Individual differences in linguistic prediction in native language comprehension and second language learning

Abstract

Individual differences ERP research shows that grammatical errors violate expectations of form (P600 effects) in some individuals and expectations of meaning (N400 effects) in others. These differences have not been fully accounted for by variability in working memory capacity or language experience. Two experiments were conducted to investigate whether the observed variability in grammar processing can be explained by differences in comprehenders’ grammar aptitude or in the way individuals build predictions during language comprehension. In Experiment 1, EEG data was recorded while native (L1) English speakers read sentences containing syntactic violations of various types, and semantic continuations that varied in expectancy. In Experiment 2, EEG data was recorded while native Chinese speakers read Chinese sentences containing one type of syntactic violation, and English sentences containing one syntactic violation familiar to the L1 and one violation unique to the second language (L2). The EEG signal was analyzed for differences in ERP response magnitudes and changes in beta power desynchronization in response to violations. English speakers’ ERP response patterns to syntactic anomalies predicted neural sensitivity to semantic anomalies. Chinese learners’ proficiency in the unique L2 rule was predicted by their grammar processing patterns in Chinese. Grammar processing variability was not explained by grammar aptitude, working memory, or language experience scores in either experiment. ERP results from both experiments and preliminary data from EEG analyses provide converging evidence that individuals may build stronger semantically or syntactically driven top-down linguistic predictions when supported by contextual constraints. Given that better predictions can lead to more efficient language processing and better language learning, these data reveal a feature of individuals’ intrinsic behavior that may be leveraged to improve spoken and written language comprehension and acquisition