W3C home > Mailing lists > Public > w3c-rdfcore-wg@w3.org > October 2001

Re: Properties not predicates (was Re: PRIMER: draft data model section)

From: Pat Hayes <phayes@ai.uwf.edu>
Date: Wed, 24 Oct 2001 16:56:23 -0500
Message-Id: <p05101070b7fce82504e0@[]>
To: Brian McBride <bwm@hplb.hpl.hp.com>
Cc: w3c-rdfcore-wg@w3.org
>Pat Hayes wrote:
>>>Just checking we are clear what is being renamed.
>>Well, I thought I was clear, but now I am completely confused.
>>>  M&S uses the term predicate for a component of a statement.
>>That means a triple, right? Or a [node/edge/node] combination in an 
>>RDF graph. Or does 'statement' mean something else? Right now, as I 
>>understand it, there are triples in Ntriples,  pieces of graph in 
>>the graph syntax, and more complicated pieces of syntax in RDF/XML. 
>>Which of these is called a statement?
>You have put your finger on a key question, that I think we need to 
>get clear if  we are to reconcile the language of the old M&S with 
>the new model theory.
>We are dealing here with two formal models.  M&S has a formal model, 
>and we have the new model theory.  I think that M&S has the concept 
>of statement and the new model theory does not.
>My *personal* reading of M&S is that statement and triple meant two 
>different things.

That is how I read it also.

>   o a statement is an abstraction; its a tuple with three 
>components, subject, ...
>   o a triple is a concrete representation of a statement, e.g. state 
>in a computer memory, markings on paper etc.
>[I'm not defending this model; I'm just trying to express it.  It 
>may well be full of contradictions]

No, it seems very coherent to me.

>How do these concepts relate to the new model theory.  Well, a 
>triple in n-triples is pretty clearly still a triple.  What is an 
>arc in a graph?  Not a statement, I think.  If I write:
>   <:sky> <:is> <:blue> .
>   <:sky> <:is> <:blue> .
>I clearly have two triples.  I would also have two arcs in one of 
>your graphs (until it got tidied).  But to M&S, (as I interpret it, 
>others would disagree) there is only one statement, because both 
>triples denote the same statement.
>I find no concept similar to statement in the new model theory.

It would be a class of triples which satisfy the same Ntriples BNF, 
or something like that, right? The MT would work smoothly if we were 
to add this notion. The point being that the meaning of any triple in 
any interpretation depends only on its syntax, so we could just talk 
of the interpretation of a statement or of a triple-token of that 
statement and get the same answer.

This all works fine until we get to reification, at which point it 
all goes horribly wrong, because the only useful sense of reification 
would be about tokens (triples) , but the only one we have actually 
got is about statements.

IHMC					(850)434 8903   home
40 South Alcaniz St.			(850)202 4416   office
Pensacola,  FL 32501			(850)202 4440   fax
Received on Wednesday, 24 October 2001 17:56:27 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 20:24:05 UTC