[sc34wg3] SAM-issue psi-generics (was: SAM-issue term-scope-def)

Bernard Vatant sc34wg3@isotopicmaps.org
Thu, 4 Jul 2002 15:28:53 +0200


Marc

I think I understand better your concern through your challenging example. Let's see if my
viewpoint can survive it :)

> But can you still distinguish between the following statements in a Topic Map, where
'Marc' is a topic id:

1.  'Marc has a name "Marc de Graauw" '
2.  'Marc has one wife'
3.  'Marc has two sons'
4.  'Marc is an instance of class topic'
5.  'Marc is serialized in XTM'
6.  'Marc contains 6 assertions'

> According to you, all of these would be statements about topic 'Marc',
> not about subject 'Marc'.

Well, there is an implicit assumption there, it is that the topic 'Marc' in the six
assertions represent the same subject, which seems to me inconsistent. Clearly in 1, 2 and
3, the topic 'Marc' represents the same subject (you over-there) but 5 and 6 deal with
another one. Assertion 4 is valid for any of both.

> However, the first three seem to me to be of a
> fundamentally different nature than the latter three.

The difference of nature comes from different semantic levels for the subjects ...

> If we say that assertions are statements about subject 'Marc', not topic 'Marc', then
the
> latter three as stated above are simple false. If we want to make them, we
> would have to reify topic 'Marc' and then - correctly - say:

> 'Marc has a name "Marc de Graauw" '
> 'Marc has one wife'
> 'Marc has two sons'
> 'Topic "Marc" is an instance of topic'
> 'Topic "Marc" is serialized in XTM'
> 'Topic "Marc" contains 3 assertions'

Yes, sort of. Except for 4. But that does not change the view that assertions are about
topics.

Let M be the topic representing in a topic map the subject
"Marc-de-Graauw-the-guy-outhere"
Let M1 be the topic representing the subject "M-element-of-this-serialized-topic-map"

The difference should show clearly when thinking that M represents a non-addressable
subject, whereas M1 represents an addressable subject, i.e. the <topic> element M. If you
refer to them in XTM, M will be referenced through <topicRef>, whereas M1 will be
referenced through <resourceRef>.

Now you can rewrite without problem nor ambiguity your assertions, replacing your original
topic 'Marc' by M in 1,2, and 3, and by M1 in 5 and 6. And of course 4 is valid for both M
and M1 ... a topic is a topic :)

> This seems quite logical and intuitive to me, but it does mean the former
> three statements are about subject 'Marc' and the latter three about subject
> 'Topic "Marc" ', so assertions are about subjects, not topics.

No. All assertions are about topics, but about two different topics (representing
different subjects)

> It seems to me in your way of looking at it we cannot distinguish anymore between saying
> anything that is supposed to relate to me, as I am sitting here, and saying
> things about a topic in a Topic Map someone has written about me.

Of course you can distinguish, if you follow the basic rule "one subject, one topic".
There you have two different subjects, so you should represent them by two different
topics, allowing you to make distinct sets of assertions, at distinct semantic levels.

Bernard