More on evolution as doctrine

Mon Nov 11 09:10:57 CST 2002

-----Original Message-----
From: Richard Zander

>Some evolution science is philosophy-based, e.g. cladistics s.str. Other
>evolution science uses standard null hypothesis testing. 
>... One basic null in evolutionary science is that...
>(i.e. all branches are sister groups and the support for
>the optimal branch is 1/3 that of the sum of the support for the other two
>branches) and all resolution is randomly generated. Rejection of the null at
>some level of confidence is an example of scientific evolutionary [ ! ] analysis.

Perhaps it is a side point but I see some misunderstanding here: the hypothesis that can be tested against null is that the characters-state pattern is random (given some underlying distribution function). However, even if you find a non-random pattern, you are still unable to say if this was generated by evolution, i.e. you are testing a hypothesis about pattern, not a hypothesis about process.
This has in common the cladistics s.str. and any other phylogeny reconstruction method.
Best!
Zdenek Skala
skala at incoma.cz