LINGUIST List 3.952

Fri 04 Dec 1992

Qs: Grammar Eval.; Monostratal; Roommate; Software

Editor for this issue: <>

Directory

1. , Inquiry on Meeting on Grammar Evaluation
2. Rick Wojcik, Re: Monostratal Theories and Tricky Subcategorization
3. Mari Olsen, Roommate wanted
4. Cari Spring, computer program that picks out words

Is there (yet) available any report of the Meeting on Grammar
Evaluation at UPenn in September? I'd appreciate a reference
if so, even a copy if practicable.
Thank you.

Karen Sparck Jones
Computer Laboratory, University of Cambridge
New Museums Site, Pembroke Street
Cambridge CB2 3QG, England

I originally posted the following on sci.lang, but got only one reply to it.

In connection with our development of a GPSG-like parser at Boeing, Phil
Harrison and I have been looking at constructions like the following. The
common thread is that these constructions have an adjective phrase which
subcategorizes the final element and forms a discontinuous constituent with
it.
 How interested are you in solving the problem?
 How tired is he of this subject?
 How eager is John to visit his grandmother?
 I know how eager John is to visit his grandmother.

The problem that we see for monostratal syntax is that the final elements
represent adjective phrases that contain an A-bar gap as the head. For a
GPSG- (or HPSG-)like analysis to work, the gap has to inherit the subcategor-
ization information from its antecedent. The analysis is pretty straightfor-
ward in a transformational theory, since those theories have a level of
syntactic representation that lacks discontinuous constituents. But does
anyone know of work in monostratal syntax that addresses this issue
specifically?

I want to make clear that I am not suggesting that these sentences are in any
way showstoppers for G/HPSG-like theories. In fact, we can think of a few
solutions for them in our system. We already have a special mechanism for
handling left-corner gaps, as in

 Who do you know from Arkansas?
 When, precisely, did he move in 1963?

But we are interested in other people's work that may have focused on the
problem of discontinuous subcategorization dependencies in nontransformational
theories. (Apparently, Dan Flickinger and John Nerbonne's article on "easy"
adjectives in the most recent Computational Linguistics does not address these
kinds of dependencies.) Any comments are welcome.

 -Rick Wojcik (rwojcikboeing.com)

I'm looking for a woman to share a room at the Biltmore for LSA.
Nonsmoker preferred. I will be out of the country Dec. 8-19, so
please reply ASAP.

Thanks.

Mari Broman Olsen
molsenastrid.ling.nwu.edu
molsencasbah.acns.nwu.edu

A friend is looking for a computer program where you can enter a
sequence of letters and the program returns all the words containing
those letters. Program can be Dos or Mac based.

Thanks.

Cari Spring
springbend.ucsd.edu