Spoken and written language interaction in normal readers and dyslexics – SWIND

The debate on the contribution of nature and nurture to the evolution of the cognitive system and brain organization has played a central role in research in cognitive science. The interaction between spoken and written language is one of the most illustrative examples of their reciprocal influence. Many research groups provided behavioral evidence showing that, in some language processing situations, 1) the ability to read is affected by knowledge of spoken language and more surprisingly, 2) the ability to process speech is affected by knowledge of written language. Yet, the underlying cognitive and neural mechanisms of this interaction remain unclear. This lack of knowledge continues to fuel a long-standing debate on the autonomy and dependency between the two language systems.
The project proposes an integrative approach that explores both spoken and written language within the same study and by using comparable experimental paradigm and material. It aims at providing behavioral and neural evidence that addresses the theoretical issue of whether the spoken and written language systems are autonomous, in interaction or becoming fully integrated through reading acquisition.
A Stage-of-Processing approach will be adopted throughout the study. It consists in considering separately each cognitive stage that comes into play during language processing. This offers a possibility to investigate the impact of task demands on the phenomenon of interest. The research program is separated into 3 parts. In Part 1, fMRI experiments will be performed using visual and auditory language tasks that tap the main processing stages: prelexical, lexical and postlexical. Within each task, the relation between the phonological and orthographic representations of French words will be manipulated in order to obtain evidence for the influence of cross-modal information on speech and written word processing. This will provide us information on the extent and limit of the dynamic between the two language systems and the underlying neural basis. In the presence of evidence for interaction or integration between the two systems, it is crucial to examine whether the contribution of the cross-modal system is essential or is merely a natural consequence of reading acquisition without functional significance. Part 2 addresses this issue by investigating the causal role of the brain areas identified in Part 1. To do so, a transcranial magnetic stimulation (TMS) perturbation protocol will be applied on the phonological and orthographic areas at different time points while the participants perform reading and speech processing tasks. If the activity of the stimulated area is essential, interrupting its function should affect performance. In addition to the information on the causal role of these areas, our chronometric protocol will also tell us about their activation time-course. In Part 3, TMS and EEG will be combined. Since we are not aiming to induce a perturbation at the behavioral level but to probe a change of activity within a neural network, only a minimal stimulation will be used. This change of activity induced by TMS will propagate to interconnected areas, which can be tracked by EEG. This will provide us information on functional connectivity between the areas involved in the task and direction of the flow of neural activity (bottom-up vs. top-down) between these areas.
Our interest goes beyond the functioning of the language system in normal readers. An exploratory case study will be performed in dyslexic adults. After an examination of the language performance profile of each participant, Part 1 of the protocol for normal readers will be applied. Investigating the interaction between the spoken and written language in dyslexics will provide insightful information on the compensatory mechanisms. Based on these preliminary data, we consider the possibility to further explore this issue in a therapeutic perspective.