Re: TMQL - Proposal: Removal of reifier axis

Hi Patrick,

[...]
>> If topic T reifies association A and T is an instance of TT. Is A an
>> instance of TT?
>>
> Well, starting from the TMDM, we know that type/instance is *not* 
> transitive:

I think this has nothing to do with the question if type-instance
relationships are transitive or not, but with the question what
"reification" implies.

If reification implies that A and T can be used interchangeable (since
A has become an explicit subject represented by T), the answer if A is
an instance of TT should be "yes". Why should reification (within a
topic map) be treated differently from our other topic
represents/reifies subject X mechanisms (via subject
identifier/locator)?

We accept that an IRI becomes magically a topic/subject and that we
can make statements like

    <http://psi.example.org/Bremen> isa onto:city

but if we reify something within a topic map, an extra step is needed
and A is not an instance of TT if T reifies A and T is an instance of
TT.

Maybe I misunderstood reification, but in my understanding A should be
an instance of TT iff T reifies A and T is an instance of TT.

[...]
>> That is, if B is an instance of the type A, and C is an instance of
>> the type B, it does not follow that C is an instance of A. 

> If you consider reification to be a form of association, then I think a
> following clause from the TMDM clears this up:
[...]

Yes, if reification is interpreted as an association and we're not
able to reify something directly, A wouldn't be an instance of TT.

Best regards,
Lars
-- 
Semagia 
<http://www.semagia.com/>

<https://twitter.com/larsheuer/> Twitter
<http://www.topicmaps.de/mailinglist/> German Topic Maps mailing list
<http://tinytim.sourceforge.net/> Open Source Topic Maps engine
<http://mappa.semagia.com/> Mappa - Python Topic Maps engine