Weights

Tom DiBenedetto tdib at UMICH.EDU
Fri Mar 7 20:06:43 CST 1997


 Guy Hoelzer wrote:

>So you conclude that, if your assumption of parsimonious evolution is correct,

My WHAT? C'mon Guy, are you trying to be provacative or what?

>your tree will accurately identify homologies and homoplasies in your data.  I
>can see this approach to refining your hypotheses of homology, but that is
>different from leaping to conclusions about phylogeny.  I am also concerned
>about the assumption of parsimonious evolution,

The WHAT?

>because unparsimonious evolution
>could lead your algorithm to confirm your homology hypothesis in instances where
>it should be rejected.  The advantage of a statistical approach is that it can
>indicate specific instances where the data suggest you should be concerned about
>such an error.

To be brief,,,we do not assume "parsimonious evolution". There were
some statistical folks back in the sixties who tried to use a
parsimony algorithm for such purposes, but such assumptions play no
role in our system. Where do you get this? Certainly not from
anything I have ever said, nor from anything I have ever heard of a
cladist saying or writing. Is this one more of the canards that you
stats folks love to hurl around?
We are parsimoniously combining a select group of corroborated
hypotheses of homology; we are not claiming that evolution is
parsimonious.




More information about the Taxacom mailing list