No subject

R. Zander bryo at PARADOX.NET
Wed Sep 2 12:56:51 CDT 1998


David Lindberg wrote:

> We are living in one of the most important and exciting times that the
> discipline of systematics has experienced since the 1700's.  Methods and
> techniques are now available to make Darwin's vision of constructing
> classifications that reflect hypotheses of relatedness a reality.  The benefits
> of such classifications have been demonstrated and championed by many
> (e.g., see the Systematics Agenda 2000, Technical Report).  The "cost" of this
> advancement in our science is instability in nomenclature.

The cost of this "advancement" in our science is a lost generation that produces highly
resolved classifications based on a statistical gimmick. Optimality criteria like maximum
parsimony, maximum likelihood, maximum posterior probability and minimum falsifiability
all result in phylogenetic reconstructions that are "best" but individually improbable.
ANY reasonably parsimonious or reasonably likely tree has predictive qualities and can be
a basis for classification. Only strong Bremer support or very high posterior probability
result in an absence of reasonable alternative trees that allow a potential
"reconstruction" of a phylogeny, and (except for subclades that are usually already
intuitively "accepted") such reconstructions are few in the literature. I recommend less
sanguine enthusiasm. I wonder what statisticians in the fields of physics, economics and
psychology might think of the way modern computerized phylogenetic analysis reconstructs
historical events by singling out a least-wrong scenario?


> This instability does
> not result from the capricious or pious actions of "authorities" scouring obscure
> literature for earlier names or usage, but rather from modern research programs
> that publish data and the hypotheses of relationships that are discovered therein.

cut here...

--


Richard H. Zander
Curator of Botany, Buffalo Museum of Science
1020 Humboldt Pkwy, Buffalo, NY 14211 USA
bryo at paradox.net   voice: 716-896-5200 ext. 351




More information about the Taxacom mailing list