B.J. Tindal, logics and terminology

[Pierre Deleporte]

B. J. Tindall <bti at DSMZ.DE>  wrote (no title):

(...)
>Thus I can sort organisms into groups by overall
>similarity, or based on character sets, but neither the resulting
>phenogram, nor the resulting cladogram is the "phylogenetic tree" unless I
>start to try to explain the resulting dendrogram in terms of evolution.

I agree completely, and I even suggest that, if your goal is phylogeny
reconstruction, you first check for evolutionary justification before
trying blindly any kind of classificatory method.
Phenetics of the first ages (numerical taxonomy in the sixties-seventies)
was effectively not intended as phylogenetic; only under the assumption of
a clocklike evolutionary process do overall similarity patterns may become
fit for phylogeny reconstruction. Your remarks could be extended to other
approaches as well. A "maximal likelyhood" approach based on a "brownian
motion" model for instance could not be clamed to produce a phylogeny
unless you assume that the general evolutionary process IS
brownian-motion-like. I you are not ready to assume this, you better not
waste your time using this method in the first place.
But how many people still just begin their analyses like this: "try
anything and see what happens"?

Any taxonomic system may potentially be taken as either a "formal"
arbitrary classification, or a phylogeny reconstruction (to coin only some
extreme alternatives). In the latter case, assumptions about the
evolutionary process should be explicitely stated, which give support for
considering the method as phylogenetic. To put it short, there is logically
no possibility of reconstructing the evolutionary pattern without
implementing some notion (however general and careful) of the evolutionary
process.

As for the ambiguity of many terms designating the different methods in
systematics, you are right again. These terms have both a general
acception, and a restricted meaning when naming a particular school in
systematics. More than this, some so-called "schools" themselves may be
melting pots, e.g. see the recent debate in "Cladistics" about the "three
taxon analysis" (3TA) method: is it an extreme form of "pattern
cladistics", or something completely different, throwing away the
contiguity criterion? By the way, 3TA has no explicit evolutionary
justification, and was originally presented as a better use of "parsimony":
another term with general and specific acceptions! You can read in the
litterature "parsimony methods" taken for "cladistics" of all kinds, and
"parsimony criterion" meaning the optimization of contiguity of character
states only (as in "standard cladistics") versus "parsimony criterion in
general" (Okham's razor)...

Original hennigian "phylogenetic systematics" is (generally speaking)
evolutionary, but "evolutionary systematics" (eclectic) is not strictly
"cladistic" as for nomenclatural rules(not strictly monophyletic), while
so-called "cladistics" is either evolutionary or formal ("pattern" kind),
as well as "phenetics"-overall similarity methods, and so on... and you
noticed that some people tend to confuse "phenetic" with "phenotypic" (I
encountered some myself outside microbiology)...
Well, we are not out of the wood indeed, but at least we could pay great
attention to specify our acception (general or school-specific) of all
these terms. To my sense we optimally should also carefully specify the
goals of our classificatory projects (phylogenetic or else) and the
underlying evolutionary models or other justifying assumptions, in order to
lower the risk of lasting "dialogues of the deaf" so frequent in taxonomic
debates.

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