RE: Blank Node Equality Question

No, I am right, Dan's wrong.
(tact, as always, one of my strengths).

Of coruse Dan is right to say that they might represent two different
individuals, so in particular

G
does not entail
_:a owl:sameIndividualAs _:b .

and Dan's message makes the case for that. (Also allowing for the special
cases where G does entail
_:a owl:sameIndividualAs _:b .
)

However, the question presupposed a graph and two blank nodes that were
indistinguishable within it.
We cannot distinguish these blank nodes on the basis of some other graph
since only urirefs and literals are global. Thus we cannot rule out
_:a owl:sameIndividualAs _:b .
and the twins might in fact be the same person.

Given the question statement, and assuming the initial graph G is
consistent, then

G
_:a owl:sameIndividualAs _:b .

is consistent.

Jeremy


> -----Original Message-----
> From: www-rdf-logic-request@w3.org
> [mailto:www-rdf-logic-request@w3.org]On Behalf Of Graham Klyne
> Sent: 21 July 2003 11:11
> To: Jeremy Carroll; jimbobbs@hotmail.com; www-rdf-logic@w3.org
> Subject: RE: Blank Node Equality Question
>
>
>
> Er, I don't think that follows... yes, on the information
> provided you can
> prove the same thing(s) about _:a and _:b, but that doesn't necessarily
> mean they're the same thing.
>
> See also DanBri's response.
>
> #g
> --
>
> At 10:53 21/07/03 +0200, Jeremy Carroll wrote:
>
>
> >I believe so.
> >
> >If the nodes are _:a and _:b
> >If we have a proof:
> >
> >G
> >|-
> >t   (where t is a triple containing _:a)
> >
> >then we can lexically swap _:a and _:b in the proof to get a new
> proof of t'
> >which contains _:b instead of _:a.
> >
> >Jeremy
> >
> >
> >(Note that the hypothesis graph means exactly the same as one in
> which one
> >of the two nodes is excised completely)
> >
> >
> > > -----Original Message-----
> > > From: www-rdf-logic-request@w3.org
> > > [mailto:www-rdf-logic-request@w3.org]On Behalf Of Jimmy
> > > Sent: 21 July 2003 03:02
> > > To: www-rdf-logic@w3.org
> > > Subject: Blank Node Equality Question
> > >
> > >
> > >
> > > If I have two blank nodes - each connected to identical nodes via
> > > identical predicates - then do the two nodes share the same
> properties?
> > > I.E.  Any inferences made about one node are true for the other node.
> > >
> > > --
> > > Jimmy Cerra
> > >
> > > ] "A good decision is based on knowledge
> > > ]  and not on numbers" - Plato
> > >
> > >
> > >
>
> -------------------
> Graham Klyne
> <GK@NineByNine.org>
> PGP: 0FAA 69FF C083 000B A2E9  A131 01B9 1C7A DBCA CB5E
>
>

Received on Monday, 21 July 2003 06:08:41 UTC