Correction re IR.
You say
> In view of Krovetz and Croft's well-known findings, one might ask, why
> is there so much IR work in WSD?

The fact is, that in relation to IR research as a whole, there's rather
little work on WSD just precisely because, as you remark, mechanisms that
are independently desirable for improving query quality, namely (in general)
adding more query terms to increase the number of conjoint matches, also
incidentally achieve disambiguation. Saying one should have explicit
WSD in IR is a classic example of the case of outsiders (typically
NLPers) wading in and saying X must be the case, without having any
knowledge of the real nature of the task. I am not saying that *no* IR
application requires explicit WSD, and one point of interest to the IR
community is precisely the implications of full text searching for this
issue. But one has to look at the overall evidence, and not look just
at individual tests or types of application. Thus I look forward
to seeing the context for Schutze and Pedersen's forthcoming results.

Karen Sparck Jones