Re: reification test case

From: "Seth Russell" <seth@robustai.net>
Subject: Re: reification test case
Date: Wed, 6 Feb 2002 07:42:17 -0800

> From: "Peter F. Patel-Schneider" <pfps@research.bell-labs.com>
> 
> > This debate does bring up an interesting question:  What is RDF?  My view
> > is that RDF is what is stated in RDF M&S
> (http://www.w3.org/TR/REC-rdf-syntax)
> > and that other documents (including email discussions) are not part of
> > RDF.
> 
> Well M&S does not explicitly state that there *cannot* be more than one
> reification of a given statement.  If it did, then you might have a case.

Well this is yet another problem with M&S.  M&S is, to put it bluntly, a
mess --- incoherent, inconclusive, and inconsistent.  As far as I am can
tell, however, the wording of M&S consistently indicates that there is
*the* reification of a statement, i.e., not several possible reifications
of a statement.  (M&S also indicates that there is at most one statement
with a given subject, property, and object.)

> > So I'm still looking for wording in RDF M&S that indicates that there can
> > be more than one reification of a given statement.
> 
> People can say anything they want about statements,  we cannot smush all
> those together on one big RDF node.  It doesn't work.   And for the record,
> Libby made an excellent summary of this long standing debate at:
> 
> http://ilrt.org/discovery/2000/11/statements/
> 
> I think any reasonable person who reads that dialogue would come to the
> conclusion that there is far more support for statings being multiples than
> the other way around.  

This may be.  I'm not arguing against this view, at least not here.

> In fact, outside of legalistic arguments, I cannot
> find a single serious argument for your position.

Well one might argue that

1/ there is at most one statement with a given predicate, subject, and object 
2/ each triple, i.e., each potential statement, has at most one reification
   in the form of a member of rdf:Statement

Why?  Well precisely so that one can uniquely identify a potential statement
and make statements about that statement.  For example, one could in this
way relate the statement to log:false via log:truthvalue, or use it in a
log:implies construct.

Is this a good way of doing this?  Not in my book, but it appears to have
proponents.

> What is really motivating you to lobby against fixing this?

Well I don't see this as *broken* so I'm not lobbying for any change.  I'm
not lobbying against the change also.  However, I am arguing that it would
be a *change*, not a fix or a clarification.  It may be that the RDF Core
WG will decide on the change.  If so, fine, or at least I would go along
with the change.  If not, also fine, at least by me.

> Seth Russell


Peter F. Patel-Schneider
Bell Labs Research

Received on Wednesday, 6 February 2002 11:27:29 UTC