Elsevier

Cortex

Volume 49, Issue 10, November–December 2013, Pages 2772-2787
Cortex

Research report
The “handwriting brain”: A meta-analysis of neuroimaging studies of motor versus orthographic processes

https://doi.org/10.1016/j.cortex.2013.05.011Get rights and content

Abstract

Introduction

Handwriting is a modality of language production whose cerebral substrates remain poorly known although the existence of specific regions is postulated. The description of brain damaged patients with agraphia and, more recently, several neuroimaging studies suggest the involvement of different brain regions. However, results vary with the methodological choices made and may not always discriminate between “writing-specific” and motor or linguistic processes shared with other abilities.

Methods

We used the “Activation Likelihood Estimate” (ALE) meta-analytical method to identify the cerebral network of areas commonly activated during handwriting in 18 neuroimaging studies published in the literature. Included contrasts were also classified according to the control tasks used, whether non-specific motor/output-control or linguistic/input-control. These data were included in two secondary meta-analyses in order to reveal the functional role of the different areas of this network.

Results

An extensive, mainly left-hemisphere network of 12 cortical and sub-cortical areas was obtained; three of which were considered as primarily writing-specific (left superior frontal sulcus/middle frontal gyrus area, left intraparietal sulcus/superior parietal area, right cerebellum) while others related rather to non-specific motor (primary motor and sensorimotor cortex, supplementary motor area, thalamus and putamen) or linguistic processes (ventral premotor cortex, posterior/inferior temporal cortex).

Conclusions

This meta-analysis provides a description of the cerebral network of handwriting as revealed by various types of neuroimaging experiments and confirms the crucial involvement of the left frontal and superior parietal regions. These findings provide new insights into cognitive processes involved in handwriting and their cerebral substrates.

Introduction

Writing is a major cultural invention and an everyday communication tool for mankind, the earliest forms of which date from approximately 6 thousand years ago. Writing to communicate has since played a central role in the dissemination of culture and concepts. Handwriting is the most common way of learning and using written language even though typing has now taken on a central role in western societies. Whatever the orthographic system, the process of writing implies the contribution of several cognitive and motor functions. A complex set of neural underpinnings support this highly specific skill.

Most of our knowledge of the neuroanatomy of writing comes from neuropsychological studies of dysgraphias or agraphias seen after some brain lesions or diseases. A variety of writing disabilities have been described, from dysorthographias, affecting lexical or phonological components – i.e., central processes – with relative preservation of letter formation, to apraxic agraphias, affecting grapheme tracing – i.e., peripheral processes (Roeltgen, 2003). Neuropsychological studies have suggested that lexical processes of writing relate to the angular gyrus (AG) (Roeltgen and Heilman, 1984) or the precentral gyrus (preCG) (Rapcsak et al., 1988) and phonological processes to the left perisylvian regions (Alexander et al., 1992b, Rapcsak et al., 2009). The motor components are believed to be linked to the left superior parietal or premotor regions (Alexander et al., 1992a, Anderson et al., 1990, Auerbach and Alexander, 1981). However, these results suffer from the usual limitations of lesion-based studies: spontaneous lesions are variable, difficult to delineate and the lesion-symptom relationship may be strongly influenced by neural plasticity and functional recovery.

Compared to other language skills, writing has long been neglected in functional brain imaging studies of healthy subjects. It is therefore difficult to establish a neurofunctional model of typical writing skills that might match the results reported in the literature on agraphia. More recently, several imaging studies have aimed to localize brain areas implicated in spelling or handwriting. The aim of the present paper is to present a meta-analysis of functional imaging studies reporting brain areas related specifically to handwriting and to isolate brain territories that characterize the written modality of language from those that also support other (motor or linguistic) abilities. Beyond the description of the network of brain areas involved, the importance of methodological aspects of these imaging experiments will be addressed.

A first meta-analysis has been recently provided by Purcell et al. (2011b), who aimed to distinguish central and peripheral processes of spelling. Central writing processes refer to the retrieval of abstract orthographic word-forms, via orthographic lexicon or phoneme-to-grapheme conversion mechanisms, and their temporary storage in working memory (in the “graphemic buffer”, see Hillis and Caramazza, 1989). Peripheral processes involve letter production (selection of allographs or letter-shape conversion processes), planning and ordering of the sequence of letters and execution of specific motor programs (Ellis, 1982). Purcell et al. (2011b) limited their data exploration to alphabetic writing systems, and the studies considered did not necessarily involve tasks requiring the actual production of script. They focused on tasks eliciting the activation of an orthographic representation of a word, for example deciding if a particular letter was present or not in a heard word or if different words were spelled the same. While rhyme spelling tasks or spelling judgement have been used to study central spelling processes, one may consider that such tasks stray too far from the processes involved in everyday handwriting and their brain substrates. Here, we rather aimed at identifying handwriting-specific processes allowing actual written production (with the exception of “mental writing” tasks). We included various writing conditions (writing from dictation, written naming, generative writing), and different writing codes, both alphabetic and ideographic (Japanese Kanji). We also considered results from experiments that were not necessarily designed to study the neural substrates of writing per se (e.g., word retrieval, handedness, creativity, comparison of different types of writing).

Word writing involves many processes such as analysis of the input sensory information (visual or auditory), access to the orthographic representation of the word to be written (either directly or via a sublexical processing; see Rapp et al. (2002) for an example of such a ‘dual-route’ model) and its temporary storage into the graphemic buffer. These central stage processes are followed by allographic processes, i.e., the specification of the format in which letter series will be produced, including the idiosyncratic way each individual actually produces graphic scripts, and this involves the programming and neuromuscular execution of appropriate motor sequences (van Galen, 1991). As mentioned earlier, several of these processes are not specific to writing and can also be involved in other tasks, whether linguistic (e.g., reading) or motor (e.g., drawing). Neuroimaging studies should be able to disentangle writing/spelling processes (i.e., conceived as the preparation of a message and its conversion into a graphic form) from unrelated input or linguistic processing, and from or non-specific motor movements. The interpretation of their results thus relies strongly on the understanding of the experimental and control tasks involved.

A first objective in experiments involving handwriting gestures is to tease apart the respective influences on brain activation of finger/hand motor activity (holding pen, moving joints) and visuospatial control of these movements. The most relevant control task consists of drawing non-linguistic stimuli such as circles (e.g., Roux et al., 2009), abstract symbols (e.g., Omura et al., 2004) or pseudo-letters (e.g., Longcamp et al., 2003). Yet another strategy is to attenuate the influence of the sensorimotor and visuospatial components in the experimental handwriting task itself. It is generally assumed that writing skills are independent of the tool or the effector since performance in agraphic patients is impaired regardless of the effector used. Some authors (especially in studies of the Japanese writing system) asked subjects to write with the finger on a board or in the air, with a limited amount of graphemes, thus removing the constraints of pen grasping (that implies maintaining adequate pressure on the paper and controlling the topokinetic aspects of the usual handwriting, i.e., the spatial dynamics of the hand in a delimited graphic space). ‘Writing in the air’ was used by Katanoda et al. (2001), in association with a visually-cued finger tapping task as a control matched in duration, speed and amplitude. Finally, another way to control for non-specific motor activity may be not to write at all (e.g., Harrington et al., 2007). These authors asked their subjects to mentally draw or write the name of a visually presented object. Although it is very difficult to control the subject's activity during the task, this approach seems to be valid since the only difference of activation reported with a second condition of actual writing was located in the motor cortex. The comparison between studies that involve sensorimotor control tasks and those not involving such control may reveal which area in the handwriting brain network is really specific to writing and which ones are not i.e., support non-specific or motor-related functions.

As stated by cognitive models (i.e., Ellis, 1982, Rapp et al., 2002), handwriting tasks may involve many processes starting by the analysis of the input sensory information (either visual as in word copy or auditive as in dictation), its processing through a visual analysis system and an orthographic input lexicon (visual input) or through a phonological input lexicon (auditive input), allowing access to the semantic system integrating knowledge about the item. Control tasks are also designed to manage such linguistic aspects of writing or processes intervening up to the writing stage. An overt oral or subvocal naming task has often been chosen to isolate linguistic processes specific to written naming (e.g., Sugihara et al., 2006), often assuming that written naming of a (visually presented) item necessarily implies prior covert oral naming. Subtracting the covert control naming from the writing condition would thus isolate activations specifically related to the “central” process of writing (i.e., orthographic and graphemic planning). The use of an oral task identical to the writing task seems to be a simple and valid approach to control for amodal central effects and isolate writing modality [for Brownsett and Wise (2010)]. Motor activations related to oral articulation may also be further “suppressed” by using another control task (such as meaningless syllable repetition; Roux et al., 2009). Input-matched conditions, in which the same stimuli as in the writing task are presented (written or aurally presented words, pictures) are obviously the simplest way to abstract from sensorial, but also linguistic (i.e., to process the stimulus), brain functions. Here again, the results from comparing experiments which did or did not manage these issues may isolate writing-specific areas in the writing network from non-specific linguistic areas. More focused studies aiming to specifically delineate writing-specific areas have used a combination of both kinds of control tasks; presentation of stimuli during both the written spelling task and the motor control task (Rapp and Dufor, 2011, Segal and Petrides, 2012), or conjunction analyses (e.g., Katanoda et al., 2001). Sugihara et al. (2006) and Roux et al. (2009) chose to resort to a conjunction between the activations from a writing task with the right hand and a writing task with the left hand, in the same group of right-handed subjects. The resulting activity was considered as independent of the motor activity of the hand and as reflecting only a specific and critical process of handwriting.

Here, we used published data (as reported in tables of activated clusters) from papers identified in the current brain imaging literature. Studies were considered when they explicitly involved handwriting processes. Our general objective was to identify the areas commonly activated by tasks thought to involve the same psycholinguistic processes. As a first step, we aimed to identify the global brain network of regions activated across all studies and contrasts (through a quantitative method). Two secondary meta-analyses were then conducted to isolate brain regions arising from contrasts that controlled for non-specific motor or linguistic processes.

Section snippets

Studies included

Given the relatively small number of imaging studies having a direct interest in writing as such, we included studies whose main objective may have be different but that reported at least one contrast involving written production. Thus, we considered experiments involving the graphic production of any linguistic content-element, i.e., from basic components such as the repeated production of a single letter, to elaborate productions such as a coherent narrative, with alphabetic, syllabic (e.g.,

Discussion

Results from the main meta-analysis allowed us to identify 12 cortical and sub-cortical functional regions involved to different degrees in written spelling tasks. Some of them were also present in the results of the secondary analyses conducted to distinguish a general writing network from non-specific input and output confounds. In an attempt to clarify the role of each component of this cerebral network, we will discuss these results with respect to the experimental contrasts and studies

Conclusion

Different paradigms have been used to study the neural correlates of handwriting thanks to neuroimaging. Their results vary and do not always agree. The results of this meta-analysis addressing the points of agreement across all of them are quite consistent with those of a recent work focusing only on spelling processes (Purcell et al., 2011b), but sometimes differ with knowledge derived from lesion studies. An essentially left-hemisphere network is recruited for various handwriting tasks, with

References (78)

  • T. Ohyama et al.

    What the cerebellum computes

    Trends in Neurosciences

    (2003)
  • T. Paus

    Location and function of the human frontal eye-field: A selective review

    Neuropsychologia

    (1996)
  • C.J. Price

    A review and synthesis of the first 20 years of PET and fMRI studies of heard speech, spoken language and reading

    NeuroImage

    (2012)
  • J.J. Purcell et al.

    A combined fMRI study of typed spelling and reading

    NeuroImage

    (2011)
  • S.Z. Rapcsak et al.

    Phonological dyslexia and dysgraphia: Cognitive mechanisms and neural substrates

    Cortex

    (2009)
  • F.E. Roux et al.

    Who actually read Exner? Returning to the source of the frontal “writing centre” hypothesis

    Cortex

    (2010)
  • P. Scarone et al.

    Agraphia after awake surgery for brain tumor: New insights into the anatomo-functional network of writing

    Surgical Neurology

    (2009)
  • C.J. Stoodley et al.

    Functional topography in the human cerebellum: A meta-analysis of neuroimaging studies

    NeuroImage

    (2009)
  • G. Sugihara et al.

    Interindividual uniformity and variety of the “Writing center”: A functional MRI study

    NeuroImage

    (2006)
  • P.E. Turkeltaub et al.

    Meta-analysis of the functional neuroanatomy of single-word reading: Method and validation

    NeuroImage

    (2002)
  • G.P. van Galen

    Handwriting: Issues for a psychomotor theory

    Human Movement Science

    (1991)
  • M. Vigneau et al.

    Meta-analyzing left hemisphere language areas: Phonology, semantics, and sentence processing

    NeuroImage

    (2006)
  • M.P. Alexander et al.

    Lesion localization in apractic agraphia

    Archives of Neurology

    (1992)
  • R.A. Andersen et al.

    Multimodal representation of space in the posterior parietal cortex and its use in planning movements

    Annual Review of Neuroscience

    (1997)
  • S.W. Anderson et al.

    Troubled letters but not numbers. Domain specific cognitive impairments following focal damage in frontal cortex

    Brain

    (1990)
  • S.H. Auerbach et al.

    Pure agraphia and unilateral optic ataxia associated with a left superior parietal lobule lesion

    Journal of Neurology, Neurosurgery, and Psychiatry

    (1981)
  • M.F. Beauvois et al.

    Phonological alexia: Three dissociations

    Journal of Neurology, Neurosurgery, and Psychiatry

    (1979)
  • P. Beeson et al.

    The neural substrates of writing: A functional magnetic resonance imaging study

    Aphasiology

    (2003)
  • S.L.E. Brownsett et al.

    The contribution of the parietal lobes to speaking and writing

    Cerebral Cortex

    (2010)
  • L. Cloutman et al.

    A neural network critical for spelling

    Annals of Neurology

    (2009)
  • L. Cohen et al.

    The visual word form area: Spatial and temporal characterization of an initial stage of reading in normal subjects and posterior split-brain patients

    Brain

    (2000)
  • M. Corbetta et al.

    Control of goal-directed and stimulus-driven attention in the brain

    Nature Reviews. Neuroscience

    (2002)
  • S. Dehaene et al.

    The visual word form area: A prelexical representation of visual words in the fusiform gyrus

    NeuroReport

    (2002)
  • S.B. Eickhoff et al.

    Coordinate-based activation likelihood estimation meta-analysis of neuroimaging data: A random-effects approach based on empirical estimates of spatial uncertainty

    Human Brain Mapping

    (2009)
  • A.W. Ellis

    Spelling and writing (and reading and speaking)

  • S. Exner

    Untersuchungen über die Localisation der Functionen in der Grosshirnrinde des Menschen

    (1881)
  • H. Goodglass

    Understanding Aphasia

    (1993)
  • G.S. Harrington et al.

    Comparison of the neural basis for imagined writing and drawing

    Human Brain Mapping

    (2007)
  • K. Katanoda et al.

    A functional MRI study on the neural substrates for writing

    Human Brain Mapping

    (2001)
  • Cited by (188)

    View all citing articles on Scopus
    View full text