Who is the postivist?

Richard Zander bryo at COMMTECH.NET
Thu Dec 11 10:05:36 CST 1997

Ted Schultz wrote:
>  Personally, I require lots of
> character support from lots of different character systems.  When
> completely unrelated systems keep telling me the same answer over and over,
> I start to think that maybe there's phylogenetic signal coming through.  It
> certainly can't be due to chance, and it strains my credulity to think that
> such congruence (nuclear genes, mitochondrial genes, adult characters,
> larval characters, etc.) could be due to convergence/parallelism as well.

No problem with this. The gist of my contribution to this thread is
perhaps best illustrated in the following analogy/parable:

A group of archeologists have a method of dating artifacts. They publish
a date for artifact A of 6547 years b.p. They say this value approaches
the true date and their method logically guarantees convergence on the
truth. A second group of archeologists develops a different method. They
publish a date for the same artifact as 6832 b.p. You read further in
that second publication and discover that the second group's method
allows them to estimate confidence limits and their date, buried in the
paper, is actually 6832 plus or minus 650 years. Why did they feature
the single value? Probably because they are competing for grants with
the first group. What is the significance of the confidence limits? It
throws light on what the words "approaches" and "converges" really
should mean to the first group.

The first group are cladists. The second group are statistical
phylogeneticists, and the confidence limits are equivalent to the number
of trees comprising the "credibility zone" of trees whose posterior
probability adds to .95. If someone could tell me how this analogy
misses the boat, I'd appreciate it.

Say, can anyone with PAUP* beta tell me if posterior probabilities are
calculated in the likelihood analysis said to be part of that
soon-to-be-released-any-time-now program?


Richard H. Zander, Buffalo Museum of Science
1020 Humboldt Pkwy, Buffalo, NY 14211 USA bryo at commtech.net

More information about the Taxacom mailing list