Paraphyly and names

[Thomas DiBenedetto]

-----Original Message-----
From: SKÁLA Zdenek 
Tom, your arguments are circular. You are saying that holophyletic taxa
(including *by definition* only clades and based solely on
synapomorphies) are best *because* they are based on
clades/synapomorphies (see above: "...evidence of the divergence=good
data").
***********
I don't see any circularity. Yes, monophyletic taxa are based on
synapomorphies, and that is what makes them "best". 
Becuase synapomorphies are character distributions that indicate real
historical groups., whereas plesiomorphies and convergences do not.
************* 
For you, I believe, the taxa (or better system) should contain
information on the phylogenetic divergence pattern. Then, holophyletic
taxa and exlusive use of synapomorphies gives perfect sense and you are
really on the best way.
*********
Yes, I do think so
************
However, I would like to recall that taxonomy/system (at least in its
original and most widespread meaning) is intended to summarize
information on the existing organisms' diversity (i.e. character
diversity) in a natural way. The "natural way" of course currently means
phylogeny;
*********
Yes
*******
however I doubt that we can simply re-define taxonomy as
phylogeny reconstruction and leave the character-information aspect.
Perhaps the ways of systematics (in the original sense) and
phylogenetics now diverge; ....
(It does not mean that holophyletic taxa have no character-information
value; they only are not *optimized* in this way.)
********
Classification as phylogeny reconstruction does not lose the
character-information aspect - to the contrary it organizes that information
in its most useful and natural manner. Characters have evolved within the
lineages that cladistics seeks to identify. Cladistic classification
presents not only a reprenstation of the lineages themselves, but also the
characters, referenced to the appropriate lineage branch, recognized as
apomorphies. The eclectic classifications conflate lineage divergence and
character resemblance, and thus obscure the real relationship between
characters and the lineages in which they evolved. For instance, did
feathers evolve twice? Once in Class Aves, and once in Class Reptilia (in
those feathered dinosaurs)? And why is it that genetic evidence shows
crocodiles to be so much more closely related to critters in a completely
different class (Aves) than to other "reptiles".  If you wish to learn
anything about the evolution of characters, you need a classification of
taxa that represents real lineages - not artificial constructs. That is why
eclectic classifications utterly fail as scientific statements.

Tom DiBenedetto

P.S. apologies to all my conversation partners, but I will be travelling
overseas for the next week or so, and will thus have to withdraw from these
discussions.