Thierry Olive

Working Memory in Writing: Empirical Evidence From the Dual-Task Technique

The dual-task paradigm recently played a major role in understanding the role of working memory in writing. By reviewing recent findings in this field of research, this article highlights how the use of the dual-task technique allowed studying the processing and short-term storage functions of working memory involved in writing. With respect to processing functions of working memory (namely, attentional and executive functions), studies investigated resource allocation, step-by-step management, and parallel coordination of the writing processes. With respect to short-term storage in working memory, experiments mainly attempted to test Kelloggs (1996) proposals on the relationship between the writing processes and the slave systems of working memory. The dual-task technique proved fruitful in understanding the relationship between writing and working memory because researchers exploited its major advantage, namely, its flexibility.

The Triple Task Technique for Studying the Process of Writing

In this chapter, we first present two variants of a technique of secondary reaction time task and verbalization task that allow researchers (1) to estimate the general temporal organization of the writing process, (2) to analyse the recursiveness of writing and (3) to measure the amount of resources allocated to the writing processes (Kellogg, 1987b; Levy & Ransdell, 1994, 1995). Next, we present a series of experiments that evaluated the validity of the method. We then synthesize studies that used the triple task method to address questions concerning the way by which situation-specific or writer-specific factors affect functional characteristics of writing. We describe results from experiments that investigated the role of writers’ knowledge, type of text planning, writing medium and cognitive capacity on resources allocation to the writing processes and on their temporal organization. Finally, we delineate how the triple task technique can be varied to answer future research questions.

Verbal, visual, and spatial working memory demands during text composition

Two experiments examined whether text composition engages verbal, visual, and spatial working memory to different degrees. In Experiment 1, undergraduate students composed by longhand a persuasive text while performing a verbal, visual, or spatial concurrent task that was presented visually. In Experiment 2, participants performed a verbal or spatial concurrent task that was aurally presented. Writing performance was not disrupted differentially across the three tasks. Performance on all concurrent tasks showed fewer correct responses and longer RTs relative to single-task, baseline data. However, the demands on visual working memory were as high as those on verbal working memory, whereas demands on spatial working memory were minimal. The findings help to delineate the roles of the verbal, visual, and spatial working memory in written composition.

Execution and pauses in writing narratives: Processing time, cognitive effort and typing skill

At the behavioural level, the activity of a writer can be described as periods of typing separated by pauses. Although some studies have been concerned with the functions of pauses, few have investigated motor execution periods. Precise estimates of the distribution of writing processes, and their cognitive demands, across periods of typing and pauses are lacking. Furthermore, it is uncertain how typing skill affects these aspects of writing. We addressed these issues, selecting writers of low and high typing skill who performed dictation and composition tasks. The occurrences of writing processes were assessed through directed verbalization, and their cognitive demands were measured through interference in reaction times (IRT). Before writing a narrative, 34 undergraduates learned to categorize examples of introspective thoughts as different types of activities related to writing (planning, translating, or revising). Then, while writing, they responded to random auditory probes, and reported their ongoing activity according to the learned categories. Convergent with previous findings, translating was most often reported, and revising and planning had fewer occurrences. Translating was mostly activated during motor execution, whereas revising and planning were mainly activated during pauses. However, none of the writing processes can be characterized as being typical of pauses, since translating was activated to a similar extent as the other two processes. Regarding cognitive demands, revising is likely to be the most demanding process in narrative writing. Typing skill had an impact on IRTs of motor execution. The demands of execution were greater in the low than in the high typing skill group, but these greater demands did not affect the strategy of writing processes activation. Nevertheless, low typing skill had a detrimental impact on text quality.

Suppressing visual feedback in written composition: Effects on processing demands and coordination of the writing processes

�� The goal of this experiment was to investigate the role of visual feedback during written composition. Effects of suppression of visual feedback were analyzed both on processing demands and on on-line coordination of low-level execution processes and of high-level conceptual and linguistic processes. Writers composed a text and copied it either with or without visual feedback. Processing demands of the writing processes were evaluated with reaction times to secondary auditory probes, which were analyzed according to whether participants were handwriting (in a composing and a copying task) or engaged in high-level processes (when pausing in a composing task). Suppression of visual feedback increased reaction time interference (secondary reaction time minus baseline reaction time) during handwriting in the copying task and not during pauses in the composing task. This suggests that suppression of visual feedback only affected processing demands of execution processes and not those of high-level conceptual and linguistic processes. This is confirmed by analysis of the quality of the texts produced by participants, which were little, if at all, affected by the suppression of visual feedback. Results also indicate that the increase in processing demands of execution related to suppression of visual feedback affected on-line coordination of the writing processes. Indeed, when visual feedback was suppressed, reaction time interferences associated with handwriting were not reliably different in the copying task or the composing task but were significantly different when visual feedback was not suppressed: They were lower in the copying task than in the composition task. When visual feedback was suppressed, writers activated step-by-step execution processes and high-level writing processes, whereas they concurrently activated these writing processes when composing with visual feedback.

Cognitive processes in writing during pause and execution periods

The present study investigated how writing processes are activated during pause and execution periods. In two experiments, handwriting demands were manipulated by asking participants to compose with their familiar handwriting or with a high-demanding cursive uppercase calligraphy. Experiment 1 investigated narrative writing, a task with low planning demands. Experiment 2 addressed essay writing, a task with stronger planning demands. Occurrences of processes and their cognitive effort were analysed by asking participants to respond to random auditory probes and then to report their ongoing mental activity according to learned categories referring to the planning, translating, and revising writing processes. All together, the findings indicate that demands on planning did not affect how writing processes were activated during pauses and execution periods but automaticity of handwriting did. When handwriting was effortless, translating was mostly activated in parallel with motor execution, whereas revising and planning were mainly activated during pauses. However, none of the writing processes could be characterised as being typical of pauses, since translating was activated to a similar extent as the other two processes. By contrast, when handwriting was effortful, participants shifted to a more sequential functioning and activated translating mainly during pauses.

Processing Time and Cognitive Effort in Revision: Effects of Error Type and of Working Memory Capacity

Cognitive effort and temporal organization of the sub-processes of revision (reading to define problems, searching a solution and transforming the text) were investigated when writers revised either spelling, syntactical or coherence errors. Differences in working memory capacity of participants were also examined and cognitive effort of reading to understand and of reading to define problems was compared. Participants revised a text while being submitted to a secondary auditory probe task and to a directed retrospection task. Results showed that the type of errors did not affect the cognitive effort and temporal organization of the sub-processes of revisions. Efficiency of the revisers depended mainly on the type of errors and partially on their WM capacity. Finally reading to define problems appeared more effortful than reading to understand.

Memory for words location in writing.

In two experiments, we investigated memory for words location after writing a text. Experiment 1 demonstrated the existence of a memory for words location in writing by showing that participants who first composed a text and were then asked to locate words extracted from their text performed above a chance level established using a computer simulation, and better than participants who did not compose a text but were told the subject of the text. Experiment 2 showed that memory for words location in writing is mainly supported by a visuospatial representation of the text, as indicated by the lower recall of words location by participants who performed a visuospatial concurrent task at the time of the composition, compared with participants who performed a verbal concurrent task. The findings highlight the role of a spatial representation of the physical layout of the text and the role of such a memory in the writing process.

Lexical spelling in children and adolescents with specific language impairment: variations with the writing situation.

The goal of this study was to compare the lexical spelling performance of children and adolescents with specific language impairment (SLI) in two contrasting writing situations: a dictation of isolated words (a classic evaluative situation) and a narrative of a personal event (a communicative situation). Twenty-four children with SLI and 48 typically developing children participated in the study, split into two age groups: 7-11 and 12-18 years of age. Although participants with SLI made more spelling errors per word than typically developing participants of the same chronological age, there was a smaller difference between the two groups in the narratives than in the dictations. Two of the findings are particularly noteworthy: (1) Between 12 and 18 years of age, in communicative narration, the number of spelling errors of the SLI group was not different from that of the typically developing group. (2) In communicative narration, the participants with SLI did not make specific spelling errors (phonologically unacceptable), contrary to what was shown in the dictation. From an educational perspective or that of a remediation program, it must be stressed that the communicative narration provides children-and especially adolescents-with SLI an opportunity to demonstrate their improved lexical spelling abilities. Furthermore, the results encourage long-term lexical spelling education, as adolescents with SLI continue to show improvement between 12 and 18 years of age.

The Visuospatial Dimension of Writing

The authors suggest that writing should be conceived of not only as a verbal activity but also as a visuospatial activity, in which writers process and construct visuospatial mental representations. After briefly describing research on visuospatial cognition, they look at how cognitive researchers have investigated the visuospatial dimension of the mental representations and processes engaged in writing. First, they show how Hayes’s research integrated the visuospatial dimension of writing. Second, they describe how the written trace can serve as a visual resource. Third, they focus on the visuospatial processes involved in constructing an overall representation of the text and its physical layout. Finally, they review findings on the visuospatial demands that planning places on working memory. All the data and theories presented in this article support the idea that writing is indeed a visuospatial activity.