- From: Seth Russell <seth@robustai.net>
- Date: Sun, 11 Mar 2001 10:58:44 -0800
- To: <jos.deroo.jd@belgium.agfa.com>
- Cc: <GK@NineByNine.org>, <aswartz@upclink.com>, <www-rdf-interest@w3.org>
From: <jos.deroo.jd@belgium.agfa.com> > > > [Graham] > > > Sooner or later, methinks, it is needed that statements are grounded in > > > "real-world" knowledge. How do you suggest that such grounding may be > > > introduced into a system? > > > > > > (I propose it is through axiomatic facts and inference rules.) > > > > [Seth] > > I agree, it is needed. I propose that practical "real-world" knowledge is > > grounded in the effective procedures of interacting active processes. > > > > Logic is great, but survival is better :) > > Mathematics provides methods for reasoning: for manipulating > expressions, for providig properties from and about expressions, > and for obtaining new results from known ones. This reasoning > can be done without knowing or caring what the symbols being > manipulating mean. [David Gries & Fred B. Schneider] Yes, but symbol manipulation is not grounding. Just manipulating symbols without grounding them to reality is but an academic game. If the semantic web (or cloud) is to be useful to us for anything, then it must be grounded in what people want to do. I believe that was the thrust of Grahm's question, but perhaps not. > Escaping into that space, and returning from it in a much > better shape can be a very effective survival mechanism :-) The question is what is the space from which we "escape" into logic and to which we return... methinks that space is the dialogues that users have with their computers and with each other ... those active processess including the intentions and contexts of which they consist must be our grounding. Seth
Received on Sunday, 11 March 2001 14:01:40 UTC