W3C home > Mailing lists > Public > www-rdf-comments@w3.org > July to September 2002

Re: Datatyping

From: Patrick Stickler <patrick.stickler@nokia.com>
Date: Fri, 27 Sep 2002 10:05:24 +0300
Message-ID: <002a01c265f4$40759af0$e180720a@NOE.Nokia.com>
To: <seth@robustai.net>
Cc: <www-rdf-comments@w3.org>

I think that we can get the MT to say just about anything
we want. I'll leave it to Pat and others to comment more about
that.

I'm working under the assumption that terms in RDF statements
are intended in some way to reflect the world (not components
of structured markup) and that assertions such as those made by 
rdfs:range are saying something about the thing in the world 
denoted by the object of the property.

If the object of the property has fixed meaning, then clearly
there is a fundamental conflict in the core machinery.

It is true that one way to "punt" on the whole issue is to
say that inline literal nodes have no interpretation in RDF
whatsoever and are simply semantic "wildcards" for applications
to interpret as they like. I.e. they denote neither strings
nor values. They denote nothing, mean nothing, and any assertions
in RDF regarding their meaning are vacuous. They are just syntactic 
shadows in the abstract graph and the RDF MT does not license
any entailments whatsoever for statements containing them.

Now, I don't consider that to reflect the goal and purpose
of RDF as a language for expressing statements about the
world such that the truth of those statements can be tested
and serve as the basis of decisions, but perhaps that's the
best we can do for now...

[Patrick Stickler, Nokia/Finland, (+358 40) 801 9690, patrick.stickler@nokia.com]


----- Original Message ----- 
From: "ext Seth Russell" <seth@robustai.net>
To: "Patrick Stickler" <patrick.stickler@nokia.com>
Cc: <www-rdf-comments@w3.org>
Sent: 26 September, 2002 19:10
Subject: Re: Datatyping


> Patrick Stickler wrote:
> 
> >>Why does the MT *need* to make the triple drawn to the LexicalNode 
> >>invalid in prescence of a range constraint ?   
> >>    
> >>
> >
> >Because the range assertion says that the object of the property
> >is a member of the particular class, and in the case of a datatype
> >class, its RDF Class extension is the value space. And a lexical
> >node is not a datatype value, but a string.
> >
> It seems to me that the MT's range entailments cannot be applied to 
> LexicalNodes at all.   This is because however hard we try we simply 
> could not draw  the arrow {uuu [rdfs:type]  zzz }  as prescribed by the 
> entailment rule [rdfs3] in the case where uuu is a LexicalNode.    Since 
> we cannot draw an arrow from a LexicalNode, I propose to change [rdfs3] 
> to exclude such an erronous entailment .... something like I have 
> depicted in my new diagram [3].
> 
> [3] http://robustai.net/mentography/jenny_mt_rdfs3.jpg
> 
> I think this works if  you'all consider a LexicalNode not to be or 
> rdf:type rdfs:Resource.   Do you?   ... and could the MT be changed as I 
> propose?  ... and would that solve your concern above?
> 
> .... and thanks for your comments in my blog :-) .
> 
> Seth Russell
> http://radio.weblogs.com/0113759/
> 
> 
Received on Friday, 27 September 2002 03:06:56 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 21 September 2012 14:16:30 GMT