Re: Comment: Representing Classes As Property Values on the Semantic Web

Ivan,

I am not sure what you are trying to say. 

public-swbp-wg-request@w3.org wrote on 05/04/2005 09:15:50 AM:

> My (editorial) comment refers to Approach #2, but is also valid 
elsewhere.
> 
> The text (as well as the code) says:
> 
> :LionSubject a :Lion.
> 
> and, sort of, makes the implicit conclusion that this also means
> 
> :LionSubject a owl:Thing.
> 
> However, as far as I understand, this is not true per se *unless* we 
know in advance that 
> we are in OWL DL or OWL Light. 

It is true in all dialects of OWL.  Is there a document somewhere that 
states otherwise?

> If we do not know that about an ontology, then there is 
> nothing that precludes :LionSubject to be also an owl:Class. 

Yes, that much is true for OWL Full, but this does not lead to your 
conclusion above.

> Because the essence of the 
> text is to emphasize using individuals at that point to make it clearly 
DL, I think it 
> would be better to explicitly say:
> 
> :LionSubject
>    a owl:Thing;
>    a :Lion .

Well, that is not incorrect, but it is redundant.  You do not need to 
explicitly say this in any dialect of OWL.  In OWL Full *every* resource 
is an OWL:Thing, in OWL DL being an OWL:Thing is entailed by being and 
isntance of a class.

-Chris