This Language Log post collects some interesting statistics from a couple of corpus studies into the empirical differences between male and female language use. The caveat at the end about the culturally bound, public-space construction of appearances -- rather than any 'natural' differences -- is important to note.
