Regression in Second Language Acquisition and Loss

Given that over half of the world’s population is bilingual or multilingual, and that this population includes many “late learners”, the study of multilingual acquisition and loss has become critical, especially from the perspective of brain function.

Claims of fundamental differences between first and second language acquisition persist, with explanations of brain plasticity loss and closures of the language acquisition mechanism in adulthood. These hypotheses align with claims of a Critical Period for language acquisition and a “First-in/Last-out” view of language change in the individual. But how solid are these claims?

In this presentation we begin with a review of recent findings on the neurophysiological organization of multiple languages in healthy individuals, showing that there are similar underlying structures and broad convergence of brain areas for language processing, regardless of age of acquisition, but aligned to language proficiency and use.

We then present findings on heritage language speakers, i.e. children of immigrants who have been exposed to a minority home language (Spanish) since birth, retain fluency in that language, but have become dominant in the majority language (English). We examine whether and how the first-learned language, Spanish, is vulnerable to changes that may be traced to influence from the later-learned language, English. Using EEG (Electroencephalography) methodology, which measures voltage changes on the scalp during a listening task in the home language (Spanish) we test two bilingual populations: heritage speakers of Spanish who are dominant in English; and native speakers of Spanish who are fluent in English but still dominant in Spanish.

Our findings show that brain responses in both populations exhibit changes in the processing of the home language. This contradicts the view that first-learned languages, acquired in childhood, should show more robustness vis-a-vis later-learned languages, a view that is consistent with both Critical Period and First-In/Last-Out hypotheses. Instead our results show that changes in Spanish can be predicted by amount of current use of English: the greater the amount of English used, the greater the change in brain response to Spanish. Our results support other psycholinguistic and neurolinguistic studies that have found permeability in the first-learned language and point to neural flexibility, rather than closures of language abilities, even in adulthood.