[topicmapmail] Subject Identifiers metadata

[Thomas B. Passin]

Kal Ahmed wrote:

> On Mon, 2004-05-03 at 17:28, Dan Corwin wrote:
> 
>>*** Kal Ahmed:
> 
> What could you possibly want to say about it ? I can understand you
> wanting to talk about the subject. I can understand you wanting to talk
> about an assertion made about the subject (i.e. a name, occurrence or a
> role in an association). I can understand you wanting to talk about the
> <topic> element in XTM (or equivalent in some other serialisation). But
> I can't see what you can possibly say about the "thing that we hang
> assertions off of" - its essentially a non-entity that comes into being
> only when we make an assertion about a subject.
> 

Don't forget, Kal, there *isn't* any such thing "we hang assertions off 
of", at least not outside a topic map, which is a collection of computer 
data structures.  That's assuming you mean "assertions" in the topic 
map/TMDM/RM (as appropriate) sense.  Outside a topic map, we (imagine 
that we) have concepts, and we somehow understand that these concepts 
bear some relationships with one another.  Also, we somehow have an 
understanding to some degree about the nature of these relationships.

How this all works is still pretty mysterious, but it is only for the 
purpose of modeling such things in a computer that we have invented such 
things as "binding points", "topics", "assertions"/"associations", and 
so on.

But still, any of these concepts can be represented in a topic map by a 
topic, given a suitable definition or PSI.  As always, we have to be 
clear about the difference between the general notion of one of these 
things, and a specific instance.

I agree with Kal, if this is one of the things he meant, that the 
*notion* of a subject or subject identifier does not really have 
instances.  But that's OK, if we can talk about it as a distinct 
concept, we can have a topic for it.

I am coming to think that the term "assertion" in the context of topic 
maps should be ejected as quickly as possible from polite society.  It 
is too overloaded and has too many connotations, and since we have 
exactly *NO* theory of the "meaning" of a topic map (no model theory, or 
what have you), we need to get rid of the connotations by getting rid of 
the term.

Back to "association", say I!

Cheers,

Tom P