[Taxacom] Pasimony and base alignment

[J. Kirk Fitzhugh]

Zander writes:
>"Ah, but what is Truth? Apropos of this is the case of the rechargeable 
>LED flashlight."

It has been a while since I read the literature on the subject, but I seem 
to recall that there are at least six theories of truth. The most commonly 
applied in the sciences is, I believe, the correspondence theory.

Zander then states,
>"By analogy, Fitzhugh is implying something like the coil is the data and 
>the magnet is the method. Thus, phylogenetists have been vigorously 
>shaking their flashlights for 30 years. As the cladistic set of batteries 
>fade, a new flashlight, parsimony, was taken up, then maximum likelihood, 
>then Bayesian methods, and so on. Fitzhugh, however, does not detail the 
>"mechanics" for properly testing phylogenetic hypotheses."

A simple demarcation is in order. Inferring some explanatory hypothesis 
from observed effects is not the same as testing that hypothesis. Parsimony 
and likelihood go hand in hand in the inference of hypotheses - they cannot 
be separated. This becomes obvious when we acknowledge that hypothesis 
inference is a form of reasoning known as abduction. Bayesian methods deal 
with testing, not hypothesis inference. I have outlined the mechanics of 
testing phylogenetic hypotheses in this forum before, so will not do it 
again. For the interested reader, I can suggest the following papers, where 
I talk about testing phylogenetic hypotheses in some detail:

Fitzhugh, K.  2005.  Les bases philosophiques de l'inférence 
phylogénétique: une vue d'ensemble.  In: Deleporte, P. and Lecointre, G. 
(Eds.) Philosophie de la systématique. Biosystema 24: 83-105. [I can 
provide an English translation]

Fitzhugh, K.  2006.  The abduction of phylogenetic hypotheses.  Zootaxa 
1145: 1-110.

And finally, Zander states,
>He [Fitzhugh] asserts that "shared similarities can never serve as test 
>evidence." Why not? Surely the barnyard observation that like produces 
>like, plus the Darwinian theory allows a framework for testing, with 
>simulations providing confidence in the comparative method."

Having a framework for testing is not the same as actually testing a 
hypothesis. A phylogenetic hypothesis makes vague claims about character 
origin and speciation events to account for shared similarities. What must 
be tested are those two classes of causal events. Shared similarities are 
irrelevant to such testing for they provide no evidence that the specific 
causal events occurred, relative to any other sorts of events. The 
hypothesis was inferred to account for shared similarities. Far be it for 
such characters to then test the very hypothesis intended to explain those 
characters.

Kirk

-----------------------------------------------------
J. Kirk Fitzhugh, Ph.D.
Curator of Polychaetes
Invertebrate Zoology Section
Research & Collections Branch
Los Angeles County Museum of Natural History
900 Exposition Blvd
Los Angeles CA 90007

Phone:   213-763-3233
FAX:       213-746-2999
e-mail:   kfitzhug at nhm.org
http://www.nhm.org/research/annelida/staff.html
http://www.nhm.org/research/annelida/index.html
----------------------------------------------------