Taxonomy issues

[jcm at head-cfa.harvard.edu]

A small comment about conflicting definitions:
this is closely related, I feel, to the 'specific vs generic'
UCD issue: we have the general concept of a V magnitude, and
the specific concept of a Johnson V or someone else's V filter.
Sometimes we want to ask a loose question about V mags and 
sometimes we want to be precise.
In the same way, in astronomical practice we sometimes loosely
refer to BL Lac objects (or red fairies, or...) and sometimes
precisely refer to BL Lac objects (by Weedman's 1978 criterion)
or to red fairies (by Richards(2002) velocity criterion). I think
our ontology can define many different flavors of precisely defined
red fairy, tagged by the reference which defines them, as subclasses
of a more broadly defined concept of red fairy. Then we can ask:
is Mu Draconis a red fairy by anyone's definition (is it in the union
of all red fairy definitions), and we can also ask: is Mu Draconis
a Richards(2002) red fairy? Or, more exactly, has someone claimed
that it is a red fairy... and as Tony pointed out, different authors
may have classified it in conflicting ways, and the data they used
may have uncertainties on it or be wrong, but our cloud of red-fairy-related
concepts is still precise. 
 It's when organizing our theory-laden classes in a relational
tree that we'll have to be very careful, and I do think we will
need to support alternate ontologies. Is an SGR (soft gamma repeater)
a nearby neutron star or an object in a distant galaxy? Even when
the question is 99.999 percent decided: is a quasar an active
galactic nucleus (and therefore is a subclass of 'objects contained in
a galaxy') or is it a nearby thing ejected from a galaxy? When Chip
Arp uses the VO, he should be able to impose his redshifts-aren't-cosmological
ontology even though we all know he's wrong. I don't think that fear
should stop us from developing one or a few default ontologies
to represent our model of the universe, although perhaps it would
be dangerous for the IVOA to officially bless one. Gotta have them, though 0
- it's the only way we will be able to go beyond what SIMBAD et al do
and search for the component of the binary star inside a globular
cluster that we are interested in, a situation in which positional
matching is not going to work.
 OK, this wasn't a small comment after all. Sorry!
   - Jonathan McDowell

 .-------------------------------------------------------------------.
 |  Jonathan McDowell            |  phone : (617) 495-7176           |
 |  Harvard-Smithsonian Center   |  fax   : (617) 495-7356           |
 |   for Astrophysics            |                                   |
 |  60 Garden St, MS6            |                                   |
 |  Cambridge MA 02138           | email : jcm at cfa.harvard.edu       |
 |  USA                          | email : jmcdowell at cfa.harvard.edu |
 `-------------------------------------------------------------------'