Assumption grammars

[John Fry]

John C. Paolillo <johnp at LING.UTA.EDU> wrote:
> Veronica Dahl and Paul Tarau ?
> >    Linkname: Assumption Grammars: Parsing as Hypothetical Reasoning
> >         URL: http://clement.info.umoncton.ca/html/JAGS/html.html
>
> Aparently this is a linear logic-ish extension to
> the Bin-Prolog of Paul Tarau.  See also
>
> http://www.binnetcorp.com
>
> I'm wondering (a) if people know about this, and
> (b) if anyone can comment on its suitability
> for the kind of things that people (such as
> Dalrymple et al.) want to do with linear logic
> in LFG.

Hi,

I too saw the Dahl/Tarau system on comp.lang.prolog but I haven't
looked into it.  In fact there are a number of programming languages
based on linear logic now, I could point you to references if you'd
like.

Dick Crouch and I have been implementing the Dalrymple et al. system,
and it turns out we don't need a full linear logic programming system
(we've been using standard prolog).  That's because in the Dalrymple
et al. system (1) only a small fragment of linear logic is used, and
(2) the proofs turn out to be highly constrained in certain ways (in
fact, a form of horn-clause resolution suffices).  This is discussed
in a paper by Gupta and Lamping:

Efficient Linear Logic Meaning Assembly. Vineet Gupta and John
Lamping.  COLING-ACL '98, Montreal, Canada. August 1998.
http://www.parc.xerox.com/istl/groups/nltt/semantics/gupta-lamping-coling98.ps

If you're interested in connections between LFG and programming
languages you might also check out the last three chapters (by Vijay
Saraswat, Dick Oehrle, and Mark Johnson, respectively) of the new
volume edited by Mary Dalrymple (1999).

Hope that helps!

Best,

John Fry