"composite" subject RE: [sc34wg3] New SAM draft

[Bernard Vatant]

Martin

| >I still have problems with the "one, and only one," bit. If you said
"a
| >single" I would have no problem, its the "and only one" that kills
the
| >ability to have subjects that combine other subjects as per my
example.

Steve

| The way I think about this is that by creating such a topic you have
| implicitly "brought into being" a new subject which encompasses both
| Astronomy and Quantum Physics. Thus the topic really does just
represent
| a single subject (which happens to be a composite). This is no
different
| than the subject "Steve's daughters" - which is also "one and only
one"
| subject (although it encompasses three other subjects).

I second Steve on that. "Single" does not mean "simple", and in fact the
"simple" (atomic?) or "composite" nature of a subject is just a question
of viewpoint. Representing Martin Bryan by a single topic makes sense,
although he's not "simple" AFAIK, no more than Astronomy BTW ...

Either you consider the subject as a whole, or you look at its
structure, and in that case, *any* subject can be easily proven complex
or composite. It's in the fractal nature of things that no thing is
simple .. not even "nothing" ... those who were in Seattle KT 2002 might
remember brilliant "quantic" explanations of Nikita about complexity of
"nothing". 

What has to be clear is the one-to-one matching of the "subject as a
whole" to its formal representation by a topic. The complexity of the
subject can further on be asserted by e.g. "whole-part" associations or
any other relevant means.

| On that basis, the current wording is acceptable. But if you think it
| might be misunderstood, perhaps there should be a note saying that
| a subject may be a composite subject?

I'm not sure it is relevant to do that. Simplicity or
complexity/composition of the subject being not what is at stake here,
IMO speaking about it at all would be more confusing than anything else.


Bernard
_____________________________________

Bernard Vatant
Senior Consultant - Knowledge Engineering
www.mondeca.com
_____________________________________