Over at one of her other blogospheric homes, Channel N, fellow ScienceBlogger has posted a link to a great talk on modeling qualitative physics by Ken Forbus. It was one of the earliest of the Cognitive Science Society's virtual colloquia, a series that it has, for some reason, discontinued. "Qualitative physics" is a semi-fancy name for everyday qualitative reasoning, and Forbus focuses on things like spatial reasoning, causal reasoning, and motion, and reasoning about physical processes. It's cool stuff.
When you're done with that, you might want to check out this paper, in which Forbus and Sven Kuehne explore the possibility of using similar modeling techniques to model political reasoning.