[Taxacom] Phylogenetic Game

[John Grehan]

Seems to me that ancestor-descendant portrayals were effectively no
different in method, just interpretation. One might say that A is more
closely related to B than C as a pattern of relationship, and then one
might interpret this to mean that C is the ancestor of A & B, or not,
depending on one's theory of what a pattern of similarity may or may not
be able to tell you about whether one taxon was descended from another.

If cladistics is just a method for arranging taxa according to what is
most closely related to what then perhaps you are correct about your
assertions that it cannot do other things. But then is there anything
wrong with that? Perhaps those other things cannot be done properly
unless one knows (or predicts) patterns of relationship in the first
place.

In biogeography one must also start with the identification of
homologous patterns before one can start inferring evolutionary events
so I am not sure that there is a distinction to be made here.

As for theory free, I suppose if one is interested in just clustering
objects without reference to history I suppose it would be theory free.
It would just be clustering according to a set of arbitrary rules. But
if history is to be invoked (whether descent with modification or
anything else) I would think that a cladistic method (or any other
method of pattern analysis whether for form or space/time) would not be
theory free, otherwise the pattern would have not necessary historical
inference.

And as usual on this topic, I plead mea culpa for any demonstration of
gross ignorance.

John Grehan

 

-----Original Message-----
From: Richard Zander [mailto:Richard.Zander at mobot.org] 
Sent: Sunday, December 12, 2010 3:05 PM
To: John Grehan; taxacom
Subject: RE: [Taxacom] Phylogenetic Game

Phylogeneticists and most others exposed to the cant of the past 30
years think only in terms of clustering. Long ago, back in the day,
evolutionary theoreticians theorized from perceived clusters of extant
and fossil taxa possible stem-groups (caules), best exemplified as
Besseyan cacti diagrams, that is, taxon A was derived from Taxon B, and
so on, kind of clustering but through time (diachronic). 

Nowadays, all clustering is synchronic, the relationships of extant taxa
and extant fossils, no theorization allowed. Cladistics is a method of
non-ultrametric clustering. You can use the results to generate
inductively theories of evolution for the group. Evolution has been
redefined as relationships based on inferred shared ancestry, not
descent with modification. 

Cladistics is a nice method of helping make a natural key (using
morphology), or getting a really broad, general idea of lines of gene
descent (using molecular data), but without progenitor-descendant
theories, the pattern presented by the data remains unanalyzed. Strongly
supported gene trees, to be sure, but unanalyzed under the rubric of
descent with modification. 

Yes, any clustering method can be done without theory. That is why some
nervy but honest cladists call cladistics theory-free. 

Okay, a noncladistic method of developing a theory of evolution is, for
instance, estimating direction of macroevolution by changes of
chromosome number, where polyploidy seldom precedes diploidy. Or
biogeographically where a widespread species with known ancient fossils
is accompanied by narrowly distributed similar species adapted to
recently emerged environments. Or judging from a complete taxon (say,
polar bear) that its strong set of unique traits makes it doubtfully
progenitor of a similar species that is more plesiomorphic
morphologically (Dollo's Rule). Or all of the above to contribute to a
well-supported ancestor-descendant tree (caulogram).

Cladistics does not make theory. The cladogram shows inferred
relationships of shared ancestry, which is axiomatically a fundamental
pattern in nature. No theory is necessary. All other traits are mapped
on the cladogram to create theorems, which must be true because
(remember the big push for the hypothetico-deductive method of Popper?)
all deductions from a first principle must be true (see Thomas Aquinas).
We then have an absolutist analytical procedure.

Paraphyly is a cladistic term for cladograms in which taxonomic names do
not nest neatly in coherent clusters. The whole bit about paraphyly is
preposterous because of course names often are not contiguous in a
cladogram. They are not because of macroevolution, particularly
ancestral taxa surviving even multiple speciation events through
fixation by stabilizing selection. Macroevolution does not necessarily
have to occur only as a sister group of a different taxon at the same
taxonomic level. Paraphyly is common and thus one can theorize that . .
. and so on. The cognitive dissonance, given clear evidence that the
cladistic method (non-ultrametric clustering) has been co-opted by
structuralist absolutists, is a mental vise.

 
* * * * * * * * * * * * 
Richard H. Zander 
Missouri Botanical Garden, PO Box 299, St. Louis, MO 63166-0299 USA 
Web sites: http://www.mobot.org/plantscience/resbot/ and
http://www.mobot.org/plantscience/bfna/bfnamenu.htm
Modern Evolutionary Systematics Web site:
http://www.mobot.org/plantscience/resbot/21EvSy.htm


-----Original Message-----
From: taxacom-bounces at mailman.nhm.ku.edu
[mailto:taxacom-bounces at mailman.nhm.ku.edu] On Behalf Of John Grehan
Sent: Sunday, December 12, 2010 7:44 AM
To: taxacom
Subject: Re: [Taxacom] Phylogenetic Game

I'm confused by the restriction of cladistics to extant specimens. Does
the inclusion of fossils mean that the method is no longer cladistic? 

I'm also confused by the absence 'of a theory of evolution of the groups
involved' Hennig (and Rosa I think) postulated an unequal divergence
from the ancestor. One can implement cladistics without caring either
way, but that would seem to be true of phenetics or any other clustering
method.

"the pattern of evidence is never used by sadists to create a theory of
evolution of the groups involved." Meaning what? Please give an example
of a non-cladistic method of systematics that does this.

On the bear speculations - it would seem that could be applied to any
theory of relationship, cladistic or not.

On paraphyly - I respectfully disagree. Some people are ok with
paraphyletic groups, others are not. So why lose sleep over it? Heck,
only a handful think that morphological evidence can potentially falsify
molecular evidence. But there's no point in castigating anyone for that.

John Grehan (a cladist according to some, an extreme cladist according
to some, a Hennigian cladists according to some, a non-cladist according
to some).