I've just been doing some work on the representation of meaning and whilst Pinker has some interesting things to say he is, at bottom, a Chomskyist. There is a lot of interesting work going on at the moment on the perceptual foundation of thought -
Lawrence Barsalou is one of the main proponents and one the many advantages of his Perceptual Symbol Theory is that it solves the symbol grounding problem - which is a well known difficulty for many accounts of language to explain how logical operations over arbitrary and meaningless symbols can come to mean, (well, he doesn't really solve it, it just doesn't arise on a perceptual account of meaning).