MATURATIONAL CONSTRAINTS ON FUNCTIONAL SPECIALIZATIONS FOR LANGUAGE PROCESSING - ERP AND BEHAVIORAL EVIDENCE IN BILINGUAL SPEAKERS

Changes in several postnatal maturational processes during neural deve lopment have been implicated as potential mechanisms underlying critic al period phenomena. Lenneberg hypothesized that maturational processe s similar to those that govern sensory and motor development may also constrain capabilities for normal language acquisition. Our goal, usin g a bilingual model, was to investigate the hypothesis that maturation al constraints map have different effects upon the development of the functional specializations of distinct subsystems within language. Sub jects were 61 adult Chinese/English bilinguals who were exposed to Eng lish at different points in development: 1-3, 4-6, 7-10, 11-13, and af ter 16 years of age. Event-related brain potentials (ERPs) and behavio ral responses were obtained as subjects read sentences that included s emantic anomalies, three types of syntactic violations (phrase structu re, specificity constraint, and subjacency constraint), and their cont rols. The accuracy in judging the grammaticality for the different typ es of syntactic rules and their associated ERPs was affected by delays in second language exposure as short as 1-3 years. By comparison the N400 response and the judgment accuracies in detecting semantic anomal ies were altered only in subjects who were exposed to English after 11 -13 and 16 years of age, respectively. Further, the type of changes oc curring in ERPs with delays in exposure were qualitatively different f or semantic and syntactic processing. All groups displayed a significa nt N400 effect in response to semantic anomalies, however, the peak la tencies of the N400 elicited in bilinguals who were exposed to English between 11-13 and >16 years occurred later, suggesting a slight slowi ng in processing. For syntactic processing, the ERP differences associ ated with delays in exposure to English were observed in the morpholog y and distribution of components. Our findings are consistent with the view that maturational changes significantly constrain the developmen t of the neural systems that are relevant for language and, further, t hat subsystems specialized for processing different aspects of languag e display different sensitive periods.