Monadic predicates and disjointness of properties and classes

This is a random thought, maybe irrelevant.

I seem to recall that there has been some discussion about whether or not 
Properties and Classes should be disjoint (e.g. in ontology languages built 
upon RDF).

I note that enforcing this may be problematic in light of our previous 
decision to implement monadic predicates using rdf:type.  Effectivey, the 
class extensions of RDF classes are monadic predicate value sets, and the 
relational extensions of RDF properties are dyadic prediciate value-pair sets.

I don't know if there are any potential "gotcha"s lurking here.

#g


------------
Graham Klyne
(GK@ACM.ORG)

Received on Monday, 4 March 2002 14:27:59 UTC