Re: [topicmapmail] Re: [topicmaps-comment] Notions have existence .....

[Piotr Kaminski <piotr@ideanest.com>]

Steven R. Newcomb writes:
> The bilingual person who writes the third topic map is,
> of course, expressing her opinion that the two subjects
> are identical.  Topic maps make such opinions
> representable, interchangeable, and useful.

Interchangeable and useful, yes.  Not representable, though.  While the
syntax to make such statements exists, PMTM4 (and XTM processing rules as
well, I believe) causes the two topics to be merged with no trace
remaining of the reason for the merging.  The "merging" itself is not (and
cannot) be represented by a topic.

In other words:  If I merge in a  "correspondences" map (one that
expresses opinions about subjects being identical) from an author whom I
do not trust, what characteristics in the resulting model inform me about
the reason that topics got merged?  There can't be any since, to speak
about topics being merged (with e.g. a "corresponds" association that has
multiple "sources" and one "target"), we would need to reference the
original topics which, be definition, no longer exist.

I'm sure somebody will be kind enough to inform me if I've missed
something obvious.  Or perhaps nobody cares about preserving information
about merges in a topic map?

        -- P.

--
  Piotr Kaminski <piotr@ideanest.com>  http://www.ideanest.com/
  "It's the heart afraid of breaking that never learns to dance."