Introduction

The National Center for Education Statistics (2003) in the U.S. reported that only about 28% of fourth graders could write at a proficient level or above (i.e., create an effective response to the task in form, content, and language, demonstrate an awareness of the intended audience, use effective organization appropriate to the task, use sufficient elaboration to clarify and enhance the central idea, etc.), 58% wrote at a basic level (i.e., demonstrate appropriate response to the task in form, content, and language, use some supporting details, demonstrate organization appropriate to the task, and demonstrate sufficient command of spelling, grammar, punctuation, and capitalization to communicate to the reader), and 14% wrote below the basic level. These data are compelling in that they indicate the national significance of writing problems in elementary education. In this regard, the development of writing skills in students presents a significant challenge for educators, particularly in this day of high-stakes testing and heightened accountability. While our understanding of the underlying cognitive components of reading has proliferated over the past 20 years, we have only begun to understand the basic neurocognitive and neurobiological factors that may underlie the development of written expression in the formative elementary school years (Edwards, 2003; Graham & Harris, 2005).

Key Research Questions

1. What theoretical models guide the study of the neurobiological processes underlying written language acquisition?
2. What neurocognitive functions are associated with the acquisition of written language and how do they affect writing at different developmental time points?
3. What neurological systems contribute to the development of written language?

Recent Research Results

What theoretical models guide the study of the neurobiological processes underlying written language acquisition?
There have been several theoretical models proposed to describe the cognitive functions involved in written expression (e.g., Kellogg, 1996), with these models generally invoking various language, short-term and long-term memory, and executive functions. The model proposed by Hayes and Flower (1980), and subsequently revised by Hayes (2006), has been one of the most influential in the field of written expression; however, the various subcomponents of the current Hayes model are unconstrained, particularly from a developmental perspective, and their interactions over time remain relatively unknown.
 
A Developmental Model of Written Expression. A more contemporary model of written expression is based on the work of Juel (1988) – the Simple View of Writing. Berninger and colleagues (Berninger & Amtmann, 2003; Berninger & Winn, 2006) have expanded this model into the Not-So-Simple View of Writing which includes three major components: transcription (for handwriting, letter production, spelling, and word production); executive functions (for planning, monitoring, and revising); and text generation. The latter component, text generation, occurs at the word, sentence, and text levels; consequently, automatic production of letters is necessary, but not sufficient, as spelling words via the alphabetic principle and related orthographic elements are necessary for writing proficiency (Jones & Christensen, 1999). In addition to moving from picture processing and production to writing processing and production (Adi-Japha & Freeman, 2001), beginning writers have immature self-regulation functions, thus requiring considerable teacher modeling and scaffolding (Berninger & Richards, 2002). Beginning writers at this stage also benefit from integrated reading-writing instruction, although it is important to note that proficient reading also is not sufficient for competent written expression (Abbott et al., 1997).

According to this developmental model, lower level neuropsychological functions are related directly to initial writing skills, and they appear to have their largest impact on the transcription subcomponent of the writing process. The writing of early elementary school students will be constrained by factors related to graphomotor output (e.g., letter formation), memory for letters and words, emergent working memory capacity, and linguistic capabilities (Berninger & Winn, 2006). By the middle of elementary school many, but not all, children have sufficient transcription skills, and their writing will progress with increased emphasis placed on text generation and executive functions. Explicit instruction aimed at both the syntax and discourse levels (e.g., content, organization, clarity, etc.) also is required (Graham & Harris, 2000; Graham et al., 2001). Further, developing writers continue to need explicit instruction in the planning (De La Paz, 1997; Graham & Harris, 2000) and reviewing/revising (e.g., Graham, 1997) components of composing as most students are not fully self-regulated writers at this developmental period (Graham & Harris, 2003). In general, the Not-So-Simple View of Writing Model postulates that neuropsychological, linguistic, and related cognitive constraints may be recursive throughout the development of the writing process, but that each of these constraints will exert relatively greater influence at different points in the developmental process.

What neurocognitive functions are associated with the acquisition of written language and how do they affect writing at different developmental time points?
Over 15 years ago Levine et al. (1993) theorized about the importance of a variety of neuropsychological functions in the writing process (e.g., memory, attention, graphomotor output, sequential processing, higher-order cognition, language, visual-spatial functions); however, no data were available to support when or to what degree these functions influenced the writing process. At the same time, Abbott and Berninger (1993) noted that oral language/verbal reasoning, including such functions as sentence memory, word finding, phonological processing, and reading, contributed to composition fluency. Berninger and Rutberg (1992) also described the importance of fine-motor planning and control to the development of writing in early elementary grade children.

More recent studies have begun to document the cognitive abilities associated with "expert" writers (Berninger et al., 2002; Gregg & Mather, 2002; Hayes, 2000; Torrance, Fidalgo, & Garcia, 2007). Recent work by our research group with fourth and fifth grade students with and without writing problems documented not only working memory deficits, but broader memory problems (Hooper et al., 2009) and executive dysfunctions (Hooper et al., 2002) that can undermine the writing process. Our group also has examined the multidimensionality manifested in elementary school students with and without writing problems (Hooper et al., 2006; Sandler et al., 1992; Wakely et al., 2006), with results suggestive of the presence of multiple written language subtypes in the regular education setting—including both typical and atypical cognitive-linguistic profiles. Consistent with cognitive components described in the Not-So-Simple View of Writing, these studies collectively have shown the importance of specific linguistic factors (e.g., semantics, grammar), along with academic functions such as reading and spelling, as key dimensions of written expression.

Perhaps one of the most studied neurocognitive functions in the area of written language is working memory (Lea & Levy, 1999; McCutchen, 2000). Working memory is important to written expression because it underlies the active maintenance of multiple ideas, the retrieval of grammatical rules from long-term memory, and the recursive self-monitoring that is required during the act of writing (Kellogg, 1999). Working memory contributes to the management of these simultaneous processes, and a breakdown may lead to problems with written output (Levy & Marek, 1999). McCutchen (2000) noted that poor writers typically have reduced working memory capacity when compared to good writers, and Swanson and Berninger (1996) reported that working memory has both general and domain-specific contributions to the writing process.

What neurological systems contribute to the development of written language acquisition?
Based on the neurocognitive functions that purportedly contribute to the development of written language, it has been speculated that there are many more brain areas involved during the act of writing than for other academic domains (Hale & Fiorello, 2004). Based on the Not-So-Simple View of Writing, it is likely that different brain systems are primarily involved at different developmental periods. To date, however, there are few studies that have identified the various neurological systems contributing to the development of written language.

Berninger and Richards (2002) provided a list of possible brain structures that could be involved in various writing functions. For example, they noted that the supplementary motor region, primary and secondary motor areas, and anterior cerebellum were brain regions that could be involved in graphomotor output (i.e., writing), and that the supramarginal gyrus, insula, and anterior cingulate were key structures important to the phonologic-orthographic (i.e., sound-writing) component. For written expression, they postulated the involvement of the left dorsolateral prefrontal cortex for the goal setting, planning, monitoring, and revising functions; the prefrontal cortex, left inferior parietal cortex, and left supramarginal gyrus for phonological working memory; the hippocampus for the long-term memory aspects of written expression; and aspects of the limbic system for the motivational and emotional components of written output.

At present, empirical studies have begun to verify these structure-function relationships in written expression. Initial lesion studies (i.e., studies of patients with deficits that arise following a brain injury) implicated the superior parietal lobe and middle frontal regions as critical to intact written expression (Roeltgen, 2003). In addition, neuroimaging studies have begun to demonstrate linkages between various fine-motor or graphomotor movements and various brain structures such as the premotor cortex, supplementary motor cortex, cerebellum, basal ganglia, left temporal cortex, thalamus, left anterior cingulate, and orbitofrontal cortex (Deus, Junque, Pujol, Vendrell, Vila, & Capdevila, 1997; Filipović, Papathanasiou, Whurr, Rothwell, & Jahanshahi, 2008; James & Gauthier, 2006; Jenkins, Brooks, Nixon, Frackowiak, & Passingham, 1994; Katanoda, Yoshikawa, & Sugishita, 2001; Menon & Desmond, 2001; Nicholson et al., 1999; van Mier, Temple, Perlmutter, Raichle, & Petersen, 1998; Richards et al., 2009). For example, Richards et al. (2009) showed that sequential finger movements were associated with key brain regions deemed important to written expression in good but not poor elementary school writers. These brain regions included the left superior parietal, right inferior frontal orbital, right precuneus, right inferior temporal, and left inferior temporal areas. Additionally, activation in these regions was significantly correlated with handwriting, spelling, and composing. In one of the few studies where participants were required to write (i.e., sentences from dictation), Menon and Desmond (2001) used functional magnetic resonance imaging (fMRI) to show the left superior parietal lobe and left inferior frontal gyrus to be key brain regions involved in written output. An exciting line of research also has begun to examine the neurological substrates involved in higher-order writing skills. Using fMRI, Berninger et al. (2009) demonstrated greater activation in prefrontal cortical regions associated with working memory during an idea generation task. Finally, Pugh et al. (2006) noted that there likely is a high degree of overlap between the neural systems serving both oral and written language and to some extent, reading (Caplan, 2004), and that the study of these associated systems will be important for increasing our understanding of the neurological bases of written language. Taken together, these findings have begun to target the basic neural systems important to written expression.

Future Directions

The study of the neurobiological foundations of written language acquisition is an exciting area of scientific inquiry. Over the past two decades significant advances have been made with respect to our understanding of written language from a cognitive perspective, particularly with respect to neurodevelopmental models that parallel cognitive growth and development. Future studies should continue to examine the various nuances of these models and how they relate to specific evidence-based interventions for children struggling with written output. Additionally, a downward extension of these models to preschool populations may provide critical knowledge with respect to early intervention efforts. Similarly, the field has begun to make gains in our understanding of the neurological substrates involved in written expression. Additional neurologically-based studies of impaired, at-risk, and typical writers at different developmental time points will also be useful in advancing this emergent knowledge-base. Finally, as evidence-based educational interventions begin to surface, perhaps by employing subtype X treatment methodology (Hooper et al., 2006), it will be interesting to determine if such interventions can modify brain activation patterns in conjunction with improvements in written expression.

Conclusions

This encyclopedia entry has provided an overview of the basic biological processes that contribute to the acquisition of written language skills. An important neurodevelopmental model of written language was discussed and its associated neurocognitive abilities detailed. Additionally, the available neurological findings to date were noted, with a specific focus on the emergent neuroimaging findings. Although the basic notion of how written language skills develop remains somewhat elusive, the field has begun to assert findings that have increased our understanding of this critical output function, and future research efforts undoubtedly will further our understanding over the next decade.