The past few days, I'd been thinking that maybe you could group words into semantic categories based on the syntactic contexts . . . something exactly like this.  Prof. Dekang Lin and his former student Prof. Patrick Pantel's websites have some pretty useful tools for retrieving semantic relationships learned from text.  Don't know how I haven't come across this before.  Here's the paper.

It's something I should have come across by now, but better late than never.

I've also been reading a couple of books lately: Fuzzy Grammar and Grammatical Constructions: Back to the Roots.  Fuzzy Grammar includes excerpts that make a good case for fuzzy syntactic categories.  Also, it seems that Noam Chomsky himself suggested degrees of grammaticality even back in 1965.  The other book I'm still working on.