In reference to Joe Felsenstein's recent comments, I would like
to _briefly_ clarify my earlier posting and then I will shut
up. I do think it is pertinent to discussion on molecular
evolution so will not worry too much if some don't agree. I
think the basic issue is how inclusive one wishes to be in
characterizing "cladistic" methods. Joe tends to be more
pluralistic than me and I want to make it clear that I can
tolerate different viewpoints as well as the next person. I'm
do not, however, see what morality has anything to do with
justifying one's assumptions in phylogenetic inference.
I was only claiming that there is more to "cladistics" than
the rule of monophyly. This is hardly a view that I hold in
isolation, although I fully realize that I probably hold a
minority view on this particular newsgroup. While it is with
some hesitation that I refer to the common categorizations of
systematists, it is my understanding that the heated debate
between "pheneticists" and "cladists" back a couple of decades
ago did not especially focus on this "rule of monophyly" issue.
In fact, both schools tried their best to make their
classifications conform as much as possible to their best
supported branching diagrams, so in a sense they both were
striving for "cladistic" classifications in Joe's sense. This
is why I originally felt compelled to respond. They did not
really differ that much on this particular issue (cf. the
so-called "evolutionary systematists").
There are many systematists, admittedly myself included,
who view cladistics as a paradigm. They use parsimony as the
criterion to choose between competing hypotheses. Although not
_all_ "cladists" would subscribe to this view, most use
parsimony methods to maximize explanatory power of similarity
due to shared ancestry by minimizing hypotheses of homoplasy
(convergence or reversals). Others seem to use "parsimony" in a
broader sense to favor one model of evolution over another.
They are entitled to their view.
Back when I was an undergraduate at the Univ. of Washington,
I remember Joe telling me that he viewed the future of
systematics as moving towards in the direction of statistical
testing of phylogenies, just as the field of physics had
already moved. Perhaps he is right. I do not think I am totally
alone, however, in viewing phylogenetic inference as more of a
problem in historic reconstruction than in statistical
inference. Sober (1988: Reconstructing the past: Parsimony,
evolution, and inference) has written on this subject much
better than I can. I do not see a difference in morphological
vs. molecular characters with respect to this issue. One can
certainly get into trouble with over-reliance on parsimony
methods, especially when evolutionary change is rapid relative
to cladogenesis, but the same is true for other methods as
well.
Well I guess that wasn't so _brief_ after all. I'll see you
at Woods Hole again in a couple of months Joe.
Doug Eernisse
Museum of Zoology
Univ. of Michigan
