W3C home > Mailing lists > Public > www-rdf-logic@w3.org > August 2002

Re: Classes and predicates as first class objects

From: Ian Horrocks <horrocks@cs.man.ac.uk>
Date: Fri, 16 Aug 2002 09:02:15 +0100
Message-ID: <15708.45447.742414.15278@merlin.oaklands.net>
To: "R.V.Guha" <guha@guha.com>
Cc: Bob MacGregor <macgregor@ISI.EDU>, Deborah McGuinness <dlm@KSL.Stanford.EDU>, www-rdf-logic@w3.org

Guha,

You chose to ignore my second point, so I will repeat it:

    The ability to treat classes/predicates as arguments to other
    predicates is of secondary importance. The crucial thing with RDF
    is that it treats the vocabulary of the language itself as
    standard classes/predicates that can be arguments to other
    predicates. This is beyond the ability of almost all logics. It is
    relatively harmless for a language as weak as RDF, but causes
    fatal complications when more expressive power is added.

Ian



On August 15, R.V.Guha writes:
> 
> Ian,
> 
>  I am sorry, I didn't mean to make this into an RDF vs DL issue. The 
> only reason I focused on DLs is that the proposed model theories for OWL 
> mirror that of DLs rather closely. I apologize if I created the 
> impression that I was trying to pick on DLs.
> 
>  While it is true that text book renderings of FOL do not allow 
> predicates as arguments to other predicates, they also don't make any 
> mention of classes. For RDF, a class is not anything special. So, we 
> should probably separate our treatment of classes from that of 
> predicates.  Also, as the Hayes & Menzel paper shows, it is quite easy 
> to create slight variants of FOL that allow for predicates as arguments 
> to other predicates. Further, as the Fikes McGuinness axiomatization 
> shows, this can be quite easily mapped into a very traditional FOL.
> 
>  Therefore, I humbly submit that allowing predicates & classes as 
> arguments to arbitrary predicates does not take us into dangerous 
> territory. This decision (whether to allow such constructs or not) 
> should, in my opinion, be based solely on their utility to the semantic 
> web. We should probably do a survey of some sort to collect examples 
> where folks have made use of or plan to make use of this feature.
> 
> guha
> 
> Ian Horrocks wrote:
> 
> >First point:
> >
> >Why we suddenly obsessed with DLs? As I have mentioned, they are
> >nothing more than a particular class of decidable subsets of FOL. The
> >ability to treat classes/predicates as arguments to other predicates
> >is beyond the ability of ANY subset of (standard) FOL, decidable or
> >otherwise.
> >
> >Second point:
> >
> >The ability to treat classes/predicates as arguments to other
> >predicates is of secondary importance. The crucial thing with RDF is
> >that it treats the vocabulary of the language itself as standard
> >classes/predicates that can be arguments to other predicates. This is
> >beyond the ability of almost all logics. It is relatively harmless for
> >a language as weak as RDF, but causes fatal complications when more
> >expressive power is added.
> >
> >Ian
> >  
> >
> 
Received on Friday, 16 August 2002 04:04:44 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:52:43 GMT