The attitrion of taxonomic expertise

Alan Harvey aharvey at AMNH.ORG
Fri Oct 18 11:01:05 CDT 1996

James Lyons-Weiler wrote:

>My point is that if statistical techniques such as maximum likelihood were
>not developed, the criterion of discordance would apply to both
>morphological and molecular data.  Because the tecniques apply to
>molecules, the fault of discordance went to the untested partition of the
>data, namely, morphology.

Maybe it's just me, but favoring partition A over partition B simply
because you applied a statistical test to A but not B seems to violate the
most basic premises of statistical testing.  In any case, _my_ point was
that statistical methods came into being because more direct ways of
analyzing the results of molecular data sets were unavailable.
Morphologists seldom use these statistical methods because they prefer the
direct approach, and because of concerns about the statistics themselves,
which, for example, seem to address questions that, in practice, are seldom
relevant to morphological data sets (e.g., "Is there a significant
phylogenetic signal in this data set?"). In this view, the mere
availability of Statistics does not automatically confer superiority to one
kind of data.

>It would be nice to not have to rely on the tree to tell you about the
>quality of the data, because it smacks of circularity.  A recent
>perspective on reciprocal illumination is tied to suspicion of inductive
>reasoning, and preference is given measures of true probabilistic support,
>which induction simply cannot provide.

Circularity arguments are logical landmines. I for one am skeptical of the
argument that "because you use the characters to build the tree, you can't
use the tree to study characters." And in any case, I described using the
tree to identify characters that should be re-evaluated, not to do the
actual evaluation. There are lots of interesting issues here!




Alan W. Harvey (aharvey at
Assistant Curator of Invertebrates
American Museum of Natural History
Central Park West at 79th Street
New York, NY 10024
(212) 769-5638; fax (212) 769-5783

More information about the Taxacom mailing list