Writing, working memory and dual-TASK
implies that researchers be particularly attentive on
their implemen-tation. One well-documented source
of problem to be avoided relates to structural
interference (Kahneman, 1973; Kerr, 1973).
Namely, mechanisms engaged in perception of the
secondary task or in its responses must be distinct
from those engaged in the primary task. Regarding
writing, responses of writers to secondary stimuli
(e.g., probes) cannot involve the hand they use for
handwriting.
By contrast, much more authors supported the
use of the dual-task technique as long as some
methodological precautions are respected when
designing these experiments. For example, Fisk,
Derrick, and Schneider (1986-87) proposed three
criteria that may validate the interpretations of dual-
task data when respected. According to the first
criterion, the primary and secondary tasks must draw
from a single pool of resources so that a trade-off
between the two tasks can take place. The
community of resources assumption can be tested by
increasing the difficulty of the primary task. This
should lead to decreases of performance at the
secondary task. By contrast, if each task draws from
a specific pool of resources, then - the secondary
task should never be affected. The second criterion
suggested by Fisk et al. (1987-88) concerns the
intrusive character of the secondary task. When the
primary task engages large amounts of resources, the
addition of even a low-demanding secondary task
might exceed the cognitive capacity and subjects are
likely to carry out the secondary task with the
detriment of the primary task. If so, data inform on
the primary task only in this overloaded situation.
This issue can be addressed simply by verifying
whether or not performance at the primary task
differs in single- or dual-task condition. The third
criterion concerns the stability of resources devoted
to the secondary task throughout the experiment.
Indeed, changes in performance at the secondary task
resulting from a change in its demands can be
falsely allotted to changes in resources devoted to
the primary task. In other words, no incidental
learning must be observed at the secondary task. A
simple test of this criterion consists in examining
block effects in the experiment. In sum, when
carefully designed, the dual-task technique allows
psychologist to further answer very important
question about the processing limits of the cognitive
system and the role of working memory in
cognition.
Working memory
Working memory is aimed at explaining how
information is temporary stored and processed
during the realization of cognitive activities, and in
particular in complex cognitive activities such as
learning, language production and comprehension.
One of the most influential model of working
memory has been proposed by Baddeley and Hitch
(1974) and then revised by Baddeley (1986, 2000).
This theoretical framework represents a development
of earlier models of short-term storage (such as the
model of Atkinson & Shiffrin, 1968). According to
Baddeley and Hitch (1974), working memory
should be conceived as a multi-component system.
One attentional component, the central executive, is
aided by two peripheral and independent systems -
the phonological loop and the visuospatial
sketchpad - that temporary store verbal, and
visuospatial information. Evidence for the
separability of the subsystems of working memory
are now abundant. For instance, verbal concurrent
tasks affect short-term storage of verbal information
but not that of visuospatial information.
Conversely, maintenance of visuospatial information
is disrupted by concurrent visuospatial tasks but not
by verbal secondary tasks (Cocchini, Logie, Della
Sala, MacPherson & Baddeley, 2002). Brain-
imaging studies also indicates that different neural
areas underlie verbal and visuospatial short-term
storage (Smith & Jonides, 1997). Finally, studies
of brain-damaged patients and of genetic disorders
report selective loss of either verbal or visuospatial
short-term memory (Jarrold, Baddeley & Hewes,
1999).
The phonological loop is specialized for the
short-term storage and processing of verbal and
acoustic information. It is fractionated into a passive
and temporary phonological store and an active
rehearsal system. Because the information stored in
the phonological store decays with time, the
rehearsal system can refresh this information. This
fractionation of the phonological loop is supported
by neuropsychological findings and by
neuroimaging data (Baddeley, 2000). The
phonological loop gives account of several
phenomena, such as the articulatory suppression and
the phonological similarity effects. By contrast, the
sketchpad is assumed to hold visuospatial
information. Although the theoretical development
of this system lags behind that of the phonological
loop, and that there is less consensus about its
functions, it is generally accepted that the sketchpad
can be fractionated into two visual and spatial
components (Logie, 1995). The visual component
(the visual cache) is a passive system that stores
visual information and spatial locations in the form
of static visual representation. The spatial
component (the inner scribe) is an active spatial
rehearsal system that maintains sequential locations
and movements and that also serves to refresh
decaying information in the visual cache. Smith
and Jonides (1997) provided findings supporting the
fractionation of the sketchpad. In a PET study, they
shown that different neural circuits and distinct
cortical regions mediate the spatial and visual
working memory systems.
Compared with the phonological loop and the
visuospatial sketchpad, the central executive has
long been the less specified system of working
memory. Baddeley (1986) first borrowed the SAS
model of attentional control developed by Norman
and Shallice (1986) to explain how cognitive
activities are managed. The conceptualization of the