Assessing sensitivity to semantic and syntactic information in deaf readers: An ERP study

Abstract

An ERP study by Mehravari et al. (2017) found deaf and hearing readers had similar N400 effects to semantic violations, but only hearing readers showed a P600 effect to verb agreement violations. To assess whether lack of sensitivity to agreement was due to early language deprivation, we presented the same stimuli (and a phrase structure violation condition) to deaf readers with early ASL exposure and a reading-matched hearing group. We also examined N400 “wrap up” effects for sentence final words. Both groups exhibited an N400 effect for semantic violations (larger negativities for errors), although the effect was unexpectedly weak for hearing readers. At the violation, only hearing readers exhibited a P600 for verb agreement violations (larger positivities for errors), indicating that the previous findings are not explained by language deprivation. Crucially, both groups exhibited an N400 effect on the sentence-final word for agreement violations – despite poor performance on the acceptability judgment task. Both deaf and hearing readers showed a robust P600 to phrase structure violations. We interpret this overall pattern as indicating that deaf readers engage in monitoring strategies whereby English phrase structure, but not verb morphology, is anticipated during reading, and that both groups exhibit neural sensitivity to grammatical errors during sentence final integration processes. The amplitude of the P600 for grammatical violations was correlated with reading skill only for hearing readers. We conclude that reading strategies for skilled deaf readers who are bilingual in ASL differ from hearing monolingual readers with similar reading skills.