERso,m
f - - .\\"5'EI TtiH C(J(IHTs'
WE J DELETED
15 TIllS PROBLEIi
f--
I!POI1Too DETAILS
15 IT MI PERsom
NEi'
I
fROIl TIlE PIIOOIIJI
COl!LD DIlGRtlS
1-
W!:!IEIKE TII.\\!
SUTE1lEm
RELP'
f - - wm .t~ I REDUIRED 1
COl.lMTS?
~
TO 001
1=1
HilI I HIDE
I--
f -
OOES TIllS I'ROBLEII
mE I ADJUSTED !1
I.' omlm
I-- H.HI I rDMlflED tI
CLASH ilTII xr "EllS?
PL\\IINING mln:cr
L IIIIEIE I sHOOLD
TO TIlE DEII.\\NDS Of
COVLD / mD',OTIIER f - -
smr FIOH ?
IXIES 11115 PIDRLEM
THE lISl'
mm 1'0EIs INESS
iH Df
TIlE REPREsElffING~~~~~~1
PIOBL[I'
1
I~ .1£'
11111111
i
~\\
I I
HOW .IM I COPING
loolmrro
1
ilTH
DiffERENT
FRDDin I PIECE
IsOl'RrES OF 'NHm?
IoIlR11I READI,G'
\\
\\
! 15 THE .'~XlETI IN
i ME DIIR'HEI.IlING'
'HIT C.\\N I 00 TO
\\
TDIJI TIllS PROBLEM
·1
'- iH\\T (~N I 00 TO DEAL
1.-)/ INTO OHE TII.\\T'S
I i1T11 THE IHIIIBITlHG l---V
iDRTII SOLrl,G'
r fEEWG DF fAILURE?
Figure 6.1
A descriptive diagram of the composing process
361
.'.

---

Given the above assumptions,
figure 6.1 has been
articulated so as to synthesize the main theoretical
discussion that has evolved around the nature of a given
problem and the interaction between such a problem and a
potential problem solver.
The diagram in Fig.
6.1 is
meant to provide a descriptive characterization of the
messy nature of the composing process.
By no means is it
intended to fulfil a pedagogic function.
Having made this point, we shall now attempt to
depict the overall structure of Fig.
6.1 so as to
facilitate its reading.
The first feature to be
highlighted is its centripetal organization,
moving from
the inner layer outward to the peripheral layers.
The
diagram in Fig.
6.1 comprises four layers,
one nuclear
and three peripheral.
The nucleus or inner layer
(darker
in shade)
constitutes the entry point and represents the
problem to be solved.
The second layer or the first
peripheral zone is schematized by horizontal interval
lines.
The third layer or the second peripheral zone is
characterized by vertical interval lines.
The fourth
layer or the third peripheral zone is signalled by white
plain bands.
These peripheral layers illustrate the
different times when the problem solver activates
cognitive processes such as problem representation,
planning of solution and evaluation of solution.
specifically, these outer layers indirectly suggest and
highlight the need for the problem solver to juggle with
362

---

different kinds of constraints such as task
characteristics
(ill-defined vs well-defined,
time limit,
available knowledge)
and affective load (resulting from
anxiety and other factors)
likely to interfere with the
problem solving process.
It is to be stressed that although some form of
ordering and grouping can be observed in the lay-out of
Fig.
6.1,
this diagram does not presuppose any logical
organization.
Attempt at imposing a logical structure
would have been self-defeating,
given the messy nature of
composing.
Put differently,
this four-layer zone diagram
by no means indicates a predetermined hierarchization in
its internal structure.
structuring and ordering will
ultimately result from the interaction between the
problem solver and the problem at hand.
In other words,
as a result of this interaction,
a question may evolve
inwardly,
closer to the nucleus,
or outwardly,
away from
the nucleus.
In sum,
potential relocation of framing and
structuring questions to anyone of the peripheral layers
may be observed in the process of composing.
This
phenomenon is captured by different reformulation of the
same questions and their occurrence in different areas of
Fig.
6.1.
The overlapping in zoning accommodates
different kinds of flexibility conducive to satisfactory
-problem solving by optimizing and maximizing sUbjects'
performance
(Kirkland & Saunders,
1991).
363

---

The second salient feature of Fig.
6.1
is the open
structure of the diagram, which denies that there is only
one way of solving a creative problem.
Such a structure
permits the accommodation of mental operations such as
incubation,
illumination and verification.
These mental
operations ,
it has been suggested,
function as check-
points intervening throughout the composing process.
In
so doing,
these cognitive check-points are bound to
permit proper 'allocation of mental capacity'
(Bereiter &
Scardamalia,
1983) and the association of 'spontaneity
and moment by moment interpretative process'
(Br i tton,
1983) .
The predominantly self-questioning structure has
been articulated so as to remind the problem solver,
if
need be,
that exploring and surveying the problem
situation involves considering alternatives.
Extracts
from different areas of the diagram may help illustrate
the point being made here:
Is it my personal answer that counts?
Does
this problem clash with my personal views? How
can I use my existing knowledge to interpret
the problem?
What can I do to help me get to the heart of
the matter?
Could I find another way of
representing the problem?
What is the most
satisfying solution to the problem?
Have I
identified where I should start from?
Is this problem new?
Have I added to the
restrictions in the problem statement?
How am
I coping with different sources of anxiety?
364

---

As these questions and the other contained in the
diagram show,
the need for the problem solver to
reformulate the problem statement is instrumental in
ensuring understanding of the problem at hand.
Alternatively,
the problem solver might have to turn a
trivial instructor-initiated problem into one that is
likely to drive him/her to seek an adequate solution that
will eliminate the difficulty.
Assuming that the ultimate goal of any written
composition programme is to help the student writer to
become an autonomous problem solver,
i t is suggested that
an indirect 'interventionist' approach would be
appropriate.
Such an approach is to be designed so as to
allow the student-writer to
(1)
develop awareness of
his/her composing process,
from the time when he/she
becomes aware of a potential writing problem or when
he/she understands the nature of a given writing problem
to the solution of that problem;
(2)
internalize the
problem solving strategies of skilled problem solvers so
as to activate these in any problem situation.
An
awareness of the open structure of the diagram in Fig.
6.1 should facilitate flexibility and self-monitoring of
planning and evaluating strategies so as to adjust to the
demands of the problem at hand,
irrespective of the field
of specialization.
365

---

Appendices
366

---

Headings
Appendix I:
Biographical and academic data
Appendix 11:
Raven Progressive Matrices
(PM 38):
Illustration
Appendix Ill:
Writing tasks
Appendix IV:
Technical equipment and elicitation of
protocols
Appendix V:
Questionnaire eliciting biographical and
academic data
Appendix VI:
Thinking about writing
Appendix VII:
First class assignment:
basis for
selecting SUbjects
Appendix VIII: Writing tasks: assessing degree of
familiarity
Appendix IX:
Evaluation of written products:
Instructions to assessors
Appendix X:
Composing timing data
Appendix XI:
Observation/segmentation scheme
Appendix XII:
Agreement among jUdges
Appendix XIII: Verbal protocols: transcripts
Appendix XIV:
Excerpts from marginal comments by
holistic markers
367

---

Appendix I: Biographical and academic data
FRENCH
ID
SEX
AGE
PLACE
PH
UNCI
TEEP
ENG
FL2
7.00
1
o
F
19.3300
DABOU
42
4.0
12.50
9.00
o
10.44
2
F
18.9200
ABIDJAN
46
2.0
12.71
13.25
o
12.16
3
F
21.7500
TRANSUA
41
2.0
12.25
10.76
10.30
4
15.35
14.00
M
23.3300
BLEDI
40
6.0
1
8.75
5
12.70
F
13.9200
ABIDJAN
46
11 .5
2
12.00
6
H
22.1700
ADIAKE
42
4.0
o
14.50
13.00
10.00
7
F
19.3300
JARNY F
52
12.5
2
17.00
15.10
10.20
8
13.85
14.50
1·1
19.3300
ADKAKRO
45
11 .5
2
12.00
9
o
] 3 . 33
11.00
11
23.2500
AGBOVL
14
1.0
10
13.95
9.00
F
20.5800
AGBVL
46
11. S
2
16.00
11.00
11
18.6000
BOUAKE
47
10.0
1
14.00
11.00
12
13.00
10.30
F
20.8300
BOUAKE
48
10.0
1
11.15
13.00
13
14.62
11
20.3300
NGSBRDKRO
47
12.5
15.43
1/,
15.02
11. 66
M
20.0000
ADZOPE
51
12.5
16.49
15
14.00
14.50
12.50
11
20.5800
KATIOl~A
46
12.5
15
14.50
15.00
11 . 60
M
21. 9200
ASSIKMSI
36
11.5
2
16.00
17
1',1
21. 2500
KORHOGO
43
11.5
13.36
15.23
18
o
10.37
11
21.9200
FERKS
16
1.0
14.25
11.82
19
10.50
11. 00
F
21.1700
KORHOGO
33
10.S
1
10.80
~O
o
12.00
10.00
F
22.2500
DIMBKRO
49
1.0
14.00
11
o
10 00
l·j
20.3300
TRANSU;'"
1:'
:1. 0
11. 00
10.00
22
o
12.37
10.20
F
23.5000
ABID.JAN
50
1.0
13.91
23
19.7500
ANOUMABA
47
11 .5
2
]1.00
11. 00
11.00
2'i
20.7500
SOUBRE
4.7
14.5
3
15.00
11. 00
9.00
2S
23 .0000
ZEPREGUHE
47
G.O
o
13.80
7.00
9.00
26
.20.7500
PORTBOUET
6.0
1
12.72
13 . SO
11.43
27
2].5800
D.JATEKET
44
6.0
1
11.50
12.00
11.00
28
lS.S800
YAKRO
50
6.0
1
15.00
13.00
13.00
29
.:'.CJ.DOOO
ABIDJAN
48
6.0
1
.15.00
12.00
.11.00
30
20.4200
ABIDJAN
52
2.0
o
12.90
11.00
13.00
:n
20.4200
SOUBRE
43
11.0
2
16.31
14.50
12.00
32
21.2000
ABIDJAN
51
10.0
1
12.00
14.00
10.00
33
21.6700
BOUAKE
50
4.0
o
11.85
12.02
10.00
34
23.]300
KATIOLA
40
) .0
o
13.00
10.00
10.50
35
21. 3300
TEAPLEU
51
13 .0
3
13.00
14.00
10.00
36
21.8300
ABIDJAN
52
11. 5
2
15.00
11.00
11. 00
37
20.0000
LOGOVALE
19
4.0
o
11.00
12.00
10.00
38
24.1700
DABOfJ
8
11.0
2
10.25
11. 7::1
12.00
39
24.0000
BINGERVLE
11
11.5
] 4 . 30
9.00
11.00
40
20.0000
OULAKPABLI
43
4.0
o
12.25
12.37
9.26
Age'
21.16
Pl1,
41.57
UNCI,
7.60
TEEP,
1. 07
ENG'
13 .55
F'L2:
12.33
French:lO.84
Note that the National Commission for the screening of students for further
education
(Commission Nationale d'Orientation)
has been recently practising
a three stage-screening procedure.
On that basis,
subjects with mean scores
in English
(ENG),
FL2
(second foreign language)
and French above 10 were
selected after the first
stage.
Those with one mean score below 10 were
considered at the second stage and those with two scores below 10 were
considered last.
ID:
subjects'
identification;
PM:
Raven progressive
matrices;
UNCI:
Universite Nationale de Cote d'Ivoire;
TEEP:
Test
in
English for Educational Purposes.
368

---

Appendix II: Raven Progressive Matrices
(PM 38):
Illustration
t\\ ex: ()
t\\
1
2
3
4
5
6
PM 38 is a non-verbal
intelligence test,
which
is based on relational
reasoning.
A testee must fill the blank above with a geometrical figure.
he is to choose the most appropriate figure from the ones given
(1-6).
369

---

Appendix Ill: Writing tasks
Referential discourse
Personal experience
Writing task 1
Write a
two-page article in which you explain the role of
television as an educational tool.
This article is to be published in the Nigerian newspaper West
Africa.!
Writing task 2
Write a
two-page text in which you explain the differences
between an African extended family system and a modern conjugal
nuclear family system.
This article is to be published in the English Club Magazine.
Academic experience
Writing task 3
In secondary school,
you may have been exposed to a variety of
English languages.
At the University,
your initiation to Theoretical
Linguistics may have given you new insights into th0 analysis of
different linguistic systems.
Write a
two-page text in which you attempt to explain the
differences and similarities between British English and American
English.
This article is to be published in the English Club Magazine. 2
World Experience
Writing task 4
Recent misbehaviour of football
fans
in Europe as well as
in
Africa have led respective Government officials to take drastic
measures against
-irresponsible hooligans",
as some journalists refer
to them.
Write a
two-page text in which you explain
how the
misbehaviour of football
fans may affect the future of international
football games.
This article is to be published in an English speaking
magazine.
Writing task 5
Computers are now fulfilling important functions
in
individuals'
everyday lives.
To emphasize the importance of
computers,
the Ivorian Government have even taken a political
decision to introduce computers in primary school.
Yet,
a number of
parents are not informed of the potentialities of computers.
Write an essay explaining the importance of knowing how to use
computers nowadays.
Remember that the purpose for writing this
article is not to express your position against or in favour of the
370

---

introduction of computers in primary school.
You are to inform a
non-specialist audience,
including both parents and educators who are
eager to learn more about
the use of computers.
Argumentative discourse
Personal experience
Writina task 1
P. number of former Ivorian students \\;lho majored in English as
their area of specialization expressed the following view:
"The
English studies syllabus at the University of Ccke d'Ivoire should
only include works by African anglophone writers such as Wole
Soyinka.
Since British and/or American writers are not concerned
with African culture and civilization,
their works should be excluded
from the sy llabus. "
In a
two-page essay argue for or against the ideas in the above
quotation.
This article is to be publi~;hed in the English Club Magazine.
Writinq task 2
As a result of the increasing number of unemployed university
graduate students,
many Ivorian parents have come to question the
value of higher education.
These disheartened parents think it is a
waste of time and money to send their children to University since
further education does not ensure employment.
Instead,
these parents
strongly feel that newly graduate secondary school pupils
(secondary
school teen-agers who hold a baccalaureat)
should be sent to the
fields where they can have an on-the-job-training and can be rapidly
productive.
Write a
two-page essay in which you argue for or against those
parents' position.
This article is to be published in an English speaking
magazine.
Writing task 3
A survey conducted in 1987 has revealed that first year
university students,
at the Faculty of Letters and Social Sciences,
consider that,
unlike in secondary school,
at the University,
students are left to their own devices,
the relationship between
teachers and students is poor,
and that great emp)lasis
is placed on
critical thinking,
and on personal and independent work.
Write a two-page essay in which you argue for or against the
results of the survey.
This article is to be published in the English Club Magazine .
. Academic experience
Writing task 4
"We hold these truths to be self-evident,
all men are created
equal,
that they are endowed by their creator with certain
unalienable Rights,
that among these are Life,
Liberty and pursuit of
Happiness.
That to secure these rights,
Governments are instituted
among men,
deriving their just powers from the consent of the
governed.
That whenever any form of Government becomes destructive
371

---

of these ends,
i t is the Right of the People to alter or to abolish
it,
and to institute a new Government,
laying its foundation on such
principles and organizing its powers in such form,
as to them shall
seem most likely to effect their Safety and Happiness."
This excerpt is extracted from the Declaration of Independence,
written and revised by Jefferson,
and later revised by the
Continental Congress that met in Philadelphia.
This Declaration of
Independence contributed somewhat to lay the foundations of the
American Democracy.
Write a
two-page essay in which you argue for or against the
claim that the concept of American Democracy is a myth.i
World experience
Writing task 5
The June 1988 summit of the leaders of the seven biggest
capitalist countries resulted in an agreement to provide new debt
relief for the poorest third-w'orlcl countries.
Specif ically,
the
leaders agreed on a menu of options from which governments can
choose:
rescheduling of debt over longer periods,
concessional
interest rates and debt write-offs.
Eligible countries include sub-Saharan African countries,
which
are benefiting from the World Bank's special assistance programme,
and whose income per head is less than US $ 425.
Given these restrictions,
only 17 countries would qualify,
but
another 15 could be eligible.
These restrictions also exclude Latin
American countries.
Write a
two-page essay in which you argue for or against the
major decisions taken by European and Japanese Heads of State.
This essay is to be published in an English speaking magazine.
1. This task was retained for the experiment and was referred to
as the referential,
familiar task
(Reff or task 1)
2. This task was retained for the experiment and was referred to
as the referential,
unfamiliar task (Refu or task 2) .
3. This task was retained for the experiment and was referred to
as the argumentative,
familiar task
(Argf or task 3).
4. This task was retained for the experiment was referred to as
the argumentative,
unfamiliar task
(Argu or task 4).
372

---

Appendix IV:
Technical equipment and elicitation
of protocols
Description of Technical equipment for protocols collection
A multisignal distributor MD-1600P,
designed for use with a
colour special-effects generator in mUltiple camera hookups.
A microphone mixer MX-710
One special-effects generator SEG-1210P.
It produces a wide
variety of special effects,
such as switching, mixing,
wiping,
keying and colour matting
(Instruction manual,
2).
One V-matic video recorder.
One VHS video tape recorder.
One trinitron colour video monitor PVM-18S0PS.
It reproduces
the playback picture and sound of video
tapes recorded on the
PAL,
SECAM or NTSC4.43 system (Instructions manual,
2).
Two monochrome video monitors PVM-90CE,
designed especially for
professional use with camera switcher, wiper,
special effects
generator or other multiple-camera applications to provide
complete monitoring facilities
(Instructions manual,
2)
Techniques of protocol gathering based on Swarts et al
(1984)
Context:
Students work in an experimental room
Provide:
Writing materials,
a tape recorder and a tape
Instructions:
To be delivered in the following order:
First,
the experimenter gives general instructions and time
limit
(usually about an hour)
so as to provide the sUbject with
a rough idea of what the task is about;
Second,
the following specification is to be brought to the
attention of the subjects:
'The most important thing about this
experiment is that we want you to say everything out loud as
you are thinking and writing your essay.
Even if i t has
nothing to do with the task.
Stray remarks and irrelevant
comments are to be included';
and
Third,
subjects are to be given the specific assignment.
An
illustration provided by Swarts et al.
is expressed as follows:
'write about abortion,
pros and cons,
for Catholic Weekly'.
Presence of experimenter:
The experimenter may choose to remain in the experimental room
to check on the tape recorder,
change the side of the tape,
and
remind subjects to verbalize thoughts.
The experimenter may
choose to leave to avoid causing distraction or undue self-
consciousness on the part of the subjects.
Procedure after completion of task
Collect essays and all written notes numbered in order.
Use
transcribing machine to transcribe verbal protocols.
Typed
manuscripts must be double-spaced
(about 10-15 pages).
373

---

Appendix V:
Questionnaire
eliciting
biographical
and academic data
Biographical data.
Name :
Date of birth:
Place of birth:
Sex:
Tick
) as appropriate
M
F
Where do your parents live?
(village,
town,
neighbourhood)
What does your father do?
What does your mother do?
What does your guardian do?
Do you have a scholarship?
(Circle your answer)
Yes/no.
How many brothers and sisters have you got?
Are you financially responsible for them?
How many of them hold only the CEPE?
How many of them hold only the BEPC?
How many of them hold the BAC?
How many of them hold a university degree?
Pedagogical data.
Where did you attend kindergarten?
Where did you attend prlmary school?
Form
School
Village/town
Year
CPI
CPII
CEI
CEll
CMI
CMII
374

---

Where did you attend secondary school?
Form
Year
School
vi llal]e/to·..'ll
Residence;
(Parents' ,
sibl ings,
relat:.ives'
guanlians'
bo.-,rding
school)
6e
5e
4e
3e
2e
lere
Terminale
BAC
(Stream)
Year:
19
What was your English average grade 1n 3e?
What was your English average grade 1n Terminale?
What was your second language average grade 1n Terminale?
What was your French average grade 1n 2e?
Terminale?
List any English speaking countries you have visited.
Country
Year
Length of stay
Frequency
Linguistic Data
Mother tongue
(s)
Other languages spoken:
2nd.
3rd.
4th.
5th .
.Languages spoken at home:
375
'.

---

Appendix VI: Thinking about Writing
1.
In your opinion how important are the following ln
helping a person to write well academically?
(Circle your choice)
(5-1) 1
- Read a lot
5
4
3
2
1
-
Study grammar
5
4
3
2
1
-
Study vocabulary
5
4
3
2
1
-
Imitate other writers
5
4
3
2
1
-
Write a lot
5
4
3
2
1
-
Rewrite any written
work
5
4
3
2
1
2 .
Give your 0plnlon of the following as definition of
academic writing.
(Circle your choice)
(5-1) 2
-
A mechanical exercise
5
4
3
2
1
-
Sets of grammatically
correct sentences
5
4
3
2
1
-
The clear expresslon
of ideas,
knowledge
and information
5
4
3
2
1
-
Self-discovery
5
4
3
2
1
3. (a)
Do you take the needs of the reader into account
when you write?
(Circle your choice)
- Always
- Sometimes
- Never
3. (b)
Try to say why.
4. (a) When you write do you spend time thinking about
grammatical correctness?
(Circle your choice)
Always
- Sometimes
- Never
4. (b)
Try to say why.
I
5:
Extremely important;
4:
Quite important;
3:
Important;
2: Not very
important;
1: Not at all important.
2
5:
Very
good
description;
4:
Good
description;
3:
Reasonable
description;
2:
Poor description;
1:
Very bad description.
376

---

5. (a)
When you write do you focus on correctness and
meaning at the same time?
(Circle your choice)
- Always
- Sometimes
- Never
5. (b)
Try to say why
6(a)
When you write something for your teacher do you ask
yourself what he or she probably expects from you?
(Circle your choice)
- Always
- Sometimes
Never
6.
(b)
Try to say why
7(a)
Do you pass your written work around to friends for
comment?
(Circle your choice)
- Always
- Sometimes
Never
7(b)
Try to say why.
8.
Try to describe how you carry out an academic
writing task
(from reading the topic to the final
piece of writing)
9.
In carrying out an academic writing task what do you
usually find to be the most difficult thing?
377

---

Appendix VII:
First class assignment: basis for
selecting subjects
Assignment one
Instructions for first draft
Write to your best friend telling him or her how you
felt after your first contact with the National
University of the Cote d'Ivoire.
Instructions for second draft
Rewrite your first draft taking the following into
account:
You may want to talk about how you felt:
-after your first contact with the registrar's
office;
-after your first meeting with all the teachers of
the Department of English;
-after your first meeting with individual teachers
in different subject disciplines;
-when you had to solve accommodation problems.
Note: These are suggestions that you mayor may not want
to consider.
You may also want to include other points
of interest.
Instructions for third.
draft
Rewrite your second draft,
watching this time for the
following:
-use of language
-grammar
-vocabulary
-spelling
-level of formality
(formal vs.
informal)
Assignment two
Rewrite the same letter to a former secondary school
teacher,
following the same instructions given for
assignment one.
378

---

Appendix VIII: Writing tasks
a)
Assessing degree of familiarity:
Instructions
A task
is familiar to you if you J.,:nm~· about
I t ;
that
is,
if you can
talk or '..Jnt02 about
it with o?tdequace and suft:icient knov.'lec1ge or
in[onnat 10n.
Read the attached selections of writing t2sks.
Using the two point
scale of familiarity: Most
Earniliar/Least fillniliar.
Say which tasks are
most
farnilial- to you and which tasks are least
familiar to you.
(Tick
(v)
as appropriate)_
Referential
Discourse
Writing Tasks
Host.:
C:dlllj ~<,,:r
Least Famili3r
1
2
3
'1
5
1
A t:guffi2nta t i ve cli seau rse
Writing Tasks
!';OS[
familj ar
Least familiar
1
2
J
4
5
I
b)
Assessing degree of familiarity·
Results
Tasks
Referential Discourse
Argumencative Discourse
1
2
3
"
5
1
2
3
4
5
Si
F
F
U
U
F
U
F
U
U
U
52
F
F
U
F
U
U
F
F
U
F
53
,
F
F
F
U
U
F
F
U
F
5~
F
U
U
U
F
U
U
u
S5
F
F
U
F
U
F
F
F
U
U
S6
F
F
U
U
U
F
F
U
U
U
57
F
U
U
U
U
U
F
F
U
U
S8
U
F
U
F
U
U
F
F
U
U
59
U
F
F
F
F
F
F
F
U
U
510
F
F
U
U
U
F
F
F
U
U
General observations:
As a
~ule,
task 1 in che set of referential discourse
type
is the most familiar
(:-)
( 0
all th,::, subjecl..s;
task 3 in the same set is
the least
familiar
(U)
to c.ll
the subjects_
Task 2
in the set of
argumentative discourse is the most familiar while task 4 in the same
discourse type is the least
familiar.
379
.'.

---

Appendix IX:
Evaluation of written products:
Instructions to assessors.
a)
Instructions to holistic raters.
Please mark these scripts holistically,
giving equal weight to
organization/content
(50%)
and to granrnar/vQcabulary
(50%).
Mark for
relevance,
organization,
and language.
Use Liz Harnp-Lyons
'FORMATIVE
FEEDBACK PROFILE'
(1986)
as a reference point
(see attached copy) .
Marginal con~ents on students' papers would be appreciated.
(Scripts
are to be marked as ordered.)
b)
Instructions to analytic raters.
The attached four sets of scripts are essays written by ten
University students majoring in English as a Foreign Language.
1.
Please mark them,
using the attached set of 7 analytic
criteria
(see attached copy) .
2.
Scripts are to be marked as ordered.
3.
Enter students'
scores on the attach2d attribute scales.
c)
Instructions to raters of introductory paragraphs.
The attached four sets of scripts are introductory paragraphs
of essays dealing with four different writing tasks,
written by ten
different University students majoring in English as a Foreign
Language.
A.
Please read them and briefly answer the following
questions:
1.
In each introductory paragraph, what problam (s)
does the
writer identify?
2.
What is the problematic situation (if any)
in which the
writer finds himself?
3.
What is the unknown solution (if any) which the writer
has identified as likely to eliminate the problematic
situation?
Note:
A problam is created by a sense of uneasiness experienced by a
writer.
A problematic situation results from an awareness of an
inconsistency,
a clash or a conflict between the writer's belief and
a belief in the world.
The unknown solution is the solution that is
likely to eliminate the problematic situation.
B.
Assign an overall mark--on a
five-point marking
scale--which you think best describes the quality of
individual scripts.
1--------1--------1--------1--------1
1 2 3 4 5
1.
Very bad introductory paragraph.
Totally
inadequate.
Problem not stated.
Problematic
situation and unknown solution not described.
No overall pattern underlying the unit.
2.
Poor introductory paragraph.
Partly inadequate.
Problem
not clearly stated.
Description of problematic situation
380

---

and unknown solution 110t clear.
Overall pattern not
sufficiently clear.
3.
Reasonable introductory paragraph.
Adequate for the most
part.
Statement of problem relatively clear.
However,
no
clear distinction between description of problematic
situation and description of unknown solution.
Overall
pattern blurs the 1ssues.
4.
Good introductory paragraph.
Adequate.
Clear statement
of the problem.
Exhaustive description of problematic
situation and unknown solution.
Overall organizational
pattern adequate.
5.
Very good introductory paragraph.
Adequate and
effective.
Problem clearly and effectively formulated.
Description of problematic situation and unknown solution
exhaustive.
Organizational pattern contributesto
effectiveness in stating problem as well as in describing
problematic situation and unknown solution.
Note:
Sets are to be marked in order given. Marks are to be written
on individual scripts.
d)
Procedure followed by Kelt raters when marking the TEEP Test.
(1)
UNCI
(National University of the Cote d'Ivoire)
mark out
of 20.
TEEP mark on scale of 0-4
(2)
To get 10 or more on the UNCI scale students must:
(a)
show they have undertood that even good university
qualifications
(in the Sciences,
what
is more)
provide
little guarantee of a
'suitable'
job-at least in the
short run
(in the UK at this time).
(b)
comment on this,
either by adding further
observations on the situation in UK (or the developed
world)
or by drawing a comparison between graduate
employm~nt opportunities in UK (or in the developed
world)
and in Cote d'Ivoire
(or the third world).
381
..

---

APPENDIX X: COMPOSING TIMING DATA
I
Di=:::course
I
1---------------------------------------------------------------1
I
ARC
I
REF
I
\\ - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 1 -------------------------------1
0
I
Familiarity
I
Fallliliacir.y
I
1-------------------------------+-------------------------------1
I
FAr''lILIAR
I
UNFAMILIAR
I
FAHILI.i\\R
I
UNFAMILIAR
I
1----------------+---------------+---------------+---------------1
IPhlEpl
P
I A
IPhlEpl
P
I A
IPhlEpl
P
I A
IPhlEpl
P
I A
I
I-------+--~--+----+----+--+--+----+----f--+--+----+----+·--+--+----+----1
I Subjecc I
I
I
I
I
I
I
I
I
I
I
I
I
I
11
IP
I
11
6081
IP
11
5851
IP
11
2621
IP
11
4841
I
I
1'1'
121
3811551'1'
21
4312911'1'
21
71
]GII'I'
21
8712371
I
1'1'
I
31
512971'1'
31
201
841'1'
31
351
511'1'
31
lSI
681
I
1'1'
I 41
211
1541'1'
41
1411011'I'
41
611601'1'
41
11/
231
1
1'1'
151
111
931'1'
51
471
631'1'
5122611481'1'
512071
1731
I
1'1'
1611011
17DI'I'
61
]41
871'1'
61
311
661'1'
6/
131
1291
I
1'1'
I
71
441
321'1'
71
101
971'1'
71
7211531'1'
71
271
431
I
1'1'
I
81
III
3951'1'
BI
111
11'il'l'
81
191
801'1'
81
151
891
/
1'1'
I
91
111
1811'1'
91
4211221'1'
I
91
991
611'1'
191
51
871
I
1'1'
1101
441
21211'
101
41
801'1'
1101
91
981'1'
1101
71
281
I
1'1'
1111
151
771'1'
111
81
341'1'
1111
1212461'1'
1111
712221
I
1'1' 1121
201
701'l'UI
61
901'1'
1121
913861'1'
1121
2712241
I
/'I'1131
711201'1'
1)1
41
391'1'
1131
141
571'1'
1131
81
831
\\
liT
\\141
III
6011'
lA1
7i
1161'T' 1141
51
IJ41T
1141
21
241
I
1'1'
151
1311021'1'
1151
111
1761'I'
1151
121
981'1'
1151
61
561
IT
161
521
9511'
1161
281
881'1'
11GI
171
591E
1161
911
1861
1'1'
171
41
541'1'
1171
1211231'1'
1171
291
1551'r
1171
51
471
1'1'
181
121
791'1'
1181
411
951'1'
1181
21
161'1'
1181
91
S61
lE
1911381
881'1'
1191
31
G41'1'
1191
711991£
1191
251
111
I
I
I
1'1'
1201
111
921'1'
1201
91
851'1'
12DI
871
191
I
I
I
,lE
1211
611.45611'
i211
191
2691T
17.11
71
901
I
I
I
I
I
I
1''1'
1221
31
381'T'
1221
171
501
I
I
I
I
I
I
I
I
I
I'I'
1231
81
1011
I
I
I
I
/
I
I
I.
1'1'
124 I
5 I
94 I
/
I
I
I
I
I
I
I
I'I'
1251
201
401
I
I
I
/
I
I
I
I
1'1'
1261
101
361
I
I
I
I
I
I
I
I
I E
127 I
24 I
152 I
-------+--+--+----+----+--+--+----+----+--+--+----+----.--+--+----~----I
SubjecclP
11
3041
IP
11
5401
IP
11
4921
IP
11
4901
I
12
IT
2110411891'1'
2117016031'1'
211331
741'I'
211301
701
I
1'1'
311411
631'1'
31
551
751'1'
31
501
901'1'
31
581139/
1
1'1'
41
851
2511'
41
2871
381'1'
41
891
1841'1'
41
761
ll81
I
1'1'
5134511541'1'
51
431
79 E
51
10
271'1'
511361
111
I
1'1'
61
411
941'1'
61
171
70'1'
6/
157
451'1'
61
3391
2391
I
1'1'
71
7911751'1'
71
241
75 'I'
71143
4671'1'
71
441
861
I
1'1'
81
1331
1181'1'
81
91
8B'I'
81
10
511'1'
81
3251
1581
I
1'1'
91
481
2001'1'
91
2GI
107'1'
I
91
29
601'1'
I
91
1481
3051
I
1'1'
101
251
751'1'
101
391
30'1'
1101
22
ll91'I'
1101
401
3141
I
11'
111
27112211'
111
61)1
58 T
1111
7
271'r
1111
2911831
I
1'1'
121
781
ll71'1'
121
291
4(; 'I'
1121
88
1021'1'
1121
31
841
I
1'1'
131
111
551'1'
131
211
59'1'
1131
8
441E
1131
151
601
I
1'1'
141
281
621'1'
141
131
87'1'
1141
249
1181
I
I
I
I
I
I'I'
151
llGI
631'1'
151
311
57'1'
1151
89
751
I
I
I
I
I
1'1'
161
141
831'1'
161
81
DJ 'I'
1161
13
481
I
I
I
I
I
1'1'
171
201
491'1'
171
301
711'1'
1171
7G
751
I
I
I
I
1'1'
1811211
931'1'
181
31
1l91T
1181
45
1031
I
I
I
I
1'1'
191
251
46/'1'
191
481
451'1'
1191
31
451
I
I
I
I
lE
/201
71
651'1'
201
71
631E
1201
3
271
I
I
I
I
I
I
I
I
1'1' 1211
321
791
I
I
I
I
I
I
I
I
I
I
I
i'I'
122 I
107 I
G11
I
I
I
I
I
I
I
I
I------~+--+--+----+----+--+--+----+----i--+--+----+----+--+--+----+---- I
( . . . continue)
382

---

\\-------+--+--+----~----.'--~--+---_+----+--.--+----+----+--+--+----+----1
ISubJecclP
I 11
167 I
IP
11
2811
I?
11 135\\
JP
11
1521
1
13
IT I 21
51 2731T
21
31
lo1~ IT
21
61
2011T
21
261
771
I
IT I 31
31 2881T
31
1121
67 IT
31
191
11"; IT
31
41
67 I
I
IT I 41
2:11
2561T
41
561
4 IT
21
171 1021'1'
41
91
1331
I
IT I 51
"I
21311'
51
861
791T
31
51
1381'1'
51
91 2591
I
IT I 6
1721
241'1'
61
17 I ]201'1'
41
41
1571'1'
61
41
1031
I
IT I 7
151 1071'1'
71
2 I 18G IT
51
2 I 1101'1
71
731
144 I
I
IT I 8
41 200lT
31
61
791'1'
61
21
97 IT
81
301
66 I
I
IT I 2
261
3211 '1'
81
431 2011 l'
71
111
160 IT
21
51
301
I
IT I J
231 3061T I 91
71
4991T
81
7/
351T
31
12 I ]231
I
IT I 4
61 3371'1' 1101
61 I 3481'1' I 91
271
1281'1'
5
171
63 I
I
IT I 5
201
1781 T 1111
31
113 IT 110 I
13 I 25211' I 9
121
881
I
IT I 9
J 0 I 137 IT
112 I
211
GGIT I 21
531
1~:'l IT
1]0
21
113 I
/
lE 110
3~1
41911' 1131
111
G) IT I 31
61
3911' III
61
82 I
I
I
I
I
IT I 21
3)1
3,11 l'
1101
151
"';i!T
112
17 I
95 I
I
I
I
I
IT I 61
61
J !j8 IT
I 5 I
2el
9,; 1'1'
113
8 I 494 I
I
I
I
I
IT I 3 I
101
:;23 IT
I <11
.; I
90lT 114
531
57 I
1
I
I
I
IT I 71
151
163 IT I 51
21
5,) IT
I 15
371
1181
I
I
I
I
IT I 81
291
BOlT I 61
811
~S 1'1'
116
101
311
/
I
1
I
IT I 91
81
5011' III I
191
1821'1' 117
341
281
I
I
I
1
lE
I 14 I
~t
811'1' I 91
1 I
:iG IT 118
801
591
I
I
I
i
I
I
I
I
IT
I 12 I
141 271 (1' 119
71
1071
I
I
I
I
I
I
I
I
lE 113 I
461
l57 lE 120
521 6041
1
I
I
I
I
I
I
I
IT 114 I
111
1 GO I
I
I
I
I
I
I
I
I
I
I
I
lE 113 I
121
1921
1
1
I
I-------i--+--+----~----+--+--+----.- ___ +--~--~----.----+--.--+----+----\\
ISubjectlP
11 2281
IP
11
6161
IP
11 233 I
IP
11
434 I
I
14
IT
21
251 2621'1'
21
371
3831'1'
21
23(
77 IT
21
1681
258 I
I
IT
31
271
971'1'
3 I
241
1401'1'
31
16 I I
711'1'
31
641 2721
I
IT
41
481 1411 '1'
'11
2081
1111'1'
41
41
14811'
41
137 I 2931
I
IT
51
4';1
22)11'
51
151
59)'!
51
921
1781 T
51
241
841
I
IT
61
221
,10 IT
61
13 I
47 IT
61
241
8011'
61
271
91 I
I
IT
71
I I I
691T
71
46
1281'1'
71
1481
ID IT
71
91
] 50 I
I
IT
81
171 20311'
81
27
1291T
8 I
22f
1521T
8 I
371 1041
I
IT I 91
511
15811' I 91
2,11
GBIT I 91
12 I 11211' I 91
521
82 I
I
IT 1101
471
921'1' 110 I 217
1781'1' 110 I
51
G4 1l' 1101
131
511
I
IT 111 I
27 I
271'1' 1111
41
) 62 1'i' 1111
691
481'1' 111 I
111
771
1
IT 112 I
20 !
BOlT I 12 I 145
1021£ 112 I
41
5711' 1121
191
1511
I
IT 113 I
111
261'1' 1131
40
12711' I 13 1
47 I
361'1' 1]31
12 I 115/
I
IT 114 I
251 170 IT 114 I
3
1091'1' 114 I
51
236 IT I 141
41
1661
I
IT I 151
291
951T 115 I 118
1071'1' 115 ) 25GI
15211' 115 I
51
361
I
1'1'
161
151
142,1
I
I
IT I 16 I 1181
2811 l' t 16)
71
581
I
IT
17 I
631
871
I
I
/'1' 117 I
831
891'1' 117 I
10 I
761
I
IT
18 I
171
781
I
I
IT 1181
81
G41T 118 I
71
141 I
I
IT
191
4\\
1391
I
I
IT I 19 I
731
1481'1' 1191
401
55/
I
IT
201
51
1081
I
I
lE 1201
81
991'1' 1201
551
711
,
I
IT
211
771
721
I
I
I
I
I
IT 121 I
81
421
I
IT
22 I
191
94,'
I
I
/
I
I
lE 1221
91
751
I
IT
23 I
1 J I
681
I
1
I
I
I
I
/
I
I
I
1
IT
24 I
"I
921
I
I
I
I
I
/
I
I
I
I
I
IT
251
4 I
811
I
I
I
I
I
I
I
I
I
1
I
lE
26f
101
911
/
1
I
I
1
I
I
I
I
I
I-------+--~--+---------+--+--~----+----+--+--r----+----+--+--+----~----l
( , . - continue)
383

---

I-------*--+--~----+----+--i·--+----+ ____ T--T--+----+----+--~--i·----+----I
ISubjectlP
111440
IP
1 110 I;
I P I 1 1194 I
IP
1114761
I
15
1'1'
21
215
24611'
2
ii
20?IT I 2
521
1361'1'
21
1051 1211
I
1'1'
J I
13
3211'
3
27
2521T I 3
131
731'1'
31
331
1731
I
1'1'
41
16
11511'
4
32
';9 IT
1 4
681
3831'1'
4
141
83 I
\\
IT
51
48
2611'
5
1~ "
91 \\'1' I 5
311
58 1'1'
5
821
561
I
\\'1'
61
5
8511'
6
20
1051T I 6
31
197 1'1'
6
281
561
I
1'1'
7\\
25
2411 '1'
7
8
'iD 11'
I 7
91
721'1'
7
31
2561
I
1'1'
8\\
15
451'1'
8
3
1971'1' I 8
321
1621'1'
8
1201
178
1'1'
91
2
2431'1'
9
22
2801'1' I 9
231
217 1'1'
9
101
88
1'1'
10 I
14
991'1'
10
9
811 '1' 110
261
65 1'1'
10
14 I
72
1'1'
III
7
3611'
11
37
871'1' III
20 I
97 1'1'
11.
61
73
1'1'
12 I
4
65 1'1'
12
15
1491'1' 112
81
84 1'1'
12
81
48
1'1'
131
3
1031'1'
13
18
791'1' 113
461
1561 '1'
13
211
25
1'1'
14 I
9
ll81'1'
14
171
69 Il' 114
31
65 1'1' 114
121
110
1'1'
151
16
36 1'1'
15
201
29(;11' 115
61
491'1' 116
21
41
1'1'
161
4
37 I
I
1'1' 116
21
ll61'1' 117
31
83
1'1'
171
16
721
I
1'1' 117
2,11
1891 '1' 118
81
192
1'1' 1181
6
123 I
I
I
I
I
I
I
I
1'1' 119 I
5
251
I
I
I
I
I
-------.--,--i----,----~ __ T--· _________ ~-----~----+----+--+--+----+----
I
Subjectlp
1 I 7291
IP
1 I ~72
?
11
5621
IP
11
491
6
\\'1'
21
14 I 29'11 T
2/
G,;
270 '1'
21
151
16411'
2 I
181
57
1'1'
31
61
941'1'
31
B
27 '1'
31
61
2481'1'
3/
61
57
/'1'
41
91
541'1'
41
105 l'
41
121 11111'
41
371
127
~
IT
51
101
28 Il'
51
-,
129 '1'
51
31
17 1'1'
51
81
103
IT
61
141
65 1'1'
61
2
32 '1'
61
21
1251'1'
61
221
173
IT
71
481
1601'1'
71
4
83 '1'
71
101
36 1'1'
71
131
60
1'1'
8/
31 2941'1'
81
8
155 l'
81
81
9211'
81
441
409
1'1' I 91
31
931'1' I 91
5
40 l' I 91
31
60 '1'
I 91
71
94
1'1' 110 I
41
22311' 110\\
7
78 l' 1101
111
168 '1' 1101
57 I
48
1'1' III I
61
3811' III I
5
284 l' 1111
201
25 l' 1111
81
50
1'1' 112 I
III
45711' 112 I
12
199 l' ! 12 I
271
83 '1' 112 I
57 I
74
1'1' 1131
21
15111' \\13/
10
6':;JT 113 I
31
62 l' 1131
96 !
80
1'1' 114 I
-il
427 1'1' 11'11
6
3317
1~ I
11
60 '1' 114 I
41
75
IT 1151
51
66 Il' 1151
LI
GlIT
15 I
201
378 '1' 115 I
21
97
IT
IIGI
71
1201£ 1161
921
96511'
16 I
261
95 '1' 1161
41
170 I
IT 117 I
31
4GI
I
I
\\
IT
171
241
39 '1' 1171
211
52 I
IT 1181
81
117 I
I
I
I
IT
181
41
95 '1' 1181
3/
1421
I
I
I
I
I
1
I
I
1'1'
191
71
135 '1' 1191
71
64 I
I
I
I
I
I
I
I
I
1'1'
201
61
265 '1' 1201
71
331
I
I
I
I
I
I
I
I
1'1'
211
51
45 '1' /211
61
861
I
I
I
I
I
1
I
I
IT
231
81
81/'1' 1221
71
108 I
I
I
I
I
I
I
I
I
1'1'
241
211
1301'1' 123 I
401
67 I
I
I
1
I
I
I
I
I
1'1'
251
41
322 1'1' 1241
641
1091
/
I
I
1
I
I
I
I
I
I
I
1'1' 1251
861
901
I
I
I
I
1
I
I
I
I
I
I
I
lE
126 I
191
4831
I-------+--+--+----+----+--+--*----+----+--~__ +----+----+--+--+----+----1
( ...
continue)
384

---

I-------+--+--+----+----i---+--+----+----+--~--~----+----+--+--+----+----1
SubjectlP
1
157 I
IP
I 11
6831
IP
I 11 214
IP
11
5041
I
7
1'1'
2
71
1841'1' I 21 1l2\\ 31BI'I' I 21
29
93 1'1'
21
1631
1B5 I
1'1'
3
261
501'1' I 31
511
169JT I 31
6
10811'
3
8:J I
721
1'1'
"
21
1181 T I 21
421
7311'
I ill
51
1731'1'
"
171
901
IT
2
541
23 I'f I 31
7 '
-,
9] IT
I 51
9
G81E
5
11
511
IT
4
81
501'1' I 41
3661
571T I 61
56
15,; IT
6
411
4401
11'
5
1 I
911T I 5 I
101
561'1' I 71
97
1511T
7
601
364 I
IT
6
91
211'1'
61
111
11811' I BI
G
291T
8
151
1571
IT
7
4201
951'1'
71
51
27 Il' I 91
5
31BIT
9
91
B8 I
IT
8
.;1
791T
BI
51
6J IT 1101
6
17 lE
10
SI
1251
1'1'
9
31
53 IT
91
BI
561T 1111
20
2421'1'
11
31
117 I
IT
10
101
HI'I'
10 I
61
5611' 112 I 113
1531'1'
12
701
911
IT
11
61
771'1'
111
41
If10IT 1131
6C
GOIT
13
21
103 I
1'1'
12
56!
170 IT
12 I
81
49 1'1' 1141
37
221'1'
1,1
III
4GI
1'1'
131
7
501'1'
131
101
1L?IT 115 I 102
ti4 lE
151
51
461
1'1'
12 I
20
13 1'1'
14 1
121
771E 1161
" 2611
I
I
I
11'
131
10
461'1'
15 I
31
49 I
I
I
I
I
I
I
IT
14 I
8
961'1'
161
251
1071
1
I
I
I
I
I
1'1'
151
1
90 1'1'
171
91
107!
I
I
I
I
I
1
1'1'
161
"
191'1'
18 I
! 51
2051
I
I
I
I
I
I
IT
171
B
451'1' 1191
171 1071
I
I
I
I
I
I
IT
181
2
97 lE 1201
L(1 I .D11
I
I
I
I
I
I
IT
19 I
13
181
I
I
I
I
I
I
I
I
I
I
1'1'
201
7
57 I
I
I
I
I
I
1
I
I
I
I
IT 1211
2
1191
I
I
I
I . I
I
I
I
I
I
I
lE
1221
7
3241
I
I
I
I
I
I
I
I
I
I
I
I-------I·--+--+----+----I--+--+----~----+--+--.----I.----+--+--~,----+----I
ISubjectlP
I 11
891
IP
11
1091
IP
11
321
IP
I 11
551
I
18
1'1' I 21
211
361'1'
21
501
611 'I'
21
6
1901T I 21
14 I
33 I
I
1'1' I
31
3!
65 1'1'
31
27
731'1'
31
19
2451'I' I 3
251
113 I
I
1'1' I '11
37 I
47 1'1'
4 I
11
'/0 11'
41
27
1531'1' ! -1
271
2931
I
1'1' I 51
281
2211 'I'
51
24
301'1'
51
17
27 IT
I 5
531
2G8(
I
1'1' I 61
211
24 IT
61
37
15217
61
41
441'1'
I .,
4 l i I 187[
I
IT I 71
51
1201'1'
71
54
1041'1'
71
2B
12011' I 7
671
261
I
1'1' I 81
201
67 1'1'
BI
14
2511'
81
17
1021'1' I 8
71
139 I
I
1'1' I 91
SI
4611'
I 9
<1l
;::011' I 9 I
1B
67 IT
I 9
871
361
I
IT
1101
231
261'1' 110
6
511'1'
101
49
951'1' I 10
411
62 I
I
1'1' 1111
GBI
235'1 l' III
7
86 11'
111
73
187 ['I' I 11
431
1431
I
1'1'
] 2 I
B7 I
901'1'
I 12
4
3181'1'
121
39
1321'1' 112
841
1161
I
11'
131
113 I 121 Il' 113
30
8611'
131
]0
21 1'1' 113
61
212 i
I
IT
111 I
621
1861'1' 114
26
2101'1'
141
35
1211 'I' 114
61
52 I
I
1'1'
15[
211
6511' 115
4
4211'
15 I
73
921'1'
15
31
263 I
I
11'
161
431
1201'1' 116
5
13211'
161 105
671'1'
16
331
2661
I
1'1'
17 I
14 I 1401'1' 117
16
6611'
171
36
3231'1'
17
381
1511
I
1'1'
18 I
1)61
101 1'1' 118
541 1321'1'
18 I
26
2611 'I'
lB
301
2721
I
1'1'
19 I
7G I 2231'1' 119
13 I
5111'
191
33
1631'1'
19
521
3221
I
1'1'
201
61
661'1' 120
361
B41'1'
201
55
2411
I
I
I
1'1'
211
721
451'1' 121
251
221'1'
21 I
71
17 6 I
I
1
!
1'1'
22 I
231
1741'1' 122
281
901'1'
221
111
161
I
I
1
1'1'
231
13 6 I
441'1' 1231
131
561
I
I
I
I
I
I
1'1'
241
51
37 1'1' 1241
451
1021
I
I
I
I
I
I
1'1'
25)
91
701'1' 1251
261
701
I
I
I
I
I
I
IT
261
71
16811' 1261
101
381
I
I
I
I
I
I
I
I
I
1'1' 1271 1651
94 I
I
I
I
I
I
I
I
I
I
1'1' 1281
B31
1181
I
I
I
I
I
I
I
I
I
1'1' [291
691
74 I
I
I
[
I
I
I
I
I
1'1' 1301
27 I
19 I
I
I
I
I
I
I
I
I
1'1' 1311
181
491
I
1
I
I
I
1-------+--+--+----·----.--+--+----+----+-- +--+----+----+--+--+----+----1
( .. ,
continue)
385

---

1-------i---+--+----~----i·--+--j----+----i--~--4·---- +----+--+--10 ----+----1
ISubjectlP
I 1 I 2491
IP
I 1 I 4601
II' I 1 I
251
II'
I li
127 I
I
19
1'1' I 21
71
6 G11' I 21
321 1081'1' I 21
11
3G611' I 2
981
1161
I
1'1' I 31
291
2011' I 31
SI 2391'1' I 31
10\\
liD 1'1' I 3
541
158\\
I
11' I 41
151
41 1'1' I 41
22/
2301'1'
I 41
,131
89 11' 1 4
91
136,1
1
11' I 51
351
2311' I 51
591
34GI1' I 51
SI
2311' 1 5
761 2291
1
11' I 61
191
13411' I 61
1561
111 I'T'
1 61
151
1031'1' I G
251
771
1
11' I 71
101
7111' I 71
341
1971'T'
I 71
111
47 11' I 7
81
721
I
1'1' I 81
161
1161'1'
I 81
251
531'1' I 81
81
13811' I 8
151
3041
I
11' I 91
141
j 03 IT
I 91
101
671'1'
I 9 I
351
8111' I 9
G5
110 I
I
1'1' \\ 10 I
41
S41E 1101
201
37811' 1101
601
15011' 110
225
701
\\
11' 1111
111
791
I
I
I
11' 1111
521
1091'1' \\ 9
15
691
I
11' 112 I
SI
85\\
I
I
I
11' 112 I
151
43 11' 110
7
1161
I
1'1' 113 I
31
44 I
I
I
I
11' 113\\
38\\
G311'
III
16
181
1
11' 1141
281
17 3 I
\\
1
\\
11' ) 14 I
8\\
921E I 12 I 312
3131
I
1'1' 1151
131 297 I
I
1
1
1'1' 1151
18 I 1741'1' 113 I
22
84 I
/
\\1' 1161
37/ 2701
I
I
I
1'1'
I 1GI
131 1211'1' 114 \\
18
2111
I
11' 117\\
111
2251
I
I
/
\\1' 117 I 140 I 1941E 115 I
4
2121
\\
lE 1181
121
3SGI
I
I
I
11' 1181
93/
245\\
I
I
\\
I
I
I
I
I
I
I
I
I
IT 1191
331
2001
I
I
I
I
I
I
I
I
I
I
I
I
IT 1201
591
2431
I
I
I
I
1
I
I
I
I
I
I
I
1'1' 121 I
871
77 I
1
I
1
I
I
I
1
1
I
I
I
IT 122 I
81
1551
I
I
I
/ - - - - - - - 1
-+--~ ____ ~----1--+--+----1----1--~
__
+----+--+--i· ----+----1
4
_ _ _ _
ISubjecl I'
1 I 3431
I P
11
6391
II'
11
G101
\\1'
11
5051
I
110
l'
21
701
10111'
21
22)
1321'1'
21
471
12511'
21
221 2081
I
'1'
31
101
83 IT
3 1
41
31 11'
31
31 1171'1'
31
6\\
1551
I
l'
41
81
SLIT
41
121
[>811'
41
181
G9 11'
41
71
1911
\\
l'
SI
241
264 1'1'
SI
71
221'1'
SI
41
74 1'1'
5 I
SI
61 I
/
l'
61
111 1401'1'
61
81
100 1'1'
61
161
1451'1'
61
21
3311
I
l'
71
16
87 11'
71
121
71 1'1'
71
111
59 11'
71
191 1141
I
l'
81
4
71 11'
81
61
57 IT
81
15'
75 1'1'
81
5\\ 1861
I
l'
I 91
16
74 1'1' I 91
91
,10 1'1'
1 91
9
1081'1' I 9 I
121
711
I
11' 110 I
5
13711' 1101
SI
11011' I 10 (
10
16411' 1101
3
2181
I
11' 1111
4
lSGl1' 1111
61
901'1'
( ill
8
GO 11' 1111
6
1841
I
1'1' 1121
5
18911' 1121
] 0 I
'i7i'1'
12 I
8
1021'1' 112 I
39
761
I
1'1' 1131
7
2GI1' 113 I
261
:;31'1'
lJ I
79
2091'1'
1131
5
1 SS I
I
11' 114\\
58
411'r 114 I
131
lOG 11'
It, 1
5
47 IT
114 I 159
531
I
IT I 151
76
63 11' I ] 5 I
81
1651'1'
15 I
7
78 11' 115\\
11
1081
I
11' 1161
16
83 11' Ilfil
111
8711'
1GI
15
98 1'1' liGl
31
761
I
1'1' 1171
3
561'1' I 17 1
231
1081'1'
171
6
79\\1' 117 I
6
69 I
I
\\'1'
118 I
iG
13011' 1181
121
571'1'
18 I
8
70\\1' 1181
15
401
I
lE 1191
36!
40111' 1191
181
22·~ 1'T'
191
4
GGIE 119 I
14
432 I
I
I
I
I
I
1'1' 1201
4/
235 lE
201
25
312 I
I
I
I
I
I
1
1
I
11' 121 I
41
1831
I
I
I
I
\\
I
I
I
I
I
I
11' 1221
101
921
I
I
I
I
I
I
1
I
I
I
I
11' 1231
211
601
I
I
I
I
I
\\
I
I
I
I
I
11' 1241
181
1421
I
I
I
I
I
I
I
I
I
I
I
lE 1251
341 3481
I
\\
I
I
I
I
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
386

---

Appendix XI: Observation/segmentation scheme.
Identifying units of concentration:
Instructions to
judges
1.
Study carefully this video protocol.
2.
Look for units of concentration in the writer's
process.
Composing episodes can be defined as
'units of
concentration'
or 'periods of sustained focus'
ln
the writer's process.
These episodes may be divided
into 'major'
composing episodes
and
'minor'
composing episodes.
'Major'
composing episodes are
'clearly autonomous episodes with strong
boundaries'.
'Minor'
compos ing episodes have
'weaker boundaries'.
As such they have stronger
connections to adjacent episodes.
In sum,
major
composing episodes are independent from other
composing episodes while minor composing episodes
are subordinate to other composing episodes.
In any of the composing episodes,
periods of pauses
in manual activity alternate with periods of
(event)
or scribal activity.
3.
Mark off a boundary if you intuitively feel that a
composing episode or a
'unit of concentration' is
clearly autonomous.
4.
Note the following:
i)
Thinking process behind 'text'
lS to be given
top priority;
ii)Linguistic and/or syntactic cues,
such as
anaphora,
deixis,
connectives and the like may be
misleading;
iii)
Re-coding might be necessary as first attempts
are re-examined.
Segmentation Procedure
1.
Mark off a boundary when you observe the writer is:
- shifting in focus
(change in object being attended
to: audience;
content; setting up goals);
- changing a train of thought; or
- or setting up a new objective.
2.
In the third colunm,
on the attached observation
sheet,
write down the first three words indicating
the beginning of a translating period and the last
three words signalling the end of the unit of that
translating period.
387

---

3.
Use pauses
(in manual activity of the writer)
as
objective indications of composing episode
boundaries.
4.
Press the pause button of the video player,
and in
the first column, on the observation sheet,
enter
the time appearing on the monitor screen,
indicating
the beginning of a unit of concentration.
Immediately below that recorded time,
enter the time
indicating the end of that
'unit of concentration'.
Both recorded time boundaries are to appear between
square brackets as follows:
[T b ]
[ T e ]
5.
Repeat steps 2-6 until you reach the very end of the
composing process.
6.
In making final decisions on the boundaries of the
units of concentration in the writer's process,
try
to match observed shift in focus
(e.g. audience) or
in setting up new objectives,
or change in a train
of thought, with pause in manual activity.
(Such
pauses may be filled or unfilled.
Filled pauses are
signalled by verbalized thought;
unfilled pauses are
complete silence.)
Note that after second thought, you may intuitively feel
that a previously coded major composing episode
is in
fact a minor composing episode.
If this turns out to be
the case,
draw a line linking the composing episode to
the adjacent one.
7.
Number 'observed' units of concentration as you go
along.
Numbers are to appear in the second columm,
on the observation sheet.
8.
Write down your observation
(s),
if any,
In the
fourth column,
on the observation sheet.
9.
Enter the total number of units of concentration on
the observation sheet.
388

---

Appendix XII: Agreement among judges
Judge
ArgIJlIlentative
Reterential
1
Untamiliar
Familiar
Untamlliar
Familiar
,
,
,
,
To!,; a 1
N-
Total
T
Tot.al
W
Total
tJ'
1
J6
, 1
~
~
"
~
35
, .,
2
"
::!':;
BD. 5
"
,,"
.:6 . 3 ~
17
Jr; . 53
"
B
" .E5
,
3
37
17
47. n
26
17
.: 1 • <1 b
"
.\\ . 65
1 ,
11
:;1 .42
,
27
18
50.00
29
15
J 5. S8
~2
18
, 1 .El~
"
J 9
" .2El I
N'
Number of agreement with judge 1
(This investigator).
The extremely low agreement on segm-::!nting some of the protocols
calls for the following
t-emarks:'
Extended
EOl:mal
tl-aining
in segmE-nting verbal protocol
is
needed.
A formalized grammar likel,!' to distinguish bet~.Jeen major
and minor composing epi30des should improve agreement on
establishing the boundary of composing episodes.
Distinguishing between cognitive operations such as
geIlerating ideas dlld illustrating assertions might help
incr~ase objectivity in segmenting composing episodes.
389

---

Appendix XIII: Verbal protocols:
transcripts
Conventions:
S4T4:
SubJect 4,
task 4
(argumentat i ve,
unfamiliar)
S7T4:
SubJect 7,
task 4
(argumentative,
unfamiliar)
S2T4:
Subject ")-,
task 4
(argumentative,
unfamiliar)
S4T3:
Subject 4,
task 3
(argumentative,
familiar)
S5T3:
Subject 5,
task 3
(argumentative,
familiar)
[text]
Reading task
[ text]
Reading or re-reading of notes 1n
verbalization of thought.
text
Verbalization of thought
text
vh-itten text
(as product)
by sub:iect
[text]
Reading of written text
text
Emphasis in subject's verbalized thought
( text)
Investigator's observation
390

---

1
Subject 4:
argumentative,
unfamiliar task (S4T4)
2
["We hold these truths to be self-evident,
all men are
3
created equal,
that they are endowed by their creator
4
with certain unalienable Rights,
that among these are
5
Life,
Liberty and the pursuit of Happiness. That to
6
secure these rights, Governments are instituted among
7
men, deriving their just powers from the consent of the
8
governed. That whenever any form of Government becomes
9
destructive of these ends,
it is the Right of the People
10
to alter or to abolish it,
and to institute a new
11
Government,
laying its foundation on such principles and
12
organizing its powers in such form,
as to them shall seem
13
most likely to effect their Safety and Happiness." This
14
excerpt is extracted from the Declaration of
15
Independence,
written and revised by Jefferson,
and later
16
revised by the Continental Congress that met in
17
Philadelphia.
This Declaration of Independence cons
18
contributed somewhat to lay the foundations of the
19
American Democracy. Write a two-page essay in which you
20
argue for or against the claim that the concept of
21
American democracy is a myth. I All right so this is my
22
topic
I.)
umm I
read it over ["We hold these truths to be
23
self-evident,
all men are created equal,
that they
24
endowed by their creator with certain unalienable Rights,
25
that among these are Life.
Liberty and the pursuit of
26
Happiness.
That to secure these rights, Governments are
27
instituted among men,
deriving their just powers from the
28
consent of the governed.
That whenever any form of
29
Government becomes destructive of these ends,
it is the
30
Right of the People to alter or to abolish it,
and to
31
institute a new Government,
laying his its foundation on
32
such principles and organizing his its powers in such
33
form,
as to them shall seem most likely to effect their
34
Safety and Happiness." This excerpt is extracted from the
35
Declaration of Independence, written and revised by
36
Jefferson,
and later revised by the Continental Congress
37
that met in Philadelphia. This Declaration of
38
Independence contributed somewhat to lay the foudations
39
of the American Democracy. Write a two-page essay in
40
which you argue for or against the claim that the concept
41
of American Democracy is a myth. )
I.} All right
I.} SO
42
what am I wanted to do really in this topic?
What am I
43
wanted to do?
[Write a two-page essay in which you argue
44
for or against the claim that the concept of American
45
democracy is a myth.
Is a myth]
{silent reading] All
46
rightl.} So am I
wanted I am wanted I am wanted I.}
to
391

---

47
judge whether the concept of American democracy has an
48
effect consisting (_) Wbether in reality in actuality
49
people may think or people may discover that (.) the term
50
the concept of American Democracy is real
(.) and this
51
demonstration is to be laid on this extract here (.)
52
extract written and revised by Jefferson (.) All right um
53
this is quite a tricky subject (.) Do I know anything
54
about the American democracy? (.) oh do I know anything
55
about it? (.) I've never been interested in politics (.)
56
AA It serves me righ£ toddy.
Wbat am I going to say? (.)
57
['We hold these truths to be self-evident,
that all men
58
are created equal, that they are endowed by their creator
59
with certain unalienable Rights,
that among life that
60
among these are Life, Liberty and the pursuit of
61
Happiness. That to secure these rights, Governments are
62
instituted among men deriving their just powers from the
63
consent of the governed. That whenever any form of
64
Government becomes destructive of these ends, it is the
65
Right of the People to alter or to abolish it. and to
66
institute a new Government,
laying its foundation on such
67
principles and organizing its powers in such form. as to
68
them shall seem mat likely to effect their Safety and
69
Happiness." This
excerpt is extracted from the
70
Declaration of Independence, written and revised by
71
Jefferson, and later revised by the Continental Congress
72
that met in Philadelphia. This Declaration of
73
Independence contributed somewhat to lay the foundations
74
of the American Democracy.
Write a two-page essay in
75
which you argue for or against the claim that the concept
76
of American democracy is a myth.
] All right. So in other
77
words 1.1 in other terms I.) I am asked to wonder (.1 or
78
to judge whether this declaration of independence by
79
Jefferson is really respected in the course of the
80
American policy I.) All right.
This is what
(.)
I am
81
asked (.) All right I.) I see what I am asked now. What
82
they want me to do is to know or to tell whether (.) the
83
American democracy takes into account (.) what Thomas
84
Jefferson I.) yes what Toma Thomas Jefferson said in his
85
declaration I.) All right [silent reading]
So what am I
86
going to say? Which aspects are to be considered in
87
dealing with a such a question? (_I Which aspects are to
88
be considered?
1.1 All right!_) So no matter no matter
89
the ideas will come by themselves.
What I must first be
90
concerned with is to try to write an introduction on the
91
basis of the faint idea I already have upon the subject.
92
(.) So I have eh What time is it? (.) Seven ten minutes
93
past (.) Okay it doesn't matter.
(subject writes down his
94
name) All right but I think that before trying to write
95
anything I must make sure I have grabbed the main idea
96
the general meaning of that statement by Jefferson (.)
97
All right (.1 So 1_) Jefferson recognizes that 1.1 every
98
human being on earth is equal to anyone else and that all
99
the government all over the world must take into account
100
this fact
(.) must consider that all men are free all men
392

---

101
are equal
I.) and I.J must accordingly respect their
102
dignity I.) All right
(.) and this is now the basis of
103
what can be called Democracy (.) So if a government that
104
does not respect
(.)
all these criterias
I.J is not a
105
democratic one and the people have the right to overthrow
106
it and to institute a government of their will
(.) All
107
right
(.) Okay. Sure. Okay. What am I going to begin by
108
in writing my introduction? (.) I'll begin by a general
109
introduction. What am I saying? I'll begin by a general
110
observation (.) a fact which everybody knows and thereof
111
trying to get my subject an interrogated. Okay.
That's
112
right (.) 1776 is a year deeply marked in the collective
113
memory of the American people;
for in this for in the
114
course of that year they got their independence, got rid
115
of any domination. Ever since a very acute sense of
116
liberty took hold of all their deeds; besides. Thomas
117
Tbmd or Tho~s I
don't know the pronunciation of it.
118
Jefferson had fairly
told or initiated this sense in the
119
Declaration of Independence: All right 1.1 he laid a
120
particular focus on the fact that all governments should
121
mind preserving their people's freedom and equality. All
122
right 1.1 Now we're going to wonder we're going to
123
interrogate this subject. We're going to ask a question
124
upon the subject. The subject roses enough certain
125
question certain number of question which we are going to
126
write now. But a is not this declaration mere statement?
127
In other wo~.) are we to find
an a application or
128
let's say a concrete application of that declaration in
129
the American inner and foreign policy?
All right 1.1 So
130
I've finished with the introduction. I read it over to
131
make sure I've not strayed from the topic.
[1776 is'"a
132
year deeply marked in the collective memory of the
133
American people; for in the course of that year they got
134
their independance, got rid of any damin any domination~
135
Ever since a very acute sense of liberty took hold of all
136
their deeds. Besides. Thomas Jefferson had fairly told or
137
initiated this sense in the Declaration of Independence:
138
he laid a particular focus on the fact that all
139
governments should mind preserving their people'S freedom
140
and equality. But is not this declaration mere staten,ent?
141
In other words are we to find a concrete application of
142
that declaration in the American inner and foreign
143
policy? J (.) All right"
I've wc; tten a very interesting
144
introduction Il m sure it1s the best introduction. Okay I
145
cheer up and since Ilve no precise information about
t.}
146
the American politics. What am I going to do? 1.1 Okay.
147
But at least Ilm sure I have some information about it.
148
Some historical information (.) must prove very useful to
149
me 1.1. Some historical information 1.1 All right 1.1
ISO
When studying in the course of the American political
151
historyl can I notice here some moments when America FaLl
152
to preserve human rights human freedom?
I.)
All right But
153
when asked when asked about American democracy I. I I an,
154
precisely asked to deal with inner policy with their
393

---

155
inner policy (.)
All right. There is no mdtter. We'll
156
try to write something. Sometime it is that you don't
157
have any precise idea about what you're going to write at
158
first but once you've begun writing some sentences you
159
discover that some ideas some ideas are crossing your
160
mind and you then realize that you knew something about
161
the topic so thus (.) I must cheer up I must never feel
162
desperated even i f some sometime at first I don't feel
163
(.) very learned about the topic. All right t.} America
164
claims herself to be a very democratic country; and
165
indeed the abundant work pieces about the American
166
Democracy would have us believe so. I.) But are we to
167
trust these essays or rather to take into account what
168
the facts claim? All right (.) In our particular work or
169
essay it's better essay
we'll take we'll ~ yeah ~
170
fair analysis of the facts
All right. ~ that's to say
171
that we'll make a great effort to express what is read
172
what is to be found in America. Of course we've never
173
been to America but through the infomation we've received
174
we have read on a newspaper or why not by hearsay we'll
175
try to discuss this topic.
[So In our particular essay
176
we'll lead a fair analysis of the facts)· and discover
177
that like most countries all over the world which claim
178
themselves to be democratic.
America owns several Parties
179
with different tendencies. If we are to believe or to
180
take into account the fact that these Parties All right I
181
think that parties must be written with capital p (.) it
182
must begin by a capital letter.
[These parties] mean the
183
people's opinions, then Democracy is to be found in
184
America. All right (.) So I can feel the ideas crossing
185
my mind now. How strange it is when I realize that at the
186
beginning I didn't have no idea about the American
187
democracy. Okay. So cheer up cheer up cheer up cheer up.
188
Never let yourself go to despair even i f you don't have
189
any idea try to cheer up the ideas will come. All right.
190
So multi-partism is a criteria a character of Democracy.
191
And i f we are to believe that these parties in America
192
convey a meaning a general opinion then this is a step
193
towards democracy. All right (.) Okay. So we have spoken
194
about opinion. Opinion is the word which is to be found
195
(.) in the sphere of politics especially in the field of
196
Democracy (.) because in a country that claims itself to
197
be politic polically democratic must take into account
198
the opinion and thus in this country one is to find
199
everyday or when necessary some voicing of opinion. All
200
right.
Is there in America any voicing of opinion? From
201
time to time (.) All right (.) er (.) All right (.) Right
202
So what can I say about the fact? What can I say about
203
the fact? (.) Is there press freedom? Is there any free
204
expression in the United States? All right. So we'll
205
begin a third paragraph saying that Press freedom and
206
liberty liberty of expression are meant by every country
207
that claims itself to be democratic.
(.) All right. I
208
must pay attention to something very important.
Okay.
I
394

---

209
must from time to time wonder whether my topic or
210
eve~thing I'm writing here or that I have written is
211
linking up with the topic. In this case I h~ve to wonder
212
whether the topic or whether the (.) my composition is
213
the right answer to the text here (.) Is my composition
214
taking into account everything that is said in the text?
215
Okay. Let's read it over.
("We hold these truths to be
216
sel-evident, all men are created equal,
that they are
217
endowed by their creator with certain unalienable Rights,
218
that among these are Life, Liberty and the pursuit of
219
Happiness. That to secure these rights, Governments are
220
instituted among men, deriving their just powers from the
221
consent of the governed. That whenever any form of
222
Government becomes destructive of these ends,
it is the
223
Right of the People to alter or to abolish it, and to
224
institute a new Government.
laying its foundation on such
225
principles and organizing its powers in such form. as to
226
them shall seem most likely to effect their Safety and
227
Happiness." This excerpt is extracted from the
228
Declaration of Independence, written and revised by
229
Jefferson, and later revised by the Continental Congress
230
that met in Philadelphia. This Declaration of
231
Independence contributed somewhat to lay the foundations
232
of the American Democracy. Write a two-page essay in
233
which you argue for or agaisnt the claim that the
234
American democracy is a myth.] All right. Let's have ~
235
look at everything we. have written so see whether we're
236
not astray trom the topic. (Silent reading ot text
237
produced so tar) All right. So atter re~ding our
238
composition we're very happy to discover that we've not
239
strayed trom the topic. All right. What time is it? Now
240
uuuuh eleven minutes.
I only have nineteen minutes left.
241
Time is running by. What a pity! [Press freedom and
242
liberty of expression are meant by every country that-
243
claims itself to be democratic.) On that account we can
244
notice that America is democratic when considering the
245
huge number of newspapers that abound there and also in
246
view of the great range of private channels of TV and
247
radios. Oh this is a ·very clumsy construction. I should
248
have written and also in view of the great range of
249
private TV and radio channels. Okay.
It doesn't matter.
250
It's only a mistake ot style(.)a tew points ott ... it
251
doesn't matter. Okay. So we'll try to (.) analyze now the
252
economic;,,]
level
because also Democracy can be seen
253
through a certain economy. All right. The last thing
254
we'll try to analyze the economical atmosphere in the
255
United States. The economic atmosphere convey the true
256
democratic mind. We'll say that At the level of the
257
economy Americans preserve a great flexibility as to
258
peoples' undertakings; the economical system, namely
259
capitalism is a bright reflect of a certain liberty which
260
the government tries to maintain.
(.) Am I really doing
261
what I am asked to do? What am I asked . ..
really asked
262
...
Haven't
I
stray from the topic? Hdven~t I
stray from
395

---

263
the topic? [Silent reading of text produced so far] All
264
right. Why should I here be speaking of capitalism when
265
asked about American Democracy? Is there any link between
266
capitalism and American Democracy? I.J Can we deny that
267
politics are the reflect in the economy of the country?
268
Can't we say that the economy I.J the economy is the
269
bright reflect of the politics in the country? All right.
270
In the case of America for example this liberty this
271
economical liberty isn't due to the fact in politics.
272
There is a certain liberty.
Yeah maybe I am sure I am not
273
wrong I am not wrong at all
I.J
Okay I am sure I'm not
274
wrong [SRd] All right, So we'll try to make a short
275
conclusion 'cause' we have only ten minutes left. Oh
276
dear! I've once more been taken aback. Okay it doesn't
277
matterl.} So All these elements taken into account we can
278
say that America is
'paradise' inverted commas ~
279
Democracy. But as far as the American foreign policy is
280
concerned can we say the same?
All right. So we're now
281
compelled to analyze the cons of our topic 'cause' in th~
282
third part we de~t with the elements likely to make us
283
discover that to sue to d certain extent America is d
284
country where democracy is to be found. NOW we're going
285
in the second time
to analyze the elements that America
286
is not really a country where democracy prevails. All
287
right.
ISubject turns the sheetJ So we'll begin bY saying
288
that
I.J
IInvestigator's intervention:
·Start a new
289
sheet-) Unlike most countries all over the world.
America
290
is a country which secret poilicy is guite a troublesome
291
and fierce one. We daresay daresay that such organization
292
as C.r.A. meaning central intelligence agency and FBI All
293
right. What was I
going to say about these organizations?
294
(.J
are assigned the duty to lead the counterpart of the
295
daylight political atmosphere.
[Silent reading] All right
296
do I have precise information about that point? I'm
297
afraid I don't have precise informations. What am I going
298
to do? [SRd] It's rather look like a mere statement all
299
my composition.
If I could for example get information
300
from newspapers or from lectures (.J
to constitute a very
301
sound basis to my composition then I would have won I.J I
302
would have written a winning composition.
But anyway I
303
can still remember an American I.) delivered a very
304
bright a very interesting lecture last time about the
305
American foreign policy.
What's his name this guy? uuuh
306
tile bloody guy what's his name? uu"io
I"forgot,.~is name.
307
What is his name this guy? Ok,j'r
it doesn't matter. I
308
must never be . .';e$perated [Si1ent reading] All right (.)
309
These secret These secret agencies mind to preserve the
310'
survival of the political regim All right
and in so
311
doing their underground activities
(Researcher's
312
intervention to signal the end of session)
313
396

---

1
SUbject 7: arvumentative. unfamiliar task (S7T4)
2
[Topic 2
"We hold these truths to be self-evident that
3
all men are created equal that they are endowed by their
4
creator with certain una ... gosh (laugh)
really tough That
5
to secure these Rights Governments are instituted among
6
men, deriving their just powers from the consent of the
7
governed. That whenever any form of Government becomes
8
destructive of these ends.
it is the Right of the People
9
to alter or to abolish it, and to institute a new
10
Government.
laying
its foundation on such principles and
11
organizing organizing its powers in such form.
as to them
12
shall seem most likely to effect their Safety and
13
Happiness. Mon dieu. phuuh. This eks excerpt is extracted
14
from the Declaration of Independence. written and revised
15
by Jefferson, and late later revised by the Continental
16
Congress that met in Philadelphia. This Declaration of
17
Independence contributed somewhat to lay the foundations
18
of the American Democracy. Write a two-page essay tsheah
19
in which you argue for or against the claim that the
20
concept of American democracy is a ~J
(.l
Prrr (.) So
21
let·s see what is he talking about then. ("We hold these
22
truths to be self-evident that all men are created equal,
23
that they are endowed what does endowed mean? Endowed or
24
is it endawed?
t.) by their creator with certain un-un-
25
nalainable Rights,
that among these are Life, Liberty]
26
Okay. Here
(.)
the text says (We hold these truths to be
27
self-evident] So the truths are the following: all men
28
are created equal.
all have un-na-alainable Rights.
that
29
among these are Life.
Liberty/ which are life.
liberty
30
and happiness (.) Okay. So the question is that everyone
31
is created equal and everyone has the right to life,
32
liberty and happiness.
OkaY.C"That to secure these rights
33
Governments Governments are instituted among men.
That
to
34
secure these rights Governments are instituted among
35
men.] Okay. So (.) the (.) Okay I.) So the aim of the
36
government is to (.) see that these rights are respected
37
to see that these rights are respected.
Okay.
So ["to
38
secure these rights (.) hummm deriving (.1 That to secure
39
these rights Governments are instituted among men (.)
40
deriving their just powers from the consent of the
41
governed-].
That means that their power come uuh no no
42
no.
It doesn't mean that It means that they have the
43
right to to
(.)
they have the right to what?
44
{-Governments are instituted among men, deriving their
45
just powers from the consent of the governed".] Okay. Hum
46
["That whenever any form of government huuh right any
397

---

47
form of government becomes destructive of these ends,
it
48
is the Right of the People to alter or to abolish it, and
49
to
institute a new Government.-]
So umm 50 e~ch
50
government secures the right to life liberty happiness
51
with the consentment of the society but when
the each
52
cime che government does not respect unun these rights umm
53
1.1 the government does not respect these rights,
the
54
governed are allowed to a to abolish it 1.1 huuh [laying
55
its foundation on such principles and organizing its
56
powers in such form,
as to them shall seem most likely to
57
effect their Safety and Happiness."]
So they can abolish
58
i t and create another government
create a government
59
that they find more adequate find more adequate 1.1 [This
60
excerpt is extracted from the Declaration of
61
Independence, written and revised by Jefferson, and later
62
revised by the Continental Congress that met in
63
Philadelphia. This Declaration of
Independence
64
contributed somewhat to lay the foundations of the
65
American Democracy. The American Democracy phuush. Write
66
a two-page essay in which you argue for or against the
67
claim that the concept of American Democracy is a ~.J
68
Oh boy this is a tough job!
[Write a two-page essay 1n
69
which you argue for or against the claim that the concept
70
of American Democracy is a myth.]
(.)
We should argue for
71
or against argue that the claim for or against the claim
72
that the concept of American Democracy is a~.
So the
73
American democracy I guess is all this So the American
74
Democracy says
dll men men are created ...
The Government
75
allows to abolish it or create a more adequate one. So I
76
am asked to argue for or against that declaration . For
77
or against.
I
think I am for it because it is true that
78
everyone is created equal
1.1 and everybodY has
79
unalainable rights. Okay So now how am I going to start?
80
I think I understand it more.
All I am asked to do is
81
1.1
to say i f I.) I agree with 1.1 If this declaration
82
made by (.) What's his name? Jeffshon Jefferson. Humm
83
{silent reading; check against problem statement,] [Write
84
a two-page essay in which you argue for or against the
85
claim that the Amer,can concept of democracy is a myth]
86
Icough, shuffles papersl.
["We hold these
truths to be
87
self-evident that all men are created equal! So 1.1 the
88
Continental
1.1 How am I going 1.1 The Continental
89
Congress no the 1.1 Philadelphian Continental Congress
90
that
1.1 revised Jefferson Declaration of Independence
91
laid the foundations of the American Democracy. I start
92
with this. Huum. No I think I'll start by Jefferson's
93
Jefferson's huum write write first Jefferson's first
94
writing huum
on the declaration of Independance
somehow
95
somewhat contributed to the foundation foundations of the
96
American Democracy. uuh [Jefferson's first writing on the
97
declaration of Independence somewhat contributed to the
98
foundations of the American Democracy]
In this
99
declaration humm Jefferson says no what did he do?-lln
100
this dec1a~ation JefEe~sonJ pointed out the fact that all
398

---

101
men are created equal and have unalienable rights.
He
102
went
[further]
to say that
huuh i f that the governments
103
no no the no no i f that the Government i f the Government
104
if they did i f they did not uum
not respect this truth
105
no they were to be abolished.
What do they mean by a
106
myth? Is it possible? Is it a dream? as self evident
107
truth,
they were to be abolished with no with
the
108
consentment of course of the people.
uum
[With the
109
consentement of the people!
with the consent
[Jefferson's
110
first writing of the declaration ... were to be abolished!
111
period.
This is not necessary.
If uuh so [This self-
112
evident truths were to be abolished]
I can't think I?)
113
kind.
Was Jefferson
I.) was Jefferson dreaming? Was
114
Jefferson seeing too far? Was Jefferson
I.) being Was
115
Jefferson going too far when writing this exsorpt? [For
116
or against the concept that the American Democracy is a
117
myth] And are those who believe
no no (crosses off and
118
are those who believe>
Uuh
[Was Jefferson going too far
119
when writing this excerpt I Giving that way the right to
120
some people to believe that the concept of American
121
Democracy is a myth? phuuh That's the introduction.
122
[Jefferson's first writing on the declaration of
123
Independance somewhat contributed to the foundations of
124
the American Democracy)
I don't think all this is
125
necessary. Jefferson's
first
writing.
I
think I
am going
126
to write this all over.
I guess is either (.)
so the
127
American Democracy says thae all men are created equal
128
all have unana1ainab1e rights which are life liberty and
129
happiness
I.) so that for the government must see that
130
these rights are respected. Jefferson's first writing on
131
the declaration [on the declaration] of Independance
132
somehow somewhat or somehow somewhat
[contributed)
133
contributed to the foundations of the foundations of the
134
~~erican demo Democracy. And I am not going to rewrite
135
all this part.
IThe subject-writer indicates with her pen
136
the part that is omitted. Comparing the first draft and
137
the final draft makes i t clear.) Was Jefferson going too
138
far when writing this excerpt no not excerpt
139
declaration? Giving uuhl.) giving therefore giving
140
therefore the right giving therefore the right to some
141
people to claim that the American Democracy is a myth?
142
Okay.
That's the introduction. And now do I
think that
143
the American Democracy is a myth? Is a myth. What do they
144
mean by is a myth? Is it possible? Is it a dream? [for or
145
against the claim that the-American Democracy is a myth]
146
I.) SO I'm supposed to say I.)
to give my opinion about
147
I.) what Jefferson said. He said that all men are created
148
equal equal
I.) all have I.) certain rights and i f the
149
government didn't respect this
I.) the government could
150
be abolished. If I agree I.)
uh
I.) certain I.) i f I
151
think it's possible
I.)
Idraws a horizontal line across
152
the page) First of all I say
I.) yes
yes. No
I.) umm a
153
myth because ot what is happening in certain countries
154
like South America
I.) which I.) even in here in Africa
399

---

155
f.)
South Africa or
(.)
no not South Africa in Africa in
156
genlUcll that all rights are
I.) humm No it isn't a myth
157
I.) No why is
I.) Wait
I.)
Yes it's a myth because it's
158
possible
for a society to abolish the government.
It
159
isn't possible because the gover~~ent I.) has I.) the ~mJ(
160
.nCl.. '.
the power I,)
er Why isn't Why is i t a myth? uum
161
I.)
I'll
talk I'll talk about dictator dictators that
162
reigned I.)
for years for years,
And I.) Okay so why it
163
isn't a myth?
I.) What examples can we take? It depends
164
on the the degree of emancipation of d country of the
165
development umm I.) of the development of a country.
166
Okay.
For example in the United States uum everyone is
167
~4give their opinion and f.) what else in certain
168
European countries in certain European countries too they
169
have the right to express to expression.
They have the
170
right
to self-expression
{re-read notes}
It is
It is
171
Icough)
un-nalainable that all men are created equal urn
172
We all have the right to I.)
life liberty and happiness
173
I.) and Jefferson I.) and Jefferson
I.) no and the
174
Continental Congress went as far as to say that the aim
175
Chat each and every government was to secure
[silent
176
reading of task}
It is undeniable that all men are
177
created equal. We all have the right to Life, Liberty and
178
happiness.
And the Continental Congress went as
far as to
179
say that the aim of each and every government must be to
180
secure these rjghts.
181
If If that government should fail in its task,
the people
182
or the government should have the right to abolish it.
183
And sew a more compromising and securing umm
one.
184
But uuh but but what but it is not we notice but it it is
185
not rare to see that certain governed are victims of-----
186
their own their own uuh their Own choices uum
uuh.
'rhey
187
soon become
[they soon become]
Objects of destructive
188
governments and find their rights completely [completelY.L
189
abolished (completely abolished).
190
In certain countries
I.) uuh dictators have for a long
191
time contributed to the suffering and lost of unalienable
192
Rights of the citizens. er an example.
In Spain, during
193
the era er of Franco in the begining in the beginning of
194
the twentieth century umm In Spain
I.),
the spanish or
195
what did the Spanish do' suffered a lot.
Umrn Instead of
196
instead of securing no of seeing to their safety and
197
happiness, Franco instead made a lot of blood spill.er Of
198
course th~ people had the right to to
alter er this
199
government
this desp despotic this despotic government
200
and what
I.) did he say? and despotic government
201
reorganize
[and]
reorganize its powers but with the army
202
being under Franco's control it seemed almost
203
impossible. er (.) In Africa too In Africa, after the
204
declaration of independance certain er certain
people
205
thought that thought that their lives would be easier and
206
that their rights might be morefmight be more)
207
considered.
But unfortunately
[unfortunately)
it was not
208
the case.
And thousands of emprisonments
[emprUlonments]
400

---

209
and killings were soon were soon what?
Were soon
210
registered [were soon registered]. Of course here again.
211
the people had the right to abolish and create a more
212
adequate government.
Like the continental congress had
213
proclaimed in Philadelphia delphia . But usually the
214
reality is somehow different.
umm The citizens of
er
215
such countries. are usually too preocupied of gaining
216
their daytime meal or simply or sometimes simply too
217
afraid to stand up and dare and dare and dare what? [too
218
afraid to stand up and dare]
speak up to alter the
219
government.
[To alter the government]
(.) What next ?
220
(Subject-writer refers to notes).
In certain countries,
221
such as European or North American it seems possible to
222
adopt this declaration. Is it because of the degree of
223
emancipation of the inhabitants? Is it because of of the
224
of the the state of of the state of development of the
225
[Is it because of the state of development] of the
226
country?
Or is it simply because of the hundred of ye9r~
227
of independance of these countries. hundred of
228
independence of these countries making them therefore
229
more mature and less
less a apt to accept certain
230
situations?
[To accept certain situations?}
urn.
231
Whatever the reason,
whatever the reason er whatever the
232
reason what whatever the reason wait
[In certain
233
countries.
such as European or North arnerican it Seems
234
possible to adopt this declaration. Is it because of the
235
degree of emancipation of the inhabitants? Is it because
236
of the state of development of the country? Or is it
237
simply because of the hundred years of indepedance of
238
these countries, making them therefore more mature and
239
less apt to accept certain situations? Whatever the
240
reason. 1 it appears that in certain parts of the world
241
the concept of American democracy d appears to be nothi.!ill.
242
else but er a an impossible no an unimpossible dX'e~JR 110
243
nothing but a myth.
[Jefferson ..... more mature]
244
401
'.

---

1
Subject 2: argumentative,
unfamiliar task
(S2T4)
2
3
4
5
6
["We hold these truths to be self-evident, all men are
'7
created equal, that they are endowed by their creator
8
with certain unalienable Rights, that among these are
9
Life, Liberty and the pursuit of Happiness, That to
10
secure these rights, Governments are instituted among
11
men, deriving their just powers from the consent of the
12
governed. That whenever any
form of Government becomes
13
destructive of these ends,
it is the Right of the People
14
to alter or to abolish it,
and to institute a new
15
Government,
laying its foundation on such principles and
16
organizing its powers in such form as to them shall seem
17
most likely to effect their Safety and Happiness." This
18
excerpt is extracted from the Declaration of
19
Independence, written and revised by Jefferson. and later
20
revised by the Continental Congress that met in
21
Philadelphia. This excerpt is extracted from the
22
declaration of Independence written and revised by
23
Jefferson. This Declaration contributed somewhat to lay
24
the foundations of the American Democracy. Write a two-
25
page essay in which you argue for or against the claim
26
that the American Democracy is a myth.]
First of all I
27
must say it'S a very nice topic tor it's dealing with the
28
concept of democracy I,J First of all what this this word
29
mean? I.J a kind of policy in which the power belongs to
30
the people the sovereign power belongs to the people and
31
this policy is specifically directed I,J is specifically
32
directed by the people and the people seems to be more
33
important than those who are ruling the country and (.J
34
and what I know also is that this kind of policy tends to
35
bring a certain kind of stability in a society [silent
36
re-reading of task) which means that it's one of the most
37
1,1 uh important policy in the world for it prevents
38
people from making troubles from protesting from going
39
against the rules which has been established and so on,
40
In conclusion it brings d certain stability in the
41
society and 1.1 I have to argue for or against the claim
42
that the concept of American democracy is a myth 1.1
43
First of all I must try to understand the word myth
44
[silent re-reading of task} What I know is that myth is
45
different from reality [silent reading of task} Like I
46
think in com con I must read the topic once more So as to
47
undertand it ["We hold this to be self-evident, all men
48
are created equal, that they are endowed by their creator
49
with certain unalienable Rights,
that among these are
50
Life, Liberty and the pursuit of Happiness, 0 0"1
well if I
402

---

51
understand this passage is
(.J
laying the stress on
the
52
equality among people 1.1 equality in right equality 'n
53
everything.
In conclusion they're created
54
equal so i t ' s dealing with equality ['That to secure
55
these rights gover~ment has are instituted among men
56
deriving their just powers from the consent of the
57
governed deriving their just powers
from the consent
of
58
the governed(.)
That whenever any government becomes
59
destructive at these ends it is the right at the people
60
to alter or to abolish it,
and to institute a new
61
government
laying its foundations in such principles and
62
organizing its powers in such form as to them shall seem
63
most likely to express their safe Satety and Happiness']
64
I.)
well i f I
understand this passage of the topic 1.1
65
the right is given to the people to go against
uh a kind
66
of poli policy which is not
gua/Tanteeing ies safety dnd
67
happiness.In other words
1.1 i f the people come across
68
1.1 a way of ruling the country which is not
1.1 uh in a
69
normal way they can protest
they have the right to
70
protest they can go against ie dnd to institutionalize
71
another kind ot government which Cdn bring d certain
72
stability
I. I
[This excerpt is extracted crom the
73
declaration of Independence written and revised by
74
Jefferson,
and later revised by the Continental Congress
75
that met in Philadelphia ... J
I. I I think I need a real
76
analysis of the American society.
In other words I must
77
try to know how the American society is governed and see
78
i f 1.1 what is being said in that topic is 1.1
I mean
1.1
79
respected 1.1 or i f what is being said is reality or 1.1
80
a myth 1.1 What's a myth and reality? [silent re-reading
81
of task} Well
first what I know is that the topic lays
82
the emphasis on equality the rights of the people,
83
equality of the people as everybody 1.1 and 1.1
the right
84
of the people 1.1 who is at the same time controlling I-I
85
the government which has been established in the society
86
and has the right to go against it i f this this the way
87
of ruling this country is not in d nonnal way_
In other
88
words I
can conclude that
the sovereign power belongs
to
89
the people and this is what democracy is 1.1 But to build
90
my my introduction I
have
first
to
(.)
to define the word
91
democracy which I know is a kind of policy in which
1.1
92
the power belongs to the people 1.1 the power belongs to
93
the people of course the sovereign power and the whole
94
country it is seems to be deem directed by the people for
95
every decision comes from him dnd what I know generally
96
speaking is that
1.1 this kind of policy tends to bring a
97
certain stability
(.J
in the country it prevents people
98
from making troubles for democracy is a kind of
(.)
d
set
99
of laws which has which have to be well respected and i f
100
it is not respected of course as the topic reveals i t is
101
the right
for the people to protest or to go against i t
102
and uninistitutionalize and institutionalize another kind
103
of policy which may bring happiness and happiness in
104
life.
So
first
I have to define the word democracy which
403
.'.

---

105
I repeated again is a kind of democray in which the
106
sovereign power or in other words people are the more
107
powerful those who are governed so to scart
the
108
introduction
(.) Democracy is a kind of policy in which
109
[in which)
the power lays in na in the people society the
110
people surroundings I mean the power belongs to the
111
people in which the Sove<eign power belongs to the
112
people,
And the people have all rights I.) and as I said
113
at the beginning this kind of policy tends to bring more
114
stability in comparison with I.) the a the other kind of.
115
policy in the world I.) and it seems to be true because,r
116
we analyze the American society today we may conclude
117
that it's I.) a certain stability its tends to be well
118
organized well ruled with I,) every right that the people
119
the people possess I.) equality freedom I.) Well let'S
120
come back to the introduction [Democracy is kind of
121
policy in which the sovereign power belongs to the
122
people) In fact,
this policy [this policy!
tends to brinq
123
a real stability in a society [a real stability), for
124
everything is well organized and simbo1ized by an
125
equality, among people of course. So [Democracy is a kind
126
of policy in which the sovereign power belong to the
127
people. In fact.
this policy
tends to bring a real
128
stability in a society Eo~ everything is well organized
129
and simbolized by an equality! and above all we come
130
across freedom I.)
liberty uh equality in rights and so
131
on [silent re-reading of task} Now what can I say? [
132
silent re-reading of task] Well now I have to argue for
133
or against the claim that the concept of a- American I.)
134
democracy is a myth. Well personally I don't think that
135
it's a myth I.)
It tends to be a certain reality. Or i f
136
we analyze as I said the American society the principles
137
which have been established gua/rantee people's safety
138
and I.) life gudTrantee a certain stability and so on. 30
139
I have to argue against that claim I.) which says that
140
the American democracy is a myth I.) I don't think so for
141
it is it this kind kind ot democra~ in America it's is d
142
myth I.) it means that it's far from a certain realism.
143
And I don't think so.
Any way if i f I.)
the principles
144
what I know is that the principles which I know is that
145
the principles are established which are established in a
146
society I.) can't be respected perfectly but the American
147
have done their best to respect what their have
148
established as principle basis of their own politics.
149
This urges me to claim that it:6not ...
myth.
My problem
150
is that I don't understand the real
meaning of myth I
151
only know that
I.) myth is different
from realism and as
152
far as I'm concerned I don't think that the American
153
democracy is a myth it's ve~ close to realism.
So I
154
have to go on this way. So I have to end the introduction
155
for
{Where everything is organized and sirnbolized by an
156
equality]
(.) now Can this I.) no wait a minute [where
157
everything is well organized and simbolized by an
158
equality) as it's evidenced in the topic.
by what~s the
404

---

159
name Jefterson
[silent re-reading of cask] Where is it?
160
[written aah and revised by Jefferson and later revised
161
by}
okay and so on. Jas it
is evidenced in the topicl
162
[silent re-reading of task]
How can
How can I
end?
Does
163
arMr.dlysis of this kind of policy be urge us to claim
164
chat
the A~erican democracy is d myth or not?
Okay. h
165
brlef
[ a brief]
analysls of the American
[of the
166
American}
(.)
society
fA brief analysis of the American
167
society]
(.Jean help us to argue for or against that
168
concept.
[Write a tWO page essay in which you argue for
169
or against the claim that
the American democracy is a
170
myth.)
And as I said I
don't
believe that this kind of
171
policy in America is a myth
for
i t ' s er very close to a
172
certain realism now I have co prove i ~.
[silent reading
173
of task]
And to do i t I
think I
have first
to analyze the
174
American society briefly to see if[silent reading of
175
task]
the concept of democracy defined in the topic is
176
r~ected in the ~~erican society if the way {silent
177
reading of task]
i t has been defined dit defined pardon
178
can be applied to the American society as a kind of
179
policy which is respecte respected
f.J
Well
the topic
180
reveals that
[silent
reading of task] all men are created
181
equal and I
think this is one of the most importanc
182
principle of American politic policies and (·endowed by
183
the creature wlth certaln unalienable Rights
(.l
that
184
among these are Life.
Liberty and the pursuit of
185
Happiness
(.l
and That to secure these rights,
186
Governments are lnstituted among men,
deriving their JUSt
187
powers
from the consent of
the governed government.
That
188
whe~ever any fo~ of Government becomes destructive of
189
these ends.
lt is the Rlght
ef the people to alter or to
190
abolish it.
and to institue a
new Government ... ·)
WeLl
191
i f I
undeystand ~his kind of policy has been created in
192
order to bring
finally a certain stability for it is
193
ruled by the people
[silent
reading of task] and as I
194
said the sovereign power belongs to [he people everywhere
195
every people have the same right and every
f.)
every pe
196
every man can say what he wants
[silent reading of task]
197
First as I said I have to analyze
f.)
the American
198
soc"ety and pe,sor.ally I'd like to th,ow I.)
the emphasis
199
on the social
f"eld
I.) and on the political
field
I.)
200
to show thac
f _J aJnerican democracy is no longer a myth
201
is r.ot a myth ~o~ no longer
r.J [silent reading of text]
202
As
far
I'm concerned
(.l
the American democracy is not a
203
myth
(.)
lv1d to start with this demonstration
(.)
what
204
can I
say?
To proceed to demonstate
f.} my opinion t.}
205
To JustiiY.-mv OPinion,
I would like to throw a certain
206
analyis on the social field and the political field.
On
207
the social
field.
the principles seem to be respected
208
(seem to be resoectedJ
so as to auarantee the stability
209
of
the coun~~lent reading of textl The word freedom
210
seems to constitute on the basis of their society.
People
211
mav under,:ake
(. I
f neople may u!1dertakeJ
everything they
212
want.
We may ta~e for examole
[~e ~dV take for example!
405

---

213
the problem of religion. There 15 a variety of religions
214
which have been established in the society [silent
215
reading of text! This freedom of undertaking may be
216
perceived on the economical field where (.) the people is
217
given the right to create so as to reinforce the national
218
economy [the national economy!.
What else can I say
219
apart from religion and freedom of undertaking? I
think I
220
can move on the polibial field (.1 The political life of
221
the American tends to be well organized [well organized].
222
On this level too (.l the principles which have been
223
established (.l seem to be maintained to bring order !.l
224
There is also [there is also t.) I a certain equality 1n
225
rights [There is also a certain equality in rights!.
226
Everyone is given the opoortunity to express his opinion.
227
[Silent reading of taskl Most of the time. people fight
228
to keep the respect of the laws which have been
229
created. [silent re-reading] This policy leads [This
230
policy leads) people (.) to keep their right and behave
231
freely at each level of the country [at each level 'of the
232
country!
(New page) The opportunity is given to the
233
governed to protest and to go against!.)
[The
234
opportunity is given to the governed to protest and go
235
against]
things which don't appear normal in the process
236
of ruling the country or the whole society [silent
237
reading of task] I
think I
have no more to say but to
238
conclude and try to see to sum up and try to see to
239
convince my reader that this kind of policy in.Jhrerica IS
240
not a myth but it's very close to reality
So lrr
241
conclusion.
this concept of democracy in America
is not
242
a myth but on the contrary it's very close to a certain
243
realism. The analysis on those fields above-mentioned
244
[The analysis above-mentionedl
prosit proves i t proves
245
it. t.) We may conclude that the americans Eight to
246
guarantee what they have established as principles [fi2!lt
247
to guarantee what they have established as principlesL
248
and this arouses a society based on a real stability
249
[silent re-reading] Their democ~(rtends to be perfected
250
perfectly respected and based on the eguality of rights,
251
freedom and so on. The whole people seems to derive their
252
a profit from it [The whole people seems to derive the;"",
253
profit from it] and appears as the real leader of the
254
country ready to bring stability if (.1 bad behaviour
255
contradict [if bad behaviour contradict the policy What
256
else can r add? [The whole people seems to derive a
257
profit from it)r mean from the . . . democracy and
258
appears as the real leader of the country.
ready to bring
259
stability if bad behaviour contradict that policy) or go
260
against what has been institutionalized I.) or
261
instituted (.1 Then this democracy can't be applied to
262
the concept [can't be applied to the concepti of myth for
263
the Americans do their best to maintain a careful
[a
264
careful] respect of the components of this so called
265
democracy [Scanning of text produced so tar]
What is
266
striking is that a great part of the political decisions
406

---

267
is put on vote so as to obtain a common consent from the
268
people In conclusion [this concept ... peoplel
In that part
269
ot the world we may without a doubt In that part of the
270
world we may without a doubt!
say that the concept of
271
democracy is jusLified (.l
[Democracy is a kind of policy
272
in which the sovereign power belongs to the people.
In
273
fact this policy tends to bring a reall~lity in a
274
society
for everything is well organized .... 1
275
407

---

1
SUbject 4:
argumentative,
familiar task
(S4T3)
2
[As result of the increasing number of unemployed
3
university graduate students,
many Ivorian parents have
4
come to question the value of higher education. These
5
disheartened parents think it is a waste of time and
6
money to send their children to University since further
7
education does not ensure employment. Instead. these
8
parents strongly feel that newly graduate secondary
9
school pupils
(secondary school teen-agers who hold a
10
baccalaureat)
should be sent to the fields where they cc.'"
11
have an on-the job-training and can be .rapidly
12
productive. Write a two-page essay in which you argue for
13
or against those parents' position. This article is to be
14 .
published in an English speaking magazine.]
(.)
All right
15
So I
read my topic once more.
[As a result of the
16
increasing number of unemployed university graduate
17
students, many Ivorian parents have come to question the
18
value of higher education.
These disheartened parents
19
think it is a waste of time and money to send their
20
children to University since further education does not
21
ensure employment.
Instead these parents strongly feel
22
that newly graduate secondary school pupils
(secondary
23
school teen-agers who hold a baccalaureat) should be s(':nt
24
to the fields where they can have an on-the-job trainin9
25
and can be rapidly productive . Write a two-page· essay,,,
26
which you argue for or against those parents' position,
27
This article is to be published in an English speaking
28
magazine.]
All right
(cough)
So what is this topic about?
29
So this topic is to
(.) This topic is meant for us to
30
critize the value of higher education.
Which use is it?
31
Which use is it? Which use is which use is it to get a
32
higher education? All right.
So this is the main question
33
that underlies all
the topic.
So I
begIn this topic with
34
a question of the following one. What use is to get a
35
higher education? What use is it? (.) All right So I
must
36
pay attention I
must not forget that my study is to be
37
laid with reference to the Ivorian realities. The Ivorian
38
realities must lead me in my topic in my composition.
39
{Re-reading of task] All right. We'll begin by an
40
introduction a very short one.
(.) All right. The Ivorian
41
government {The Ivorian government] as well aR. many other
42
countries in Africa felt concerned with [The Ivorian
43
government as well as many other countries in Africa felt
44
concerned with]
having a class of scholars.
But unlike
45
the other countries our motto should I
say our
(.) or the
46
Ivorian? The Ivorians' motto I'm Ivorian sure but it'S
408

---

47
very clear that I should write our [But unlike the other
48
countries)
the Ivorian motto or watchword or motto is
49
better watchword or motto was that school should serve Or
50
should be a tool for development
[should bel a tool for
51
development. Right so that is to say That is to say that
52
any scholar is meant to get a job so as to partake or to
53
contribute to contribute is better [tol
contribute to the
54
development.
[Silent re-reading) All right.
(The Ivorian
55
government as well as many other countries in Africa felt
56
concerned with having a class of scholars. But unlike the
57
other countries the ivorians' motto was that school
58
should be a tool for development. That is to say that any
59
scholar is meant to get a job so as to contribute to the
60
development All right. How shall I
keep on? How shall I
61
continue the introduction? (to get a
job so as to
62
contribute] So the question I
must never forget it. What
63
use is it to get a higher education? [Scanning) But when
64
the scholars fail to be given a job,
(refers to task)
65
education,
(refers to tdsk)
especially a higher one
66
education is education is not ~hought worthy is not
67
thought worthy. All right
[The Ivorian government .. is
68
not thought worthy. But when the scholars Eail to be
69
given a job,
education is.
[education]
especially a
70
higher one is not thought worthy. I All right What use is
71
it then to get a higher education
[what use is it then to
72
get a higher education]
when a
scholar or when a pupil
73
could get an on-the-job [refers to task] training and ~
74
productive?
All right that's a too long introduction.
75
[The Ivorian government as well as many other countries
76
in Africa felt concerned with having a class of scholars.
77
But unlike the other countries the Ivorians I
motto was ,._-
78
that school should be a tool for development. That is to
79
say that any scholar is meant to get a
job so as to
.
80
contribute to the development. But when the scholars fail.
81
to be given a
job,
education,
especially the higher one- ~.
82
is not thought worthy. What use is it then to get a
83
higher education when a pupil could get an on-the-job-
84
training and be rapidly productive?! All right. So [reads
85
the task)
To which extent could one prefer a
(.) a COnlllWJ!1
86
Job to a higher education? What does higher education
87
provide,
to someone (.) is very keen on studying? All
88
right
(refers to task) All right. As it is expressed in
89
the concept,
higher education means that at University
90
one is endowed with [endowed withl a teaching [with a
91
teaching] which is
[with a teaching which is] an
92
accomplished one an accomplished one [As it is expressed
93
in the concept.
higher education means that at the
94
unive~ity one is endowed with a teaching which an
95
accomplished one. 1 (.) All right. That is to say that
96
after such an education
after getting such an education
97
the student has a mind has got a mind has
got a mind
98
[has got a mind] open to many realities. All right let's
99
remember Let's remember that Sir Robert Francis Bacon
100
said that knowledge is power.
All right.
So we'll now t~
~09

---

101
to analyze which kind ot power i t is.
Someone who has d
102
knowledge what kind kind of power it is that he's got ?
103
Power to think,
the power to choose the power to decide.
104
All right Power to choose aEter analysis.
after analysis
105
after analysis power of self-determination [power to
1D6
choose aEter analysis] oE se1E-determination [power oE
1D7
se1E-determination!
All right unlike any financial power
1D8
such a power is got for ever.
All right.
That's a very
1D9
important
u4e' of a higher edycation.A11 right. We'll
110
also be about to say that A higher education CA higher
111
education] is
la higher-education isl the one got at
112
universities. The latter are places [the latters the
113
latters are places! of very important very important
114
human relationships. All right Thus
the one or the
115
seeker aEter such an education has many chances
[has
116
many chances! to go beyond the national
[national!
117
boundaries boundaries).
I.t Such chances are not awarded
118
to someone who would stop his studies [are not awarded.10
119
someone who would stop his studies] earlier. All right.
120
So we've made a kind of analysis of the use of a higher
121
education what is tokretained is that a higher educatio~
122
can provide (.) the knowledge knowledge is power as said
123
Sir Francis Bacon and knowledge higher education can also
124
provide tor someone who who's really a keen student who
125
is d keen student he can get a certain nwnber of contact
126
and have a mind open to the other realities beyond his
127
boundaries beyond his country and the realities of other
128
countries. So ~ can say that (.) All right.
A higher
129
education
provides both psychological and social benefit
DD
All right. A higher educated is self-confident is selE-
131
conEident and can afford and can aEEord many re1ationshiq<
132
[many re1ationshi!l!]
likely1ike1y[A higher educated man ;.:>
133
self-confident and can afford many relationshiP!
to
134
discover many horizons [many horizons] All right. So [~
135
higher education provides
... social benefits A higher
136
educated man is se1E confident and can aEford many
137
relationships likely to make to make him:
discover rna)!¥
138
horizons.] All right. Thus we've ana1yzed which use it is
139
for someone to get a higher education. And now we'are
140
trying to ana1yze the use it is for someone to get a job.
141
We are for example having a very high education. So we'll
142
try to (.) to introduce that idea ... part of our work (.)
143
All right.
[many horizons many horizons] Now what about
144
someone[someoneJ all right a graduate, a secondary school
145
pupil [someone] a graduate secondary school pupil,who
146
would get a job rapidly? All right.
lA higher education
147
provides ... Now what about
_ someone a graduate secondary ~"''''Dol
148
pupil who would get a job rapidly?! All right Which
149
benefits can he derive from such a job? All right. So (.)
15D
It
is quite sure[lt is quite sure it is quite sure
it is
151
quite sure]
that socially speaking this person [that this
152
person] will feel secure !will feel secure will Eeel
153
secure] because he has got a job and is self-supporting
154
is self-supporting. All right.
[is self-supporting] Many
410

---

155
economic benefits also spring from a job from a job [Many
156
economic benefits also spring from a job in our African
157
society especially the Ivorian in our Ivorian society or
158
the Ivorian society
(Ivorian society}
the average man
159
bears in mind
(In the Ivorian society the average man
160
lii\\""~ bears in mind] that the criteria of success is for
161
example [is] affording for example a big car, [is
162
affording for example a big car] a big house [In the
163
Ivorian society the average man must bear in mind that
164
the criteria of success i . affording for example a big
165
car, a big house.]
Thus someone who jobs someone who works
166
'not who jobs that's quite slangy someone who works [Thus,
167
someone who] works indeed is indeed on the right it's
168
better it sounds better indeed is on the right way is
169
indeed [works] is indeed on the right way Inew page) to
170
that kind of success to that kind of success. All right.
171
So these are the different benefits one derives
from d
172
( . )
d
worker statute d
worker appearances in the socie~y.
173
But we should not forget Nevertheless, we should not
174
forget that such on-the-job-training [such on-the-job
175
training) as thought of by the Ivorian parents by the
176
Ivorian parents are somewhat very crippling crippling ones
177
[crippling ones]
All right.
[very crippling ones
178
crippling ones]
People who go there are not meant to have
179
£ paramount level of education education. Thus
[People
180
who go there are not meant to have a paramount·,·education.
181
Thus! They may thus working they may thus working and is
182
better and is better they may stop worrying wor~ing
183
about getting highly educated or even getting a minimum
184
of education
All right. So we're now compelled to make
185
comparison between all these values all these values
186
we've learned we've discovered that someone who is hi~11y
187
educated can get
some psycholo..-gical and social benetJ~f:s
188
different from the others for example which someone I.)
189
who gets a job who have these benefits. Who both benefits
190
are ve~ different.
So we're now compelled to make d
191
comparison between these kinds of benefits and to express
192
Qu/rselves to express our point of view.
So we are going
193
to say that What we are afraid of is the sight of is the
194
man [What we are afaid of is ~he sight of] a man who has
195
got who has a big belly symbol of success [symbol of
196
success who has a big belly] and yet who is not learned
197
at all. The best sight would be that of a man who may as
198
much stuffing in his belly well this doesn't sound very
199
good English as much s~uffing in his belly as in his
200
head. It's obvious that a higher educated nowadays has
201
difficulties to get a job. But by the same time by the
202
same time with his education he can try to fend for
203
himself to fend for himself and find something to do
204
which may suit him [which may suit him! All right. Such
205
is not such is not the case Such is not the case [Such is
206
not the case] Qi a pupil who has got a job, because
207
unconsciously he has been imposed an on-the-job-training
208
and after.and afterwards a productive job. All right
I.)
411

---

209
[productive job)
All right. So we've finished making
210
comparison between the different benefits one derives
211
from
(.) a high education and a job rapidly got a job on
212
the spot. All right. So we'll now move to the conclusion.
213
We'll try to make a kind of synthesis of everything we've
214
been saying.
We'll
express our own point of view.
What
215
we'll say doing in so doing what we'll say. In so doing
216
we'll say that Compared to the status [the statusl of
217
worker with a low level of education
a high education a
218
high educated a hightducated is better high educated is
219
far [is farl
preferable. All right.
[is far preferable
220
[silent re-readingl
far preferable) Even i f the latter i f
221
the latter apparently doesn't provide many social and
222
economic benefits, at least it makes man feel
[feel]
223
secure secure [feel secure) in the world in the world.
224
(.) From his knowledge [his] ability to aaquire to
225
adquire knowledge man has towered up towered up from the
226
other beings of the universe. Conversely to the animala-b
227
for example, [conversely to the animals for example) VJbich
228
consumes without thinking [without thinkingl man has the
229
power of knowing how to create opportunities for himself.
230
All right. In other words [in other wordsl, in a way or
231
another, high education the high education the high
232
education provides the means to overcome whatever
233
difficult situation. All right. Proofreading. All right
234
So we've finished dealing with this topic. We have not
235
finished completely but we'll try now to check everything
236
we've been writing.
We may have made some mistakes of
237
writing or spelling. We may have for example forgotten to
238
write some certain words we'll try to check everything
239
before finishing the text completely.
(.) All right
240
(silent proofreading) All right [in the worldl
(Africa is
241
crossed out and replaced by in the world). to the
242
development of the country.
(of the country is added).
243
All right that was very important.
(Silent reading.
244
Investigator asks subject to read out aloud. Subjects
245
keeps on reading silently, muttering follows.
Inaudible).
246
412

---

Subject 5:
argumentative,
familiar task
(S5T3)
2
[As a
result of the increasing n'~er of unemployed
3
unIversity graduate students,
many Ivowian parents have
4
come to question the value of higher education.
These
5
disheartened parents think i t ' s a waste of time and money
6
to send their children to University since furth~r
7
education does not ensure employment.
Instead,
these
8
parents strongly feel that newly graduate secondary
9
school pupils
(secondary school teen-agers who hold a
10
baccalaureatl
should be sent to the fields where they can
I I
have an on-the-job-training and can be rapidly
12
productive. Write a two-page essay in which you argue tor
13
or against
those parents'
position.
This article is to be
14
publ~shed in an English speaking magazine.) Okay. Here
15
what are the key words or the broad idea of the topic ?
16
[As a
result of the increasing number of unemployed
17
university graduate students,
many Ivorian parents have
18
come to question the value of higher education.
These
19
disheartcu-ad parents think it is a waste of time and nloney
20
to send their children to University since further
21
education does not ensure employment.
Instead this paren
22
these parents strongly feel
that newly graduated newly
23
grad secondary pupils secondary graduates newly graduate
2~
secondary school pupils
(secondary school
teen-agers who
2S
hold a baccalaureatl
should be sent to the fields where
26
they can have an on-the-job-training and can be rapidly
27
productive.]
Okay.
what is the broad idea of this topic?
28
[should be sent. Okay.
[Write a
two-page essay in which
29
you argue for or against those parent's position. This
30
article is to be published in an Eng11sh speaking
31
magazine.]
Those parents who are arguing that high that a
32
pupil
who don't have a
job should be sent to the field.
I
33
have to argue or for or against those parents.
Okay I
34
think that I ' l l
deal with Dy topic in two parts parts.
35
First of all
I ' l l
t'V to show what can be
(reads task]
36
what can be the good of such idea.
I try to then to argue
37
for them.
Arguments arguments
for the parents.
And on the
38
second point of view from a second point of view i f they
39
are grounds to say such such a a thing we must also
40
recognize that
the higher education is also important
41
then on the second part I
t~ to see I'll
try to point
42
out
to argue then against
those parents against those
43
parents those parents [written notes generated during
44
preparatory process].
[Instead,
these parents strongly
45
feel
that
newly graduate secondary school pupils
46
(secondary school teen-agers who hold a baccalaureat)
'.

---

47
should be sent to the fields where thy can have an on-
48
the-job training and can be rapidly productive.] So send
49
children.
[Instid instid of au lieu de cela these parents
SO
strongly feel that newly graduate secondary school pupils
51
(secondary school teen-agers who hold a bacca1aureat)
52
should be sent to the fields where they can have an on-
53
the-job-training and can be rapidly productive.]
In fact
54
the parents are not wrong in saying that because it's 8
SS
common truth now in above all in the third world
56
unemploy~ment grow this unemploy~ment grow this last year
57
let's say these last decade since last as a matter of
58
fact until 1970 the problem of unemploYrl11ent was not so
59
heavy then unemploy~-ment grow this last decade and
60
moreover we can say that with education with education
61
the children the children let's say the pupil can be
62
helpful helpful for their parents for their parents let's
63
say because of because of because for instance because of
64
for instance in the agricultural field they learn about
65
the different grounds the different grounds the differ&lt
66
grounds and the way to the way to let's say to process
67
the way to process in order to better crops in order to
68
better the crops. Okay. On the second point of view (they
69
can have an on-the-the job training and can be rapidly
70
productive]
Yeah.
They can also be productive as far as
71
as far as in the health ground is concerned I mean health
72
ground is concerned that although for instance you have
73
followed literal studies although let's say health ground
74
health ground health ground. This will develop during my
75
on during the body of my work. Umm. What else? What else?
76
Unemployzment grow this last decade and this let's say
77
back up the assertment of parents.
Let's say this is an
78
evidence of this assertment.
Let's say this is an
79
evidence of this assertment.
Let's say the parents'
80
assertment umm unemploY£ment
grow this last decade
81
evidence umm the parents with their education (inaudible!
82
Yeah umm Many Ivorian parents parents many Here the topic
83
is concerned with all the world. Is it for the let's say
84
or is it
for university or many Ivorian parents. Here the
85
case lies on Africa.
Ivory coast, COte d'Ivoire the Ivory
86
Coast let's say Cote d'Ivoire I have nO choice.
Let1s ssy
87
it can be extended to the rest of Africa because these
88
third world countries are always.
It is the same gap to
89
be filled.
Let's come to the matter in hand. let's say
90
that on the health ground it's it meanS also hygienic
91
they'll try to better the hYgenic the hy~nic er the
92
hygenic the hygenic the hyg~ic ways les habitudes les
93
habitudes the hygenic habits of those parents. With
94
educational the children can be helpful let's say because
95
of in the agricultural field they learn about the
96
different ground and the way (.J In fact the parents the
97
parents are not so wrong Hygienic habits of their
98
parents.
Urrun.
What else? I
for instance what can I
teach
99
to my parents? If I happen to happen to return to the
100
village?
As I have already mentioned I could tell them
414

---

101
that for instance 1'11 show them the way how to process
102
better crop to teach them.
Also with my education I can
103
be able I can be able to tell hmm them I can be able to
104
teach them. What else? To teach them crops agri that's
105
mean hmm Okay. [Unemp10y..-rnent grow this last decade
106
evidence tor the ~rents because tor instance agriculture
107
hygenic hygenic that can lead to the hy~nic habits
108
hy9knic habits (see notes) that can lead la hausse to the
109
rise la hausse can lead to the rise hmm of life
110
expectancy. This is quite important. hmm This is quite
III
important.
And doing so the pupils are very productive
112
are very productive and doing so the pupils are very
113
productive even i f not on the spot on the long run. In
114
the long run can be quite productive. Okay.
A fourth
115
point a fourth point can it be for instance they have
116
they can be helpful to deal with I have already mentiont-cl
117
it this is not a point they have yeah as far as medecine
118
are concerned to deal with as far ds"medecine are
119
concerned. hmm A fourth point that can be mentioned a
120
fourth point that can be mentioned a fourth point that
121
can be mentioned a fourth point that can be mentioned a
122
fourth point that can be mentioned urn [Write a two-page
123
essay in which you argue for or against those parents in
124
which you argue for or against those parents.] Or the
125
topic. (read the task) Okay. Am I to choose or to deal
126
with the two? This is quite intellectual. 1'11 deal with
127
the parents arguments against the parents since I have
128
for the first part 1'11 Come to the second Okay the
129
fourth point is that the future struggle future struggle
130
the fourth point is that future struggle between
131
countries between countries even continents will be will
132
be will be based on the number of higher educated ms1 of
133
higher educated men of higher educated lIlIlD. Under these
134
circumstances under these circumstances 1'11 try to show
135
1'11 try to 1'11 try to tell them that our country our
136
country tor instance,
our country [muse not remain
137
indifferent, must not remain indifferent to this
138
struggle] to this fact to this fact,
let's put it into
139
brackets into inverted commas.
The fourth point is that.
140
And the second point is that the second point is that the
141
second point is that hmm the more a country the'more a
142
country let's say the more higher educated men a country
143
will have the more he'll be he'll be let's say honorified
144
honorified honorified, glorified to pronounce a glory on
145
somebody
will be a eulogy hmm or let's say they'll be
146
honorified ... The second point is that with higher
147
education the second point is that with higher education
148
the second point is that with higher education the second
149
point is that with higher education the level of life the
150
level the level the level of life will be will strengthen
151
will strengthen will strengthen will strengthen and let's
152
refer to let's refer to the European the European by
153
European I mean
all those white people European western
154
les occidencaux as we say in French the western
415

---

155
countries. As a matter or fact health ground
156
transplantation of heart transplantation of heart that
157
transplantation of heart energy energy or let's say means
158
of transportation means of transportation hrnm so on.
159
Third level. What time is it? I have begun nearly 20
160
minutes.
{Level of life will be strengthened. Level of
161
life will be strengthened.} What else? {Means of
162
transportation}. What else?
What else education can
163
bring as er a good thing as a good thing. Okay Here I can
164
mention that the level of life can be strengthened I CBn
165
mention for instance that the crops can be transformed
166
can be transformed can be transformed in our country in
167
our country thanks to these higher educated men educated
168
persons. Okay.
[a two-page essay in which you argue fo,
169
or against those parents' position.] positions those
170
parents those parents' position. Okay. I have 40 minutes
171
left I think I must begin with the introduction not to be
172
in order to finish my the I.) [As a result of the
173
increasing number of unemployed university graduate
174
students,
university graduate students university
175
graduate students many Ivorian parents have come to
176
question the value of higher education. These
177
disheartened parents think it is a waste of time and
178
money to send their children to University since further
179
education does not ensure unemployment.
Instead,
these
180
parents strongly feel that newly graduate secondary
181
school pupils
(secondary school teen-agers who hold a
182
bacca1aureat) should be sent to the fields where they can
183
have an on-the-job-training and can be rapidly
184
productive.
(.) on-the-fie1d)
In fact bY the way in
185
Africa parents often parents often think that sending
186
their children to school is an investment that is to say
187
they send their children to school i f ever he becomes
a
188
higher man a VIP. Let's be begin now Iread task then
189
translate bit: ne les assure pas pour l'emploi) oh the
190
problem I mention on-the-fie1d. I think this is to
191
represent the villages. Yeah th villages not the farm sur
192
les champs aux champs I.) On the fields [to the fields
193
aux champs where they can have on on-the-job-training and
194
can be rapidly productive. (.1 Write a two-page essay in
195
which you argue for or against those parents' position.]
196
Hmm okay Introduction I must go from a general idea to
197
specific. hmm Okay. I think that I must first of all
198
think of the fact that unemployment has increased even in
199
those so-called developed countries these so-called
200
developed countries. Okay Unemployment unemployment
201
unemployment. No a unemployment unem let's say let's
202
begin. This last decades
this last decades this last
203
decades this last decades of our century our century.
204
This is irrelevant this last decade this last decade of
205
er our century [this last decade of our century]
IStart a
206
new sheet) No let's say during During this last decade of
207
[During this last decade of)
Is it necessary to put our
208
century? Okay let's put it. I'll change it later.
[During
416

---

209
this last decade of our century] unemployment has grown
210
[has] been considered as a plague no not as a plague ~
211
one of the greatest problem in the world in the world
212
particularly particularly in the
so-called why let's say
213
simply [in the] Third.'lorld countries [in the Third -
214
World countries. In the thidl:o.World countries] where the
215
number of [where the number of where the number of where
216
the number of]
er s there any thing? let's check a
217
synonym of unemp1qyment where unemployment where the
218
employed the unemployed persons [where the unemployed
219
person] are more and more numerous Okay fare more and
220
more numerous are more and more numerous]. This let's say
221
this fact is even extended [this fact is even extended]
222
to the higher educated person let's say men to avoid
223
repetition [higher educated] ~ Hmm Consequently,
224
[Consequently Consequently Consequently] parents are have
225
become [parents) let's say [parents] who send their
226
children to school [school who send their children to
227
~chool] have become anxious [become anxious] .lmJ1 ask and
228
let's say and think that [and think that)
[newly graduate
229
secondary school pupils (secondary school teen-agers who
230
hold a baccalaureatl should be sent to the fields where
231
they can have an on-the-job training and can be rapidly
232
productive.]
[and think that] the pupils who don't have
233
jobs should be led [should be led! I must [should be ledl
234
to the fields. [Consequently, parents who send their
235
children to school have become anxious and think that the
236
pupils who don't have jobs should be led to the fields.
237
Should be led to the fields. Should be led to the field;
238
Should be led to the fields.
I must try to insert the
239
fact that this is also in C te d'Ivoire or Ivory Coast I
240
have no choice in Cote d'Ivoire.
And Ivorian parents.
241
This is also what Ivorian parents think. To what extent
242
[To what extent To what extent] Dans que11e mesure [To
243
what extent To what extent) have this idea have this
244
idea [To what extent have No [To what extent have) those
245
parents are those parents right? Or to what extent is
246
this idea. To what extent is this idea. This idea uh To
247
er. To what extent are those parents. Let's come to the
248
topic now.
Let's come to the development now.
Host of the
249
as at what time is it? Before going urn At the beginning
250
[At the beginning) we must mention that
[we must mention
251
that) generally in Africa [generally in Africa) parents
252
always
(parents always]
expect
[always expect always
253
expect always expect]
something [always expect something]
254
from their children [parents always expect something from
255
their children from their children) they send to school
256
To be precise [to be precise] they firmly believe that
257
[they firmly believe that) those children will become
258
someday a VIP. VIP is academic? Let's put it into
259
inverted commas.
and provide them with money_ But when
260
the contrary occur or these children can't have jobs they
261
are depressed and want them [and want them) to come back
262
to the villge
[and want them to come back to the village]
417

---

263
where they think they can be helpful. As a matter of
264
fact. This assertment [this assertment this assertroentJ
265
has some. This.Hmm Let's say As a matter of fact[as ~
266
matter of fact! there are good ground [there are good
267
grOund there are good ground there are good ground I ~
268
assert that. First of all [First of alII Ob I think that.
269
Oh only twenty-five minutes left.
(First of alll thank~
270
to their higher Let's say let's put it between brackets.
271
[ thanks to their higher! education the pupils [the
272
pupils the pupils! can be helpful as far as agricultur~
273
is concerned [as far as agriculture is concernedl. Indeed
274
since [indeed sincel they have learnt about [about I the
275
different grounds let's say the different the differenk
276
grounds (even if they are specializedl specialised with s
277
or z:? where they can know how the can know how to process
278
in order to better the crops. This indeed This help is
279
quite important let's say useful in that almost [almost)
280
all the countries of [almost all the countries of! those
281
the 'fhiri
World have their economics [all the Thiran
282
Weld countries have their economicsl based on
283
agriculture. [all the Thirg world countries have their
284
economics based on agriculture. Agriculture) Through
285
FurthermoreFurthermore Furthermore Furthermorelpupils can
286
be[can be]helpful[helpfull in the health ground[in the
287
health ground] in that they can urge their parents
288
s'habituer to adopt to adopt more hygenic habits[more
289
hygenic habits more hygenic habits more hygenic habitsJ
290
This fact [This fact this factI er [this factI obviously
291
allow(This fact obviously allow) the rise of the life
292
expec expectancy(life expectancy) On tbe second point of
293
view on de plus In addition if parents[if parents) happen
294
to deal with er admlnistativeradministrative
295
administrative administrative) formalitls administratives
296
[administrative] matters their children can be useful . .In
297
fact [In factl the parents in fact the parents in fact "e
298
can we may argue [In factI we may argue with those
299
parents to assert
[we may argue with those parents to
300
assertl that the pupils who [the pupils who) don't get a
301
job after [don't get a job after) their studies
302
should(should be sent) be sent to the field [should be
303
sent to the fields]. Under those circumstances education
304
is seen as something
[as somethinglun-necessa~
305
unnecessary.
But we may wonder whether education is not
306
also [whether education is not alsoJ has not also some
307
other goals. In other words, is education only useful to
308
employment? lonly useful to unemployment?l It's a common
309
saying that [It's a common saying that! that the future
310
struggle will be laid on intellectuals. Thus it is
311
important [it's important I to have higher educated men to
312
meet this these standards [these standards these
313
standards these standards]
Then sending back educated to
314
the fields may make them lose their abilities. Now our
315
country mustn't be indifferent to this reality.
(Subject
316
glances at notes) Moreover
[moreover]
higher educated
418

---

•
317
person can allow to strengthen life level [to strengthen
318
life level! To be convinced let's refer to the western
319
countries (refers to notes) [to the western countries)
320
where higher educated men even though they don't have
321
jobs reallze goods things, I mean the researches.
(.) ~
322
the last resort we can't be fully
323
419

---

Appendix XIV: Excerpts from marginal comments by
holistic markers
P.
421-422: Task 2, produced by subject 6
P.
423-424: Task 1, produced by subject 6
P.
425-426: Task 2, produced by subject 8
Task
1:
referential,
familiar discourse
Write a
two-page article
in which you explain the role of
television as an educational
tool.
This article
is to be published
in the Nigerian
newspaper West
Africa
Task 2:
referential.
unfamiliar discourse
In secondary school,
you may have been exposed to a variety of
English
languages.
At the University,
your
initiation to Theoretical
Linguistics may have given you new insights
into the analysis of
different
linguistic systems.
Write a
two-page text
in which you attempt to explain the
differences and
similarities between British English
and American
English.
This
article
is to be published in the English Club Magazine.
420

---

Since the six form at grammar school
we have been
learning English
/
but we didn't know wlldt kind of
English It was.
But oow at
the Unl.
verslty we
ar~ trying to make different between American and BritiSh
---- .- --
English.
Alld
In order Lo help us
In our studies we hove som~ dlsclplln
like phonetics and
I inguistlcS. In
linguistics/or instance the way of
r
reading a word,
the sound/~!p us to know If
I t ' .
lJ,rltlsh or America
~
~
But before trying to
search out the dlfferenc~s which exist between
-~"
""'-----
that two ways of speaking
let's see their background ant It's conse·
Quences.
British for
instance will
always remember the_dat.e_l0~6 because
It is et that'moment that Wllliam the conQueror conQuered England.
He -w-:;~ fr'o~ No~~'~ndje and so spoke FreJch. Afterthey'trled to~pply
their
language to the countries that th~y had conquered. So that gave
'~'(L;"'" birth to different way of speaking in, the-now" Great BrIL~in. That
-.......----- --
."
--
' -
~
why they have got,
sometimes, words which resen,ble !.rench.
They
lcl':- had been colonized by someone coming fromf ~.ek..
Americans blunt way of speaking Is not really their fault.
Becau
?
se AmerIca
Is a big town
in which you Can encounter people of everywhe
L {;.. ~ Indians, !V"itish, African and so o<v.~ all those people t;;~'-
c..
? " ,
~~ ..L-t..-Jc,....<!,.llA
~ ~
)",'J;;.r.- speak English,L...-But lt IS -,,0(" an
~cademic one. And American are oblt-
I
ged to search a language,alw?vs.Fnali.<~ but a language everyone of those
enumerated above can ~nderstand.-'
--- _.
- •..
C!hat
is why
in order to say good morning to someone they won't repeat
,jr';'; that Quite long sentence but they will just say"Hi". And also when
:t......A~_the~ enter a shop for instan~e they have not enough time to say
.r~'. politely to the grocer that hiS goods are elft:ensive. They tell hUll";O in or
• ! " .
""y
that prople fird 111lfXlUte, rude.
~.
In that case
briti5h are mo:e p01ite and ~ll bred. But I think Ame-
"r"">< ricans way of speaking not depend~ on themselves: They are always
__ r::
runing after scientific developl1'~nt so they J~..~_\\!'..e_not enough time to
4'~~
~-
,
pay attention to polite ways.
After that 5call
anal{;sis/
let try to see the words that depend
I
on their cultural
background.
American for
instance call the house
.f\\'-C \\'~jJ.ere people live a "flat" and the British call the ~ame thing an
n~.I' "_>Jr:'<'rtrrent"_ ~nd whi'Jt. i~ ~ur~ is th;lt they have qat dilferent way of feeding them-
l&>V~lves.AmeriCCln.5for instance eat thiflCl.s that British hardly know or don't like .5imrl
c''''~
While
learning the.. la~.9u.~. to come.J>a..c~ to our studies, we. Can
-.1, .• 1'-<..•
notice that our teache~61ave different ."'~ of speak.ing which ~!.~~_.
',,_. r~ .."
':If" \\.;r,.-.r('
thr>v t:l,-r..~turll':'C tl-;("'mseJ\\.·C'.c;. Thrlt'~ to ~':>\\, Ih~ ..,..,Jr of
sr'?.::lkina 5~lime.s
i:lrv:t '.,:,ry oft"!, r1('!""Cry]s en ....11(\\r'" \\,:e
li,:e, der:end<;;. m the r!;'octi.tion et that thina.
I· ....... "
' f
'n'r ,"",.., ;>"..-1 f'ri I i ~h
h."v<:> (Jot rli ((ernnt \\,;;;lV r-f .r.r....,~kina 1t. is not
!,. . . I;.·y' ',.lli,:I,
J~ -:-':'rl(-C:'"11'·-~. ·PIPr,'" Cl- n
f'~~
sl:€Cifi
.... \\:0r-:j~
like t!':;~e oi
~l.. ,.
,1,; ..'. --;'-':-.' ~,r t'I,~ ....·.l~turrl]
~1·-:k-
421
'.

---

ground. Apart from that the language It the same. An Amr,rlcan won't
feel forel9nl In Great Britain because he can't understand what people
say. Nor will a British In America. That's to say that If there are
some differences between th<>ir way of speaking. their are also many siml-
L.·ot':"larlties. And these latest are the more Important.
'l1~ .. t:J:JJ<- U
Iv ••1. tL.~ ~
;~
Let's precise before completed tn~t essay that In'bOth counfries .
. It Is the language "English" which Is spoken. Nonetheless it is not
~
(,.:zt:....-~-to know that this remarks have been done with the helD of dlSCl-<~
rt""" pllnes like linquisticSand phonetic5'whlch try to unveiled the difficul-
ties of a language to his
students. But even If differences are slight.
theY,re
even Important for the student who must know the kind of
~.v:: English he speak. And so these differences make his studies a little
bit difficult. But let's
say they can't be at the basis of someone
failing in an exam because it Is the language. &> everything considered
these differences which make
the study of English difficult are not
so important because British or American, all is English and one of these country
can be understood by his neigbour. But even if we are able to speak every kind of
r:na.l ish it is too qood be able to make the difference between the
two ways of speaki~
As a conclusion we can say that American and British speak English
but disciPlines
like phonetic and linguistic for
having studies the
difficulties the language
have discovered tha t there
are some diffe-
rences in their way of speaking. And that
differences whatever their
type must be considered to keep the originality of the language in the
1 country which concerned. American for instance while apologizing for
their rude way of speaking, also ask
people to accept them
as they are
because
they are convince that they can't change that phenomen.
422

---

--
G...-~'"
)
Television
is-'Hle-ef--itle
invention! which show,th,t man-lnlnd
Is developed.
Qr--"1an after0-J"har/.;;J:k .t'~nothlng to do than sleep or...S'~ 5~~~'
/
ffieoPle
in t~eir sclentlflc research find It suitable to establish some-
'""
thing which will help
man to spend his
time off without boring himself.
,J
7
h.I
,<-,S
these
<ci~..u.fi..c5found machin ... tJLJlrjlje_CJfilm~. On~those means
\\P i~But if TV first reason of being is to distract people. it IS
~als~ true that after years people found that it ~ a good mean to reach
)!.ll---!~':.-w_~~I~_c~w_d~. Therefore authorities of countries take
TV as their
mean to tell
their ideas to their people.
7
In developing countries there ar~!!!l~. So
~o Television is not the problem of the government alone. Many people are
intere/sed
in TV production. But how and what do they do with TV. We saw
'1.lntrodu_-
that TV was a mean of distraction but after years people try
..---- - -
-----
1
l,-
to give their 4-~.:.~ideas, their proposalOby TV. And the see it is a good
.."e-idea because
at the same time many and many
people can te:.eard the news.
After that experiences, people beceme more Interested
in lV.
It
is useful to watch TV because
it is the only way of seeing
the president
or it
is the only way of knowing things which concern other countries. But
apart from news tell
to whole country by the help of TV, we can remark the
Irole of TV in politlcal
fights, war.
At that moments,
TV is used to tell
to people what~they have to ~up-~
I.nort. And also during
legislative
"fights" Polltlcian"'use TV to teach
"-')
U
/___
people
how to judge a politician and
~-". to choose t~_~_est_of._them.,n<;}~f-'f
one by one they give arguments to convince their public. But by the way whb .
is concerned by TV.
I~ Developing and free-countries, everybody can havr
a TV
if he has got the mean "'If
buying
i t . AM
one is free to follow the erni~-.:;H'll
he
w~nY:·. So TV in th~t count~ies TV teach the oth@rs. That is to say how to live well.
'But let':i precise that doing so TV ~ti 11 being Cl great mean of distraction. What about
undevelopment countries ?
In
Ivory Coast the arrival
of TV was
in the
ears 1960 with the
in-
dependance.
First TV was
consiJered as as a\\d-i-s"traction
ean-s r{-It was a way
of making people forget their troubles and their
-t-i-1'e1l1ent- of job. Buti" the
years
1965 to
19"0,
people thought
that TV would be a good mean to teach
the kind of development they want to the
small
children.
Therefore TV
was introduced in educational
system.
But in primary school.
I...tJ...g~~our~
at which people looked some events and try to make their mind on
it. That
;-~;thodwas at that approve by everybody but after people real ized that the
~ Jellel of chi ldren was becoming ptorsL and worst. So what to do ?
·,They realized Lrlere 'Was nothing else to do than ClE--~!'"'-ed TV from the educa-
-'
tlonal
system. Just because
instead of being a tOG.!
as the wanted. TV :Jar~
liclped
in l-iC'r.-.:'""If1? their
srSLe.'i:1.
GuL that
IS only a~_l?rce of the rrobl'2m.
le\\ '55'?':'
1V in thiS cl1 unt!-y· all-.'ays tlut
In other occasion TV after hiS fal-
423

---

, -
in educational system became a mean of transmitting news to people.
~e,erybodY who ha'e got something to say (which concern
all the coun-
~ come to TV and say it. And also politicians ta<e TV as their
fight mean. They try to do kin" of "lu.,. ~.=c.~",,'
~'!.-.~~d
~.~.,o." TV, promising earth and he.,en to poor people and making·
speeches they will ne'er apply to the realltles.
That .Iso Is a king of education of the -mole country. 8 ecause TV Is the
only thing apart from Radio which
permit u. to ha'. news from other
countries. So telltngJO people that in other countries war Is going
on and also the lec< of good is killing people Is at the same time a way
of a5<lng the then to consider their own stituatlon as the best. And
to forget their many problems and not pay attention to their country j
polltic~ [,erything considered when decisions will be ta<en, they will
be related to them by TV. So TV in those countries. educate people for
the ad,antage of another one. The ad,antage that TV
gi,es in those cuun-
tries are not too much just a moment of distraction or a way of learning
how to coo< some foods.
And It gl'es emissions to small people which Is to praise because chil-
dren have to be considered. We have for instance IITantie Lea ll
in
Cote d' I,oire which permit the union of many different children. But was
TV rea 1y successfu I in his wor< of educator?
I think that
It a Question which must be asked to specialists,
in order to point out ad,antages we retire from TV and the drawbacks
also. But
'hat [ can say concerning [,ory Coast is that~politi­
cians very much to maintain the peace in the country. Because, today
if something happen and people stike . if the watch TV and heard their
president telling, teaching them the position of the country, praisillg
the best men of the. country, tell ing them how it wi 11 ashamful to see
the only country people find peaceful in war
or in fight they will
stop their slike.
424

---

__ I{', VTTl-l
;;;1.."-
r-1-0- 1,1 secondary SChooll we have e<posed tO~j.of ","gl ish languages.
The purpo·. of teaching
this varl&>Y ~ngllsh language (0 the pupils In
,~ . I
17-
~
·cA-I)-..'secondary
Is ~ tl""" mM0r the different ~ of tile Iwguages and
falso to make tllem kno~ tIle language. rIle aim is
not to make tIle
J,.,~puPlls t;;;;-ow the language very well, but to make them know the lan-
,
; guage a little bit. Everything
is done at the same
time,
it means
~ that the teachers do not make the difference between the American
English and the British ~ngllsh. So the pupils used to learn the engllsh
~;,,::. language without knowing Its oriolne. It mellns that the pupils learns
J;:~ the words
without knowing
if
they are
American's or Brit1sh l s. So we
can say that
in secondary school
the pupils study the languages as
the
~_~ teachers tell them to do, use the wordS without knowing their e~act
~ meaning, because you can find a word In the American English which does
.fJ,.~ not mean the same thing or does not refer to the same 35 in British.
~. So they use to make sentences which sometimes make nO sense .
.'J;:-,n.-.'
At the univerSity/ I inguistics is made to make p~-..Q.ple understan~ the
procedure
of the language,
it means
hoW'
the language itself i~_~_:
the evolutIon WhlCh has
occured In the language and see ha ..
to use
words
In a sentence. We can
ask
thiS
Question: ~
) ;'1::ole oL.li.ngulStic~[..I1::.1,Loo tt..- .... _:1';'_ ·-·~·1-tL:. y~)
/ ~
-v
7r;~~
",-,_ 17: X
FIrst of all/ we can say
that !.lnglCfwc:..:ls a sClence..'li.u:h_all..9,w people
:.t~ ~~_~~e gevelopment of their o.. n languag~~~!.!!L~. ~_~_Le}l~~ studi,es
z1....~."'" humaio_ It is why it is used to make the di!f"~r~~nce bet .. een the British
language and ~~mer~all englIsh.
We know that the two languages are
_
~
.::;-
y-t.....~
~gl ishe-tut there_iS_ Ifference bet .. een them. The American
was colo-
nized by the Brit1.6h they gave tllelr~to the americans. If the AmerIcan
L
J
5
..-In,..(
decid~ to
baptize their
language as "American u
instead of OI~glish".
?
it
is because
they want to
sho .. their independance
to prove the Bri-
.~~)C tish that they ~~~.i.r._own_mastersand so on-tYl.!.!-ack english also
constitute a difference between the two
languages, because the black
._tc.....::. American used to
create
words which don1t ell;ist in British language,
the way of writing the words is sometimes differ~nt. So lin9ui~-.!tf can
allow the students to understand or to get the difference between the
~,F~O languages in making some analysis between them. Ellen if we say that
~
there
IS a difference between the two
languages, ..e can also say that
'-(
there are simi larities.
As
I said.
the American was colonized by the British, of course they
gdlle them tl,elr
\\anguag~. If tile British gave the Amerlcan their langua-
ge.
it was Q~ssarly tIle same thing they are going to gIve to
them and
teach
th'?m how to use
~h'? "Piords. So
if far
instance we took
2 books
on'?
..... rJt~'?n
t':
<'ill
,,\\:l:erl":a!l and
th'? otller ..... ritten by a British ...."e are
~~"j"""1 '.' ~l~':'
l"!l~~:~ tnr;
<;.jr-::;-:
~:)ol<; In t1l'? t ..... Q [)":'('':5,
If
the Ar.lerlcan
425
'.

---

.../~
is a black. man,
we are going to have some words from ~-e-.k--A-tnerlctln
,.
l~age·.--trmean, black engllsll. OuL we arc yolllg La ulldersLallu olle
book as well 4S we understand the other. '}<>""UJ
v3«)<. So we can say that ther{()dlfference between British engllsh and the
American one. But the differences are not very Impo;(ant for we can
understand the language.
'Even if there are differences between/ them~'-lWt-the grammar Is the same/
/1'~IS not American grammar and British grammar. They have the same'
y: ri,Ves. ThE" way to make the . 5ente~ce5 i~ the sam~•. Gramm~r is the same
hinq
in Enqlish lanquage which
llves
the
two
clvllisatlons.
And
this problem of differencE" between the two
language
is
a problE"m of ~
.:L-t.',- I....- ~
(
mentality for
the
two people speak
the same
language.
And
the~~mer!ca~
do not want
people to confuse
them with
the British and
to show
the
British
that they
are
no longer under
their control,
for
they are
free.
l cl, "'"
. /
, \\:tY e-:v-t
~
~~
~~t:~
r
",,-1:..,./4 ~-eJ J.. ~c-L .. Lt i--o..r:&., .
o.L-., ~JLt ,rr~ f~i:, «..<'1-
r;~(! /. -I- ~
IV"
~ ~ ~ ~~f ~ ~
426

---

---

BIBLIOGRAPHY
Abrams, M. 1953.
The mirror and the lamp.
New York:
Oxford University Press.
Alderson,
J.C.,
K.J. Krahnke and C.W.
Stansfie1d 1987.
Reviews of English language proficiency tests.
Washington,
D.C.: Teachers of English to Speakers of
other languages.
Allport,
D.A. 1985.
Distributed memory,
modular
subsystems and dysphasia.
In Newman and Epstein
(eds.)
1985.
Anderson,
J.R.
1980.
Cognitive psychology and its
implications.
San Francisco: Freeman and Company.
Anderson,
R.C. 1977.
The notion of schemata and the
educational enterprise: General discussion of the
conference.
In R.C. Anderson,
R.J. Spiro,
& W.E.
Montague (eds.)
1977.
Anderson,
R.C.,
R.J. Spiro and W.E. Montague
(eds.)
1977.
Schooling and the acquisition of knowledge.
Hillsdale,
N.J.: Lawrence Erlbaum Associates.
Applebee,
A. N. 1982.
Writing and learning in school
settings.
In M. Nystrand (ed.)
1982.
Arapoff,
N. 1967.
Writing: a thinking process.
TBSOG
Quarterly 1: 33-39.
Arndt, V.
1987.
Six writers In search of texts: a
protocol-based study of Ll and L2 writing.
ELT
Journal 41: 257-267.
Atkinson,
P.
1979.
Research design In ethnographic
research.
In Research methods in education and the
social sciences.
Part 5 Block 3 DE 303.
Milton
Keynes: Open university.
Augustine,
D. 1981.
Geometries and words:
linguistics
and philosophy: a model of the composing process.
College English 55: 221-231.
Baddeley,
A.D. 1976.
The psychology of memory.
London:
Harper and Row.
Bakan,
D.
1954.
A reconsideration of the problem of
introspection.
Psychological Bulletin 51: 105-118.
427

---

Bartlett,
E.J.
1981.
Learning to write: some cognitive
and linguistic components.
Linguistics and
literacy series:
2.
Papers in Applied Linguistics.
1982.
Learning to revise:
some component processes.
In Nystrand
(ed.)
1982.
Bartlett,
F.C.
1932.
Remembering.
Cambridge University
Press.
Beach,
R. 1976.
Self-evaluation strategies of extensive
revisers and non-revisers.
College Composition and
Communication 27:
160-164.
Beach,
R. and L.S.
Bridwell 1984.
New directions in
composition research.
New York: The Guildford
Press.
Beattie,
G.
1983.
Talk:
an analysis of speech and non-
verbal behaviour in conversation.
Milton Keynes:
Open University Press.
Bereiter, C. 1980.
Development in writing.
In L.W.
Gregg and E.R. Steinberg
(eds.)
1980.
Bereiter, C. and M.
Bird. 1985.
Use of thinking aloud in
identification and teaching of reading comprehension
strategies.
Cognition and Instruction 2:
131-156.
Bereiter,
C. and M.
Scardamalia 1981.
How Children cope
with the cognitive demands of writing.
In C.
F.
Fredericksen and J.F.
Dominic
(eds.)
1981.
1983.
Does learning to write have to be so
difficult?
In A Freedman,
1.
Pringle and J. Yalden
(eds.)
1983.
1987.
The psychology of written language.
Hillsdale,
New Jersey: Lawrence Erlbaum.
Bernardt,
E.B. 1986c.
Proficient texts or proficient
readers?
ADFL Bulletin 18: 25-28.
Berlin, J. 1984.
Writing instruction in nineteenth
century american colleges.
Carbondale: Southern
Illinois University Press.
1988.
Rhetoric and ideology In the writing class.
College English 50:
477-494.
Bialystok, E.
1990.
Communication strategies: A
psychological analysis of second-language use.
Oxford:
Basil Blackwell.
428

---

Bizzell,
P. 1982.
Cognition,
convention and certainty:
what we need to know about writing.
Pretext 3:
213-243.
Blass,
T. and A.W. Siegman 1975.
A psycholinguistic
comparison of speech, dictation and writing.
Language and Speech 18: 20-34.
Block,
B.
1946.
Studies in colloquial Japanese.
11:
syntax.
In Language 22:
200-248.
Block,
E.
1986.
The comprehension strategies of second
language readers.
TESOL Quarterly 20:
463-491.
Bloomfield,
L.
1933.
Language.
New York.
Bobrow,
D.G. and A. M. Colin
(eds.)
1975.
Representation
and understanding:
studies in cognitive science.
London: Academic Press.
Boomer,
D.S.
1965.
Hesitation and grammatical encoding.
Language and Speech 8:148-158.
1970.
Review of F. Goldman-Eisler,
Psycholinguistics:
Experiments in spontaneous
speech.
Lingua 25: 152-164.
Boomer,
D.S. and A.T. Ditmann 1962.
Hesitation pauses
and juncture pauses ln speech.
Language and Speech
5: 215-220.
Bourne,
L.E.Jr.,
Ekstand,
B.R. and R.L. Dominowski,
1971.
The psychology of thinking.
Englewood Cliffs,
N.J.:
Prentice-Ha11,
Inc.
Bracewell,
R.J. 1980.
Writing: a cognitive activity.
Visible Language 14: 400-422.
Braddock,
R.R. L1oyd-Jones,
and L.
Schoer 1963.
Research
in written composition.
Urbana,
Illinois: National
Council for the Teaching of English.
Bransford,
J.D.,
Barclay, J.R. and J.J.
Franks 1972.
Sentence memory: a constructive versus interpretive
approach.
Cognitive Psychology 2:
331-350.
Bridwell,
L.
S.
1980.
Revising strategies in twelfth
grade students'
transactional writing.
Research
the Teaching of English 14:
197-222.
Bridgeman,
B. and S. Carlson 1984.
Survey of academic
writing tasks.
Written Communication 1:
247-280.
Britton, J. 1978.
The composing processes and the
functions of writing.
In C.R. Cooper and L. Ode11
(eds.)
1978.
429

---

1983.
Shaping at the point of utterance.
In A.
Freedman,
I.
Pring1e and J.
Yalden
(eds.)
1983.
Britton, J.,
T.
Burgess,
N. Martin,
A. McLeod,
and H.
Rasen,
1975.
The development of writing abilities
11-18.
London and Basingstoke: Macmil1an Education
Ltd.
Brookfield,
S.D.
1987.
Developing critical thinkers:
challenging adults to explore alternative ways of
thinking and acting.
Milton Keynes: Open University
Press.
Brown,
E. and Miron,
M.S.
1971.
Lexical and syntactic
predictors of the distribution of pause time in
reading.
Journal of Verbal Learning and Verbal
Behaviour 10:
658-667.
Bruffee,
K.A.
1984.
Collaborative learning and the
conversion of mankind.
College English 46:
635-652.
Bruner, J.S.
1960.
The process of education.
Cambridge,
Mass.: Harvard University Press.
1962.
On Knowing:
essays for the left hand.
Cambridge,
Mass.:
Harvard university Press.
1967.
Towards a theory of instruction.
Cambridge,
Mass. : The Belknap press of Harvard University
Press.
1974.
The relevance of education.
Harmondsworth:
Penguin.
Bruner J.S. and J.M. Anglin 1974.
Beyond the information
given:
studies in the psychology of knowing.
London: A1len and Unwin.
Bruner, J.S.,
J.J Goodnow and G. A. Austin 1956.
A study
of thinking.
New York: Wiley and Sons,
Inc.
Burt,
M.,
H. Dulay and M. Finochiaro (eds.)
1981.
Viewpoints on English as a second language.
New
York: Regents.
Butterworth,
B.
1975.
Hesitation and semantic planning
In speech.
In Journal of Psycholinguistic Research
4: 75-87.
1980.
Some constraints on models of language
production.
In B.
Butterworth (ed.)
1980.
(ed.)
1980.
Language production,
volume I:
speech
and talk.
London: Academic Press.
430

---

(ed.)
1983.
Language production,
volume 11:
Development,
writing and other language processes.
London: Academic Press.
Butterworth,
Band G. Beattie 1978.
Gesture and silence
as indicators of planning in speech.
In R. Campbell
and G. T. Smith (eds.)
1978.
Byrne,
R.
1977.
Planning meals: problem-solving on a
real data-base.
Cognition 5:
287-332.
Campbell,
C.
1990.
Writing with others' words: uSlng
background reading text in academic compositions.
In B.
Kroll
(ed.)
1990.
Campbell,
C., V.
Flashner, T. Hudson,
and J.
Lukin
(eds.)
1982.
Proceedings of the Los Angeles Second
Language Research Forum Volume 11.
Los Angeles:
University of California at Los Angeles.
Campbell,
R.
and G.T.
Smith (eds.)
1978.
Recent advances
in the psychology of language:
formal
and
experimental approaches.
New York:
Plenum.
Carlman,
N. 1986.
Topic differences on writing tests:
How much do they matter?
English Quarterly 19: 39-
49.
Carlson,
B.
1988.
Cultural differences in writing and
reasoning skills.
In Alan C.
Purves
(Ed.)
1988.
Carrell,
P. L.
19 83a.
Some issues in the role of
schemata, or background knowledge,
in second
language comprehension.
Reading in a Foreign
Language 1: 81-92.
1983b.
Three components of background knowledge in
reading comprehension.
Language Learning 33: 183-
207.
1984a.
Evidence of a formal schema in second
language comprehension.
Language Learning 34:
87-
112.
1989a.
Metacognitive awareness and second
language reading.
Modern Language Journal 73: 121-
134.
Carrell,
P.L. Pharis,
B.G. and Liberto,
J.C. 1989.
Metacognitive strategy training for ESL reading.
TESOL Quarterly 23: 647-674.
Carrell,
P.,
J.
Devine and D. Eskey
(eds.)
1988.
Interactive approaches to second language reading.
New York: Cambridge University Press.
431

---

Carson,
J.E. 1990.
Reading-writing connections:
toward a
description for second language learners.
In B.
Kroll
(ed.)
1990.
Carson,
J.E.,
Carrell,
P.L.,
Silberstein,
B. Kroll and
P.A. Kuehn 1990.
Reading-Writing relationships in
first and second language.
TESOL Quarterly 24:
245-
266.
Casanave, C.P.
1988.
Comprehension monitoring in ESL
reading: a neglected essential.
TESOL Quarterly 22:
283-302.
Chase, C.I. 1968.
The impact of some obvious variables
on essay test scores.
Journal of educational
measurement 5:
315-319.
Chelala,
S.I.
1981.
The composing process of two
Spanish-speakers and the coherence of their texts: A
case study.
International doctoral dissertation
abstracts 42/12 5045-A.
Cheskey,
J. A.
1984.
The effects of prior knowledge and
audience on writing.
Dissertation Abstracts
International 45: 2740A.
University Microfilms No.
DA 8428407.
1987.
The effects of prior knowledge and audience
on high school students' writing.
Journal of
Educational Research 80:
304-313.
Chosson,
J.F.
1975.
Entrainement mental.
Paris:
Editions du Seuil.
Claparede,
E.
1933.
,
1
"
La genese de
'hypothese: etude
experimentale.
Archives de psychologie 24: 1-155.
Coffman,
W.E.
1971a.
Essay examinations.
In R.
Thorndike
(ed.)
1971.
Coh8n,
A.D.
1987.
Using verbal reports in research
language learning.
In Faerch, C.
and G.
Kasper
(eds.)
1987.
Cohen,
A.D. and C. Hosenfeld 1981.
Some uses of
mentalistic data in second language research.
Language Learning 31: 285-313.
Co1e, M. and S. Scribner 1974.
Culture and thought: a
psychological introduction.
New York: Wiley.
Coles, W. Jr. 1969.
Freshman composition:
the circle of
unbelief.
College English 31:
134-42.
432

---

Collins,
A. and Gentner,
D.
1980.
A framework for a
cognitive theory of writing.
In L.W. Gregg and E.R.
Steinberg
(eds.)
1980.
Connor,
U.
1987.
Research frontiers In writing analysis.
TESOL Quarterly 21:
677-696.
Cook,
M.
1969.
Transition probabilities and the
incidence of filled pauses.
psychonomic Science 1:
191-192 .
Cooper,
M.M.
1986.
The ecology of writing.
College
English 48:
364-375.
Cooper,
C.P. and A. Matsuhashi 1983.
A theory of the
writing process.
In M. Martlew (ed.)
1983.
Cooper,
C.R. and S. Greenbaum (eds.)
1986.
Studying
writing: linguistic approaches.
Written
communication annual: an international survey of
research theory.
Volume I.
Beverly Hills:
Sage
Publications.
Cooper,
C .R. and L. Odell 1977.
Evaluating writing:
describing,
measuring,
judging.
Urbana,
Illinois:
NCTE.
Cooper,
C.R. and L. Odell
(eds.)
1978.
Research on
composlng: points of departure.
Urbana,
Illinois:
NCTE.
Cooper, M.
and M. Holzman 1983.
Talking about protocols.
In College Composition and Communication 34:
284-
293.
Croll,
P.
1986.
Systematic classroom observation.
Social Research and Educational Studies Series:
3.
London: The Falmer Press.
Crowley,
S.
1977.
Components of the composing process.
College Composition and Communication 2:
166-169.
Crowhust,
M.C. 1978.
Syntactic complexity In t,,,o modes
of discourse at grades 6,
10 and 12.
Brandon,
Manitoba,
Canada: Brabdon University.
(ERIC No.
ED
168037).
Crowhust, M. and G.L.
Piche 1979.
Audience and mode of
discourse effects on syntactic complexity in writing
on two grade levels.
Research in the Teaching of
English 13: 101-109.
Daiute, C.
1981.
Psycholinguistic foundations of the
writing process.
Research in the Teaching of
English 15: 5-22.
433

---

d'Arcais,
G.B.F. and W.J.M. Levelt
(eds.)
1970.
Advances
in psycholinguistics.
New York: American Elsevier.
de Beaugrande, R.
1982.
Psychology and composition:
past,
present and future.
In M. Nystrand
(ed.)
1982.
1984.
Text production:
toward a science of
composition.
Norwood,
N.J.: Ablex Publishing
Corporation.
Denzin,
N.K.
1978.
The research act: a theoretical
introduction to sociolinguistic methods.
Second
edition.
New York: McGraw-Hill.
Diederich,
F.B.
1974
Measuring grmvth in Ivriting.
Urbana,
Illinois:
National Council of Teachers.
Dobrin,
D.N.
1986.
Protocols once more.
College English
48:
713-725.
Donaldson,
M.
1976.
Development of conceptualization.
In V. Hamilton and M.D. Vernon
(eds.)
1976.
Donovan,
T.R. and B.W. McClelland
(eds.)
1980.
Eight
approaches to teaching composition.
Urbana,
Illinois: National Council of Teachers of English.
Dornick,
S.
(ed.)
1977.
Attention and performance IV.
London: Academic Press.
Duncker,
K.
1927.
A qualitative
(experimental and
theoretical)
study of productive thinking
(solving
of comprehensible problems)
The pedagogical
seminary 33:
642-708.
1945.
On problem solving.
Psychological monograph
5
(Whole No.
270).
Dvorak,
T.
1986.
Writing in the foreign language.
In
B.H. Wing
(ed.)
1986.
Elbow,
P.
1973.
Writing without teachers.
New York:
Oxford University Press.
1981.
Writing with power.
New York: Oxford
University Press.
E1desky,
C.
1982.
Writing in a bilingual program:
the
relation of L1 and L2 texts.
TESOL Quarterly 16:
211-228.
El1is,
R. 1985.
Understanding second language
acquisition.
Oxford: Oxford University Press.
434

---

Emig,
E.
1964.
The uses of the unconscious in composing.
College Composition and Communication 16: 6-11.
1969.
Components of the composing process among
twelfth-grade writers.
Unpublished doctoral
dissertation.
Harvard University.
1971.
The composing process of twelfth graders.
NCTE Research Report No.
13.
Urbana,
Illinois:
National Council of Teachers of English
1977.
writing as a mode of learning.
In R.C.
Gebhardt
(ed.)
1977.
Ericsson,
K.A. and H.A. Simon 1980.
Verbal reports as
data.
Psychological review 87:
215-251.
1984.
Protocol analysis.
Cambridge,
Mass.: MIT
Press.
Estes,
W.K.
1978.
Handbook of learning and cognitive
processes.
Volume 5: human information processing.
Hillsdale,
New Jersey: Lawrence Erlbaum Associates.
Evans,
B. and B. Waites 1981.
IQ and mental testing: an
unnatural science and its social history.
London
and Basingstoke: The MacMillan Press Ltd.
Faerch, C. and Kasper, G.
(eds.)
1987.
Introspection ~n
second language research. Clevendon: Multilingual
Matters.
Faigley,
L. 1985.
Non-academic writing:
the social
perspective.
In Odell,
L. and D. Goswami
(eds.)
1986.
Competing theories of process: a critique and
a proposal.
College English 48:
527-542.
Faigley,
L.,
R.D. Cherry,
D. A. Joliffe and A. M. Skinner
1985.
Assessing Writers' knowledge and processes of
compos~ng.
Norwood, N.J.: Ablex.
Faigley,
L. and S. Witte 1981.
Analyzing revision.
College Composition and Communication 32: 400-414.
Fielding,
N.G. and J.L.
Fielding 1986.
Linking data.
Qualitative Research Methods, Volume 4.
Beverly
Hills, Ca.: Sage Publications,
Inc.
Flower,
L.S. and J.R. Hayes 1977.
Problem solving
strategies and the writing process.
College English
39: 449-461.
1980a.
The cognition of discovery: defining a
rhetorical problem.
College Composition and
Communication 31: 21-32.
435

---

1980b.
The dynamics of composing: making plans and
juggling constraints.
In L.W. Gregg and E.R.
Steinberg
(eds.)
1980.
1981a.
A cognitive process theory of writing.
College Composition and Communication 31: 365-387.
1981c.
The pregnant pause: an inquiry into the
nature of planning.
Research in the Teaching of
English 15: 229-244.
1984.
Images,
plans and prose:
the representation
of meaning in writing.
Written Communication 1:
120-160.
Fredericksen, C.F. and J.F. Dominic
(eds.)
1981.
Writing:
the nature,
development and teaching of
written communication volume 2.
Hillsda1e,
N.J.:
Laurence Erlbaum Associates.
Freedle,
R.O.
(ed.)
1979.
New directions in discourse
processing.
Norwood,
N.J.: Ablex Publishing
Corporation.
1985.
Achieving cognitive synthesis of separate
language skills:
implications for improving
literacy.
In C.N. Hedley and A.N.
Barrata
(eds.)
1985.
Freedle,
R.O.
and P. Duran (eds.)
1987.
Cognitive and
linguistic analyses of test performance.
Volume
XXII in the series Advances in discourse processes.
Freedle,
R.O. andG. Hale 1979.
Acquisition of new
comprehension schemata for expository prose by
transfer of an narrative schema.
In R.O.
Freedle
(ed.)
1979.
Freedman,
A.,
I.
Pringle and J. Yalden
(eds)
1983.
Learning to write:
first
/
second language.
London:
Longman.
Freedman,
S.W.
1983.
Student characteristics and essay
test writing performance.
Research in the Teaching
of English 17:
313-324.
(ed.)
1985.
The acquisition of written language:
response and revision.
Norwood,
N.J.: Ablex
Publishing Corporation.
Friedlander,
A.
1990.
Composing in English: effects of a
first language on writing in English as a second
language.
In B.
Kroll
(ed.)
1990.
Garman, M.
1990.
Psycholinguistics.
Cambridge:
Cambridge University Press.
436

---

Garrett, M.F.
1988.
Processes in language production.
In F.J, Newmeyer
(ed.)
1988.
Gaskill, W.H.
1986.
Revising in spanish and English as
a second-language: a process-oriented study of
composition.
International dissertation abstracts
47/10 3747-A.
Gass,
S.
1984.
A review of inter language syntax,
language transfer and language universals.
Language
Learning 34: 115-131.
'
Gass,
S. and E.M. Varonis 1984.
The effect of
familiarity on the comprehensibility of nonnative
speech.
Language Learning 34:
65-90.
Gebhardt, R.C.
(ed.)
1979~
Composition and its teaching:
articles from College Composition and Communication
during the editorship of Edward P.J. Corbett.
ohio
council of Teachers of English Language Arts.
Godshak,
F.I.
F. Swimford and W.E. Coffman 1966.
The
measurement of writing ability.
New York College
Entrance Examination Board.
Goldman-Eisler,
F.
1958a.
Speech production and the
predictability of words in context.
Quarterly
Journal of Expperimental Psychology 10: 96-106.
1958b.
The predictability of words in context and
the length of pauses in speech.
Language and Speech
1:
226-231.
1958c.
Speech analysis and mental processes.
Language and Speech 1 59-75.
1968.
Psycholinguistics: experiments in spontaneous
speech.
London: Academic Press.
Goody,
J.
(ed.)
1968.
Literacy in traditional societies.
cambridge: Cambridge University Press.
1977.
The domestication of the savage mind.
Cambridge: Cambridge University Press.
1980.
Technology,
tradition and the state in
Africa.
London: Oxford University Press.
Green,
D.W. and P.C. Wason 1982.
Notes on the psychology
of writing.
Human Relations 35: 47-56.
Greenfield,
P.M. and J.S. Bruner 1966.
Culture and
cognitive growth.
International Journal of
Psychology 1: 89-107.
437

---

Greeno,
J.C.
1978.
Natures of problem-solving
abilities.'
In VJ.K.
Estes
(ed.)
1978.
Gregg,
L.W.
and E.R. Steinberg
(eds.)
1980.
Cognitive
processes in writing.
Hillsdale,
N.J.:
Lawrence
Erlbaum Associates.
Hairston,
M.
1982.
The winds of change: Thomas Kuhn and
the revolution in the teaching of writing.
College
Composition and Communication 33: 76-88.
Hall,
C. 1987.
Revision strategies in Ll and L2 writing
tasks:
a case study.
Unpublished doctoral
dissertation.
university of New Mexico.
1990.
Managing the complexity of revising across
languages.
TESOL Quarterly 24:
43-60.
Halliday,
M.A.K.
1976.
Cohesion in English.
London:
Longman.
Hammersley,
~1. and P. Atkinson 1983.
Ethnography:
principles in practice.
London: Tavistock.
Hamp-Lyons,
L.
1988.
The product before:
task-related
influences on the writer.
In P.C.
Robinson
(ed.)
1988.
1990.
Second language writing: assessment issues.
In B. Kroll
(ed)
1990.
Hamp-Lyons,
L.
and B Heasley 1987.
Study writing: a
course in written English for academic and
professional purposes.
Cambridge: Cambridge
university Press.
Harrington,
D.V.
1981.
Teaching students the art of
discovery.
College Composition and Communication 1:
7 -14.
Harris,
Z.S.
1951.
Methods In structural linguistics.
Chicago:
Hawkins,
P.R.
1971.
The syntactic location of hesitation
pauses.
Language and Speech 14: 277-288.
Hayes,
J.R.
1987.
The complete problem solver.
Hillsdale,
N.J.:
Lawrence Erlbaum Associates.
Hayes J.R. and L.S.
Flower 1980a.
Identifying the
organization of writing processes.
In L. Gregg and
E.
Steinberg
(eds.)
1980.
1980b.
Writing as problem solving.
Visible
Language 14: 388-399.
438

---

1983.
Uncovering cognitive processes in writing: an
introduction to protocol analysis.
In P. Mosentha1,
L. Tamor and S. Wa1ms1ey
(eds.)
1983.
Hayes-Roth,
B.
1979.
Flexibility in executive
strategies.
N:1170.
Santa Monica,
Ca.: The Rand
Corporation.
Hayes-Roth,
B. and Hayes-Roth,
F.
1979.
A cognitive
model of planning.
Cognitive Science 3:
275-310.
Hayes-Roth,
B. and P. Thorndyke 1979.
Decision-making
during the planning process.
N:1213.
Santa Monica,
Ca.: The Rand Corporation.
Heath,
S.
B.
1983.
Ways with words.
New York: Cambridge
University Press.
Hedley,
C.N. and A.N.
Baratta
(eds.)
1985.
Contexts of
reading.
Norwood,
N.J.: Ablex publications
Corporation.
Henderson,
A.
F. Goldman-Eisler and A.
Skarbek. 1966.
Sequential temporal patterns in spontaneous speech.
Language and Speech 9:
207-216.
Hildenbrand,
J.L.
1985.
Carmen: a case study of an ESL
writer.
International doctoral dissertation
abstracts 46/12 3637-A.
Hillocks,
G.,
Jr.
1986a.
Research on written
composition: new directions for teaching.
Urbana,
Illinois:
ERIC Clearinghouse on Reading and
Communications Skills and the National Conference on
Research in English.
1986b.
The writer's knowledge:
theory,
research and
implications for practice.
In A.R.
Petrosky and D.
Bartholomae 1986.
1987.
Synthesis of Research on teaching writing.
Educational Leadership May 1987: 71-82.
Homburg,
T. J.
1984.
Holistic evaluation of ESL
composition: can it be validated objectively? TESOL
Quarterly 18:
Horrowitz,
D.
1986.
Process,
not product: less than
meets the eyes.
TESOL Quarterly 20:
141-143.
Hughes,
A. and Lascaratou,
C. 1982.
Competing criteria
for error gravity.
ELT Journal
36: 175-182.
Hull,
G.
1987.
The editing process in writing: a
performance study of more skilled and less skilled
439

---

college writers.
Research in the Teaching of
English 21: 8-29.
Humes,
1983.
Research on the composing process.
Review
of Educational research 53: 201-216.
Irmscher,
W.F.
1979.
Writing as a way of learning and
developing.
In College Composition and
Communication 30: 241-244.
Jacobs,
H.L.,
S.A.
Zikgraf,
D.R. Wormuth,
V.F.
Harfield
and J.B.
Hughey 1983.
Testing ESL composition: a
practical approach.
English composition program.
Rowley,
Nass.: Newbury House Publishers,
Inc.
Jacobs,
S.
1982.
Composing and coherence:
the writing of
eleven pre-medical students.
Linguistics and
Literacy Series 3.
Washington,
D.C.: Cent er
for Applied Linguistics.
Jakobson,
R.
1960.
Linguistics and poetics.
In
Sebeok,
T.A.
1960.
Jeffress,
L.A.
1951.
Cerebral mechanisms in behavior.
New York: vii ley.
Jones,
S.
1982.
Attention to rhetorical form while
composing In a second-language.
In C. Campbell,
V.
Flashner.
T.
Hudson,
and J.
Lukin
(eds.)
1982.
Jones,
S. and Tetroe,
J.
1987.
Composing in a second
language.
In A. Matsuhashi
led).
1987.
Johnson,
D.M.
1955.
The psychology of thought and
judgment.
New York:
Harper and Brothers Publishers.
Johnson-Laird,
P.N.
1983.
Mental models:
towards a
cognitive sc.ience of language,
interference and
consciousness.
Cambridge: Cambridge University
Press.
Kaplan,
R.B.
1966.
Cultural thought patterns in
intercultural education.
Language Learning 16: 1-
20.
1967.
Contrastive rhetoric and the teaching of
composition.
TESOL Quarterly 3: 10-16.
1983.
Contrastive rhetorics some implications for
the writing process.
In A.
Freedman,
I.
Pringle and
J. Yalden
(eds.)
1983.
1988.
Contrastive rhetoric and second language
learning:
toward a theory of contrastive rhetoric.
In Purves,
A.C.
led.)
1988.
440
.'.

---

King,
M.L.
1978.
Research in composition: a need for
theory.
Research in the Teaching of English 12:
193-202.
Kinneavy,
J.
1971.
A theory of discourse.
Englewood
Cliffs,
N.J.:
Prentice-Hall.
Kintsch,
W. and T.A. van Dijk. 1978.
Toward a model of
text comprehension and production.
Psychological
Review 85: 363-394.
Kirkland, M.R. and A.P. Saunders,
1991.
Maximizing
students'
performance in summary writing: managing
the cognitive load.
TESOL Quarterly 25:
105-121.
Kosslyn,
S.M.
1980.
Image and mind.
Cambridge,
Mass.:
Harvard university press.
Krapels,
A.R.
1990.
An overview of second language
writing process research.
In B. Kroll
(ed.)
1990.
Krashen,
S.D.
1977.
The monitor model for adult second
language performance.
In Burt,
Dulay and
Finnochiaro
(eds.)
1977.
1981.
Second language acquisition and second
language learning.
Oxford:
Pergamon Press.
1984.
Writing research, theory and applications.
Oxford:
Pergamon Institute of English.
Kroll,
S.M.
1978.
Cognitive egocentrism and the problem
of audience awareness in written discourse.
Research in the Teaching of English 12:
269-281.
1990
(ed.).
Second Language writing: research
insights for the classroom.
New York: Cambridge
University Press.
1990.
What does time buy?
ESL student performance
on home versus class compositions.
In S.
Kroll
(ed.)
1990.
Kuhn,
T.S.
1970.
The structure of scientific
revolutions.
Chicago:
University of Chicago.
Langer,
J.A.
1984.
The effects of available information
on response to school writing tasks.
Research in
the Teaching of English 18: 27-44.
1987.
The construction of meaning and the
assessment of comprehension: an analysis of reader
performance on standardized test items.
In O.R.
Freedle and P Duran
(eds.)
1987.
441

---

Lashley,
K.S.
1951.
The problem of serial order in
behavior.
In L.A. Jeffress
(ed.)
1951.
Lay,
N.D.S.
1982
Composing processes of adult ESL
learners: a case study.
TESOL Quarterly 16:406.
Lay,
C.H. and A. Paivio 1969.
The effects of task
difficulty and anxiety on hesitations in speech.
In
Canadian Journal of Behavioral Science 1: 25-37.
Levelt, W.J.M.
1983.
Monitoring and self-repair in
speech.
Cognition 14:
41-104.
Lloyd-Jones,
R.
1977.
Primary trait scoring.
In C.
Cooper and L. Odell
(eds.)
1977.
Lounsbury,
F.G.
1954.
Transitional probability,
linguistic structure, and systems of habit-family
hierarchy.
In C.E. Osgood and T.A. Sebeok (eds.)
1954.
Luria,
and F.
I.
Yudovich 1971.
Speech and the
development of mental processes in the child. J.
Simon (ed.).
Baltimore: Penguin.
Lunsford,
A.
1980.
The content of basic writers' essays.
College Composition and Communication 31:
278-290.
Maclay,
H.
and Osgood,
C.E.
1959.
Hesitation phenomena
in spontaneous English speech.
Word 15:
19-44.
Macrorie, K.
1976.
Telling writing.
Rochelle Park,
N. J.: Hayden.
1968.
To be read.
English Journal 57:
686-692.
Mahl, G.F.
1956.
Disturbances and silences in the
patient's speech in psychotherapy.
Journal of
Abnormal and Social Psychology 53: 1-15.
Manouan, A.
1979.
Le preterit anglais:
son emploi par
des etudiants ivoiriens de premiere annee
d'universite.
These pour un doctorat de troisieme
cycle preparee sous la direction de Monsieur le
Professeur Jean Guenot.
Universite Paris VII.
1989.
De la difficulte de reussir dans
l'Enseignement superieur: cas de la Faculte des
Lettres, Arts et Sciences Humaines.
Unpublished
manuscript.
1990.
Overall academic success and writing
performance: any quantifiable evidence?
Unpublished
manuscript.
442

---

1991.
Holistic vs Analytic marking:
reconciling two
irreconcilable approaches?
Paper presented at the
25th annual meeting of IATEFL,
Exeter.
Markham,
L.R.
1976.
Influences of handwriting quality on
teacher evaluation of written work.
American
Educational Research Journal 13:
277-283.
Martin-Betancourt, M.E. 1986
The composing processes of
Puerto Rican college students of English as a
second-language.
International doctoral
dissertation abstracts 47/07 2577-A.
Martlew,
M.
(ed.)
1983.
The psychology of written
language:
Developmental and educational
perspectives.
Chichester: Wiley.
Mayer,
R.E.,
Larkin,
J.H. and Kadane,
J.B.
1984.
A
cognitive analysis of mathematical problem-solving
ability.
In R.J.
Sternberg
(ed.)
1984.
Matsuhashi,
A.
1979.
Producing written discourse:
a
theory-based description of the temporal
characteristics of three discourse types from four
competent grade 12 writers.
Unpublished doctoral
dissertation.
State university of New York at
Buffalo.
1981.
Pausing and planning:
the tempo of written
discourse production.
Research in the teaching of
English 15:
113-134.
1982.
Explorations In the real-time production of
written discourse.
In M. Nystrand
(ed.)
1982.
1985.
Revision,
addition and the power of the
unseen text.
In S.W. Freedman (ed.)
1985.
1987.
Revising the plan and altering the text.
In
Matsuhashi
(ed.)
1987.
(ed.)
1987.
Writing in real
time: modelling
production processes.
Norwood,
N.J.: Ablex
publishing corporation.
Matsuhashi,
A.
and C. Cooper 1978. A video time-monitored
observational study: The transcribing behavior and
composing processes of a competent high school
writer.
Paper presented at the Division C
Symposium,
'The Writing Process',
American
Educational Research Association 1978 Annual
Meeting,
Toronto,
Ontario,
March 29,
1978.
McCutchen,
D.. 1986.
Domain knowledge and linguistic
knowledge in the development of writing ability.
Journal of Memory and Language 25:
431-444.
443

---

McLaughin,
B.
1987.
Theories of second language
learning.
London:
Edward Arno1d.
1978b.
The monitor model:
some methodological
considerations.
Language Learning 20:
309-32.
Miller,
J.
1987.
A grammatical characterization of
language disorder.
In proceedings of the first
international symposium on specific speech and
language disorders in children.
London: Association
for All Speech Impaired Children
(AFASIC).
Mohan,
B. and W.
Low 1985.
Academic writing and Chinese
students:
transfer and developmental factors.
TESOL
Quarterly 19: 515-534.
Moffett,
J. 1968
Teaching the universe of discourse.
Boston: Houghton-Miff1in.
Mosentha1,
P.,
L. Tamor and S. Wa1ms1ey
(eds.)
1983.
Research on wricing: principles and methods.
New
York:
Longman.
Murray, M.
1978a.
Internal revision:
a process of
discovery.
In C.R. Cooper and L. Odell 1978.
1978b.
Teach the motivating force of revision.
English Journal
67: 56-60.
Myers,
G.
1986.
Reality,
consensus,
and reform in the
rhetoric of composition teaching.
College English
48: 154-177.
Newel1,
A. and H.A.
Simon 1972.
Human problem solving.
Eng1ewood Cliffs,
N.J.:
Prentice-Ha11.
Newkirk,
T.
1984.
Anatomy of a breakthrough: case study
of a college freshman writer.
In Beach and Bridwell
(eds.)
1984.
Newman,
S. and Epstein,
R. 1985.
Current perspectives Ln
dysphasia.
Edinburgh: Churchill Livingstone.
Newmeyer,
F.J.
1988.
L.inguistics:
the Cambridge survey.
Volume III Language: psychological and biological
aspects.
Cambridge: Cambridge university Press.
Nisbett,
R.E. and T.D. Wi1son 1977.
Telling more than we
can know: verbal reports on mental processes.
Psychological Review 84:
231-259.
No1d,
E.N. 1981.
Revising.
In C.H.
Frederiksen and
Dominic
(eds.)
1981.
No1d,
E.N.
and B.
Davis 1976.
The discourse matrix.
College Composition and Communication 31: 141-152.
444

---

Nystrand,
M.
(ed.)
1982.
What writers know:
the',
. ;.,
language,
process and structure of written'
discourse.
New York: Academic press.
O'Connell,
D.C.,
S. Kowal,
and H. Hormann1970.
Semantic determinants of pauses.
In G.B.F.
d'Arcais
and W.J.M.
Levelt
(eds)
1970.
O'Connell,
D.C. and S.
Kowal 1972.
Cross-linguistic
pause and rate phenomena in adults and adolescents.
In Journal of Psycholinguistic Research 1: 155-164.
Odell,
L.
1980a.
The process of writing and the process
of learning.
College Composition and Communication
31: 42-50.
1980b.
Teaching writing by teaching the process of
discovery: an interdisciplinary enterp'rise.
In L. W.
Gregg and E.R.
Sternberg
(eds.)
1980.
Odell,
L.,
C.R. Cooper and C. Courts 1978.
biscourse
theory:
implications for research on composlng.
In
C. Cooper and L. Odell
(eds.)
1978.
Odell,
L. and D. Goswami
(eds.)
1985.
Writing in
nonacademic settings.
New York: Guilford:
Olson,
D.R. 1977.
From utterance to text:
the bias of
language in speech and writing.
Harvard Educational
Review 47:
257-281.
Osgood,
C.
E. and T. A.
Sebeok
(eds.)
1954.
Psycholinguistics: a survey of theory and research
problems.
Baltimore: Waverly Press.
Perl,
S. 1978.
Five writers working:
case studies of the
composing processes of unskilled college writers.
Unpublished doctoral dissertation.
New York
University.
1979.
The composing processes of unskilled college
writers.
Research in the Teaching of English 13:
317-336.
1980.
Understanding composing.
College Composition
and Communication.
31:
363-367.
Perl,
S. and A. Egendorf 1979.
The process of creative
discovery:
theory,
research, and implications for
teaching.
In D. Mcquade
(ed.)
1979.
Perron,
J.D.
1977.
Written syntactic complexity and the
modes of discourse.
Paper presented at the meeting
of the American Educational Research Association,
New York.
445
,
.-

---

Petrosky,
A.R. and D. Barholomae (eds.)
1986.
The
teaching of writing.
Eighty-fifth Yearbook of the
National society for the Study of Education,
Part
II.
The National Society for the Study of
Education. Chicago,
Illinois: The University of
Chicago Press.
piaget,
J.
1954.
The construction of reality in the
child.
New York: Basic Books.
Pianko,
S.
1979.
A description of the composing
processes of college freshman writers.
Research in
the Teaching of English 13: 317-336.
Pike,
K.L.
1945.
The intonation of American English.
Ann Arbor: university of Michigan Press.
.
"
POlncare,
H.
1946.
Mathematical creation.
The
Foundations of Science 1913.
Translated by George
Bruce Halsted.
New York:
Science Press.
polanyi,
M.
1958.
Personal knowledge.
London:
Routledge
and Kegan Paul.
Purves,
A.C.
(ed.)
1988.
vhiting across languages and
cultures: issues in contrastive rhetoric.
Written
communication Annual: an International Survey of
Research and Theory Volume 2.
Newbury Park,
Beverly
Hill:
Sage Publications.
Purves,
A.C. and W.C.
Purves 1986.
Cultures,
text
models,
and the activity of writing.
Research in
the Teaching of English 20: 174-197.
Radford,
J.
1974.
Reflections on introspection.
American Psychologist 29: 245-250.
Raimes,
A.
1985b.
What unskilled writer,do as they
write: a classroom study of composing.
TESOL
Quarterly 19:
229-258.
1987.
Language proficiency,
writing ability and
composing strategies: a study of ESL college
students writers.
Language Learning 37:
439-468.
Rea,
P.
(rev.)
1987.
Test in English for educational
purposes.
In J.C. Alderson,
K.J.
Krahnke and C.W.
Stansfield
(eds.)
1987.
Reid,
J. 1984a.
The radical outliner and the radical
brainstormer: a perspective on composing processes.
TESOL Quarterly 18: 529-533.
1990.
Responding to different topic types: a
quantitative analysis from a contrastive rhetoric
perspective.
In B.
Krol1
(ed.)
1990.
446
.'.

---

Reither, J.A.
1985.
Writing and knowing: toward
redefining the writing process.
College English 47:
620-628.
Resnick,
L.B. and R.
Neches 1984.
Factors affecting
individual differences in learning ability.
In R.J.
sternberg (ed.)
1984.
Reynolds, A. and A. Paivio.
cognitive and emotional
determinants of speech.
Canadian Journal of
Psychology 22:
164-175.
Robinson,
C.P.
(ed.)
1988.
Academic writing: process and
product. ELT Documents 129.
Hong Kong:
Modern
English Publications and the British Council.
Rochester,S.R.
1973.
The significance of pauses in
spontaneous speech.
Journal of Psycholinguistic
Research 2:
51-81.
Rohman,
D.G.
1965.
pre-writing: the stage of discovery
in the writing process.
College Composit'ion a·nd.
Communication 16: 106-112 . ~ "
Rohman,
D.G.
and A.a. Wlecke 1964.
pre-writing: the
construction and application of models for concept
formation in writing.
U.S. Department of Education,
Office of Educational Research and Improvement
(OERI) Washington, D.C.
20208.Eric Reports no.
EO
001273.
Rohwer, W.D. Jr.
1978.
An elaborative conception of
learner differences.
In W.K. Estes (ed.)
1978.
Rose, M.
1980.
Rigid rules,
inflexible plans and the
stifling of language: a cognitivist analysis of
writer's block.
In College Composition and
Communication 31: 388-400.
Rosen, H.
1969.
An investigation of the effects of
differentiated writing assignments on the
performance in English composition of a selected
group of 15/16 year old pupils.
Unpublished
doctoral dissertation.
University of London.
Rubin, J.
1981.
A study of cognitive processes in second
language learning.
Applied Linguistics 2:
117-131.
Rumelhart,
D.E.
1977.
Toward an interactive model of
reading.
In S Dornick (ed.)
1977.
Rumelhart,
D.E.
and A. Ortony 1977.
The representation
of knowledge in memory.
In R. Anderson, R. Spiro
and W.V. Montague
(eds)
1977.
447

---

:,': .'
"
,~.
Ruth, L.
and S. Murphy 1984.
Designing.topic~ f6r';
writing assessment: problems of meaning.
College
Composition and CommunLcation 35:" 410-'422.
'.
1988.
Designing writing tasks for the assessment of
writing.
Norwood, N.J.: 'Ablex Publishing
Corporation.
San Jose, C.P.M.
1972.
Grammatical structures in 'four
modes of writing at fourth-grade level.
Unpublished
doctoral dissertation.
syracuse University, N.Y . .
Scardamalia, M.
and C.
Bereiter 1985b.
Research on
written composition.
In wittrock, M.C.
(ed.)
1985.
Scardamalia, M.,
C.
Bereiter and Goelman 1982.
The role
of production factors in writing ability.
InM.
Nystrand (ed.)
1982.
Schank, R. and R.
Abelson 1977. Scripts,
plans,
goals and
understanding.
Hillsdale, N.J.: Lawrence Erlbaum
Associates.
Scribner, S. and M. Cole 1978.
Literacy without
schooling: testing for intellectual effects.
Harvard Educational Review
48: 448-461.
Sebeok, T.A.
1960.
Style in language.
Cambridge, Mass.:
MIT Press
Smagorinsky, P.
1989.
The reliability and validity of
protocol analysis.
written Communication 6:
463-
479.
Smith, J.M. 1979.
Bruner on writing.
College
Composition and Communication 30:
129-131.
Smith,
F.
1982.
writing and the writer.
London:
Heinmann Educational Book Ltd.
Sommers, N.
1978.
Revision in the composing process: a
case study of college freshman and experienced adult
writers.
Unpublished doctoral dissertation.
Boston
University.
1979.
The need for theory in composition.
College
Composition and Communication 31:
46-50.
1980a.
Response to Sharon Crowley 'Components of
the composing process'.
College Composition and
Communication 29:
209-211.
1980b.
Revision strategies of student writers and
experienced adult writers.
College Composition and
Communication 31:
46-50.
448

---

,.
- ..~
".-:.: . ~
.
. .;' '. ',;;-;. .
Spack,
R.
1985.
Literature,
reading,
writing.and' ESL:
bridging the gaps.
TESOL Quarterly 703-?25:
, '
I
Spiro,
R.J.
1977.
Remembering information'f·r,o'~ text:' the
'state of schema' approach.
In R.e. Anderson,
R.J.
Spiro,
and W.E. Montague
(eds.)
1977.
Spivey,
N.N.
1983.
Discourse synthesis: construct'ing
texts in reading and writing.
Unpublished doctoral
dissertation.
University of Texas at Austin.
Stallard, C.K.
1974.
An analysis of the wri t ing behavior'·.,
. t·.-
of good student
writers.
Research in the Teaching', ,>
. : .
of English 8:
206-218.
. "
Steinberg,
E.R.
1986.
Protocols,
retrospective reports;
I
and the stream of consciousness. ' College Eng'lish
48:
697-712.
Sternberg, R.J.
1984.
Advances in the psychology of
human intelligence Volume 2.
Hi11sda1e,
N.J.:
Lawrence Er1baum Associates.
Sterng1ass,
M.S.
1988.
The presence of thought:
introspective accounts. of reading and writing.
Norwood, N.J.: Ablex Publishing Corporation.
Stewart, D.
1969.
Prose with integrity: a primary
objective.
College Composition and ,Communication
20: 223-227
Stoehr, T.
1967.
Writing as thinking.
College English
28:
411-421.
Stosky,
S.
1986.
On learning to write about ideas.
College Composition 37: 276-293.
Succi, G.J.
1967.
The validity of pause as an index of
units in language.
Journal of Verbal Learning and
Verbal Behaviour 6: 26-32.
Swarts,
H.,
L.S.
Flower and J.R. Hayes 1984.
Designing
protocol studies of the writing process: an
introduction.
In R. Beach and L.S.
Bridwell
(eds.)
1984.
Tannenbaum,
P. H.,
F. Williams and C.S. Hillier 1965.
Word predictability in the environment of
hesitations.
Journal of Verbal
Learning and verbal
Behaviour 4:
134-140.
Taylor,
1.
1969.
Content and structure in sentence
production.
Journal of Verbal Learning and Verbal
Behaviour 8: 170-175.
449

---

Tedick,
D. J.
1988.
The effects of topic familiarity on
the writing performance of non-native writers of
English at the graduate level.
Ph.D.
dissertation.
The Ohio State University.
Order Number 8824624
Thorndike,
E.L.
1917.
Reading as reasoning: a study of
mistakes in paragraph reading.
Journal of
Educational Psychology 8: 323-332.
Thorndike,
R.
(ed.)
1971.
Educational measurement
Washington,
DC:
American Council on Education.
Urzua,
C.
1987.
'You stopped too soon': second language
children composing and revising.
TESOL Quarterly
21:
279-304.
van Bruggen, J.A.
1946.
Factors affecting regularity of
the flow of words during written composition.
Journal of Experimental Education 15:
133-155.
van Dijk,
T. A.
(ed.)
1985.
Discourse and communication:
new approaches to the analysis of mass media
discourse and communication.
Berlin: Wa1ter de
Gruyter.
van Dijk,
T.A.
and W.
Kintsch 1983.
Strategies of
discourse comprehension.
New York:
Vernon,
H. and M.D. Vernon.
The development of cognitive
processes.
London: Academic Press.
Voss,
R.
1983.
Janet Emig's the composing processes of
twelfth graders:
a reassessment.
College
Composition and Communication 34:
278-283.
Vygostky,
L.S.
1962.
Thought and language.
Cambridge,
Mass.: MIT Press.
Wason,
P.C.
1978.
Specific thoughts on the writing
process.
Presented at the Cognitive Processes in
Writing--Interdiscip1inary Symposium on Cognition,
Carnegie-Mellon University.
1980.
Specific thoughts on the writing process.
In
L.W. Gregg and E.R. Steinberg
(eds)
1980.
Weir,
C.J.
1983.
Identifying the language problems of
overseas students in tertiary education in the
United Kingdom.
Unpublished doctoral thesis.
University of London.
1988a.
Communicative language testing.
First
edition.
Exeter: University of Exeter Press.
1988b.
Academic writing: can we please all the
people all the time?
In C.P. Robinson
(ed.)
1988.
450

---

1990.
Communicative language testing.
Second
edition.
Hemel Hempstead,Hertfordshire:
Prentice
Hall International
(UK)
ltd.
.
Weiss,
R.H.
1980.
Writing in the total curriculum: a
programme for cross-disciplinary cooperation.
In
T.R. Donovan and B.W. McClelland (eds)
1980.
Whale,
K.B. and S.
Robinson 1978.
Modes of students'
writing: a descriptive study.
Research in the
Teaching of English 21:
349-355.
\\oJhite,
R.V.
and D.S. HcGovern 1991.
Pressional \\vriting
course.
Centre for Applied Language Studies.
University of Reading.
widdowson,
H. G.
1983.
New starts and different kinds of
failure.
In A.
Freedman,
I.
pringle and J. Yalden
(eds.)
1983.
Wilkes,
A.L. and R.A.
Kennedy 1970.
Response retrieval
in active and passive sentences.
Quarterly Journal
of Experimental Psychology 22:
1-8.
Winfield,
F.E. and P.
Barnes-Felfeli 1982.
The effects
of familiar and unfamiliar cultural context on
foreign language composition.
Modern Language
Journal 66:
373-378.
Wing,
B.H.
1986.
Listening,
reading,
writing: analysis
and application.
Northeast Conference on the
Teaching of Foreign Languages,
Inc.
Winterowd,
R.S. 1983.
From classroom practice into
psycholinguistic theory.
In A.
Freedman,
I.
Pringle
and J. Yalden
(eds.)
1983.
Witte,
S.P. 1983c.
Topical structure and revision: an
exploratory study.
College Composition and
Communication 34:
313-341.
1985.
Revising composing theory and research
design.
In S.W.
Freedman
(ed.)
1985.
1987.
Review of research on written composition:
new directions for teaching.
College Composition
and Communication 38: 202-207.
Witte,
S.P.,
and R.D. Cherry 1986.
Writing processes and
written products in composition research.
In C.
Cooper and S. Greenbaum (eds.)
1987.
Wittrock,
M.
(ed.)
1985.
Handbook of research on
teaching.
Third edition.
New York: Macmillan
Education Ltd.
451

---

Wolcott, W.
1987.
Writing instruction and assessment:
the need for interplay between process and product.
College Composition and Communication 38:
40-46.
Woodward, M. and L. M.A.
Frances 1988.
statistics for
health management and research.
London:
Edward
Arnold.
Young,
R.
1978.
Paradigms and problems: needed research
in rhetorical invention.
In C.R.
Cooper and L.
Odell (eds.)
1978.
Young,
R.E., A.L.
Becker and K.L.
pike 1970.
Rhetoric:
discovery and change.
New York: Harcourt Brace
Jovanovich,
Inc.
Zamel, V.
1976.
Teaching Composition in the ESL
classroom:
what we can learn from research in the
teaching of English.
TESOL Quarterly 10:
67-75.
1982.
writing: the process of discovering meaning.
TESOL Quarterly 16:
195-209.
,,4~'
1983.
The composing processes IOf~advanced ESL
students:
six case stu·d'ie-S-.
TESOL ~QUart8rlY 1: 165-
185.
(l(
M r \\ L
",
C~.
t:
. )
.
1984. In search of tn~ ~y:
rese~r-~h and practice in
•
-
\\
/
'U
U
composition.
On TESOL\\83,:
the/qy€stion
of control.
Washington D.C.: TESOL'>,,- ~,?
"---~--
.-~-
Zoellner, R.
1969.
Talk-write:
a behavioral pedagogy for
composition.
College English 30:
267-320.
452

---