[CL] XML annex: nothings and the ID namespace

[John F. Sowa]

Pat and Chris,

I'm very sympathetic to alternate views about
what "names" mean.  My preferred ontology is
to treat them like labels that can be acquired
or changed at a whim.  With that option only
variables would "denote" and names would be
related to individuals by a HasName predicate:

   (Ex)(Person(x) & HasName(x,"John")

And HasName would certainly *not* be functional
in either direction, since the same person might
have multiple aliases and multiple people might
have the same name.

Be that as it may, I would cast my vote with
Chris on the following point:

 > I don't know of any presentation of more-or-less
 > standard predicate logic that does not distinguish
 > variables from constants.

That is why I was willing to follow KIF in defining
CGIF, even though I have more sympathy for a pure
variable semantics with variables and names related
by a dyadic HasName predicate.  (And by the way,
I would use that ontology for dealing with physical
objects, such as cats and people, since their names
are too unconstant to be considered constants.)

John