Real taxa => Ranking

[pierre deleporte]

A 10:37 28/09/2004 -1000, Richard Pyle wrote :

>Well, I suppose each individual taxonomist could declare some set of
>criteria on a taxon-by-taxon basis, against which individual organisms could
>be "tested".  But is there anything intrinsic to the organism (i.e., not
>part of an arbitrary human definition) to say that one taxonomist's set of
>criteria are objectively better than another taxonomist's set of criteria?

Some "things" are physically self-delimited out there. Individual organisms 
for instance, during the time they fit together in one materially coherent 
living being or dead specimen... They exist as a consistent material system 
of connected or interacting parts, hic et nunc.
Some "things" or material systems are not so well delimited. e.g. a social 
group of interacting individuals, during the time they interact... A 
"population" is still a much less "self-delimited" material system 
(interactions are looser, and most of the members of what we call a single 
population simply do not really, materially interact with one another). But 
species, and any higher-level taxa, are clearly not self-coherent material 
systems. We cannot "identify" them out there, we must define them as 
classes of things (hence, concepts) on the basis of some criteria, 
sometimes based on elaborated scientific theories (e.g. inferring  descent 
from an exclusive common ancestor, which is clearly not written on the 
specimens).

Now, about "taxonomist's criteria better than other ones". We, human 
beings, class "things", this or that way, on the basis of our criteria, and 
hopefully according to our needs. The notion of "better", optimal criteria, 
cannot be searched for in material systems out there, because nature is not 
self-organized in material systems ready for all our imaginable 
classificatory needs. Blue birds do not come conveniently packed together 
in blue country and grooming one another to show us that they belong to the 
same class of blue birds. Optimality of classificatory criteria must be 
checked against our needs. Why do we want to class this kind of things? For 
which purpose? In the best of cases, we should first of all answer this 
question explicitly and, only then, establish the specification of the 
desired classification according to our intended use of the classsification.

"Objectivity", or precision, or, more generally, optimality of our 
classificatory criteria can be checked, but not against "nature", just 
against our needs... if clearly formulated!

A corollary is that there can hardly be a unique classificatory system (and 
possibly ranking system) optimal for all imaginable needs. If this is true, 
the enterprise of setting a unique classification of some kind of things 
for all purposes must be a source of endless conflicts, opposing differing 
and incompatible solutions: uncompromising 
one-criterion-optimal-for-one-purpose classifications, versus compromising 
mixed-criteria-optimal-for-nothing ones, with a potentially illimited 
number of possibilities in each category.
The holy grail of the unique-and-universally-optimal classification is a myth.

I could have make it shorter by quoting Popper (this time): all 
classifications are conventional.
But conventions are not "arbitrary" in the sense of "nonsensical" if they 
fit a well argumented (human-designed) specification.
Let's rank optimally for our needs. When we need ranks at all.

Best,
Pierre


Pierre Deleporte
CNRS UMR 6552 - Station Biologique de Paimpont
F-35380 Paimpont   FRANCE
Téléphone : 02 99 61 81 66
Télécopie : 02 99 61 81 88