W3C home > Mailing lists > Public > public-rdf-comments@w3.org > July 2014

Re: Errata Ch 2. Classes - http://www.w3.org/TR/rdf-schema/#ch_classes [SEC=UNCLASSIFIED]

From: Sam Pinkus <sgpinkus@gmail.com>
Date: Fri, 18 Jul 2014 18:03:45 +1000
Message-ID: <53C8D4E1.2090807@gmail.com>
To: Paul Murray <pmurray@anbg.gov.au>
CC: public-rdf-comments@w3.org
Hi Paul,

On 07/18/2014 05:05 PM, Paul Murray wrote:
> On 16/07/2014, at 3:48 PM, Sam Pinkus wrote:
>> 1. Do you really need to state that there is a distinction between a
>> class and its instances? The first paragraph already makes this as
>> clear as it needs to be.
> I think so - it isn't as obvious as it might appear (although I don't
> have that first paragraph with me - perhaps it does cover everything
> adequately - I shall plough on regardless)
If anything restating the distinction between a class and class instance
make the points you raise less obvious to reader. But are no different
pages. The specific section I am talking about does not describe the
aspects of the object model you describe.
> The problem is that normally, (or perhaps philosophically?) "dogs"
> /is/ simply all dogs. "Integers" /is/ all integers. This isn't
> problematic because the idea of dogs and actual dogs are - not sure
> how to express it - on different planes. They have different
> ontological status. "dogs" lives in the box over there, alongside
> "Australians" and "things manufactured from wood"; and fido, spot,
> princess, I and my guitar live in an entirely separate box. 
> In RDF, we pull "dogs" down into an object in its own right just like
> any other object. "dogs" in terms of triples looks exactly the same as
> "Fido" -  it is an instance of a class, it has a URI, it is a subject
> and an object of various triples. So at the graph level, it's an
> object (a URI). At the OWL level, we treat it as being quite different.
> When people misunderstand this, it goes like this:
> Primus: "but how can a class of things itself be a thing?"
> Secundus: "it just is. We have to have *some* way of discussing it"
> Primus: "But, if a class is a thing, and all things have a class, the
>  what's the class of a class?" 
> Secundus: "It's the class class, and before you ask, the class of the
> class class is the class class itself".
> Primus: "Wouldn't the class class need a class thats a class class
> class?" 
> Secundus: "No. Why?"
> Primus: "It just seems like it should."
That is a good way of putting it. The amendment are not to do with this
aspect of the model however.

I really do think people are generally confused by the fact that class
are objects, and the object model of RDF in general though! I think this
chapter could spell it out more clearly, and absolutely could do with a
diagram showing both the "instance of" and "sub class of" relationships
between the basic objects!


Received on Friday, 18 July 2014 09:12:41 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 16:59:45 UTC