Contribution of writing to reading: Dissociation between cognitive and motor process in the left dorsal premotor cortex

Abstract

Functional brain imaging studies reported activation of the left dorsal premotor cortex (PMd), that is, a main area in the writing network, in reading tasks. However, it remains unclear whether this area is causally relevant for written stimulus recognition or its activation simply results from a passive coactivation of reading and writing networks. Here, we used chronometric paired-pulse transcranial magnetic stimulation (TMS) to address this issue by disrupting the activity of the PMd, the so-called Exner's area, while participants performed a lexical decision task. Both words and pseudowords were presented in printed and handwritten characters. The latter was assumed to be closely associated with motor representations of handwriting gestures. We found that TMS over the PMd in relatively early time-windows, i.e., between 60 and 160 ms after the stimulus onset, increased reaction times to pseudoword without affecting word recognition. Interestingly, this result pattern was found for both printed and handwritten characters, that is, regardless of whether the characters evoked motor representations of writing actions. Our result showed that under some circumstances the activation of the PMd does not simply result from passive association between reading and writing networks but has a functional role in the reading process. At least, at an early stage of written stimuli recognition, this role seems to depend on a common sublexical and serial process underlying writing and pseudoword reading rather than on an implicit evocation of writing actions during reading as typically assumed.

Keywords: Exner's area; cortico-spinal excitability; functional role; sublexical process; transcranial magnetic stimulation.

Figure 1

Individual participants’ stimulation sites of the PMd (white triangles) and M1 (white circles) normalized in the MNI space and projected on a standard template. “+” and “×” showed the average MNI coordinates of PMd (−32 2 52) and M1 (−35 −22 55), respectively.

Figure 2

An illustration of the timeline of a trial in the lexical decision task. Trials began with a fixation cross, followed by a stimulus and a blank screen. The duration (in ms) of each segment is written next to the screen. In trials that included TMS, a paired‐pulse TMS was applied in one of the three time‐windows.

Figure 3

The RTs obtained for handwritten pseudowords, handwritten words, printed pseudowords and printed words in the lexical decision task. White, gray, and black bars indicate the trials in which TMS was applied in the 0/40 ms (baseline), 60/100 ms, and 120/160 ms time‐window, respectively. The vertical bars represent standard errors.

Figure 4

The percentages of correct responses (%RC) obtained for words and pseudowords in the lexical decision task. White, gray, and black bars indicate the trials in which TMS was applied in the 0/40 ms (baseline), 60/100 ms, and 120/160 ms time‐window, respectively. The vertical bars represent standard errors.

Figure 5

The RTs and percentages of correct responses (%RC) obtained in the control task when TMS was applied in the 0/40 ms (baseline), 60/100 ms, and 120/160 ms time windows. The vertical bars represent standard errors.

Figure 6

Peak‐to‐peak MEP amplitudes of the FDI muscles measured when a single‐pulse TMS was applied on M1 at 0, 60, and 120 ms poststimulus onset, during the lexical decision task. Black and gray bars indicate responses to handwritten and printed characters, respectively. The vertical bars represent standard errors.