Neurophysiological correlates of automatic integration of voice and gender information during grammatical processing

Abstract

During verbal communication, interlocutors rely on both linguistic (e.g., words, syntax) and extralinguistic (e.g., voice quality) information. The neural mechanisms of extralinguistic information processing are particularly poorly understood. To address this, we used EEG and recorded event-related brain potentials while participants listened to Russian pronoun–verb phrases presented in either male or female voice. Crucially, we manipulated congruency between the grammatical gender signaled by the verbs’ ending and the speakers’ apparent gender. To focus on putative automatic integration of extralinguistic information into syntactic processing and avoid confounds arising from secondary top-down processes, we used passive non-attend auditory presentation with visual distraction and no stimulus-related task. Most expressed neural responses were found at both early (150 ms, ELAN-like) and late (400 ms, N400-like) phrase processing stages. Crucially, both of these brain responses exhibited sensitivity to extralinguistic information and were significantly enhanced for phrases whose voice and grammatical gender were incongruent, similar to what is known for ERPs effects related to overt grammatical violations. Our data suggest a high degree of automaticity in processing extralinguistic information during spoken language comprehension which indicates existence of a rapid automatic syntactic integration mechanism sensitive to both linguistic and extralinguistic information.

Picture naming test through the prism of cognitive neuroscience and linguistics: adapting the test for cerebellar tumor survivors—or pouring new wine in old sacks?

“…More than one technique might be combined in the study: take MRI and MEG, each efficient as a standalone method, which together provide more detailed spatiotemporal information ( Alekseeva et al, 2022 ), or fMRI and dMRI used to map gray matter regions alongside fiber bundles ( Jarret et al, 2022 ). Eventually, data gathered via neuroimaging could lead to the emergence of further models of lexical retrieval.…”