Colloquium: Procedures for Computing in Syntax and Semantics

January 29, 3:30-5pm, Cobb 201
Tim Hunter, University of Maryland, College Park

In this talk I will show how expressing ideas in a computationally explicit manner makes possible (i) insights into the interactions between independently-motivated linguistic proposals that would otherwise go unnoticed, and (ii) investigations of the interactions between linguistic theory and other areas of cognitive science. The bulk of the talk will show how this approach can provide a unified explanation for three apparently unrelated facts: the inaccessibility of intermediate projections (X' nodes) for syntactic movement (Chomsky 1995), the puzzling way adjunction interacts with this restriction, and the impossibility of local movement ("anti-locality"; Abels 2003). These properties all follow naturally if we take as a starting point the computationally explicit grammatical formalism from Stabler (2006), and make the minimal modifications required to permit a neo-Davidsonian semantics where the interpretation of arguments relies on their structural positions. In the rest of the talk I will discuss some experimental work investigating the algorithms used in the verification of quantificational statements. In this work a precise understanding of *what* is being computed (provided by semantic theory) is integrated with a precise understanding of *how* such computations could and could not be carried out in the mind (provided by independent psychological results). The results suggest that a complete theory of semantic competence will go beyond the specification of recursively computed truth conditions.