Item

Abstract

Language contributes to the architecture of the human brain. Whereas some of the perisylvian white matter pathways seem to be hard wired, other develop in parallel with language acquisition. However, studies on language-related neuroplasticity have focused on developmental aspects that are transversal to all languages, ignoring possible language differences. For example, languages like German require online processing of abstract structural information (morphosyntax), anatomically supported by the dorsal arcuate fascicle, whereas languages like English more strongly engage lexical-semantic processing, which involves predominantly ventral fibre tracts. We therefore investigated how languages with different processing demands shape the language network. We compared diffusion MRI scans for three age, sex and education-matched groups with three different native languages: German, English and Mandarin Chinese. Anatomical regions of interest (ROIs) were defined in a template generated for this subject group, both in the inferior frontal gyrus and in the anterior and posterior superior and middle temporal gyri. Using probabilistic fibre tracking, we computed anterior-posterior, fronto-temporal and whole-brain connectivity of these ROIs and compared the respective connectivity strengths and maps of connection probability. We found higher dorsal fronto-temporal connectivity in the German group than in both the English and Chinese groups. Conversely, the English group showed higher ventral connectivity between the posterior temporal cortex and anterior frontal regions. In turn, the medium and short-range connectivity to the neighbouring temporal cortex and to the inferior parietal lobe was higher in Chinese speakers. These differences in connectivity indeed reflect the particular demands of the native language of the individuals as hypothesised. This study is a first indication that the wiring of language-relevant areas depends on the specific demands of each language. Further studies taking into account the genetic background and functional connectivity are needed to understand the implications of the study.