Prof. Gregory Kobele | Relating linguistics and the brain

From the perspective of a cognitive scientist, the raison d'etre of linguistics is to provide a high level theory of the human sentence processing mechanism (HSP). Viewed in this way, linguistic theory should inform and in turn be informed by work in the psychology and neurobiology of language. This has proven difficult to achieve, because it is not at all clear how to link the intentionally atemporal high level descriptions of linguistic theory to anything close to an online model needed to make predictions about the time course of language processing.

In this talk I will discuss work (both mine and others') on compiling high level linguistic grammar descriptions (written in the style of Chomsky's minimalist program) into algorithms for parsing sentences, on deriving online sentence processing complexity measures from linguistic gramamrs, and on linking parsing algorithms to fMRI data. There may be many distinct parsing algorithms (bottom-up, top-down, left-corner, etc) realizing the same grammatical specification. This provides an opportunity to explore how the different predictions made by these different algorithms line up with experimental data, as well as how differences in linguistic analyses relate to different predictions about behavioural and neurobiological data.