Farewell to Species - and Nomenclatural Stability

[Stuart G. Poss]

Thomas DiBenedetto wrote:

"To say that two groups are sister-groups is merely to say that they are more
closely related to each other than either is to any third group. We can use the
empirical data of character distributions to reach this conclusion, since the
shared presence of derived characters is evidence of exclusive common ancestry.
"

Yes, if there is no character incompatibility, no reason to doubt that the
characters are in fact uniquely derived, or no dispute on these issues then a
phylogenetic taxonomy would be most straightforward.  However, because of the
way selection acts upon particular combinations of DNA and the complex
interdependency of "characters" so encoded, interpretation of the unique
derivation of character states is a complex and seldom straightforward
excercise.  This is true even at the level of nucleic acids, since similarity
of sequence alignment is a necessary first step to establishing homology among
comparable base pairs and this is a temperature dependent process.   Further,
short of a relatively stable neighborhood of bases, there is inherent
uncertainty in identification of single substitution of a base at a given
position as homologous substitution since there is always a small probability
that a given substitution may be secondary or even tertiary in nature.

Although there may be no "characters" that are not themselves the product of
evolution,  we must keep in mind that cladistic characters are investigator
defined mappings or perceived morphological or molecular disjunction upon a
taxon (typically composed of more than one member), whose identifiable states
should at least in principle be subject to empirical measurement of some kind.
In order to make testable statements about organisms with respect to such
mappings, not necessarily soley phylogenetic in nature , we need to have a set
of nomenclatural rules that can be EASILY APPLIED and CONSISTENTLY USED used to
identify the appropriate name to be used for the taxa in question.

Constraining the genus group name to be holophyletic, while certainly highly
desirable in many circumstances, should be relegated to taxonomic revisions and
monographs and not yielded as a general nomenclatural rule.  The "phylogenetic
nomenclature"  that you propose would have the undesirable property of having
the potential to change every name in a genus whenever a new member  is
discovered or when someone looks at an alternate set of characters or a new
gene and reaches an alternative conclusion about its phylogeny.  Reflexive use
of such a general rule would certainly cast away what little progress toward a
stable nomenclature we have been able to achieve over the past 250 years or so,
with little prospect of a system emerging that will ensure the stability
necessary to do any science at all.  Lets not confuse the principles of
nomenclature and phylogeny reconstruction.  Both are complex enough already and
there is no need to do so.