UC Davis

AbstractLanguage is a quintessentially human trait. Many decades of neurolinguistic research provided evidence of neural structures which specialize in complex linguistic and cognitive processes supporting human communications. Because the world is multilingual, (Crystal, 2010; de Bot, 2019) a prominent question related to brain processes supporting language is whether the neural representation of language changes as a function of the number of languages one knows. This study attempts to depict a more comprehensive picture of brain plasticity in multilinguals, by integrating behavioral with functional, structural, and diffusion MRI data. The questions investigated stem from newer dual-stream models of language processing that frame brain architecture, supporting language function in terms of a language network (Friederici & Gierhan, 2013; Hickok & Poeppel, 2007b). Based on this framework, language representation for multilinguals compared to monolinguals is investigated within brain regions specialized for language processing (a.k.a. core language nodes; Fedorenko & Thompson-schill, 2014), and regions of domain-generality, associated with language control. Three main findings surface from this investigation. First, monolinguals and highly proficient multilinguals similarly recruit core language brain regions during the processing of native and second languages. These same regions show similar restructuring patterns in grey matter structure and white matter connectivity. Second, compared to monolinguals, highly proficient multilingual speakers show stronger reliance on the cingulo-striatal subnetwork (Dosenbach et al., 2008; Wu et al., 2021) of the cognitive control system, during language comprehension. Decreases in grey matter thickness and volume, along with changes in white matter integrity within this subnetwork, accompany changes in the responsiveness of these regions during language tasks. Finally, contrary to predictions of recent models of bilingual language inhibition and control (Green, 1986, 1998), multilinguals show different patterns of language activation and inhibition. Additionally, these seem to be modulated by language dominance. The implications of these findings on current neurolinguistic theory and models of language processing in speakers of multiple languages (Abutalebi & Green, 2016; Green & Abutalebi, 2013; Grundy et al., 2017; Pliatsikas, 2020) are discussed.