Paraphyly and names

[Pierre Deleporte]

At 15:01 18/01/2002 -0500, Tom DiBenedetto wrote:

>That depends on what kind of information you wish to represent. If you wish
>to deliver information about the history of the taxa, then no, your
>formulation is not more informative. If you wish to make some general
>subjective comment about the extent to which different groupings of species
>might resemble each other, then sure, your arrangement might work.
>Personally, I am committed to a systematics that seeks to represent the
>history of life.
>
>Tom DiBenedetto

I agree, and by the way "information content" depends of what kind of 
information one wants to convey in a classification.
If scientists disagree on this point (e.g.: names should represent only the 
hierarchy of clades, or combine this with grades some way), they will not 
agree on naming taxa. In the first case, if they are reluctant to name all 
(or any) clade, they are at risk of disagreeing on what clades are worth 
naming. And in the second case (clades combined with grades), they are 
additionally at risk of not agreeing on 'significant' gradist criteria; the 
problem lies in the tentative use of two criteria, and not a unique 
criterion for classification.

All trivial for me. But at least, discussing this way, we avoid the notion 
of "naturalness" of groups, in favor of the decision to convey this or that 
kind of information in our classifications. Which seems to me a better 
ground for setting the debate.

Maybe I'm stupid, but I don't think that "Nature" classes in any way. 
Rather, Human beings class for their convenience, so I can hardly grasp 
what a "natural classification" could be.
Maybe a "natural hierarchy" would make more sense? (I'm not sure). And then 
the separate problem of constructing of a hierarchical classification, or 
not...
I'm no mathematician, but maybe there are some around there: isnt' there a 
connection between hierarchy and uniqueness of the criterion? Can ONE 
hierarchy be based on TWO criteria? Seems that the solution could not be 
unique, depending on the rules for combining the criteria...

I can hardly swallow "natural classification" anyway. But I'm ready to learn.

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