Re: owl:sameAs use/misuse/abuse Re: homonym URIs

Bernard,

On 15 Jun 2007, at 10:30, Bernard Vatant wrote:
> But is not 'A owl:sameAs B' intended to mean, by OWL definition,  
> that in any context using the semantics of URIs, those semantics  
> are the same for A and B, so whatever assertion is true of  A is  
> true of B, so A and B can be used indifferently.

Huh? No. A and B are *symbols*. They are two different symbols. They  
cannot be used indifferently.

The symbols A and B denote two *referents* A' and B' (glossing over  
the question how exactly they denote those referents). A owl:sameAs B  
means that anything that is true for A' is true for B' and vice  
versa. You can use A' and B' indifferently (because they are the  
same, if you choose to believe the owl:sameAs claim), but not A and B.

We communicate using symbols. In natural language, in RDF, and in  
OWL. Changing the symbols changes what is communicated, even if the  
symbols denote the same thing.

> But if through http protocol  you retrieve "what the owner of A  
> declares is true of A" and "what the owner of B declares is true of  
> B" (read : RDF descriptions of A and B), with no certitude  
> whatsoever if those descriptions are consistent or not, that means  
> http protocol is not a context where A and B have the same semantics.

“A and B have the same semantics” is a very murky statement best  
avoided.

HTTP deals only with symbols (URIs), descriptions of referents (where  
the referent is a non-information resource), and representations of  
referents (where the referent is an information resource). It never  
deals with the referents themselves.

Two symbols having different descriptions doesn't preclude me from  
interpreting them as having the same referent. The same-ness of  
referents never enters the picture at the HTTP level.

> I can live with that, but it seems at least hard to understand and  
> harder to explain, if one judge by the everthread about it.
> OTOH, if one uses owl:sameAs for URIs identifying resources which  
> are *information resources*, then they should actually redirect to  
> the same document. Yes?

That would be a good idea, though technically I don't believe you  
*have to*.

Richard



>
>
> -- 
>
> *Bernard Vatant
> *Knowledge Engineering
> ----------------------------------------------------
> *Mondeca**
> *3, cité Nollez 75018 Paris France
> Web:    www.mondeca.com <http://www.mondeca.com>
> ----------------------------------------------------
> Tel:       +33 (0) 871 488 459
> Mail:     bernard.vatant@mondeca.com  
> <mailto:bernard.vatant@mondeca.com>
> Blog:    Leçons de Choses <http://mondeca.wordpress.com/>
>
>
>
>

Received on Friday, 15 June 2007 14:23:33 UTC