[Tcs-lc] Concepts of Higher Taxa

Richard Pyle deepreef at bishopmuseum.org
Tue Mar 29 18:13:37 PST 2005


Roger wrote:

> What I am saying is that if I create a concept for genus
> G1 containing species S1 through S10 and belonging to
> family F1 then I have defined a tree of relationships.

Yes, but you've also (presumably) defined (or mapped) twelve concepts, and
relationships between S1-S10 and G1, and between G1 and F1.  I don't think
it's prudent to additionally define relationships between S1-S10 and F1,
because those should be inherited by the relationship between G1 and F1.

> Each of those other 11 concepts have definitions going up
> and down the ways which are included in 'my' concept of G1.

In my world view, you would have 12 potential concepts; even if you are the
first to define only one of those twelve.  The other 11 may be mapped as
congruent to previously-existing concept definitions (earlier
AccordingTo's), but they are still defined.

> I have therefore defined a classification not a single object.

I would say that you've asserted twelve concept objects, of which one is
novel (the other 11 mapped as congruent to previously-defined concepts).

But people don't seem to like this sort of "concept proliferation" (even
though we have to index them anyway as usage instances, and even though
computers have no difficulty at all dealing with huge numbers of instances),
so an alternative is to create only one concept instance (G1), and map
directly to other, "inter-AccrodingTo" concept instances (includes S1-S1 of
earlier author; included in F1 of earlier author), without defining an
additional 11 concepts "AccordingTo" you.

> I am saying "I believe these things should be arranged in
> this way" that is very different from defining a single
> taxon under the current model.

I guess I'd have to see your instance document to understand exactly what
you mean.

Rich




More information about the Tcs-lc mailing list