Linguistics Department

Job Candidate - Jonathan Brennan (The Children's Hospital of Philadelphia) “Modeling the neural correlates of incremental sentence interpretation”.

Fri, February 10, 2012 | UTC 1.118

12:00 PM - 2:00 PM

A basic question for neurobiological models of language is understanding how the brain combines words to make complex representations. Linguistic theory divides this task into several computations including those that build sentence structure (syntactic structure-building) and those that compose word meanings together (semantic composition). The close relationship between these computations, however, has posed a strong challenge for efforts to distinguish them at the neural level. To address this issue I consider a hypothesis derived from work in computational linguistics that these computations may dissociate when considered as operations that are applied incrementally during language comprehension. Under such an approach, different theories of the relationship between syntactic and semantic computations carry distinct predictions about incremental processing load. In this talk, I show how predictions from formal, psychologically grounded, models of incremental parsing can be tested against brain data from fMRI and magnetoencephalography (MEG) experiments. Results from studies where participants perform relatively natural tasks, like reading a story, provide novel evidence concerning the localization and timing of brain activity involved in building complex linguistic representations. Furthermore, comparing brain activity with predictions from parsing models based on two different grammatical architectures offers evidence favoring theories which draw a distinction between structure-building and semantic composition, as opposed to those which do not. As such, this work forms part of a broader goal to develop new methods for using brain data to evaluate models of linguistic competence and performance.

Bookmark and Share