EEG entrainment to the prosodic structure in spoken language

International audienceA key characteristic of speech is the quasi-regular rhythmic information it contains. Recently, it has been suggested that ongoing neural oscillations take advantage of this characteristic by adjusting their phase to match the rhythmic structure in the signal, a mechanism called phase entrainment. Such a mechanism can serve many purposes, from ensuring high neural sensitivity at the most relevant parts in the speech signal to simultaneously processing and segmenting speech at multiple time-scales. A growing number of studies show evidence of syllabic encoding in the theta band (4-8 Hz) and phonemic encoding in the low-gamma band (25-35 Hz). However, the role of entrainment in the segmentation of larger linguistic units, such as words and phrases, has been relatively understudied. Here we focus on neural synchronisation in the delta band (~2 Hz) corresponding to the prosodic structure. Indeed, while languages' metrical structures have long been recognised to play a vital role in segmenting speech into words, neural alignment to the temporal regularities conveyed by the prosodic hierarchical structure has yet to be empirically tested. We outline a research plan that aims to establish whether neural oscillations align to prosodic regularities in spoken language and how such entrainment contributes to speech comprehension