Re: Subjects as Literals

On 01/07/2010 02:59, "Ross Singer" <rossfsinger@gmail.com> wrote:

> On Wed, Jun 30, 2010 at 9:36 PM, Hugh Glaser <hg@ecs.soton.ac.uk> wrote:
>> Great - more crystallization of the problem.
>> 
>> On 01/07/2010 02:14, "Ross Singer" <rossfsinger@gmail.com> wrote:
>> 
>>> I suppose my questions here would be:
>>> 
>>> 1) What's the use case of a literal as subject statement (besides
>>> being an academic exercise)?
>> I would have thought the same as a use case for a literal as object.
>> I want to say:
>> "Semantic Web Revisited" foo:isTitleOf bar:somePaper
>> Why should I be forced to say
>> bar:somePaper dcterms:hasTitle "Semantic Web Revisited"
>> ? Seems pretty arbitrary to me.
> 
> Actually, your example seems pretty arbitrary.
But I think that the requirement to do
bar:somePaper dcterms:hasTitle "Semantic Web Revisited" .
is arbitrary.
> The latter,
> <http://example.org/foo> dcterms:title "Semantic Web Revisited" is
> conventional and convention seems pretty important for the semantic
> web to work right.
So what is the argument for the convention?
And what is the thing that is wrong if I do it the other way?
That is, tell me what does not "work right" if I have literals as subjects.
> 
> Just because you feel like you should do it doesn't mean you should.
> RDF/XML is a pain in the butt to parse because there a million ways to
> serialize it.  
I don't think RDF/XML is an issue here - in fact I don't like RDF/XML much,
so any examples I use tend to use N3.
Actually, I can't quite interpret what you say, as I thought that RDF/XML
was a serialisation itself.
 >Your example is doubly uncompelling since bar:somePaper
> dcterms:title "Semantic Web Revisited" *works* and pretty much any
> agent would be able to deal with it and understand it.
Yes, it works, but only because the agents conform to what seems an
arbitrary restriction to someone like me who looks at RDF as a boring graph.
> 
> I don't really care one way or the other about this topic, but I think
> this needs to move beyond some theoretical inverse relationship to
> make the argument.
Hmm.
But I think that is moving towards a Turing machine argument.
I would not suggest (I think) that RDF cannot express what I want - but the
convenience is the issue.

There was a time when it was not permitted to use an array index as the
value for a loop limit in Fortran - it has long been understood that such
arbitrary restrictions cause trouble without significant benefit.

In fact, a question I would like to ask, but suspect that noone who can
answer it is still reading this thread ( :-) ):
For those who implement RDF stores, do you have to do something special to
reject RDF that has literals as subject?

> 
> -Ross.

Received on Thursday, 1 July 2010 02:22:02 UTC