I am most interested in how generative and computational models of language ought to be constrained by certain philosophical problems in the mental representation of linguistic knowledge. My work argues that our most serious difficulty at present is that what we identify as the 'syntax-semantics interface' has no cognitive reality, but is rather a cultural phenomenon, so we do not as yet have a genuinely cognitive theory of language.
To correct this, I believe we must try to reconcile the best of Chomskyan linguistics with the best of the later Wittgenstein. Though this is in many ways profoundly counter-intuitive, the end result is a truly cognitivist research program that lays to rest many of the antagonisms of linguistics and opens up new possibilities for computational models of human intelligence and its origins.
Justifications for a discontinuity theory of language evolution. Biolinguistics 11. 171-220.