RE: Can Resource be the top of our ontology ?

Danny Ayers wrote:
> But surely by talking of XTM, you are identifying it? When you make the
> assertions about it, you want the assertions to 'stick' to XTM,
> so in effect
> you are associating the assertions with an identifier. If you want to
> examine those assertions about XTM you need to have some space you can
> address as 'XTM' to retrieve the assertions...surely?
>
It is true that I am identifying XTM by talking about it - if nothing else,
I am identifying it by name. Hopefully I am identifying it unambiguosly so
that any reader would understand what this XTM 'thing' really is. In fact,
XTM allows me to provide a number of addressable resources which describe
the XTM concept for just that purpose.

However, what I am not doing is defining a location for XTM. The concept of
XTM is not addressable. Just like the concept of a major brand such as Nike
is not addressable. There may be lots of things which describe this concept
(the XTM Spec, the Nike website) but these resources are not the concept,
merely some description or representation of it.

Of course, the <topic> element I create in my topic map for the concept
'XTM' is a resource. But it is only a proxy for the XTM concept. The element
can be addressed, but it can only describe the concept, it cannot *be* XTM.

Being able to clearly distinguish between when a resource *is* the subject
under discussion and when a resource *describes* the subject under
discussion is crucial not only for human understanding but also for
automated processing and having a distinction between the two (both
conceptually and syntactically) is a nice feature of XTM.

Cheers,

Kal

> ---
> Danny Ayers
> http://www.isacat.net
>
> <- -----Original Message-----
> <- From: www-rdf-interest-request@w3.org
> <- [mailto:www-rdf-interest-request@w3.org]On Behalf Of Kal Ahmed
> <- Sent: 10 April 2001 02:04
> <- To: Danny Ayers; Seth Russell; RDF-IG
> <- Subject: RE: Can Resource be the top of our ontology ?
> <-
> <-
> <- Danny Ayers wrote:
> <- > <- Absolutely! Things with no identity are not nothing, they
> are simply
> <- > <- unidentifiable within the bounds of a computer system. With the
> <- > <- development
> <- > <- of new identification schemes, things may move from
> <- > <- Non-Addressable Subject
> <- > <- to Resource over time.
> <- >
> <- > I'm curious - what is the purpose of non-addressable subjects in
> <- > a computer
> <- > system?
> <- > i.e. what can you actually do with things you can't identify?
> <- > (apart from sling 'em on the pile in the corner  ;-)
> <- >
> <- > examples would be nice
> <- >
> <-
> <- Consider the XTM / XTM Specification. XTM is a non-addressable
> <- subject, the
> <- XTM Specification is an addressable resource. Using these two topics, I
> <- could create an association (say, "specified by") between XTM
> and the XTM
> <- Spec. I could then use XTM as the central topic for a bunch of other
> <- associations (e.g. of types "tutorial on", "application uses")
> <- etc. In other
> <- words, I am making a set of assertions about XTM, without requiring the
> <- subject to be addressable.
> <-
> <- Cheers,
> <-
> <- Kal
> <-
>
>
>

Received on Tuesday, 10 April 2001 04:56:29 UTC