The role of high-level processes for oscillatory phase entrainment to speech sounds

Neural oscillations adjust their phase to rhythmic stimulation, a phenomenon called phase entrainment. This mechanism seems to be of particular importance for the processing of speech: Assumed to underlie speech comprehension, phase entrainment is omnipresent in current theories of speech processing. Nevertheless, speech is a complex stimulus and both low- and high-level processes might contribute to phase entrainment as it is commonly reported in the literature. Our aim was to disentangle these processes and provide a detailed characterization of the neural mechanisms underlying phase entrainment to speech. For this purpose, we constructed speech/noise stimuli without systematic fluctuations in sound amplitude or spectral content (here termed “low-level” features), while keeping both fluctuations in high-level features (including phonetic information) and intelligibility. In human psychophysical and electroencephalographic (EEG) data as well as primate intracranial recordings, we were able to show that phase entrainment can be observed in response to speech sounds in which systematic fluctuations in low-level features have been removed. This “high-level” entrainment shows specific characteristics and seems to reflect a particularly efficient mechanism of speech processing which is conserved across species. Finally, the relation between phase entrainment and speech comprehension remains debated. Based on the data presented here and elsewhere, we discuss possible reasons (and solutions) for this controversy and propose how brain stimulation techniques can help to clarify the role of oscillatory phase entrainment for the comprehension of speech sounds.