Carl Pollard (pollard@zypern.iwbs.heidelbg.ibm.com)
Mon, 4 Jul 1994 10:43:54 +0200
Dear Lynne Murphy, As far as I'm concerned the theory of coordination is a clean slate. What little appears in Pollard and Sag 1994 is BY NO MEANS intended to be definitive (in fact to be honest, it was Ivan's idea -- I didn't want to say ANYTHING about coordination in this book since I don't thinkl we have a well-worked out theory of it yet). My basic feeling is that what is needed is a general theory that works for gapping, nonconstituent coordination, right-node raising, etc., and that so-called constituent coordination is just a simple special case. In any case what you say is right: if a coordinate NP has an index, it does not have any direct connection with the indices of the conjuncts. It seems that in some cases, the content of a coordinate structure is some kind of lattice-join of the contents of the conjuncts, whereas in other cases it is as if there IS NO coordinate node, at least not as far as the content is concerned, bnut rather the contents of the conjuncts are contributed in parallel, in some sense. I would dealy love to see a respectable theory about this subject matter, but i have not seen one yet and I believe it is a gigantic undertaking. Best wishes -- Carl
This archive was generated by hypermail 2.0b3 on Fri Dec 18 1998 - 20:31:39 PST