Hoen, M. , Dole, M. , Grataloup, C. & Meunier, F.
Centre de Recherche en Neurosciences de Lyon Equipe Dynamique Cérébrale et Cognition INSERM U1028 - CNRS UMR5292 Groupe de recherche ALP
Although people with developmental dyslexia are characterized by normal oral language skills in quiet environments or favorable listening situations, they show clear speech-comprehension difficulties as soon as this task is made more complicated. In conditions of cognitive load or when the listening situation is made less favorable by the addition of external noise for example, people with dyslexia show a reproducible and life-time lasting deficit that could relate to a subjacent phonological deficit. In the current paper we will review data from multiple studies that were run in our group in order to characterize this deficit both in terms of behavioral and cerebral correlates. On the behavioral side, we will first show how the informational content of the interfering sound strongly modulates this deficit. In particular, we will discuss the fact that concurrent speech compared to any other types of non-speech or speech-derived noises constitutes the most difficult masker for dyslexics who seem to be especially sensitive to the informational component of masking phenomena. We will then display a series of measures testing different listening configurations that show that dyslexics exhibit normal binaural listening abilities and are able to perform binaural unmasking of speech-in-speech. This result suggests that binaural listening abilities are not the cause of the deficit. We will then discuss different aspects of the speech-in-speech situation where we manipulated psycholinguistic dimensions of both target- and interfering-words as the lexical frequency or the size of the phonological neighborhood to show that inside the informational masking compound, the phonological level seems to play a crucial role in dyslexics? difficulties. Finally we will mention recent neuroimaging evidence including functional (fMRI) and morphological (VBM) data supporting the idea that one of the cortical correlates of dyslexia is an abnormal asymmetrisation of certain speech-processing areas.