revised thinking on RDF reification, imports

From: Dan Connolly (
Date: 05/15/01

Before today's discussion about daml:imports ...
and appropos of Pat's message to
of Tue, 15 May 2001 10:44:11 -0500
about what ontologists want, and quoting issues...

A while ago, I tried to translate RDF to KIF[1]...
I ran into a pile of interesting use/mention issues.

They can sorta be summarized by asking:

	Is the subject of the sentence "Mary hit the ball"
	a word that starts with 'M' or a probably-female person?

i.e. is the rdf:subject of a statement a URI or a resource?

At that time (Aug 2000) I decided that the rdf:subject
of an rdf:Statement was a URI (ala: the subject is a person).

Then I ran into TimBL's cwm design, where he was moving
terms in and out of "contexts". It didn't make sense
in this world-view, but flipping to the other view,
it does make sense.

So I'm now working under the impression that the
rdf:subject of an rdf:Statement is a resource.

So now sw:says and sw:entails can be related
without resorting to wtr... oops; crud; can't
work out the details just now... time for
the teleconference...

[1] KIF as an RDF Schema (in progress Aug 2000)

Dan Connolly, W3C

This archive was generated by hypermail 2.1.4 : 04/02/02 EST