November 15, 2017
Topic: Brain imaging in research on language
Speaker: Dr. Elissa Asp, English & Linguistics St. Mary’s University, Halifax, NS
Time: 11:30am – 1:00pm
Location: DSB -505
In presenting a rationale for studying ‘language in its reality’ in cognitive neuroscience, Roel Willems (2015: 7) describes natural language use as ‘dirty and complicated’. He contrasts it with the ‘sterile’ samples typical of experimental work and argues that the dirty stuff is worth investigating not just because it is after all the ‘real stuff’, but also because what we learn from controlled samples in the lab may not, or only partially, translate to ‘real language’ used by people and supported or impaired in real brains. In this talk, I explore the value and challenges of venturing into the relative wilderness of cognitive neuroscience of natural language use through review of imaging studies employing tasks with (varying) claims to ecological validity such as scene description and picture naming, and through discussion of current models of the neurocognitive networks supporting lexical and syntactic representations and tasks. One goal is to show that our statistical analyses of MEG data – which cluster regions together according to their time courses – can introduce a bit of elegance and simplicity into relatively natural language data by reducing ‘functional connectivity’ to temporal co-activation across the whole time course. However, I will also show that statistical analyses can only partially tame the wilderness. Grounding in theoretical perspectives in linguistics, the neurobiology of language and other domains, and psychometrics for designing studies and evaluating results are also necessary for there to be progress in the cognitive neuroscience of natural language use.
Willems, R. (ed.)(2015). Cognitive Neuroscience of Natural Language Use. Cambridge University Press.