I agree with Geoff and Lou that a corpus reader would be a good idea--
no one has yet mentioned what always seems to me a principal function
of such readers, whoich is to reprint older historical material, often lost
from view, but which was actually seminal but is now neglected. this is
a moral duty since people have short memories in any field connected to computing. Im currently involved in two readers; one on MT (with Somers and Nirenburg, in press, and already mentioned by Harry in this discussion I think) and the
other on the Lexicon with James Pustejovsky, both for MIT Press and years late, but in both cases we really did scour about for old papers that we thought people really ought to know about but usually didnt. I dont think this aim is inconsistent with the one of 'where should a beginner start?' at all.
Yorick Wilks
This archive was generated by hypermail 2b29 : Wed Jun 20 2001 - 10:27:36 MET DST