[Taxacom] clique analysis in texbooks

[Richard Zander]

"Guessing better" is much used in classical systematics. It is an
example of "bounded rationality" (see Gigerenzer's book of same name),
which is figuring out an acceptable solution with not enough
information, or optimality under constraint. Thus, "naive" taxonomists
(those who work without benefit of software) try to (1) model evolution
then (2) make a classification from that model. 

Phylogeneticists have a form of bounded rationality in which they change
the first objective to (1) model present-day evolutionary relationships.
Problematically, restricting evolutionary models to synchronic patterns
results in contradictions, such as morphology contradicting molecules,
as in the present thread. The basic presupposition of rationality is no
contradictions, so phylogeneticists treat all contradictions to a
molecular tree as homoplasy or convergence or polyphyly or whatever
excuse for the contradiction. Changing the objective to a subset that
has good statistical properties and relegating contradictions is not
bounded rationality, it is irrationality. 

"Phylogeny" is in practice not a synonym for evolution. 

 

* * * * * * * * * * * *
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 Richard Jensen
Sent: Friday, August 12, 2011 9:48 AM
To: taxacom at mailman.nhm.ku.edu
Subject: Re: [Taxacom] clique analysis in texbooks

Sometimes we can "guess better".  Years ago, before the development of
PAUP and other more sophisticated software, I used the then current
version (mainframe) of Steve Farris' parsimony program to create a
phylogeny for oaks.  I was able, by "intuitive" branch swapping, to
create a more parsimonious tree than was generated by the software.  Of
course, today's algorithms do this for us to generate (with small data
sets) THE most parsimonious tree(s) for the data.

"Fiddling" is ok, as long as one provides a sound rationale for the
fiddling.  I'm not sure what Nero's rationale was, although much was
lost because he did fiddle around!

On 8/12/2011 9:59 AM, Richard Petit wrote:
>> IMO, morphological cladograms, if you fiddle with them enough, can 
>> make a fine natural key. This would be good guidance to evolutionary 
>> relationships, and one might expect such morphological relationships 
>> to be different from the relationships in molecular trees for reasons

>> I've gone on and on about.
> This demonstrates the analogy I brought up a year or two ago about
weighing pigs. "Fiddling" with a cladogram is exactly the same as
"taking out rocks you might have guessed wrong and replacing them with
rocks you think you can guess better."
>
> dick p.