Re: Proposals? Re: use/mention and reification

On Thu, 2002-01-24 at 04:10, Jan Grant wrote:
> Can we get small amounts of words to describe these please? Currently
> there seem to be three proposals knocking around, with a possibility of
> others: (insert "this WG resolves that..." in front of all these)
> 
> 
> Proposal #1:
> 
> 	A resource R in an RDF graph G is said to be the reification
> 	of the statement
> 		S P O .
> 	iff there exist in G the following statements:
> 
> 		R <rdf:type> <rdf:Statement> .
> 		R <rdf:subject> f(S) .
> 		R <rdf:predicate> f(P) .
> 		R <rdf:object> f(O) .
> 
> 	where f is simply the identity function.
>
> Objections to this thus far appear to be (apologies if I'm putting words
> in people's mouths here):
> 
> 	danbri: objects on the basis that the intended interpretation
> 	associated with the "utterer" of S P O may assign different
> 	denotations to S, P and/or O - in other words, this is "broken"
> 	when reifying statements taken from different sources in the
> 	same graph (is that accurate, Danbri?)
> 
> 	danc: objects on the grounds that there's a use/mention
> 	problem here, that it's hard to see how to make a MTetic
> 	interpretation of this work. (again, correct me if I'm wrong).
> 	(See his earlier messages on this) (- If there isn't a way
> 	to make this make sense, then this seems to be a showstopper
> 	objection)

No, it's not a show-stopper objection; I don't see a conflict
between Proposal #1 as stated above and the model theory.

I just don't have any use for that sort of reification;
i.e. I don't see why it's in the spec.
I on behalf of the W3C Semantic Web Advanced Development group,
I think. (I should double-check with the rest of the group.)

If other folks find it useful (or if we're prepared to
accept it as the best interpretation of the current spec
or whatever), then I don't object.

Or... hmm... while I don't see a big conflict with the
model theory, I do see a bunch of stuff that should
be added to the model theory to describe this idea
of reification, and it looks really ugly.

Here's an entailment test:

input:

	<msgid:s@example> <msgid:p@example> "o".

does it rdf-entail this?
	_:ss <http://www.w3.org/1999/02/22-rdf-syntax-ns#object> "o".
	_:ss <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> 
<http://www.w3.org/1999/02/22-rdf-syntax-ns#Statement>.

?

The way I read proposal 1, it does.

In fact, it seems axiomatic that for any literal "oo", there's
a statement with that object. i.e. here's an entailment test:

input: (none/empty)

output:
	_:ss <http://www.w3.org/1999/02/22-rdf-syntax-ns#object> "oo".




> Proposal #2: (Danbri or danc to fill in the blanks..?)
> 
> 	Looks like proposal #1 but f is defined differently.

yeah... f(X) is quote(X)

> 	f(X) = X	if X is a literal

umm... I'm not sure how to quote literals.

> 	f(X) = ?	if X is a URI-labelled resource

I was thinking
  f(<http...>) = "http:..."

but I didn't consider quoting literals and blank nodes.
Blech.

> 	f(X) = ?	if X is a blank node

I dunno.

This macro-expansion form of quoting is messier than I thought.
Maybe TimBL's right... maybe parseType="??:quote" is the
only mechanism that meets my requirements.

> 	jang: might object if f isn't injective (or "one-to-one")

as well you should.

> Proposal #3:
> 
> 	this WG resolves that it's never heard of an eleven-letter
> 	word beginning with "R". (DanC's option A)

I find that acceptable.

> (Other things that have been mentioned in passing include other ways of
> including statements in a "non-assertive" fashion)
> 
> Can we give these all fancy letter names and try again?

1/2/3 are sufficiently fancy for me.

-- 
Dan Connolly, W3C http://www.w3.org/People/Connolly/

Received on Thursday, 24 January 2002 19:01:32 UTC