ARG-ST on phrases?

[Adam Przepiorkowski]

Dear HPSGers,

Some 11 months ago, there was a discussion on this list on whether ARG-ST
could/should appear on phrases, in addition to words.  (See
http://hpsg.stanford.edu/hpsg-l/1998/0124.html and other messages in the
thread.)

As far as I can see, there were no strong arguments given for either
position.  Those who prefer to keep ARG-ST on words say that this leads to
more restrictive grammars (an argument cannot be selected on the basis of
what arguments it itself selects), but as noted by Carl Pollard:

> there is already stuff in SYNSEM values that is never selected for that
> nobody ever worries about, CONT|QUANTS, CONT|NUCLEUS, QSTORE, and REL to
> mention just a few.  Why aren't there any verbs which require that the VP
> or sentential complement's CONTENT be a psoa whose QUANTS value contains
> a universal quantifier?

I've been wondering if, in the meantime, any arguments were found for or
against having ARG-ST also on phrases?

Any thoughts on this (also references to existing literature) will be very
welcome.  Please answer to the list or directly to me; if there is enough
interest, I'll post a summary.

Many thanks!

Best,

	Adam P.

--
           ,
ADAM PRZEPIORKOWSKI
----------------------------------------------------------------------
Polish Academy of Sciences    |
ul. J. Ordona 21              |office: (+48 22) 363709 ext. 43
01-237 Warszawa               |home:   (+48 22) 6438372
Poland                        |email:  adamp at ipipan.waw.pl
----------------------------------------------------------------------
URL:  http://www.ipipan.waw.pl/mmgroup/ap.html
----------------------------------------------------------------------