Content uploaded by Thierry Olive
Author content
All content in this area was uploaded by Thierry Olive
Content may be subject to copyright.
125
8Writing and Working Memory
A Summary of Theories
and Findings
Thierry Olive
INTRODUCTION
The cognitive approach to text composition, which began in the early 1980s following two chapters
from Hayes and Flower (1980) and Flower and Hayes (1980), rst investigated the cognitive pro-
cesses necessary to compose a text. In the decade that followed Hayes and Flower’s publications,
afew studies attempted to understand how limits of the human cognitive system impose constraints
on writing operations. At that time, working memory was already considered as a major limitation
on human information processing (Baddeley & Hitch, 1974). Working memory indeed entails the
ability to manipulate and to temporary store information, and is assumed to be of limited capacity.
Thus, echoing Flower and Hayes (1980) who had underlined that writers are “full-time overloaded”
(p. 33), a few studies examined writers’ cognitive effort (e.g., Kellogg, 1987, 1988), which was con-
sidered as an index of the global engagement of working memory by writers. Research on working
memory in writing fully developed in the 1990s, as a result of more elaborated theories on work-
ing memory (Baddeley 1986; Just & Carpenter, 1992), but also because models of writing that
integrated working memory were proposed (McCutchen, 1994, 1996; Kellogg, 1996). Since, both
writing acquisition and skilled writing have been shown to be strongly related to working memory.
Working memory is important to writing for different reasons. First, it provides temporary stores
for transient information created during composition. For example, while writers are transcribing a
sentence, they may need to keep in mind an idea that they just thought about. Similarly, they may
need to temporarily remember a long sentence while writing all the words down. Semantic, syn-
tactic, lexical, morphologic, and orthographic information need to be temporarily stored at some
moment during the composition process. Second, the processing capabilities of working memory
are strongly involved to coordinate and switch among the numerous writing processes. Third,
working memory is constantly engaged during writing to construct the writer’s multidimensional
representation of the text in construction. For example, when revising, writers have to compare their
already produced text with the mental representation of the text they intend to compose, both on
the linguistic dimension but also in the semantic one. These few examples show how writing entails
the different functions of working memory. They also give an idea of how this capacity limited
cognitive system may play a role in constraining operations of the writing processes, in terms of
temporary storage and supervision functions.
This chapter attempts to provide a summary of models and ndings on the role of working
memory in writing. It may also help to understand how some writers’ difculties are related to
working memory demands of the writing processes. The chapter begins with a brief sketch of work-
ing memory functions and conceptions. Then, after having described how working memory is inte-
grated in models of writing, the role of working memory in writing is discussed by summarizing
experimental ndings. The chapter aims at underlining the important cognitive demands writing
places on working memory, and at highlighting the nature of these demands. The conclusion of the
chapter proposes some ideas for future research on the involvement of working memory in writing.
AU: Please
add to refer-
ences list.
Y119829_C008.indd 125 20/08/11 1:06 AM
126 Writing: A Mosaic of New Perspectives
THEORIES OF WORKING MEMORY IN WRITING
Following research in cognitive psychology on limitation of the cognitive system, writing research
quickly pointed working memory as a function (or structure) of the cognitive architecture capable
of constraining the use of the writing processes. For instance, Flower and Hayes’ proposal about
writers in permanent cognitive overload was reformulated in the beginning of the 1990s in terms
related to working memory: the cognitive overloaded underlined by these authors was simply sug-
gesting that writers fully engage the total capacity of their working memory when composing a
text. Kellogg (1987) showed the very important cognitive cost of all the writing processes in adult
writers. Scardamalia and Bereiter (1987) emphasized that transition from the knowledge-telling
strategy to the knowledge-transforming strategy required engaging working memory to store new
representations and constraints taken into account by writers. McCutchen (1994) used the metaphor
of a writer juggling with the different writing processes and suggested that skilled writing is linked
to the ability to make the writing processes interactive in working memory. Berninger and Swanson
(1994) proposed a description of the various stages of writing acquisition in children that was based
on the constraints exerted by the writing processes on working memory. In the mid-90s, working
memory was thus fully integrated in writing research.
However, as various models of working memory have been proposed, a difculty for the researcher
interested in writing and working memory is to choose a model of working memory which allows
questioning the writing process from an operational point of view. Accordingly, before detailing the
models of writing that integrate working memory, this section begins by briey describing the main
models of working memory and how its processing and storage functions are conceptualized. An
additional function of working memory can be identied: long-term working memory, which entails
the ability to use information stored in long-term memory without activating it in working memory
(Ericsson & Kintsch, 1995). However, since the role of this function has not yet been integrated
in models of writing (excepted by McCutchen, 2000) and has not been experimentally studied
(excepted by Kellogg, 2001), this section only focuses on working memory.
Models of Working MeMory
All models of working memory have in common to distinguish temporary storage of information
from executive functions or the supervision of cognitive activities. They differ, however, in their
structure and in the postulates that underlie their conceptions. Yet two families of models can
be identied: capacity models consider that a single central resource is used for processing and
temporary storage (Cowan, 2005; Engle, 2002; Just & Carpenter, 1992); componential models
include various pools of resources devoted to particular operations (storage and processing) with
specic codes of information (verbal working memory or visual working memory; Baddeley,
2000, 2007).
Capacity Models of Working Memory
Capacity (or unitary) models of working memory postulate that a single and central resource is
shared among all the processes involved for accomplishing a particular task. For example, in Just
and Carpenter’s (1992) capacity theory of text comprehension, a unitary resource, attention, activate
the processes necessary for comprehension (e.g., decoding or inferential processes), but also main-
tains activated the intermediate representations created when integrating read information (e.g., the
situation model, the referent of an anaphora, a previous sentence). When the demands of the current
operations are more important than the quantity of available attention, then some processes and\or
temporary storage are damaged or slowed down. It results from this mechanics that the task of the
reader is to automate some of the processes involved in comprehension to be able to manage all the
demands of comprehension. In general, low-level processes such as decoding of the text are automa-
tized to keep working memory capacity available for inferential processes.
AU: Please
add to refer-
ences list.
AU: 2001a
or b?
Y119829_C008.indd 126 20/08/11 1:06 AM
127Writing and Working Memory
Capacity models are based on theories of activation of information in long-term memory
(Anderson, 1983; Cowan, 2005; Cantor & Engle, 1993). In this framework, long-term memory is
conceived as a network where activation propagates. When a threshold of activation of an element
in long-term memory is reached or exceeded, this element (a declarative or procedural knowledge)
becomes available for a cognitive operation or for a temporary storage. Working memory capacity
would correspond to the maximal quantity of available activation and the content of working mem-
ory would correspond to the activated and thus accessible part of long-term memory (Anderson,
Reder, & Lebiere, 1996; Conway & Engle, 1996). For Engle, Kane, and Tuholski (1999) or Cowan
(2005), working memory provides with a general-purpose attentional control system that allows
maintaining memory traces in a high state of activation compared to interfering information. So,
only situations that involve control of attention can be considered to engage working memory. Its
main purpose is indeed to cope with distractive and interfering situations. The Engle or Cowan
models do not distinguish specicity of coding (verbal or visuospatial). According to the authors,
working memory capacity reects efciency of an attentional component and does not concern
short-term storage capacity (Unsworth, Heitz, & Engle, 2005) and may be conceived as a measure
of general capacities of an individual, and is assumed to be predictive of general intelligence (Engle
et al., 1999).
Componential Models of Working Memory
Componential models of working memory postulate different pools of resources. This is the case
of the basic working memory architecture outlined by Baddeley (1986) and updated by Baddeley
(2000). In that framework, working memory consists of various processing and temporary storage
components. One system, the central executive, monitors the cognitive processes, while specialized
memory registers are devoted to transient storage of specic information (the slave systems). More
precisely, the phonological loop ensures retention of verbal information and the visuospatial sketch-
pad is responsible for storage of visual and spatial information. It seems, however, that the visuo-
spatial sketchpad has to be divided into two separate registers: one visual and one spatial (Logie,
1995). The episodic buffer, an intermediate structure which serves both for storage and processing,
has as its purpose to integrate information of different formats coming from the other slave systems
or from long-term memory. The central executive fulls several functions: it coordinates the two
slave systems, it allows recovery of information from long-term memory, and it functions as an
attentional system for monitoring complex cognitive activities (Baddeley, 1996, 2007).
This short presentation of the two main theoretical conceptions of working memory distin-
guishes unitary models of working memory from componential models. However, as stressed
by Miyake and Shah (1999), “all of the seemingly disparate models in [their] volume have a lot
in common” (p.443). It must be noted that according to Baddeley (1996), capacity models of
working memory supposedly assume the different functions of the central executive component.
Furthermore, according to Baddeley and Hitch (2001, p. xvii) there is “a general acceptance of
the need to assume both of a general executive system and specic verbal and visual systems.”
Similarly, Engle et al. (1999) indicate that “the working memory/attention system entirely is prob-
ably neither unitary nor entirely separable into domain-specic systems” (p. 125). Accordingly, a
mixed model of working memory may be constituted by a domain-free component or resource and
by domain-specic component (or resources) in charge of short-term storage of information. In the
eld of writing, the point is not to decide on these issues, but to adopt a model that allows better
understanding of text production.
Models of Working MeMory in Writing
Writing research has shown that working memory is essential to writing (Olive, 2004). It should be
noted that whether researchers are interested into writing acquisition or into skilled writing, their
conception of working memory differs. For instance, McCutchen (1996), who was interested in
AU: Please
add to refer-
ences list.
Y119829_C008.indd 127 20/08/11 1:06 AM
128 Writing: A Mosaic of New Perspectives
acquisition, adopted a capacity model; Kellogg (1996), who described the demands of the writing
processes in skilled or expert writing, based its proposals on the componential model of Baddeley.
Working Memory and Writing Acquisition
To describe the role of working memory from a developmental point of view, McCutchen (1996)
followed Just and Carpenter (1992) and proposed a capacity theory of writing. McCutchen consid-
ers that a single resource is shared between the storage and processing demands of writing. In this
context, the difculties encountered when composing a text, especially by young writers, would be
related to the high demands of the processes involved in writing. In that framework, skilled writ-
ing requires efcient management of working memory resources by an adequate orchestration of
the writing processes. Automatization, and more specically increased efciency and uency of
processing (McCutchen, 1988), is the main source for writers to obtain more available resources.
In addition, she proposes that writers with larger working memory capacity, that is, with more
resources available, can coordinate the writing processes more efciently and increase their inter-
activity. Thus, for McCutchen, skilled writing is reached when planning, translation, and revision
of a text can be interactively coordinated in working memory. In the early acquisition of writing,
the writing processes apparently operate as if they were encapsulated. They each occupy almost all
working memory capacity and thus cannot interact with other activated processes. This encapsu-
lated functioning is typical of the knowledge-telling strategy depicted by Bereiter and Scardamalia
(1987) that involves only local planning, by content retrieval-production cycles. Beginning writers
use this simple strategy because it does not exert strong demands on working memory. It is only
through an increased uency of the writing processes and particularly through automatization of
transcription that working memory resources will allow to fully exploit planning and reviewing as
in the knowledge-transforming strategy.
To account for the role of working memory in the different phases of writing acquisition,
Berninger and Swanson (1994) modied the original model of Hayes and Flower (1980) by detail-
ing the structure of the translating process. They distinguished text generation processes (selecting
syntactic structures and lexical items) from transcription (spelling and handwriting). Then, they
described three stages in writing acquisition, each with a different constraint exerted on working
memory by a specic writing process. Until grade 3, writing performance of children is mainly
constrained by demands of transcription. Between grades 4, 5, and 6, because transcription is more
automatized, it allows emergence of local planning and of revision.
After grade 6, development of the writing processes continues and their demands in working
memory come mainly from global planning. At that phase, the writing processes can interact and
writers are able to take into account rhetoric constraints in their productions. From identication
of these working memory constraints, progression of the three phases of Berninger and Swanson’s
model can be characterized as follows: While translating occupies the front of the stage during the
rst phase, working memory ensures its coordination with revision in the second phase, and nally
with planning in the third. Moreover, with this increased interactivity, beginning writers operate on
linguistics units whose size increases: words and then sentences, paragraphs, and nally the text as
a whole. However, it is important to notice that working memory demands of writing remain very
important across those phases.
The heuristic value of a capacity conception of working memory to understand writing acquisi-
tion is based on a fundamental mechanism underlying this approach: working memory resources
are shared between the various processes involved in a task (e.g., handwriting, text generating, plan-
ning, revising). It then becomes easy to understand how the heavy demands of an emerging process
affect activation of other writing processes necessary to achieve the written product or composition.
Meanwhile, this approach stressed the importance of practice and emphasized the role of automa-
tization, but also improvement of efciency from one phase to another during writing acquisition
(Berninger & Swanson, 1994; McCutchen, 1988).
Y119829_C008.indd 128 20/08/11 1:06 AM
129Writing and Working Memory
Componential Working Memory Models in Skilled Writing
The componential model proposed by Kellogg (1996) focuses on skilled and expert writing. It
therefore does not describe the changes in working memory demands that happen with writ-
ing skill acquisition over time. Furthermore, the componential conception of working memory
does not describe how efciency of the writing processes modulates how the resources of work-
ing memory are shared between these processes. This componential model basically describes
the nature of the demands that the writing processes poses to working memory. It therefore
clearly details the relationship between the writing processes and the multiple components of
working memory, especially in terms of the short-term storage demands. In such a componen-
tial framework, what hypotheses can be raised on these relations? The following paragraphs
are based on Kellogg’s (1996, 1999) categorization of the writing processes, and they partly
reect his theoretical proposals, which are nevertheless complemented by other theoretical
considerations.
The planning process is closely linked with the visuospatial sketchpad (i.e., visuospatial working
memory) when writers select information they want to include in their text and when they organizeit.
To nd an idea to write, writers can search for information in their environment, during reading for
example. They can process and integrate gurative drawings, pictures, and so on. They can also
retrieve concrete conceptual knowledge from their long-term memory. In these cases, the visual
component of working memory should be involved. However, when organizing this information,
they may either mentally or on paper represent organization of their knowledge in the form of spatial
structures. They can also use diagrams, plans, trees, and so on. It is likely that structuring a text then
engages the spatial component of working memory.
Translation mainly involves the phonological loop, which refers to auditory working memory.
For example, translating would require the articulatory rehearsal mechanism of the phonological
loop to temporarily maintain active segments of texts that are not yet written but which have already
been formulated by the writer. Moreover, the phonological representations of the constituents of a
sentence must be stored temporarily during syntactic processing.
Execution of a text puts only low demands on the central executive. Handwriting does not
resort to verbal working memory or visuospatial, at least among the experienced writer. It is
however probable that beginning writers control their movements by visuospatially process-
ing the letters they are handwriting. Similarly, to handwrite the various letters that make up
a word, the phonological loop of working memory might be engaged when writers mentally
spell the word. For example, Colombo, Fudio, and Mosna (in press) have recently shown that a
concurrent articulatory suppression task affects selection of the serial order of the graphemes
in a word.
Regarding reading, research on text comprehension has shown that auditory working mem-
ory, or the phonological loop, is fundamental to this activity (Gathercole & Baddeley, 1993).
Presumably, reading during writing also engages the phonological loop. Moreover, in text compo-
sition, reading implies that writers focus their attention on relevant points in the text while inhib-
iting alternative interpretations; therefore reading may also engage the central executive. Editing
is also expected to impose demands on the central executive. To nd the very different types of
errors or inaccuracies that can be present in a text, writers have to focus their attention on dif-
ferent aspects of their text at a time. The phonological loop may also be involved in detection of
errors. For example, evaluation of the linguistic formulation of the text may require intervention
of articulatory rehearsal (Larigauderie, Gaonac’h, & Lacroix, 1998). Similarly, it can be expected
that in adult writers who are also competent readers, visuospatial working memory is required,
for example, to identify occurrences of letters or of morphemes about which writers think they
may have committed errors. Thus, when searching for lexical or spelling errors, a surface reading
is sufcient to detect problems. In that case, visuospatial processes may be used (see the conclu-
sion to this chapter).
AU: Please
conrm
change from
“Section
3.4.”
Sections
are not
numbered.
Y119829_C008.indd 129 20/08/11 1:06 AM
130 Writing: A Mosaic of New Perspectives
WRITING AND WORKING MEMORY: MAIN FINDINGS
McCutchen’s (1996) and Kellogg’s (1996) models have constituted a source for many studies and
these models have allowed raising specic questions that could not have been answered without
them. A rst approach investigates the cognitive cost of writing. For example, Kellogg (1994; see
also Piolat, Olive, & Kellogg, 2005) compared the cost of text composition to that of various cog-
nitive activities (incident or intentional learning, reading, playing chess masters) and observed
that skilled writing is one of the most demanding cognitive activities. Of course, composing is
even more demanding for beginning writers who have not yet automatized the writing processes.
A second approach examines the nature of the demands imposed by writing on working memory
by investigating the relationship between the writing processes and the verbal and visuospatial
components of working memory.
Cognitive effort of Writing
The overloaded writer of Flower and Hayes (1980) raised several questions about nature of the
cognitive cost of writing. The terms of cognitive effort were rst used in research on the functional
limitation of the human cognitive system. Navon and Gopher (1979) postulated that the human cog-
nitive system has at any moment a nite amount of processing “resources” (or facilities). Nowadays,
working memory is considered as supporting this pool of resources. Additionally, a sharing mecha-
nism allows assigning the working memory resources to the different processes engaged in a task.
The amount of resources devoted to a particular process or task would depend on the computational
complexity of the task, the level of expertise, and the practice and development of an individual,
but also on his/her level of stress and arousal. In this perspective, cognitive effort would correspond
to the amount of resources devoted to the cognitive processes engaged in a task, or in other terms
to the capacity of working memory necessary to carry out a task. So, writing research quickly
attempted to assess the cognitive cost of the different writing processes.
Cognitive Effort of the Writing Processes
The cognitive effort of planning is primarily related to recovery and organization of content. For
example, cognitive effort of writers with low thematic (or domain) knowledge is more important
than that of writers with high domain knowledge (Kellogg, 1987). However, availability of knowl-
edge interacts with writing skill. Olive, Piolat, and Roussey (1997) indeed showed that writers with
relatively little skill allocated as many resources to planning, irrespective of their familiarity with
the topic of the text, while more skilled writers devoted fewer resources to planning when writing
about familiar topic. Penningroth and Rosenberg (1995), for their own part, forced writers to include
in a story a conclusion with either little information (low processing load) or with more information
(high processing load). In the latter condition, writers did not exhibit more effort but changed their
writing strategy.
Regarding content organization, creating a plan before composing, that is, separating the
demands of planning from demands of the other writing processes, also reduces the overall cost of
production and leads to producing texts of better quality (Rau & Sebrechts, 1996) containing more
ideas (Glynn et al., 1982). Kellogg (1988, 1994) experimentally tested the impact of various types of
drafting strategies and of organizing modes (e.g., plan in the form of written or mental list of topics,
in conceptual networks). In essence, he found that only use of the writing processes varies according
to the prewriting strategy. In addition, Kellogg (1993, 1994) showed that composing descriptive and
argumentative texts require writers to develop more cognitive effort than narratives.
Text revision is also highly demanding. McCutchen, Francis, and Kerr (1997) argued that a
signicant amount of available resources would be allocated to error detection, that is, to reading
(Hayes, 1996). Roussey and Piolat (2008) investigated whether reading a text in order to evaluate
it is more expensive than reading a text to understand it. They also examined how nature of the
Y119829_C008.indd 130 20/08/11 1:06 AM
131Writing and Working Memory
errors contained in the text has affected writers’ cognitive effort. They conducted an experiment in
which students had to perform a comprehension task and a revising task on a text containing either
syntax errors or spelling errors. Analysis of the cognitive effort associated with the critical reading
and comprehension reading processes, and of the participants’ comprehension and revising perfor-
mance, showed that critical reading is more effortful than comprehension reading. It also showed
that critical reading was more effortful with syntax errors than with spelling mistakes.
Translation has been shown to be the less expensive process, presumably because this process
shares several mechanisms with the more practiced and thus more automatized processes engaged
in speaking. Actually, few studies have evaluated how difculties encountered in text formulation
affect the working memory demands. More intensive research has to be conducted in this direction
to explain how translation does function and how it requires working memory. Indeed, even if some
processes are shared between oral and written language production, activities specic to writing,
such as spelling are very effortful at least in children. Estimating the cognitive cost of spelling
among beginning writers would enable to better understand the specic procedure or language for-
mulation in writing. Similarly, due to the stylistic constraints of most of the writing situations, the
choice of words (lexical selection) is far more important in writing than in speaking. Retrieval of
a lexical item, at least in some cases, should involve working memory, particularly when different
words are available and that writers have to choose the one that best suits his ideas. In the same way,
the choice of a syntactic structure should also engage working memory more strongly than when
speaking. In sum, the properties of the writing situation (absence of the recipient, high stylistic pres-
sure) should weigh on working memory demands of translating.
Finally, although handwriting is generally considered relatively automatic in adult writers, it
requires more deliberate, conscious effort for beginning writers, even if automatization of hand-
writing progresses with age (Graham & Weintraub, 1996). Actually, handwriting is the main con-
tributor of quality in beginning writers (rst primary grades), and this contribution decreases with
age (Berninger et al., 1992, 1994, 1996). The cost of handwriting also impacts the high level writing
processes (Fayol, 1999). For example, Bourdin and Fayol (1994, 2000, 2002) showed that the cost
of handwriting reduces performance on a written serial recall task compared to an oral recall. In
addition, the cost of handwriting has a clear impact on how the writing processes are orchestrated
(Berninger, 1999). For example, the use of the knowledge-telling strategy by children would be
related to the high demands of handwriting (Bereiter & Scardamalia, 1987). Similarly, automa-
tization of handwriting in adults allows them to activate the high-level writing processes (plan-
ning, translating, and revising) simultaneously to transcription whereas children can only adopt a
sequential strategy (Olive & Kellogg, 2002; Olive, Alves, & Castro, 2009). The same phenomenon
is observed according to the level of skill in typing when using keyboard (Alves, Castro, & Olive,
2008; Alves et al., 2007).
Developmental Changes in Writers’ Cognitive Effort
Despite the early interest of writing research in the role of working memory in children, few stud-
ies have analyzed cognitive effort of beginning writers and how it changes with age and practice.
However, if text composition is effortful for adults, it is especially for children who have not yet
acquired all the necessary skills to translate their ideas into a coherent text. Novice writers have
indeed to supervise several very expensive processes since these processes are only emerging. With
practice and training, some of the writing processes become less expensive (as is the case with
handwriting or spelling). This automatization provides more working memory resources, which in
turn allows children to take into account a greater number of constraints related to the writing task
and situation.
Assuming that cognitive effort reects the cognitive demands of writing, then the analysis of cog-
nitive effort of beginning writers should help to understand how the demands of working memory
production change with age, and how different factors affect these demands. In this context, Olive
and colleagues (2009) have studied how cognitive effort changes in children of grade 5 and 9 that
AU: Please
add to refer-
ences list.
AU: Year
changed
from 2008
to match
references,
please
conrm.
Y119829_C008.indd 131 20/08/11 1:06 AM
132 Writing: A Mosaic of New Perspectives
produced narrative and argumentative texts. They observed that cognitive effort decreased from
grade 5 to grade 9 only with argumentative text. This nding conrms, rst, that students continue to
improve their skills related to writing argumentative texts between grades 5 and 9 and, second, that
writing narrative texts is well mastered in grade 5. These two types of texts indeed impose different
demands on working memory because early exposure of children to narrative through reading or
through different media allows them to acquire cognitive structures that they can apply to the produc-
tion of this type of text, making the composition of narratives easier to manage and less demanding
(Donovan & Smolkin, 2006; Halliday, 1975). Argumentative composition is, by contrast, taught only
from grade 9 and a textual schema does not support its production. Producing an argumentation thus
requires the use of the high-demanding knowledge-transforming strategy. As a result, production of
an argumentative text remains more difcult and costly to beginning writers than to adult writers
(Kellogg, 2001a).
The observed reduction of cognitive effort through grades seems somewhat paradoxical. Indeed,
with age, the writing processes are in charge of more complex and sophisticated operations. For
example, older writers do not locally revise their text; rather, they adopt revision procedures that
globally evaluate the text, both at a surface level surface and at a semantic level (Allal, 2004).
Similarly, planning is more complex and no longer concerns only the next text segment to produce
but the text in its entirety (Scardamalia & Bereiter, 1987). An increase in cognitive effort should
be observed consequently. However, it is likely that at certain grades, presumably near the end of
an acquisition phase, the overall cost of production decreases, indicating a release of resources
resulting from automatization that precedes the establishment of a more complex process. This is
consistent with Berninger and Swanson’s (1994) proposals about phases in writing acquisition. In
this context, a more systematic study of the changes with age in cognitive effort of writing should
inform about the stage of acquisition a writer is in.
To conclude on the cognitive effort of writing, several ndings have validated the idea that all
writing processes are effortful and that they share common working memory resources (Brown
etal., 1989; Kellogg, 2001a). Similarly, the impact of factors related to the writing situation (drafting
strategy, type of text) or of writers’ characteristics (verbal skills, level of knowledge on the topic)
has been investigated (Olive, Kellogg, & Piolat, 2002; Piolat & Olive, 2000). A research line that
needs to be examined more extensively deals with cognitive effort of writers of different age (Olive
et al., 2009). Another unexplored issue relates to the impact of social and motivational factors, or of
emotional disposition in writers. For example, Martinie, Olive, and Milland (2010) have observed
that writers in cognitive dissonance had less working memory resources available than writers who
did not experience dissonance. Such research will surely shed light on the social and emotional
regulation of writers’ cognition.
verbal deMands of Writing
As a language activity, composing a text engages verbal working memory, the phonological loop
in terms of Baddeley’s model. Several ndings support that claim. For example, Marek and Levy
(1999) used the irrelevant speech effect to determine whether verbal working memory is involved
when composing a text. The irrelevant speech effect refers to a lower performance in presence
of an unattended discourse. This decrease in performance is assumed to result from interference
in verbal working memory. Marek and Levy showed that when writers produce sentences from
several words while being submitted to an unattended listening task, the produced sentences are
of lower quality and contain more errors. Madigan, Johnson, and Linton (1984) have also shown
that writers write more slowly with and unattended listening task. Chenoweth and Hayes (2003;
see also Hayes & Chenoweth, 2003) have used articulatory suppression (repetition of syllables) to
interfere with verbal working memory. They showed that articulatory suppression during produc-
tion of sentences affects the number of spelling errors. Additionally, when verbal working memory
is fully loaded, for example by asking writers to memorize ve or six digits, the texts contain more
AU: Year
changed
from 2008
to match
references,
please
conrm.
Y119829_C008.indd 132 20/08/11 1:06 AM
133Writing and Working Memory
spelling errors (Fayol, Largy, & Lemaire, 1994) and shorter sentences (Kellogg, 2004). All these
ndings may be interpreted as resulting from interference in verbal working memory with opera-
tions carried out by the translating processes. However, planning may also call upon the phono-
logical loop. Indeed, White, Lea, and Ransdell (1999) have found that a verbal working memory
concurrent task produced a greater effect on the duration of the pre-writing pause (a pause mainly
devoted to content planning) compared to a task overloading the central executive component of
working memory.
Revision is also in relation with verbal working memory, particularly when writers read their
text over to themselves. However, few studies have specically explored the nature of the working
memory component involved by revision. One way to answer this question may consist in asking
a writer to revise a text and to simultaneously perform a verbal working memory, for example,
articulatory suppression. Both the number of revisions and numbers of errors in the text could be
measured. If revising requires verbal working memory, then fewer revisions and more errors should
be observed. One has however not to forget that the presence of an error in a text may actually come
from a lack of knowledge of the writer that led him/her not detecting this error. Nevertheless, it can
reasonably be argued that, at least among adults, a spelling error (lexical or grammatical) is a sign
of a poor revision activity. Finally, it must be noticed that Marek and Levy (1999) did not observe
any effect of an inattentive listening task on the number of errors, suggesting that error edition may
not systematically engage verbal working memory. For instance, editing errors presumably requires
visual working memory in addition to verbal working memory, particularly in adults detecting sur-
face errors (Dedeyan, Olive, & Largy, 2006).
Research on involvement of verbal working memory in writing has shown the importance of
this component in text composition. It is highly probable that majority of the verbal demands come
from the translating process. Although it is premature to rule on the subprocesses of translating that
engage verbal working memory, it is probable that the phonological representations of the words
are stored in verbal working memory before words are written down. Interestingly, neuropsycho-
logical observations have shown that oral productions are correct even in patients whose verbal
short-term memory is severely impaired (Gathercole & Baddeley, 1993). So, it has been suggested
that writers with impaired verbal working memory might directly retrieve spelling of a word from
an orthographic lexicon, which does not rely on the phonological representations of the graphemes
(Shelton & Caramazza, 1999). By contrast, spelling processes of healthy writers may be expected
to strongly rely on verbal working memory. To conclude, revision also engages verbal working
memory. Indeed, revising a text implies reading and many studies have shown the close relationship
between this activity and verbal working memory (Just & Carpenter, 1992).
visuospatial deMands of text CoMposition
Writing a text also imposes visuospatial demands on working memory. For example, Lea and Levy
(1999) showed that a concurrent visual tracking task reduces writing uency, but less than a concur-
rent verbal task. Levy and colleagues (1999) also observed that a secondary visuospatial task inter-
feres with production of a text. In a recent research, Olive, Kellogg, and Piolat (2008) have isolated
the visual and spatial demands by asking students to write a text while performing a visual or spatial
concurrent task. They observed interference between writing and both the visual and spatial tasks.
However, performances decreased more with the visual task than with the spatial one. Dissociation
of visual and spatial demands therefore seems relevant to clarify the visual and spatial demands of
text composition.
In that perspective, Kellogg (1999) proposed that visual working memory would be engaged
when writers deal with concrete images when planning the content of their text. Passerault and
Dinet (2000) conrmed this proposition showing that production of a descriptive text (which relies
on mental imagery) engages more largely the visuospatial working memory than composition of
an argumentative text. Similarly, Kellogg, Olive, and Piolat (2007) showed that both denitions of
AU: Please
add to refer-
ences list.
AU: Order
of names
changed
to match
references,
please
conrm.
Y119829_C008.indd 133 20/08/11 1:06 AM
134 Writing: A Mosaic of New Perspectives
abstract and concrete words require verbal working memory, but that only denitions of concrete
words required visual working memory. This selective interference probably occurred during the
planning phase when participants had to form a mental picture of the referent of the concrete name
to dene. Abstract names did not conduct to such interference because they did not involve, or at a
lesser extent, mental imagery.
Kellogg (1999) also proposed that spatial working memory is engaged when writers organize
content of their text. To test that hypothesis, Galbraith and colleagues (2005) asked writers to
produce a text in three distinct phases. During the rst phase, writers had to search for ideas by
noting them briey. During the second phase, structuring, writers were asked to organize into
a plan the ideas they had previously noted. Specically, they had to write an outline. Finally, in
the third and nal phase, writers were asked to generate their text by precisely formulating it.
To test the visual and spatial demands of outlining, writers were simultaneously confronted with
several concurrent tasks, among which a visual task (projection of visual noise on background of
the computer monitor) and a spatial one (tracking task). The results of this experiment conrmed
the role of spatial working memory when structuring a text; it also showed that fewer new ideas
were created with a spatial concurrent task. Indeed, in examining change in number of ideas
between the three phases of production, Galbraith et al. (2005) observed that the spatial second-
ary task reduced the number of new ideas generated during the structuring phase. It therefore
seems that structuring ideas is important for generating new ideas (for similar results, see also
Galbraith et al., 2008).
All these studies emphasize the role of visual and spatial working memory in text composition.
Composing a text, or simply handwriting, is indeed a visuospatial activity. A graphic trace distrib-
uted across the space offered by the support of production materializes the act of writing. In this
sense, writing is a visual activity guided by the writer’s eyes, but it is also a spatial activity as it
applies to organizing the text on the page. The written text leaves a graphic trace in space. If this
visuospatial dimension of the writing seems obvious, it is surprising to notice that only Hayes (1996)
and Kellogg (1996) have incorporated it into their model. According to Hayes, writers mentally rep-
resent the spatial layout of their text, which should involve visuospatial working memory. Le Bigot,
Passerault, and Olive (2009) have conrmed that claim by showing that only a visual interference
during composition reduces subsequent memory for location of words on a page (see also Piolat,
Roussey, & Thunin 1997).
CONCLUSION
More than a decade after the publication of the two major models of writing that featured integrated
working memory, what can be said about the relationship between writing process and working
memory components? First, it seems important to distinguish between the visual and spatial com-
ponents of working memory. Second, a detailed description of the writing processes is important,
as Kellogg (1999) discussed when he distinguished between content retrieval and content structur-
ing for proposing hypotheses about visuospatial working memory. Finally, it should be noted that
future studies have to be conducted purposely to understand the nature of the demands of writing.
In particular, it seems important to extend the work on working memory and writing in a develop-
mental approach. It is indeed likely that by contrast with novice writers, more skilled writers adopt
different strategies based on more or less automatized processes, but which also may require dif-
ferent levels of representation in working memory. For example, Dedeyan, Largy, and Olive (2006)
examined detection of subject-verb errors while participants performed verbal and visual secondary
tasks. They showed that in novice writers the detection of subject-verb agreement interfered with the
verbal task, whereas it interfered with the visual secondary task in adults. They thus concluded that,
in novice writers, error detection in a text is based on an algorithmic procedure that relies on verbal
working memory; yet, skilled writers use visual search procedure to identify in their texts surface
features of morphological agreement.
Y119829_C008.indd 134 20/08/11 1:06 AM
135Writing and Working Memory
FUTURE RESEARCH
This chapter on working memory in writing highlights the diversity of issues that can be addressed
about the relationship between text composition and its demands on working memory. This relation-
ship has been examined by focusing on the cognitive effort of writing and on the different demands
of the writing processes. The studies reported have provided not only insight into the development
of writing skills but also insight into the nature of skilled writing. Although more recent and less
numerous, the studies that have explored the relationship between writing processes and the various
components of working memory have enabled us to describe the nature of the mental representa-
tions writers process when they compose a text. In sum, these studies have helped to propose more
ne-grained understanding of writing by describing the writing processes but also the mental rep-
resentation activated in the writer’s mind. The challenge of the coming years will be to incorporate
into models of production the theoretical advances in the models of working memory, particularly
those related to long-term working memory, the episodic buffer of working memory, and the execu-
tive functions.
For example, one issue to address relates to the involvement of long-term working memory
(Ericsson & Delaney, 1999). The role of long-term working memory in the construction of the
writing expertise has already been emphasized (McCutchen, 2000). Evidence in favor of a visuo-
spatial representation of a text also raises questions that might be explained by long-term working
memory. It indeed seems that the mental representation of a produced text is multidimensional.
Indeed, the writer’s mental representation of the text presumably integrates the situation (Kintsch,
1998) or mental model of the text (Johnson-Laird, 1983), a semantic level representing organiza-
tion of the content of the text (its macro- and microstructure), and a level of verbal representation
(the text base) as reected in works on text recall (Kellogg, 2001b). A representation in the writer’s
episodic memory might also be constructed allowing to nd where and when information has been
inserted in the text, particularly when composition is very long and spread over several sessions of
writing. Moreover, a visuospatial level of representation is necessary for helping precise location
of information in the text during the writing. Added to the transient representations necessary to
store the intermediate work of the writing processes, the demands on working memory are likely
to exceed the capacity of working memory. Long-term working memory provides a convenient and
elegant solution to circumvent these extreme requirements: this multidimensional representation of
a text may be stored in long-term working memory and information it contains can be recovered by
activating cues in short-term working memory.
In this framework, as proposed by Baddeley (2000, 2007), the episodic buffer, that is, the inter-
face between working memory and long-term memory, may be the memory register where these
cues are stored. Its role is indeed to store and to integrate information that comes from the short-
term components of working memory or that are retrieved from long-term memory by constituting
a writing episodic trace integrating all levels of representation. Text representations come from
various sources and may use different codes, while in the episodic buffer these various codes should
be homogenized. However, if this proposal seems attractive to explain results problematic to the
classical model of working memory, the experimental difculty to test this register did not already
afford to investigate this new component of working memory.
Future research on the involvement of working memory in text production should also take into
account the division of the central executive in various executive functions. According to Miyake
and colleagues (2000), at least three executive mechanisms can be distinguished: exibility is the
ability to change strategy or move from one cognitive operation to another; updating refers to the
ability to update working memory with new information; and inhibition is the ability to block irrel-
evant information for the task at hand (e.g., intrusive thoughts) or routine procedures automatically
activated. Moreover, executive functions are not specic to a particular cognitive activity, they also
are essential to all goal directed and adaptive behavior because they provide mechanisms for con-
trolling and regulating the cognitive activities. Executive functions should be heavily involved in the
Y119829_C008.indd 135 20/08/11 1:06 AM
136 Writing: A Mosaic of New Perspectives
production of a text because they allow writers to switch between the different writing processes.
For instance, the recursivity of writing may be thought as involving exibility and inhibition to
block operations of a writing process and activating another. Supposedly, cognitive effort of writing
may reect writers’ exibility to alternate the writing processes. It could even be argued that cogni-
tive effort is actually an assessment of switching costs.
Finally, limitations of working memory cannot alone explain the complexity and difculty of
text composition, particularly as regards the coordination of the writing process. Thus, even if a
componential model of working memory in writing may be seen as an improvement compared to
capacity models of writing because it species the operations and cognitive structures involved
in writing, some other limitations may occur. Torrance and Galbraith (2006) point out that if one
takes into account not only the macroprocesses identied by Hayes and Flower (1980) but all the
subprocesses involved in producing a text, these subprocesses may interact in various ways, and
other processing limitation of the cognitive system can impede their proper implementation. Thus,
even though some conicts between processes can come from competition on allocation of work-
ing memory resources, alternative explanations may be suggested. For example, difculty in coor-
dinating or chaining process may result from differences in processing speed. Salthouse’s (1996)
processing speed theory indeed postulates a limited time mechanism that operates in complex cog-
nitive tasks and that may constrain input of processes when processing is too slow (see also the
proposal from Navon and Miller, 2002, about a mechanism for sharing time). Structural theories of
bottlenecks already integrated in speaking (Ferreira & Pashler, 2002) may also be useful to describe
how limitation in the ability to process information intervenes in writing. The challenge of writing
research will be to successfully integrate these cognitive limitations in future models of writing in
order to better reect the production strategies used by writers to circumvent the limits of the cogni-
tive system.
In sum, as writing is one of the most effortful cognitive activities, working memory is strongly
involved when composing a text. It is thus considered as a major constrain on the use of the writing
processes. Capacity or unitary conceptions of working memory in writing have helped us to better
understand how resources are shared between the various writing processes, and how these differ-
ent processes emerge during writing acquisition according to the demands they place on working
memory. Componential conceptions of working memory in writing, which assume domain-specic
registers for storing information, have claried the nature of the demands the writing processes
place on working memory. This latter approach has shown that composing a text not only requires
verbal processing but also visuospatial processing at a large extent. Future research on the role of
working memory in writing will have to focus particularly on the executive functions of working
memory since writing requires coordinating several processes.
REFERENCES
Allal, L. (2004). Integrating writing instruction and the development of revision skills. In L. Allal, L.Chanquoy,&
P.Largy (Eds.), Revision: Cognitive and instructional processes (pp. 139–156). Dordrecht, Netherlands:
Kluwer Academic Publishers.
Alves, R. A., Castro, S. L., & Olive, T. (2008). Execution and pauses in writing narratives: Processing time,
cognitive effort and typing skill. International Journal of Psychology, 43, 469–479.
Alves, R. A., Castro, S. L., Sousa, L., & Strömqvist, S. (2007). Typing skill and pause-execution cycles in writ-
ten composition. In M. Torrance, L. Van Waes, & D. Galbraith (Eds.), Writing and cognition: Research
and applications (pp. 55–66). Amsterdam: Elsevier.
Anderson, J. R. (1983). Retrieval of information from long-term memory. Science, 220(4592), 25–30.
Anderson, J. R., Reder, L., & Lebiere, C. (1996). Working memory: Activation limitations on retrieval.
Cognitive Psychology, 30, 221–256.
Baddeley, A. D. (1986). Working memory. Oxford: Oxford University Press.
Baddeley, A. D. (1996). Exploring the central executive. The Quarterly Journal of Experimental psychology,
49A(1), 5–28.
Y119829_C008.indd 136 20/08/11 1:06 AM
137Writing and Working Memory
Baddeley, A. D. (2000). The episodic buffer: A new component of working memory? Trends in Cognitive
Sciences, 4, 417–423.
Baddeley, A. D. (2007). Working memory, thought and action. Oxford: Oxford University Press.
Baddeley, A. D., & Hitch, G. (2001). Working memory in perspective: Foreword. In J. Andrade (Ed.), Working
memory in perspective (pp. xv–xix). Hove: Psychology Press.
Berninger, V. W. (1999). Coordinating transcription and text generation in working memory during composing:
Automatic and constructive processes. Learning Disability Quarterly, 22(2), 99–112.
Berninger, V. W., Cartwright, A., Yates, C., Swanson, H. L., & Abbott, R. D. (1994). Developmental skills
related to writing and reading acquisition in the intermediate grades: Shared and unique variance.
Reading and Writing: An Interdisciplinary Journal, 6, 161–196.
Berninger, V. W., & Swanson, H. L. (1994). Modifying Hayes and Flower’s model of skilled writing to
explain beginning and developing writing. In E. C. Buttereld (Ed.), Advances in cognition and edu-
cational practice, Vol. 2. Children’s writing: Toward a process theory of the development of skilled
writing (pp. 57–81). Greenwich, CT: JAI Press.
Berninger, V. W., Whitaker, D., Feng, Y., Swanson, H. L., & Abbott, R. D. (1996). Assessment of planning,
translating and revising in junior high writers. Journal of School Psychology, 34, 23–52.
Berninger, V. W., Yates, C., Cartwright, A., Rutberg, J., Remy E., & Abbott, R. D. (1992). Lower level develop-
mental skills in beginning writing. Reading and Writing: An Interdisciplinary Journal, 4, 257–280.
Bourdin, B., & Fayol, M. (1994). Is written language production more difcult than oral language production?
A working memory approach. International Journal of Psychology, 29, 591–620.
Bourdin, B., & Fayol, M. (2000). Is graphic activity cognitively costly? A developmental approach. Reading
and Writing: An Interdisciplinary Journal, 13, 183–196.
Bourdin, B., & Fayol, M. (2002). Even in adults, written production is still more costly than oral production.
International Journal of Psychology, 37, 219–227.
Brown, J. S., McDonald, J. L., Brown, T. L., & Carr, T. H. (1988). Adapting to processing demands in dis-
course production: The case of handwriting. Journal of Experimental Psychology: Human Perception
and Performance, 14, 45–59.
Cantor, J., & Engle, R. W. (1993). Working memory as long-term memory activation: An individual-differences
approach. Journal of Experimental Psychology: Learning, Memory, and Cogntion, 19(5), 1101–1114.
Chenoweth, N. A., & Hayes, J. R. (2003). The inner voice in writing. Written Communication, 20, 99–118.
Colombo, L., Fudio, S., & Mosna, G. (in press). Phonological and working memory mechanisms involved in
written spelling. European Journal of Cognitive Psychology.
Conway, A. R. A., & Engle, R. W. (1996). Individual differences in working memory: More evidence for a
general theory. Memory, 4, 577–590.
Cowan, N. (1993). Activation, attention and short-term memory. Memory & Cognition, 21(2), 162–167.
Cowan, N. (2005). Working memory capacity. London: Psychology Press.
Dédeyan, A., Olive, T., & Largy, P. (2006). Implication des composants de la mémoire de travail dans la détec-
tion des erreurs d’accord sujet-verbe: approche développementale. Paper presented at the International
Conference on Cognitive Approach of Written Language Learning. Rennes, France, October.
Donovan, C. A., & Smolkin, L. B. (2006). Children’s understanding of genre and writing development. In
C.A.MacArthur, S. Graham, & J. Fitzgerald (Eds.), Handbook of writing research (pp. 131–143).
New York, NY: Guilford Press.
Engle, R. W. (2002). Working memory capacity as executive attention. Current Directions in Psychological
Science, 11, 19–23.
Engle, R. W., Kane, M. J., & Tuholski, S. W. (1999). Individual differences in working memory capacity and
what they tell us about controlled attention, general uid intelligence and functions of the prefrontal
cortex. In A. Miyake, & P. Shah (Eds.) Models of working memory: Mechanisms of active maintenance
and executive control. London: Cambridge Press.
Ericsson, K. A., & Delaney, P. F. (1999). Long-term working memory as an alternative to capacity models
of working memory in everyday skilled performance. In A. Miyake & P. Shah (Eds.), Models of work-
ing memory: Mechanisms of active maintenance and executive control (pp. 257–297), Cambridge, UK:
Cambridge University Press.
Fayol, M. (1999). From on-line management problems to strategies in written composition. In M. Torrance &
G. Jeffery (Eds.), The cognitive demands of writing: Processing capacity and working memory effects in
text production (pp. 13–23). Amsterdam: Amsterdam University Press.
Fayol, M., Largy, P., & Lemaire, P. (1994). Cognitive overload and orthographic errors: A study in French writ-
ten language. The Quaterly Journal of Experimental Psychology, 47, 437–464.
AU: Can
you update
this? Please
give volume
number and
page range.
AU: Please
give English
translation
for article
title.
Y119829_C008.indd 137 20/08/11 1:06 AM
138 Writing: A Mosaic of New Perspectives
Ferreira, V., & Pashler, H. (2002). Central bottleneck inuences on the processing stages of word production.
Journal of Experimental Psychology: Human Learning and Memory, 28, 1187–1199.
Flower, L. S., & Hayes, J. R. (1980). The dynamics of composing: Making plans and juggling constraints.
In L. W. Gregg & E. R. Steinberg (Eds.), Cognitive processes in writing (pp. 31–49). Hillsdale, NJ:
Lawrence Erlbaum Associates.
Galbraith, D., Ford, S., Walker, G., & Ford, J. (2005). The contribution of different components of work-
ing memory to knowledge transformation during writing. L1-Educational Studies in Language and
Literature, 5(2), 113–145.
Galbraith, D., Hallam, J., Olive, T., & Le Bigot, N. (2008). The role of visual and spatial components of work-
ing Memory in planning. Paper at the 11th International Conference on Writing. Lund, Sweden, June.
Gathercole, S. E., & Baddeley, A. (1993). Working memory and language. Hillsdale, NJ: Lawrence Erlbaum
Associates.
Glynn, S. M., Britton, B. K., Muth, D., & Dogan, N. (1982). Writing and revising persuasive documents:
Cognitive demands. Journal of Educational Psychology, 74, 551–567.
Graham, S., & Weintraub, N. (1996). A review of handwriting research: Progress and prospects from 1980 to
1994. Educational Psychology Review, 8, 7–87.
Halliday, M. A. K. (1975). Learning how to mean: Explorations in the functions of language. London: Arnold.
Hayes, J. R. (1996). A new framework for understanding cognition and affect in writing. In C. M. Levy &
S. Ransdell (Eds.), The science of writing: Theories, methods, individual differences, and applications
(pp. 1–27). Mahwah, NJ: Lawrence Erlbaum Associates.
Hayes, J. R., & Chenoweth, N. A. (2003). The inner voice in writing. Written Communication, 23(2), 135–149.
Hayes, J. R., & Flower, L. S. (1980). Identifying the organization of writing processes. In L. W. Gregg &
E.R.Steinberg (Eds.), Cognitive processes in writing (pp. 3–30). Hillsdale, NJ: Lawrence Erlbaum.
Johnson-Laird, P. N. (1983). Mental models: Towards a cognitive science of language, inference and con-
sciousness. Cambridge, MA: Harvard University Press.
Just, M. A., & Carpenter, P. A. (1992). A capacity theory of comprehension: Individual differences in working
memory. Psychological Review, 99, 122–149.
Kellogg, R. T. (1987). Effects of topic knowledge on the allocation of processing time and cognitive effort to
writing processes. Memory & Cognition, 15, 256–266.
Kellogg, R. T. (1988). Attentional overload and writing performance: Effects of rough draft and outline strate-
gies. Journal of Experimental Psychology: Learning, Memory, & Cognition, 14, 355–365.
Kellogg, R. T. (1993). Observations on the psychology of thinking and writing. Composition Studies, 21, 3–41.
Kellogg, R. T. (1994). The psychology of writing. New York, NY: Oxford University Press.
Kellogg, R. T. (1996). A model of working memory in writing. In C. M. Levy & S. Ransdell (Eds.), The sci-
ence of writing: Theories, methods, individual differences, and application (pp. 57–71). Mahwah, NJ:
Lawrence Erlbaum Associates.
Kellogg, R. T. (1999). Components of working memory in writing. In M. Torrance & G. Jeffery (Eds.), The
cognitive demands of writing: Processing capacity and working memory effects in text production
(pp. 43–61). Amsterdam: Amsterdam University Press.
Kellogg, R. T. (2001a). Competition for working memory among writing processes. American Journal of
Psychology, 114, 175–191.
Kellogg, R. T. (2001b). Long-term working memory in text production. Memory & Cognition., 29 (1), 43–52.
Kellogg, R. T. (2004). Working memory components in written sentence generation. American Journal of
Psychology, 117, 341–361.
Kellogg, R. T. (2008). Training writing skills: A cognitive development perspective. Journal of Writing
Research, 1, 1–26.
Kellogg, R. T., & Mueller, S. (1993). Performance amplication and process restructuring in computer-based
writing. Journal of Man-Machine Studies, 39, 33–49.
Kellogg, R. T., Olive, T., & Piolat A. (2007). Verbal, visual and spatial working memory in written language
production. Acta Psychologica, 124, 382–397.
Kintsch, W. (1998). Comprehension: A paradigm for cognition. New York, NY: Cambridge University Press.
Largy, P., & Dédéyan, A. (2002). Automatisme en détection d’erreurs d’accord sujet-verbe: étude chez l’enfant
et l’adulte. L’Année Psychologique, 102, 201–234.
Larigauderie, P., Gaonac’h, D., & Lacroix, N. (1998). Working memory and error detection in texts: what are
the role of the central executive and the phonological loop? Applied Cognitive Psychology, 12, 5050–527.
Lea, J., & Levy, M. C. (1999). Working memory as a resource in the writing process. In M. Torrance &
G.Jeffery (Eds.), The cognitive demands of writing (pp. 63–82). Amsterdam: Amsterdam University
Press.
AU: Please
add in-text
citation or
delete from
the list here.
AU: Please
add in-text
citation or
delete from
the list here.
AU: Please
add in-text
citation or
delete from
list here;
give English
translation
for article
and journal
titles.
Y119829_C008.indd 138 20/08/11 1:06 AM
139Writing and Working Memory
Le Bigot, N., Passerault, J-M., & Olive, T. (2009). Memory for words location in writing. Psychological
Research, 73, 89–97.
Levy, C. M., White, K., Lea, J., & Ransdell, S. (1999). Contributions of the visuo-spatial sketchpad, phonologi-
cal loop and central executive to writing and recall. In E. Esperet & M.-F. Crété (Eds.), Proceedings of
the 1998 Writing Conference (pp. 41–48). Poitiers: MSHS, Université de Poitiers.
Madigan, R. J., Johnson, S. E., & Linton, P. W. (1994). Working memory capacity and the writing process.
Paper presented at the American Psychological Society, Washington. DC.
Marek, P., & Levy, C. M. (1999). Testing the role of the phonologocal loop in writing. In M. Torrance &
G.Jeffery (Eds.), Cognitive demands of writing. Amsterdam: Amsterdam University Press.
Martinie, M.-A., Olive, T., & Milland, L. (2010). Cognitive effort of dissonance induced by the writing of a
counterattitudinal essay. Journal of Experimental Social Psychology, 46, 587–594.
McCutchen, D. (1988). “Functional automaticity” in children’s writing: A problem of metacognitive control.
Written Communication, 5, 306–324.
McCutchen, D. (1994). The magical number three, plus or minus two: Working memory in writing. In
J.S.Carlson (series ed.) & E. C. Buttereld (vol. ed.), Advances in cognition and educational practice,
Vol. 2: Children’s writing: Toward a process theory of the development of skilled writing (pp. 1–30).
Greenwich, CT: JAI Press.
McCutchen, D. (1996). A capacity theory of writing: Working memory in composition. Educational Psychology
Review, 8, 299–325.
McCutchen, D. (2000). Knowledge, processing, and working memory: Implications for a theory of writing.
Educational Psychologist, 35, 13–23.
McCutchen, D., Francis, M., & Kerr, S. (1997). Revising for meaning: Effects of knowledge and strategy.
Journal of Educational Psychology, 89, 667–676.
Miyake, A., Friedman, N. P., Emerson, M. J., Witzki, A. H., Howerter, A., & Wager, T. (2000). The unity and
diversity of executive functions and their contributions to complex “frontal lobe” tasks: A latent variable
analysis. Cognitive Psychology, 41, 49–100.
Miyake, A., & Shah, P. (Eds.) (1999). Models of working memory: Mechanisms of active maintenance and
executive control. Cambridge: Cambridge University Press.
Navon, D., & Gopher, D. (1979). On the economy of human processing system. Psychological Review, 86,
214–255.
Navon, D., & Miller, J. (1987). Role of outcome conict in dual task interference. Journal of Experimental
Psychology: Human Perception and Performance, 13(3), 435–448.
Navon, D., & Miller, J. (2002). Queuing or sharing? A critical evaluation of the single-bottleneck notion.
Cognitive Psychology, 44, 193–251.
Olive, T. (2004). Working memory in writing: Empirical evidences from the dual-task technique. European
Psychologist, 9, 32–42.
Olive, T., Alves, R. A., & Castro, S. L. (2009). Cognitive processes in writing during pauses and execution
periods. European Journal of Cognitive Psychology, 21, 758–785.
Olive, T., Favart, M., Beauvais, C., & Beauvais, L. (2009). Children’s cognitive effort in writing: Effects
of genre and of handwriting automatisation in 5th- and 9th-graders. Learning & Instruction, 19,
299–308.
Olive, T., & Kellogg, R. T. (2002). Concurrent activation of high and low-level production processes in written
composition. Memory & Cognition, 30, 594–600.
Olive, T., Kellogg R. T., & Piolat, A. (2002). The triple-task technique for studying the process of writing. In
T. Olive & C. M. Levy (Eds.), Contemporary tools and techniques for studying writing (pp. 31–58).
Dordrecht: Kluwer Academic Publishers.
Olive, T., Kellogg, R. T., & Piolat, A. (2008). Verbal, visual and spatial working memory demands during text
composition. Applied Psycholinguistics, 29, 669–687.
Olive, T., Piolat, A., & Roussey J.-Y. (1997). Effort cognitif et mobilisation des processus en production de
texte: effet de l’habileté rédactionnelle et du niveau de connaissances. In D. Mellier & A. Vam Hoffe
(Eds). Attention et contrôle cognitif: Mécanismes, développement des habiletés, pathologies (pp. 71–85).
Rouen: Publications de l’Université de Rouen.
Passerault, J-M., & Dinet, J. (2000). The role of the visuo-spatial sketchpad in the written production of descrip-
tive and argumentative texts. Current Psychology Letters: Behaviour, Brain & Cognition, 3, 31–42.
Penningroth, S. L., & Rosenberg, S. (1995). Effects of a high information-processing load on the writing pro-
cess and the story written. Applied Psycholinguistics, 16, 189–210.
Piolat, A., & Olive, T. (2000). Comment étudier le coût et le déroulement de la rédaction de textes? La méthode
de triple-tâche: un bilan méthodologique. L’Année Psychologique, 100, 465–502.
AU: Please
add in-text
citation or
delete from
the list here.
AU: Please
give English
translation
for article
title, journal
title, and
publisher.
AU: Please
give English
translation
for article
and journal
titles.
Y119829_C008.indd 139 20/08/11 1:06 AM
140 Writing: A Mosaic of New Perspectives
Piolat, A., Olive, T., & Kellogg R. T. (2005). Cognitive effort of note taking. Applied Cognitive Psychology,
19(3), 291–312.
Piolat, A., Roussey, J.-Y., & Thunin, O. (1997). Effects of screen presentation on text reading and revising.
International Journal of Human-Computer Studies, 47, 565–589.
Rau, P. S., & Sebrechts, M. M. (1996). How initial plans mediate the expansion and resolution of options in
writing. The quarterly Journal of experimental Psychology, 49A(3), 616–638.
Roussey, J.-Y., & Piolat, A. (2008). Critical reading effort during text revision. European Journal of Cognitive
Psychology, 20, 765–792.
Salthouse, T. A. (1996). The processing-speed theory of adult age differences in cognition. Psychological
Review, 103(3), 403–428.
Scardamalia, M., & Bereiter, C. (1987). Knowledge telling and knowledge transforming in written composi-
tion. In S. Rosenberg (Ed.), Advances in psycholinguistics: Vol. 2. Reading, writing and language learn-
ing (pp. 143–175). Cambridge, UK: Cambridge University Press.
Shelton, J. R., & Caramazza, A. (1999). Decits in lexical and semantic processing: Implications for models of
normal language. Psychonomic Bulletin & Review, 6, 5–28.
Torrance, M., & Galbraith, D. (2006). The processing demands of writing. In C. A. MacArthur, S. Graham, &
J. Fitzgerald (Eds.), Handbook of writing research (pp. 67–82). New York, NY: Guilford Publications.
Unsworth, N., Heitz, R. P., & Engle, R. W. (2005). Working memory capacity in hot and cold cognition. In
R.W. Engle, G. Sedek, U. Hecker, & D. N. McIntosh (Eds.), Cognitive limitations in aging and psycho-
pathology (pp. 19–43). New York, NY: Cambridge University Press.
Y119829_C008.indd 140 20/08/11 1:06 AM
