Re: "meaningless terms" verbage for Primer

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


----- Original Message ----- 
From: "ext Jan Grant" <Jan.Grant@bristol.ac.uk>
To: "Patrick Stickler" <patrick.stickler@nokia.com>
Cc: "fmanola" <fmanola@mitre.org>; "w3c-rdfcore-wg" <w3c-rdfcore-wg@w3.org>
Sent: 10 December, 2002 13:54
Subject: Re: "meaningless terms" verbage for Primer


> On Tue, 10 Dec 2002, Patrick Stickler wrote:
> 
> >
> >
> > [Patrick Stickler, Nokia/Finland, (+358 40) 801 9690, patrick.stickler@nokia.com]
> >
> >
> > ----- Original Message -----
> > From: "ext Jan Grant" <Jan.Grant@bristol.ac.uk>
> > To: "Patrick Stickler" <patrick.stickler@nokia.com>
> > Cc: "fmanola" <fmanola@mitre.org>; "w3c-rdfcore-wg" <w3c-rdfcore-wg@w3.org>
> > Sent: 10 December, 2002 12:38
> > Subject: Re: "meaningless terms" verbage for Primer
> >
> >
> > > On Tue, 10 Dec 2002, Patrick Stickler wrote:
> > >
> > > > If there is no machine interpretable interpretation, then IMO
> > > > there is no interpretation whatsoever. Eh?
> > >
> > > This seems to be a persuasive argument for dropping language tags.
> >
> > I don't follow. Though the language tags do not affect the
> > denotation of typed literals, they have consistent and unambiguous
> > interpretation by machines (even if that interpretation is
> > disjuct from the datatyping interpretation of the typed literal.
> >
> > On the other hand, if some term has no consistent machine
> > interpretation, in any way, at any level, then it is useless
> > as part of a solution for the global interchange of knowledge
> > for which RDF is supposed to serve as a foundational component.
> >
> > No?
> 
> I'm sorry, I thought you were saying that entailments were the only way
> to give machine "meaning" to RDF; the conter-position is that various
> bits of RDF have a meaning on some other level (eg, use for
> documentation in the cases of comment). That is, "meaning" as people
> generally understand the term, not as in, rules in the semantics
> document.

I have no problems with allowing documentation for humans, but
RDF is IMO first and foremost about machines and what machines
can make of RDF statements.

And yes, entailments are not the only way that machines can
interpret RDF.

> I'm still not convinced that "10"@en^^xsd:integer and
> "10"@fr^^xsd:integer would get treated the same way by all application
> programmers; 

Well, it depends on what you mean by "treated".

If they're taken to denote different values, then that's a bug 
in the application, plain and simple.

However, the language tag might be used to filter some values
from other values, for whatever reason. This is not particularly
useful in the case of the xsd:integer datatype, but for other
datatypes, which contain natural language content which is
otherwise also constrained by the lexical grammar of the
datatype, this is a very useful bit of functionality. E.g.

  some:thing x:shortLabel "Car"@en^^x:shortToken .
  some:thing x:shortLabel "Auto"@fi^^x:shortToken .

Nevertheless, the interpretation
of the typed literal both as to the value it denotes and as
to which language differentiated set it belongs to are both
unambiguous and consistent for machines according to the 
relevant specs. 

> there may be a consistent machine interpretation of
> these, but if that's because there's only one implementation that deals
> with these at the moment then I don't see that as being a particularly
> valid argument :-)

No, it's because the RDF, XML and relevant datatype specs jointly and 
explicitly say what the valid interpretations are, and every implementation
should conform to the specs (even if there is only one implementation ;-)

Cheers,

Patrick

Received on Tuesday, 10 December 2002 07:15:12 UTC