RE: Rules WG -- draft charter -- NAF

> Not so fast.  What if it is difficult to determine just what objects
> (belonging to a particular class) exist?  Then what does 
> closing (a class) mean?

There are two issues:

1) Closing a class/predicate, such as with N3's "definitiveDoc"
declaration. In such a case (of typically administrative information)
you don't have to deal with indefinite information and applying NAF 
to such a predicate corresponds to classical negation.

2) The semantics of NAF in the presence of indefinite information:
in this case there are still some research issues concerning the
proper concept of preferred/"minimal" model, but clearly NAF(P(a))
fails (cannot be inferred) on the basis of P(a) v P(b).

-Gerd