[Taxacom] Binomial Nomenclature etc.

[Richard Pyle]

David Campbell wrote:

> Hypotheses for species definition are sometimes testable, if they 
> are of the "this group of organisms shows a set of features markedly 
> different from other organisms" genre.  More samples will reveal if 
> the features in question blend in with other groups or not.  

I suppose in cases where proposed new species, as well as the most
comparable known species, are known from only a very few specimens with
limited knowledge of patterns of inter- and intra-specific variation among
the group, you could frame a species hypothesis this way.  I guess it
depends on the nature of the individual beast(s) in question.  In my world,
the lumper/splitter debates are usually always regarding well-known groups,
and the arguments tend to be much more philosophical in nature.  But I
realize there are other groups where the arguments stem from insufficient
facts, rather than differences in opinion concerning a robust set of facts.
I suspect that in most groups, there are simply not enough practicing
taxonomists to establish a lumper/splitter debate for most species -- and
this is the purest form of Darwin's species definition (effectively, a
species is what a taxonomist or community of taxonomists says it is).

> But what level of difference counts as enough is an assumption, 
> not testable.  We can test "Does this difference reflect some limit 
> to interbreeding" in many cases but not "How much of a limit to 
> interbreeding should we recognize with a species name?"

Yes!  Exactly!  Well stated!

Aloha,
Rich