Neural changes during short-term novel spoken and sign language learning: fMRI and DTI studies

Abstract

It is conventionally thought that language processing is supported by two different brain areas: Broca's area for language production, and Wernicke's area for comprehension. Recent neuroimaging data, however, has shown that the brain networks underlying language processing are more complex than initially thought and closely integrated for spoken language and sign language. In this thesis, we conducted two longitudinal studies at three different approaches, behavioural, functional and structural imaging, which investigate the neural processing underlying spoken and sign language learning. In the first study, Magnetic Resonance Imaging (MRI) brain scans were taken from 20 English-speaking healthy volunteers before and after attending a novel spoken language-training course, learning phonetically minimal distinctions in Arabic that are allophones in the participant’s first language, of one hour for three consecutive days. The same paradigm was applied in the second study for 26 Englishspeaking healthy volunteers who were taught British sign language. Again, minimal contrasts between sign language gestures were learnt. Behavioural results for both studies show performance improvement in all variables including a phonetic discrimination task, and a pronunciation task for spoken language, and a gesture discrimination task and a signing task for the sign language group. Functional MRI results illustrate significant brain activity increasing in multiple language regions including left inferior frontal gyrus (IFG), bilateral middle temporal gyrus, left angular gyrus and the right cerebellum. Moreover, functional connectivity increased significantly after training between these areas. These areas also show significant DTI changes represented by increasing fractional anisotropy (FA) and decreasing in mean diffusivity (MD) and radial diffusivity (RD). The fMRI and DTI changes are correlated with the participant’s behavioural improvement. These findings reveal a high degree of similarity in the neural activity underlying signed and spoken languages. Furthermore, the rapid neural changes noted as an impact of short-term learning oppose the conventional belief that brain changes associated with language learning require long training periods.