For video footage from past you can visit the individual event pages, or go to our YouTube Channel

To filter by event category, click on the event category link in the table below or use the menu on the right.

List of Past Events

Creating and navigating structure in real time

Matthew Wagers

Tuesday, March 04, 2008, 01:00pm - 02:00pm

University of Maryland, Department of Linguistics

Copy to My Calendar (iCal) Download as iCal file
 

 

Understanding language is generally rapid, accurate and seemingly effortless. This capacity is supported by an ability to incrementally recover the structured representations that guide interpretation. However a key challenge for encoding and using the structure of novel expressions in real time is moving between the constituents of the representation in a structure-sensitive but efficient manner. In this talk I will discuss how structurally sensitive real time processes are and present three experimental case studies that illustrate where structure-dependent grammatical constraints succeed and where they fail. In particular I will argue that there are important architectural distinctions between prospective, predictive structure building mechanisms and retrospective, retrieval-based structure building mechanisms that influence how decisions are made to analyze the input. Evidence from behavioral studies on the formation of wh-dependencies and the establishment of subject-verb agreement supports the idea that prospective processes are generally most faithful to the grammar, while retrospective processes are prone to the intrusion of ungrammatical analyses. In support of this distinction, I will situate the behavioral findings in the context of results from formal memory models as well as neurocomputational models of combinatorial representation.  Mapping the border between grammatical fidelity and grammatical fallibility in timecourse studies provides crucial information about how large, complex structures are encoded. By better understanding the encoding processes, I will argue that we can strengthen the link between linguistic theory and experiment.

 

Matthew Wagers