[sc34wg3] Documenting merging rules in TMDM .. and unmerging too

[Ann Wrightson]

LMG wrote:

Because some parts of the TMDM have semantics that require merging.
You can't define properties like [subject identifiers] without requiring
merging on their values, because that would conflict with the semantics of
the property.

The same argument applies in reverse to topic names, variant names, and
occurrences: to always require merging would be in conflict with the
semantics assigned to these constructs.

Bernard wrote:

But in knowledge management and science, a fundamental process is the
discovery that what was considered so far a single subject is in fact more
than one. Complexity growth is a process completely opposite to formation of
black holes : it creates new information about new subjects... 

I say:

It's possible for a model to have something in its semantics (i.e. convey a
certain meaning, in this case sameness) without it being (logically)
necessary to respond to that meaning with action (in this case, merging).
It's clear to me that the TMDM semantics LMG describes does include a
necessary action of merging in response to certain kinds of sameness.
However, there is also a view that this is not necessary (where necessary ==
an inseparable part of the concept), but is a separable consideration, that
can (& should) be modelled independently (US position).  

This suggests to me that some of the apparent fruitlessness of this debate
may arise from a buried difference about the kind of semantics we are
dealing with for TMDM - in particular, between an operational semantics
(LMG?) and a declarative semantics (PD/BV?). 

Hope this helps...

Ann W.