RE: from the RDF model theory: Re: [httpRange-14] What do HTTP URIs Identify?

Jonathan,

> From: www-tag-request@w3.org [mailto:www-tag-request@w3.org] 
>
> If two people using the same URI might be referring to 
> _different concepts_, then the model that you would need 
> would someone need to represent a concept as, for example, a 
> node (which is unlabelled) and has a property called "my URI" 
> and that this property is not uniquely identifying i.e. it 
> would not be an "inverse functional property" in OWL parlance 
> (oh I hope i've got this new OWL terminology correct :-)). 
> Well you _could create_ such a representation but that 
> wouldn't be the way that RDF, for example, models the world.

The model you would need is one of possible worlds, which is in fact how
RDF models the world. The two parties above can be seen to be describing
two different, but possible, worlds.


> From the RDF model theory http://www.w3.org/TR/rdf-mt/ (1.2)
> 
> "Urirefs are treated as logical constants, i.e. as names 
> which denote something (the things are called 'resources', 
> but no assumptions are made about the nature of resources.) 
> ... Similarly, the model theory given here has no special 
> provision for tracking temporal changes. It assumes, 
> implicitly, that urirefs have the same meaning whenever they 
> occur.To provide an adequate semantics which would be 
> sensitive to temporal changes is a research problem which is 
> beyond the scope of this document. "
> 
> Hope this helps.

It did help, but I don't think the MT gives credence to what you're
arguing. What you've quoted described a state of affairs that is
necessary for RDF inferencing to make sense, not a state of affairs for
distributing URIs to Resources. Thus, it is a statement of requirements
for RDF entailments to be sound ones, not a statement of what is the
case. 

If you look further down to section 1.3:

[[[
The following definition of an interpretation is couched in mathematical
language, but what it amounts to intuitively is that an interpretation
provides just enough information about a possible way the world might be
- a 'possible world' - in order to fix the truth-value (true or false)
of any ground RDF triple. It does this by specifying for each uriref,
what it is supposed to be a name of; and also, if it is used to indicate
a property, what values that property has for each thing in the
universe. Together with our assumed fixed interpretation of literals,
this is just enough to fix the truth-value of any ground triple, and
hence any ground RDF graph.(We will show how to determine the
truthvalues of non-ground graphs in the following section.) Notice that
if we left any of this information out, it would be possible for some
well-formed triple to be left without a determinate value; and also that
any other information - such as the exact nature of the things in the
universe - would, regardless of its intrinsic interest, be irrelevant to
the actual truth-values of any triple in the world being specified.

Asserting an RDF graph amounts to claiming that it is true, which is
another way of saying that the world it describes is, in fact, so
arranged as to be an interpretation which makes it true. In other words,
asserting a piece of RDF amounts to asserting a constraint on the
possible ways the world might be. Notice that there is no presumption
here that any RDF graph contains enough information to specify a single
unique interpretation. It is very difficult, and usually impossible, to
assert enough in any language to completely constrain the
interpretations to a single possible world, so there is no such thing as
'the' unique RDF interpretation.

]]]

and this paragraph (from the ellipsis of your quote by the way), the
meat of which is in the first sentence:

[[[
An interpretation assigns meanings to symbols in a particular vocabulary
of urirefs. Some interpretations may assign special meanings to the
symbols in a particular namespace, which we will call a reserved
vocabulary.We will use two reserved vocabularies in this document,
defined using the Qname syntax with the prefixes rdf: and rdfs: 

]]]

then you see that the Model Theory is very much in the business of
having ambiguity eradicated, before inferencing begins. Thus the Model
Theory acknowledges ambiguity as the case; we should expect nothing less
from a denotational semantics. Moreover, were it the case that URIs do
in fact name one and only one thing, a model theoretic semantics would
not be necessary for RDF to function; something I've claimed elsewhere
on the tag list. It seems to me what your arguing for is for the unique
RDF interpretation to only ever be the case, which the MT explicitly
acknowledges as generally impossible to come by.

We need to accept ambiguity of identifiers as a fundamental constraint
in an open information system, and move to see what can be built and is
useful under that constraint. This is in the  same way the internet
accepts latency and the web accepts the absence of hypertext backlinks.
Two immediate avenues I'm interested in are weighting RDF graphs with
probabilities and using RDDL to build a network of retrievable RDF
claims about resources from third parties.

A REST system deals with this matter by giving the naming authority the
right to say what is identified. But as others have pointed out, the
valuable metadata will come from third parties, not from the naming
authority. This is to avoid important problems with trusting the
authorities claims out of hand, but I don't think it precludes the
authority having the last word on what is being named. 

regards,
Bill de hÓra

Received on Saturday, 3 August 2002 11:56:09 UTC