Re: propositions"Stuart C. Shapiro" <shapiro@cs.Buffalo.EDU>
Date: Wed, 11 May 1994 12:13:20 -0400
From: "Stuart C. Shapiro" <shapiro@cs.Buffalo.EDU>
To: firstname.lastname@example.org, genesereth@cs.Stanford.EDU, interlingua@ISI.EDU,
Cc: email@example.com, hans@cs.Buffalo.EDU
Subject: Re: propositions
Pat, et al.:
I would modestly like to recommend my paper, "Belief spaces as sets of
propositions," Journal of Experimental and Theoretical Artificial
Intelligence 5, 2&3 (April--September 1993), 225--235, which is also
available via waters, mosaic, etc. under
The tack I take is that propositions are entities of the domain of
discourse just like people, places, things, properties, acts, times,
events, etc., and are represented (denoted) by some individual
constants, but mostly by functional terms. Moreover, a Knowledge Base
consists of a set of these terms---not sentences. The only sentences
are meta-linguistic tags indicating which propositions are "believed"
by the system, and thus can be used by the inference engine, which
produces newly explicitly believed propositions from existing ones,
including rules, which, themselves, are functional terms representing
The above paragraph just gives the flavor of the paper. It leaves out
the arguments and the formalization. I would appreciate reactions to
the paper much more than flames to this message.
The theory set out in this paper is being fleshed-out and continued by
my student Hans Chalupsky in his dissertation, Belief Ascription by
Way of Simulative Reasoning, which he should be finishing this summer.
There will be a paper based on this dissertation in this summer's
Cognitive Science meeting.
Stuart C. Shapiro, Professor
Department of Computer Science PHONE: 716-645-3935
SUNY at Buffalo FAX: 716-645-3464
226 Bell Hall INTERNET:firstname.lastname@example.org
Buffalo, NY 14260-2000 BITNET: shapiro%cs.buffalo.edu@ubvm