The Visuospatial Dimension of Writing

Abstract

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.

Keywords

writing visual and spatial processes text representation planning working memory

No one would argue with the fact that writing a text is a linguistic activity: Writers must turn their communicative intentions into clear, organized, and coherent language. However, this linguistic dimension should not be allowed to overshadow the fact that composing a text is also a visuospatial activity. Writing is a visual activity guided by the writer’s eyes; it is also a spatial activity in that it involves organizing the text by leaving a graphic trace on the page. Additionally, writing often consists in translating visual and spatial mental representations into text, and the written text may also be represented in a visuospatial format. For instance, researchers in the field of genetic criticism have found signs of visuospatial processing in the writing process. Genetic critics analyze the notes, sketches, drafts, manuscripts, typescripts, proofs, and correspondence (the avant-textes) of famous writers to understand their writing processes (Deppman, Ferrer, & Gordman, 2004). As reported in Olive, Passerault, Lebrave, and Le Bigot (2010), an example of visuospatial processing can be found in the genesis of the first page of Flaubert’s manuscript of Herodias, where the difficulty Flaubert encountered in describing a citadel stemmed from the conflict between the different mental images he had to process and the difficulty of converting these representations into a linear, verbal description (see also Grésillon, Lebrave, & Fuchs, 1991).

Studies of the visuospatial dimension of writing are surprisingly scarce, and the cognitive models of writing describe only fleetingly, if at all, the underlying mental representations and processes. As a consequence, writing research has almost completely ignored how writers manage and process the visuospatial dimension of writing. This is particularly astonishing given that early research on cognitive psychology pointed to the necessity of visuospatial representations and processes.

The idea that the human cognitive system cannot be understood solely in terms of verbal and propositional abstract features, as proposed by Pylyshyn (1981), emerged very early on in cognitive psychology, with several researchers providing evidence that cognition involves visuospatial representations with specific properties. For example, Shepard postulated that mental representations of objects behave like real objects. In a series of studies (e.g., Shepard & Cooper, 1982; Shepard & Metzler, 1971), participants were shown a pair of images comprising a reference image and a test image that had undergone a greater or lesser rotation. They were asked to decide whether the two images were identical except for their orientation. To respond, participants had to perform a mental rotation on the test image to compare it with its reference. As decision time was proportional to the angle of rotation, Shepard and his colleagues concluded that there was an isomorphism between the rotation of the real object and the mental rotation performed by participants, suggesting analogical encoding of the object. This analogy between real objects and their mental representations has also been evidenced in studies of the properties of mental maps encoding our spatial environment. Kosslyn (1975) found that the time taken to mentally scan a mental image depends on the “true distance” that has to be scanned (see also Kosslyn, Ball, & Reiser, 1978). Mental maps also reflect our beliefs about how the world is organized (Tversky, 1981). Thus, if one asks individuals to say whether Philadelphia or Rome is the more northerly city, most respond Philadelphia, even though the correct answer is Rome. This error reflects the fact that since Rome has a Mediterranean climate, we assume that it lies further south than Philadelphia. Thus, visuospatial representations are not built on visual or spatial perceptual input alone but also incorporate general knowledge. It is widely accepted that it is possible to construct a mental representation of a “visuospatial” environment without any visual input, simply by reading a text describing this environment (Kintsch, 1998).

Cognitive and neuropsychological research on working memory also distinguishes between verbal and visuospatial processes. For instance, in addition to a general attentional system (the central executive), Baddeley (1986, 2000) has proposed the existence of two specialized memory registers dedicated to the transient storage of specific information: a phonological loop ensuring the temporary retention of verbal information and a visuospatial sketchpad responsible for the storage of visual and spatial information. The strongest arguments in favor of this verbal-visuospatial dissociation come from the study of patients in neuropsychology (e.g., Farah, 1988). For example, some patients with a deficit in the temporary storage of verbal information have no visuospatial impairment and vice versa. Brain imaging has confirmed that different regions of the brain are involved in the short-term maintenance of verbal, visual, and spatial information (Smith & Jonides, 1997). In addition to neuropsychological dissociations, further arguments in favor of the visual-spatial distinction are based on the functional properties of visual and spatial storage. Logie (1995) argued that visuospatial working memory comprises not only a passive short-term visual storage component but also an active mechanism for the rehearsal of visual and spatial material. Finally, visual and spatial abilities emerge at different rates, with visual abilities developing more rapidly than spatial ones (Pearson & Logie, 1998). In sum, cognitive psychology has demonstrated not only the specificity of visuospatial processing but also the dissociation between visual and spatial processing, both functionally and structurally.

In this context, the present article examines how cognitive research on writing has described and explained the visuospatial dimension of the mental representations and processes that are engaged when composing a text. After briefly reviewing how Hayes integrated the visuospatial dimension of writing in his own research, we describe how the written trace can serve as a visual resource. Finally, we extend Hayes’s groundbreaking work by examining findings on the role of visuospatial processing in the construction of an overall representation of a text and its physical layout and on the visuospatial demands that planning places on working memory.

Hayes’s Contribution to Research on the Visuospatial Dimension of Writing

Cognitive research on writing has been greatly inspired by Hayes and Flower’s (1980) model, which was the first to describe the knowledge and cognitive processes involved in writing. Although the visuospatial dimension of writing was never explicitly addressed in this model, Hayes and Flower nevertheless identified at least one constraint—the text produced so far—liable to engage visuospatial processes (see below).

Hayes started to investigate the visuospatial dimension of writing a few years after the publication of the Hayes and Flower model, by analyzing the determinants of paragraphing (Bond & Hayes, 1984). In this study, Bond and Hayes suggested that experienced writers base their paragraphing on both the linguistic and spatial characteristics of their texts. They asked readers to paragraph a text displayed without any line breaks. In some versions of the text, content words and pronouns were replaced by sequences of the letter X, while other versions were made up exclusively of Xs, with lines representing the sentences. Paragraphing was closest to the original text when the version contained words but remained fairly close when Xs replaced the content words and pronouns and even when they formed a graphic version of the text. Results of a simulation in which paragraphing depended, first, on identifying a semantic break through the presence or absence of a formal cue (e.g., a lexical change) and, second, on paragraph length closely mirrored the paragraphing performed by the participants. Bond and Hayes therefore suggested that paragraphing depends on textual cues, such as the repetition of content words or the presence of pronouns marking a thematic change, but also on the spatial characteristics of the text, such as paragraph length.

Another line of research that led Hayes to consider the visuospatial dimension of writing was the role of the writing medium. Text is produced within a limited space (a screen or a sheet of paper), and formats and materials can vary, depending on the writing medium. For example, when a word processor is used, the size of the writing space—the computer screen—may affect the writing process, the writer’s strategy, and the mental representation of the text. Haas and Hayes (1986) suggested that reading and writing on a small screen poses at least three types of problems: locating information, proofreading, and constituting an overall representation of the text (“sense of the text”). With small screens and some types of word processors, writers have access to only a restricted area of the document and are sometimes unable to visualize the edges of the page. This difficulty may explain differences in writing revision behavior. For instance, Piolat, Roussey, and Thunin (1997) showed that a text presentation mode that does not allow the page boundaries to be visualized results in less efficient revision, with fewer errors identified, and poorer recall of the text. Additionally, idea structuring may be more difficult with word processors, as writers cannot insert short notes or create graphic effects for associating ideas. By contrast, tools allowing writers to order their ideas and shuffle sections of the text around may give them better control of the arrangement of their ideas on the page, presumably resulting in a text with more highly organized textual sequences.

By identifying reading as a fundamental process of writing, Hayes (1996) further emphasized the visuospatial dimension of writing. According to his view, reading allows writers to create three types of representations: where possible, a representation of the source text, then a representation of its author, and finally a representation of its (physical) spatial structure. It is widely acknowledged that readers incidentally encode a text’s spatial attributes and construct a spatial representation of it (Kennedy, 1992). For example, long regressive ocular saccades beyond the visual perceptual span are performed with considerable precision, as if the reader had developed a representation of the text that made it easier to control and guide these long saccades (Frazier & Rayner, 1982). Other evidence in favor of a representation of the physical characteristics of the text comes from studies of the impact of text presentation format on reading and comprehension (Baccino & Pynte, 1994; Kennedy & Murray, 1987) or on memory for word location (for a review, see Le Bigot, Passerault, & Olive, 2010). According to Hayes (1996), writers use their visuospatial representations to find and retrieve information in the text.

Finally, in a new model of writing, Hayes (1996) explicitly introduced the visuospatial dimension of text production in the form of the visuospatial sketchpad, first described in Baddeley’s (1986) model of working memory. Unfortunately, the role of this component of working memory during the writing process is scarcely described. Kellogg’s (1996) proposals compensate for this lack by describing the relationships between each component of working memory and the different writing processes (see below).

To conclude this section, two lines of research that have so far been neglected in writing research can be traced back to Hayes’s preliminary studies of the visuospatial dimension of writing. On one hand, it seems important to investigate writers’ visuospatial representations of the spatial layout of their texts. On the other, one issue that appears important for improving our understanding of the visuospatial dimension of writing is the role of visuospatial working memory during writing. Composing a text places equally heavy demands on verbal and visual working memory (Olive, Kellogg, & Piolat, 2008). However, before addressing these issues, we will examine one of the most obvious visual features of writing, namely the written trace.

Visual Processing of the Written Trace

In comparison with speaking, one important characteristic of writing is the permanence of the written trace. The text that has already been produced normally remains visible to the writer (active window of the word processor, current sheet of paper) throughout the writing process, although it may sometimes be partially or even wholly absent if, for example, the screen is particularly small (e.g., smartphone, tablet). Therefore, the writer may encounter difficulties while reading or reviewing if the text produced so far has disappeared.

The written trace is generally assumed to affect the cognitive demands of writing. Two hypotheses have been advanced to explain which writing processes rely on the written trace. The first hypothesis suggests that it helps the writer to control the transcription of the text and thus reduces the cost of graphomotor processes. The second hypothesis suggests that it serves as an external memory supporting the high-level writing processes (planning, linguistic formulation, revision).

Findings supporting the hypothesis of reduced transcription demands come primarily from research on handwriting. More specifically, the written trace clears the motor programs that have already been executed from working memory. The main function of visual feedback is to control text transcription. For example, in children learning to write, the more the visual feedback is degraded, the more the quality of their handwriting deteriorates (Zesiger, 1995). In adult writers, the same phenomenon is observed: Writers omit or add superfluous features to their handwriting and produce less well-formed letters as a result (Smyth & Silvers, 1987; Van Doorn & Keuss, 1992, 1993; Van Galen, Smyth, Meulenbroek, & Hylkema, 1989), thus exhibiting the same symptoms as patients with spatial dysgraphia (Ellis, Young, & Flude, 1987). The written trace therefore appears to provide spatial guidance for the execution of the movement sequences required for the graphomotor transcription of the text.

The second hypothesis—reduced demands of the high-level writing processes—is supported by research on text composition, which argues that the written trace acts as an “external auxiliary memory.” Thus, the text that has already been transcribed does not need to be memorized by the writer. Writers can handle more blocks of information than speakers can, which may explain the superiority of writing over speaking when recalling information (Grabowski, 2007). Studies that have analyzed the effect of removing the written trace on text quality provide data consistent with this interpretation. Thus, the suppression of the written trace has been shown to influence syntactic structure (Piolat, 1983), coherence (Atwell, 1981), and overall text quality (Hull & Smith, 1983). The effects on content are less clear-cut. Although these findings support the idea that the written trace facilitates the high-level writing processes by reducing their working memory demands, some studies have failed to find any effect of the suppression of the written trace on text quality (Gould, 1980; Teleman, 1981).

To disentangle these two hypotheses, Olive and Piolat (2002) evaluated the impact of suppressing the written trace on transcription and the working memory demands of the high-level writing processes. One group composed a text in a standard situation (with the written trace), and a second group composed a text with no written trace. The demands of the writing processes were assessed using a concurrent reaction-time task. Furthermore, the writers had to perform several writing assignments that allowed the researchers to distinguish between transcription and the high-level writing processes. This experiment showed that when the written trace is not visible, only the demands of transcription increase. In sum, the written trace supports the visuospatial guidance of transcription. The fact that suppression of the written trace did not affect the demands of the high-level writing processes suggests that while the visual presence of the text may support these processes, it is not essential to them because writers presumably have constructed a mental representation of their text.

A Visuospatial Representation of the Text

The written trace is helpful to writers not only because it is an auxiliary memory but also because it allows them to review those parts of the text that may be useful or even necessary for the continuity or quality of the writing. When we need to find information in the text produced so far (e.g., for editing), we do not generally need to read the entire text: Most of the time, we know where this information is located. But on which cognitive process is this ability to locate an idea or a word based?

Memory for the location of a word or piece of information may be based on a representation of textual content, but it may also rely on a visuospatial representation of the text. As Hayes (1996) has suggested, writers may mentally represent the spatial layout of their text by constructing a representation of their text that is “analogous” to the visual percept provided by the written trace and which therefore has structural similarities to the “real” text, such as the arrangement of the words on the page. Furthermore, memory for word location in reading is largely dependent on the text’s visuospatial characteristics. For example, Zechmeister and collaborators found poorer localization performances for items in the lower part of the page (Zechmeister & McKillip, 1972; Zechmeister, McKillip, Pasko, & Bespalec, 1975). Furthermore, several studies have highlighted the impact of different types of text presentation on location recall for a previously read text (Lovelace & Southall, 1983; O’Hara, Sellen, & Bentley, 1999). These findings can be interpreted as reflecting a gradual loading of a page-based location-tagging mechanism (Fischer, 2000). Thus, in reading, memory for word location seems to be based on a visuospatial cue (i.e., the page).

To test the hypothesis that writers construct visuospatial representations of their texts, Le Bigot et al. (2009, 2011, in press) investigated memory for word location in writing. First, Le Bigot et al. (2009, Exp. 2) observed that writers who performed a visuospatial concurrent task while writing their texts had poorer subsequent recall of word location than participants who performed a verbal concurrent task. Second, Le Bigot et al. (in press) showed that participants who performed a spatial task concurrently with text composition recalled fewer word locations than participants who performed a visual concurrent task. Finally, Le Bigot et al. (2011) observed that participants who were able to construct both temporal and visuospatial representations of their text had better memory for word location than participants who had to rely on temporal information alone (as in text dictation). This again indicated that visuospatial representations of texts are used to locate words. Consequently, Le Bigot et al. argued that, like readers, writers construct visuospatial representations of their texts which they can then access to retrieve word locations.

There is further evidence suggesting that writers construct visuospatial representations of their texts. Piolat et al. (1997) examined the effects of two types of text presentation on participants’ text revision and found that revision quality was better with a page-by-page presentation than with a scrolling presentation. Furthermore, Haas and Hayes (1986) compared the effects of computer screen versus hardcopy presentation when reordering a scrambled text. They observed that reordering was faster in the hardcopy condition than in the computer screen one.

All these findings indicate that when writers compose texts, they not only encode their semantic, verbal, and temporal dimensions but also automatically encode their spatial attributes (Logan, 1998). Depending on the tasks that writers are asked to perform, these dimensions will play greater or lesser roles. For instance, although word location appears to depend mainly on the visuospatial representation of the text, the verbal, visuospatial, and temporal dimensions probably all interact (Le Bigot et al., 2010).

More recently, Olive (2011) has suggested that mental representations of written texts may be multidimensional. A multidimensional representation would integrate the text’s situation (Kintsch, 1998) or mental model (Johnson-Laird, 1983), a semantic level representing its content and organization (its macro- and microstructure), and a verbal representation (the text base), as reflected in research on text recall (Kellogg, 2001b). In addition, a representation may also be constructed in episodic memory, to allow the writer to recall when and where the information was inserted in the text, particularly in the case of lengthy text composition spread over several writing sessions. Moreover, a visuospatial level of representation is necessary for the precise location of information in the text during the writing process.

The integration of these multiple text representations into a single, multidimensional one may be carried out by the episodic buffer of working memory, the new component that Baddeley (2000) added to his model of working memory to account for the ability of working memory to integrate information from multiple sources and to draw on long-term knowledge (Baddeley, Allen, & Hitch, 2010). The episodic buffer is a limited-capacity component that constitutes an interface between working memory and long-term memory and integrates information in different formats (e.g., verbal and visual information) coming either from the other slave systems or from long-term memory. The episodic buffer means that not only verbal and visuospatial information but also abstract or conceptual information can be manipulated in working memory. In sum, the episodic buffer provides a mechanism for binding different pieces of information in different formats into a single chunk.

Integrating the multiple codes that are processed during writing into a unitary multidimensional representation of the text, in addition to processing and storing the transient representations manipulated by the writing processes, is likely to exceed the capacity of working memory. Olive (2011) has suggested that long-term working memory may provide a convenient and elegant solution for circumventing these extreme demands (Ericsson & Kintsch, 1995). Updating the multidimensional representation of the text would thus take place in the episodic buffer, but this representation would be stored and maintained in an activated state in long-term memory. Information contained in the multidimensional representation of the text would be accessed through cues activated in short-term working memory (i.e., in the episodic buffer). This direct access to long-term memory would reduce processing demands (for more details on the functioning of long-term working memory, see Ericsson & Kintsch, 1995; for its possible role in writing, see Kellogg, 2001a, and McCutchen, 2000).

Visuospatial Processing During Planning

Hayes’s (1996) statement about the need to incorporate visuospatial working memory into models of writing echoed Kellogg’s views (1996, 1999). In his model, Kellogg describes the relationship between the writing processes and the multiple components of working memory, especially in terms of verbal and visuospatial short-term storage demands. Specifically, he proposes that visuospatial working memory is engaged when writers select and organize information for inclusion in their texts. Writers can search for ideas in their environment (e.g., while reading). They can process and integrate figurative drawings or pictures. They can also retrieve concrete conceptual knowledge from their long-term memory. The visual component of working memory is presumably involved in all these cases. By contrast, when they come to organize this information, they may represent it in the form of spatial structures either mentally or on paper. They may also use diagrams, plans, trees, and so on. Thus, in all probability, structuring a text engages the spatial component of working memory. In sum, retrieving and structuring the content of a text may well engage visuospatial working memory.

Retrieving and processing concrete content

The nature of the writing theme and the information that is processed has a direct impact on visuospatial processes. Describing a monument or writing about a friendly relationship involves different types of information of different natures: The first theme obviously engages visuospatial processes in addition to the verbal and abstract processes required when composing a text based on the second theme.

In this context, Passerault and Dinet (2000) argued that writing a descriptive text involves more mental imagery than writing an argumentative essay. They therefore hypothesized that overloading visuospatial working memory affects the production of descriptive texts more than that of argumentative texts. Accordingly, participants had to produce a text while memorizing various complex geometric figures. The memorization of geometric figures was expected to interfere with the processing of figurative content for the descriptive text but not with that of the more abstract content required for the argumentative essay. Passerault and Dinet duly observed that a visuospatial interfering task affected writers’ fluency when they composed the descriptive text. This confirms the involvement of the visuospatial component of working memory when composing a descriptive text.

Kellogg, Olive, and Piolat (2007) also tested the hypothesis that content planning requires visuospatial working memory when the writer has to deal with figurative or concrete content. They predicted that interference from a verbal task would be observed in the production of definitions of both concrete and abstract words but that interference from a visual task would be observed only when writers produced the definitions of concrete words (Kellogg, 1996). To test this prediction, they asked participants to write definitions of abstract or concrete words. In their first experiment, participants produced their definitions while performing a concurrent task that involved either verbal or visual working memory (e.g., detecting changes in syllables or shapes between trials). Results confirmed Kellogg’s hypothesis about the role of visual working memory in content processing. Moreover, in their second experiment, the authors failed to observe interference from a spatial task, suggesting that the processing of concrete content is based solely on visual working memory, specifically on the short-term storage functions of the “visual cache” (Logie, 1995). By contrast, using the same definition task, Raulerson, Donovan, Whiteford, and Kellogg (2010) found that sentence production also placed demands on spatial working memory. However, they interpreted this result in terms of handwriting demands rather than planning demands.

Content Structuring

The involvement of visuospatial processes also depends on the strategy used by the writer to structure the text. For example, we can assume that visuospatial processing is involved differently depending on whether writers directly formulate their texts or begin by producing nonlinear drafts (e.g., a summary or more generally an ordering of information). The main characteristic of these types of drafts or plans is that they require the spatial distribution of information on the sheet of paper to mark the semantic breaks. This spatial arrangement of information on the paper allows writers to build a conceptual map of the text content (Novack & Wandersee, 1991). In this context, Kellogg (1999) suggested that spatial working memory is required when writers organize the content of their texts. It is at this stage that writers can use visuospatial planning tools (e.g., charts) to order their information.

Galbraith, Ford, Walker, and Ford (2005) tested this hypothesis by asking writers to produce a text in three distinct phases. During the first phase, writers had to search for ideas by noting them briefly. During the second, structuring phase, writers were asked to arrange the ideas they had previously noted in the form of a plan. Specifically, they had to write an outline. In the third and final phase, writers were asked to generate their texts by precisely formulating them. To test the visual and spatial demands of outlining, writers were administered several concurrent tasks, including a visual task (projection of visual noise onto the background of the computer monitor) and a spatial one (tracking task). The results of this experiment confirmed the role of spatial working memory when structuring a text, for when the authors compared the three production phases, they found that the spatial secondary 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 Galbraith, Hallam, Olive, & Le Bigot, 2009).

Additionally, Olive, Crouzevialle, Le Bigot, and Galbraith (2010) found that structuring a text with a diagram requires more spatial working memory and throws up more new ideas than outlining does. In their study, participants had to compose a text first by retrieving content, second by structuring that content, and finally by formulating the text. During the structuring phase, writers were asked to use either a linear structure or a diagram, while performing either a visual or a spatial task. As expected, the spatial task interfered more than the visual task, particularly when writers structured the content of their texts in diagram form. Taken together, these studies confirm the engagement of spatial processing during content structuring.

To conclude this section, it is important to stress that writing processes other than planning may also require visuospatial working memory (Ransdell, Levy, & Kellogg, 2002). For example, although handwriting is not thought to rely on visuospatial working memory in experienced writers, it may require visuospatial processes in beginning writers, who have to control their graphomotor execution. Error detection may also engage visuospatial working memory in adult writers, as suggested by Chenoweth and Hayes (2003), who observed that error detection does not totally cease when the text is no longer visible. As already indicated, reading the text produced so far may also require visuospatial processing to construct the visuospatial representation of the overall text (for general reviews on working memory and writing, see Olive 2004, 2011, 2012).

Conclusion

All the data and theories presented in this article strongly support the idea that writing is indeed a visuospatial activity. Further research is, however, needed to foster our understanding of the visuospatial dimension of writing. The visual-spatial distinction would appear to be a fruitful one in this context, as it has already helped researchers to test specific hypotheses about planning demands or memory for word location. Moreover, a systematic investigation of the visuospatial demands of the different writing processes is particularly important: In addition to planning, translating and revising may also rely on visuospatial processes. For example, writers may engage in visuospatial processing when reading through their texts to review them or check their spelling. Detecting lexical or spelling errors engages visuospatial working memory when skilled writers scan the surface of their texts (Larigauderie, Gaonac’h, & Lacroix, 1998). Further research is also needed to tease apart the differential contributions of the visuospatial working memory processes related to spelling processes.

It is also important to extend research on the visuospatial dimension of writing by adopting a developmental approach. We would expect novice writers to use different strategies than skilled writers. For example, Dédéyan, Largy, and Olive (2006) showed that, in novice writers, detecting errors in texts mainly engages verbal working memory, whereas skilled writers use a visual search procedure to identify the surface features of morphological agreement in their texts (Dédéyan, Largy, & Negro, 2006).

Writing research also needs to explore how, and to what extent, the visuospatial resources of working memory are shared between the different visuospatial cognitive operations that are required in text composition (e.g., when retrieving content from long-term memory, constructing a mental representation of the whole text, reading and revising the text). Are the visuospatial resources available in working memory sufficient, or do writers need to manage the limited capacity of visuospatial working memory? Finally, as text composition requires both verbal and visuospatial working memory, do these resources compete during composition? Some initial answers can be found in the componential or capacity nature of the different working memory models, as well as in models of human cognitive architecture. For example, Paivio (1986) argues in favor of the systematic dual coding of information, in which independent verbal and visuospatial encoding of information cannot be avoided (for an application to writing, see Sadowski & Paivio, 2001).

In more practical terms, the emergence of new writing media in varying sizes and forms (computers, tablets, smartphones, electronic paper), and with varying text input systems (classic keyboards, electronic pens, smartphones, speech recognition), may affect the visuospatial processes involved in writing. Further research on writing tools is therefore needed to improve our understanding of their impact on writing processes and more particularly on the visuospatial dimension of writing.

Footnotes

Bios

Thierry Olive, a former lecturer at the University of Paris 8, has worked as a research scientist at the French National Center for Scientific Research since 2002. He is a member of the Center for Research on Cognition and Learning, a research unit run by the University of Poitiers and the National Center for Scientific Research. His work explores the cognitive, developmental, and emotional aspects of writing, with a special interest in the role of working memory in writing and the dynamics of the writing processes.

Jean-Michel Passerault has been professor of cognitive psychology at the University of Poitiers since 1992. He is a member of the Center for Research on Cognition and Learning, a research unit run by the University of Poitiers and the National Center for Scientific Research. He conducts research on writing (writing acquisition, working memory, and writing).

References

Atwell

(1981). The evolution of text: The interrelation of reading and writing in the composing process. Unpublished doctoral dissertation, Indiana University, Bloomington.

Baccino

Pynte

(1994). Spatial coding and discourse models during text reading. Language and Cognitive Processes, 9, 143-155.

Baddeley

A. D.

(1986). Working memory. Oxford: Oxford University Press.

Baddeley

A. D.

(2000). The episodic buffer: A new component of working memory? Trends in Cognitive Sciences, 4, 417-423.

Baddeley

Allen

R. J.

Hitch

G. J.

(2010). Investigating the episodic buffer. Psychologica Belgica, 50, 223-243.

Bond

S. J.

Hayes

J. R.

(1984). Cues people use to paragraph text. Research in the Teaching of English, 18, 147-167.

Chenoweth

N. A.

Hayes

J. R.

(2003). The inner voice in writing. Written Communication, 20, 99-118.

Dédéyan

Largy

Negro

(2006). Mémoire de travail et détection d’erreurs d’accord verbal: Etude chez le novice et l’expert [Working memory and detection of verbal agreement errors: A study in novice and skilled writers]. Langages, 164, 57-70.

Dédéyan

Olive

Largy

(2006, October). Implication des composants de la mémoire de travail dans la détection des erreurs d’accord sujet-verbe: Approche développementale [Involvement of working memory in detecting subject-verb agreement errors: Developmental approach]. Paper presented at the International Symposium on Cognitive Approaches to Learning Written Language, Rennes, France.

10.

Deppman

Ferrer

Groden

(Eds.). (2004). Genetic criticism. Philadelphia: University of Philadelphia Press.

11.

Ellis

A. W.

Young

A. W.

Flude

B. M.

(1987). “Afferent” dysgraphia in a patient and in normal subjects. Language and Cognitive Processes, 3(2), 99-127.

12.

Ericsson

K. A.

Kintsch

(1995). Long-term working memory. Psychological Review, 102, 211-245.

13.

Farah

M. J.

(1988). Is visual imagery really visual? Overlooked evidence from neuropsychology. Psychological Review, 95, 307-317.

14.

Fischer

M. H.

(2000). Perceiving spatial attributes of print. In Kennedy

Raddach

Heller

Pynte

(Eds.), Reading as a perceptual process (pp. 89-117). New York: Elsevier.

15.

Frazier

Rayner

(1982). Making and correcting errors during sentence comprehension: Eye movements in the analysis of structurally ambiguous sentences. Cognitive Psychology, 14, 178-210.

16.

Galbraith

Ford

Walker

Ford

(2005). The contribution of different components of working memory to knowledge transformation during writing. L1: Educational Studies in Language and Literature, 5(2), 113-145.

17.

Galbraith

Hallam

Olive

Le Bigot

(2009). The role of different components of working memory in writing. In Taatgen

van Riijk

(Eds.), CogSci 2009 proceedings (pp. 3028-3033). Austin, TX: Cognitive Science Society.

18.

Gould

J. D.

(1980). Experiments on composing letters: Some facts, some myths, and some observations. In Gregg

L. W.

Steinberg

E. R.

(Eds.), Cognitive processes in writing (pp. 98-127). Hillsdale, NJ: Erlbaum.

19.

Grabowski

(2007). The writing superiority effect in the verbal recall of knowledge: Sources and determinants. In Galbraith

Torrance

(Ed.), Writing and cognition (pp. 165-179). Bingley, UK: Emerald Group.

20.

Grésillon

Lebrave

J.-L.

Fuchs

(1991). Flaubert: “Ruminer Hérodias.” Du cognitif-visuel au verbal-textuel [Flaubert: “Ruminating Herodias.” From cognitive-visual to verbal-textual]. In Ferrer

Lebrave

J.-L.

(Eds.), L’Écriture et ses doubles: Genèse et variation textuelle (pp. 27-109). Paris: Editions du CNRS.

21.

Haas

Hayes

J. R.

(1986). What did I just say? Reading problems in writing with the machine. Research in the Teaching of English, 20, 22-35.

22.

Hayes

J. R.

(1996). A new framework for understanding cognition and affect in writing. In Levy

C. M.

Ransdell

S. E.

(Eds.), The science of writing: Theories, methods, individual differences and applications (pp. 1-27). Mahwah, NJ: Erlbaum.

23.

Hayes

J. R.

Flower

L. S.

(1980). Identifying the organization of writing process. In Gregg

L. W.

Steinberg

E. R.

(Eds.), Cognitive processes in writing (pp. 3-30). Hillsdale, NJ: Erlbaum.

24.

Hull

G. A.

Smith

W. L.

(1983). Interrupting visual feedback in writing. Perceptual and Motor Skills, 57, 963-978.

25.

Johnson-Laird

P. N.

(1983). Mental models: Towards a cognitive science of language, inference, and consciousness. Cambridge: Cambridge University Press.

26.

Kellogg

R. T.

(1996). A model of working memory in writing In Levy

C. M.

Ransdell

S. E.

(Eds.), The science of writing: Theories, methods, individual differences and applications (pp. 57-71). Mahwah, NJ: Erlbaum.

27.

Kellogg

R. T.

(1999). Components of working memory in writing. In Torrance

Jeffery

(Eds.), The cognitive demands of writing: Processing capacity and working memory effects in text production (pp. 43-61). Amsterdam: Amsterdam University Press.

28.

Kellogg

R. T.

(2001a). Long-term working memory in text production. Memory & Cognition, 29, 43-52.

29.

Kellogg

R. T.

(2001b). Presentation modality and mode of recall in verbal false memory. Journal of Experimental Psychology: Learning, Memory, and Cognition, 27, 913-919.

30.

Kellogg

R. T.

Olive

Piolat

(2007). Verbal, visual and spatial working memory in written language production. Acta Psychologica, 124, 382-397.

31.

Kennedy

(1992). The spatial coding hypothesis. In Rayner

(Ed.), Eye movements and visual cognition (pp. 379-397). New York: Springer-Verlag.

32.

Kennedy

Murray

W. S.

(1987). Spatial coding and reading: Some comments on Monk (1985). Quarterly Journal of Experimental Psychology, 39A, 649-671.

33.

Kintsch

(1998). Comprehension: A paradigm for cognition. New York: Cambridge University Press.

34.

Kosslyn

S. M.

(1975). Information representation in visual images. Cognitive Psychology, 7, 341-371.

35.

Kosslyn

S. M.

Ball

T. M.

Reiser

B. J.

(1978). Visual images preserve metric spatial information: Evidence from studies of image scanning. Journal of Experimental Psychology: Human Perception and Performance, 4(1), 47-60.

36.

Larigauderie

Gaonac’h

Lacroix

(1998). Working memory and error detection in texts: What are the roles of the central executive and the phonological loop? Applied Cognitive Psychology, 12, 505-527.

37.

Le Bigot

Passerault

J.-M.

Olive

(2009). Memory for word location in writing. Psychological Research, 73, 89-97.

38.

Le Bigot

Passerault

J.-M.

Olive

(2010). Le souvenir de la localisation des mots d’un texte. L’Année Psychologique, 110, 321-346.

39.

Le Bigot

Passerault

J.-M.

Olive

(2011). Implication of temporal and visuospatial information in memory for word location. Journal of Cognitive Psychology, 23(4), 522-530.

40.

Le Bigot

Passerault

J.-M.

Olive

(in press). Visuospatial processing in memory for word location in writing. Experimental Psychology.

41.

Logan

G. D.

(1998). What is learned during automatization? II: Obligatory encoding of spatial location. Journal of Experimental Psychology: Human Perception and Performance, 24, 1720-1736.

42.

Logie

R. H.

(1995). Visuo-spatial working memory. Hove, UK: Erlbaum.

43.

Lovelace

E. A.

Southall

S. D.

(1983). Memory for words in prose and their locations on the page. Memory & Cognition, 11, 429-434.

44.

McCutchen

(2000). Knowledge, processing, and working memory: Implications for a theory of writing. Educational Psychologist, 35, 13-23.

45.

Novak

J. D.

Wandersee

(Eds.) (1991). Concept mapping. Journal of Research in Science Teaching, 28, 10.

46.

O’Hara

Sellen

Bentley

(1999). Supporting memory for spatial location while reading from small displays. In Proceedings of the Conference on Human Factors in Computing Systems (pp. 220-221). New York: ACM.

47.

Olive

(2004). Working memory in writing: Empirical evidence from the dual-task technique. European Psychologist, 9, 32-42.

48.

Olive

(2011). Working memory in writing. In Berninger

V. W.

(Ed.), Past, present, and future contributions of cognitive writing research to cognitive psychology (pp. 485-503). New York: Psychology Press.

49.

Olive

(2012). Writing and working memory: A summary of theories and of findings. In Grigorenko

Mambrino

Preiss

(Eds.), Writing: A mosaic of new perspectives (pp. 125-140). New York: Psychology Press.

50.

Olive

Crouzevialle

Le Bigot

Galbraith

(2010, September). Visuospatial working memory in planning. Paper presented at the 12th International Conference on Writing (EARLI). Heidelberg, Germany.

51.

Olive

Kellogg

R. T.

Piolat

(2008). Verbal, visual and spatial demands during text composition. Applied Psycholinguistics, 29, 669-687.

52.

Olive

Passerault

J.-M.

Lebrave

J.-L.

Le Bigot

(2010). La dimension visuo-spatiale de la production de textes: Approche de psychologie cognitive et de critique génétique [The visuospatial dimension of text production: Cognitive psychology and genetic criticism approach]. Langages, 177, 33-59.

53.

Olive

Piolat

(2002). Suppressing visual feedback in written composition: Effects on processing demands and coordination of the writing processes. International Journal of Psychology, 37(4), 209-218.

54.

Passerault

J.-M.

Dinet

(2000). The role of visuospatial sketchpad in the written production of descriptive and argumentative texts. Current Psychology Letters: Behavior, Brain & Cognition, 3, 31-42.

55.

Paivio

(1986). Mental representations: A dual coding approach. New York: Oxford University Press.

56.

Pearson

D. G.

Logie

R. H.

(1998). La mémoire de travail visuo-spatiale: Fractionnement et développement [Visuospatial working memory: Fractionation and development]. In Bideaud

Courbois

(Eds.), Image mentale et développement (pp. 139-156). Paris: Presses Universitaires de France.

57.

Piolat

(1983). Localisation syntaxique des pauses et planification du discours [Syntactic location of pauses and discourse planning]. L’Année Psychologique, 83, 377-394.

58.

Piolat

Roussey

J. Y.

Thunin

(1997). Effects of screen presentation on text reading and revising. International Journal of Human-Computer Studies, 47, 565-589.

59.

Pylyshyn

Z. W.

(1981). The imagery debate: Analogue media versus tacit knowledge. Psychological Review, 88, 16-45.

60.

Ransdell

Levy

C. M.

Kellogg

R. T.

(2002). The structure of writing processes as revealed by secondary task demands. L1: Educational Studies in Language and Literature, 2, 141-163.

61.

Raulerson

B. A.

III Donovan

M. J.

Whiteford

A. P.

Kellogg

R. T.

(2010). Differential verbal, visual, and spatial working memory in written language production. Perceptual and Motor Skills, 110, 229-244.

62.

Sadowski

Paivio

(2001). Imagery and text: A dual coding theory of reading and writing. Mahwah, NJ: Erlbaum.

63.

Shepard

R. N.

Cooper

L. A.

(1982). Mental images and their transformations. Cambridge: MIT Press.

64.

Shepard

R. N.

Metzler

(1971). Mental rotation of three-dimensional objects. Science, 171, 701-703.

65.

Smith

E. E.

Jonides

(1997). Working memory: A view from neuroimaging. Cognitive Psychology, 33, 5-42.

66.

Smyth

M. M.

Silvers

(1987). Functions of vision in the control of handwriting. Acta Psychologica, 65, 47-64.

67.

Teleman

(1981). On visual feedback in writing. Proceedings of the Sixth International Conference of Applied Linguistics (pp. 424-425). Lund, Sweden: Lund University.

68.

Tversky

(1981). Distortions in memory for maps. Cognitive Psychology, 13, 407-433.

69.

Van Doorn

R. R. A.

Keuss

P. J. G.

(1992). The role of vision in the temporal and spatial control of handwriting. Acta Psychologica, 81, 269-286.

70.

Van Doorn

R. R. A.

Keuss

P. J. G.

(1993). Does the production of letter strokes in handwriting benefit from vision? Acta Psychologica, 82, 275-290.

71.

Van Galen

G. P.

Smyth

M. M.

Meulenbroek

R. G. J.

Hylkema

(1989). The role of short-term memory and the motor buffer in handwriting under visual and non-visual guidance. In Plamondon

Suen

C. Y.

Simner

M. L.

(Eds.), Computer recognition and human production of handwriting (pp. 253-271). Singapore: World Scientific.

72.

Zechmeister

E. B.

McKillip

(1972). Recall of place on the page. Journal of Educational Psychology, 5, 446-453.

73.

Zechmeister

E. B.

McKillip

Pasko

Bespalec

(1975). Visual memory for place on the page. Journal of General Psychology, 92, 43-52.

74.

Zesiger

(1995). Ecrire. Approches cognitive, neuropsychologique et développementale [Writing: Cognitive, neuropsychological and developmental approaches]. Paris: Presses Universitaires de France.