RE: DDTF/layering: weak class theory seems good enough (5.3, 5.10)

From: patrick hayes <phayes@ai.uwf.edu>
Subject: RE: DDTF/layering: weak class theory seems good enough (5.3, 5.10)
Date: Thu, 30 May 2002 14:09:07 -0500

> >
> >This would, in effect, end up with a very non-standard logic for the
> >ontology level.  We would also be doing a grave disservice to the upper
> >levels of the layer cake.  We would, in effect, be taking some of the bad
> >decisions made in RDF and reusing them, making it even harder to get to the
> >rule or logic level.
> 
> Could you briefly summarize what those bad decisions were? Seems to 
> me that there is nothing in RDF itself that prevents one extending it 
> to full FOL, in fact to full ISO-KIF if one wishes to go that far 
> (which is a subset of full Lw1w). Of course this would be a syntactic 
> extension, but there is nothing in RDF that says it cannot be 
> syntactically extended.
> 
> Pat

Well, how about:

1/ Everything in RDF is triples.  
   1a/ Every layer will use RDF syntax.
2/ Every triple in RDF is asserted.

Yes, if you can extend the syntax this all goes away, but, of course, we
have been fighting for a long time to reverse the ``all syntax is RDF
syntax'' bad decision.

peter