FW: falsifying evolution

[Panza, Robin]

If we can be sure that the "character states" were evolutionarily
comparable, character congruence (as opposed to congruence between trees
achieved using different character states) is testable by statistics.  E.g.,
are feather-covered wings and lactation more significant (in terms of
numbers of loci needing to be modified to produce) than iris color?  If so,
congruence (or lack) between the former is stronger evidence for/against
similarity by evolution than would be  congruence/lack between the latter.
In evolutionary theory, we expect phylogenetic "reversals" in the former to
be much less likely than "reversals" in eye color, because we expect the
former to require more evolutionary change, but can we "prove" it?  This
would be a testable hypothesis of evolutionary theory.

Robin

Robin K Panza                          panzar at carnegiemuseums.org
                                                ph: 412-622-3255
Section of Birds                         fax: 412-622-8837
Carnegie Museum of Natural History
4400 Forbes Ave.
Pittsburgh  PA  15213-4080  USA



-----Original Message-----
We do not (normally) get "total incongruence of characters". Certainly
there is incongruence. However, if you do a good morphological
phylogenetic analysis with a large number of characters for a certain
group, and then another one using, say, some DNA sequence, you will see
a fair amount of congruence between the results. Perfect? No (unless you
are very lucky), but a very long way INDEED from a random pattern, which
you might expect from separately created "kinds". If molecular and
morphological phylogenies (or other phylogenies derived from independent
character sets) did show no significant association across a wide range
of taxa, then the hypothesis of descent from a common ancestor would be
very seriously weakened.

This is very different from invoking ad hoc hypotheses to explain what
homoplasy is present.

--
Dr. Wolfgang W?ster  -  Research Fellow
School of Biological Sciences    Voicemail: +44 1248 382301
University of Wales              Fax: +44 1248 371644
Bangor LL57  2UW                 E-mail: w.wuster at bangor.ac.uk
Wales, UK                        http://sbsweb.bangor.ac.uk/