The neural correlates of phonological mismatch during spoken language comprehension

No Thumbnail Available
Issue Date
Journal Title
Journal ISSN
Volume Title
Listeners may predict upcoming linguistic information to facilitate the processing of spoken language. The strongest stances on linguistic prediction propose that predictions are specified at the level of word form and phonology. Identifying neural responses that are specific to violations of form predictions would lend support to these strong proposals. Early negative-going event-related potentials (ERPs) such as the Phonological Mismatch Negativity (PMN) and the N200 have been used to study the effects of phonological form mismatch during sentence comprehension. However, it has been argued that these early components are not distinct from the N400. In this EEG study, we investigate whether there are ERP components that are specifically sensitive to phonological mismatch and whether these components are separable from the N400. Participants listened to sentences that either ended with a highly predictable word, an unpredictable word which overlapped phonologically with the predictable word, or an unpredictable word without phonological overlap. The waveforms in the phonologically unpredictable conditions diverged from the predictable condition as early as 150ms following word onset. Preregistered analyses revealed a widely distributed negative-going ERP in response to unpredictable words in both the early (150-250ms) and the N400 (300-500ms) time windows. However, the effects in the early time window did not present with a scalp distribution which was distinct from that observed in the N400 time window. As such, these results are consistent with an account of the early negativity as an earlier onset of the auditory N400. Although these findings do not necessarily exclude phonological prediction during sentence comprehension, they do stand in opposition to previous reports of an observable effect that is reflective of phonological mismatch detection. As such, the results of the present study do not provide evidence in support of the strongest stances on linguistic prediction, but instead introduce skepticism regarding the reliability of the proposed early ERP components.
Faculteit der Sociale Wetenschappen