Language learning from the brain out
The typical fMRI study of auditory-verbal language functioning returns what is now so commonly found as to constitute “the usual suspects” of classic language cortex (Broca’s and Wernicke’s areas) plus fronto-parietal regions associated with attention and memory encoding. Far less is known about how this network emerges. One approach is to study children, who are in the process of acquiring or refining language skills. Large-scale study of children’s language processing has documented age-related changes in networks that already look remarkably adult-like as early as age five. A second approach is to look at language learning, unconfounded by maturational change (i.e., in adults). My lab has been implementing natural language paradigms that capitalize of principles from the statistical learning literature to examine learning as it happens in the scanner to examine the emergence of this skill in adult learners. This work demonstrates differences between brain networks that are engaged when learners are provided consistent statistical regularities in the input and when they are not. Further, the network revealed during statistical learning depends in part on the type of statistic the brain is given to process. We also demonstrate that learners who are given input that contains multiple statistical regularities (e.g., those governing identification of word units and others for word order) shift their processing strategy over time to reflect attention to different aspects of the input and that lateralization shifts, first seen in children, also occur in adults learning engaged in statistical learning of language.
Dr. Elena Plante is a Professor and Head of the Department of Speech, Language, and Hearing Sciences at the University of Arizona.
The seminar is free and open to the public.