要旨
Multilingual individuals face an ongoing challenge in managing their language system. In order to efficiently communicate, a polyglot brain must have mechanisms that permit the selection of the relevant lexical, syntactic and even pragmatic set for the communicative environment. Behavioural evidence points to an impact of bilingualism on various domains of executive function such as conflict monitoring and task switching, suggesting that domain-general cognitive functions are affected by multilingualism. Consistent with this, functional imaging work on the neural basis of language control in multilinguals has implicated a number of regions associated with the executive control system, including the dorsal striatum, the supplementary motor area and the anterior cingulate cortex. There are also brain structural consequences of multilingualism, although the findings in the literature on structural differences between monolinguals and bilinguals are heterogeneous. In a previous study we observed functional recruitment of the bilateral caudate nucleus during a challenging language control task, namely simultaneous interpreting. Not only is this structure is also a key node in the executive control system, but it has also been shown to be enlarged in bilingual individuals compared to monolinguals. We elected to investigate the relationship between caudate nucleus volume and multilingualism in individuals who speak three or more languages, which we posit may have even greater language control requirements than bilingualism. For this, we acquired structural MRI data (T1-weighted MPRAGE, 1*1*1mm voxels) in 66 multilingual individuals (39 Female, mean age: 25.85 years). Participants reported speaking three or more languages (range: 3 to 9, mean 4.36), and were interviewed on their language acquisition and proficiency levels. In order to have a quantitative, continuous measure of language experience which takes into account more than just number of languages, we calculated weighted sums of age of acquisition (lower receiving higher weight) and proficiency (more proficient receiving a higher weight) ('LEXP'). Imaging data were processed using FIRST for segmentation of subcortical structures. Volumes of the caudate nuclei were extracted and tested for a relationship with LEXP using a partial correlation, controlling for participants' age, sex and total intracranial volume. In the light of previous reports we expected a positive correlation, and a one-tailed test was therefore performed. LEXP was found to correlate significantly with the volumes of both caudate nuclei (left: r=.260, p=.022; right: r=.221, p=.043). Brainstem volume was tested using the same analysis as a null control, and showed no relationship with LEXP (r=.154, p=.117). These results indicate that the additional challenges posed by controlling, and possibly acquiring, additional languages beyond bi- or even trilingualism has an impact on the caudate nucleus. This is consistent with the increasing weight of evidence that points to the role of the dorsal striatum in multilingual language control
