Polytomies

Kip Will kww4 at CORNELL.EDU
Mon Jun 16 23:59:25 CDT 1997


Eric Zurcher wrote:
>Ah, there indeed is the rub! There is no way of KNOWING (at least not
>reliably, via any current techniques of which I am aware) when we are in
>fact looking at a true evolutionary polytomy.

We certainly can't know the truth of these historical events but we can know
that the evidence (characters) we have at hand either support complete
resolution or are ambiguous. It is necessary (and I think correct) to expect
dichotomous relationships among taxa. This way we can recognize the less
common polytomies (and reticulations) that stand out from the usual
dichotomies. If we accept polytomies as the norm, or even as occuring as
frequently as dichotomies, then how can one know when to show resolution? We
can keep trying to resolve until the conflicting evidence shows us that this
is probably a "true" polytomy. A sort of 'auxiliary principle' if you will.
No evidence is needed (in fact data is ignored) to de-resolve relationships.

>But I would regard a forced
>or apparent "resolution" of "true" polytomies as a greater pitfall than the
>failure to resolve "false" ones, for false "knowledge" is worse than mere
>ignorance.

A bush certainly must be the best since it shows no false relationships, no?
In fact, no, since a polytomy implies many possible resolution that may not
be true.  Unless you mean by "false 'knowledge'" and "forced resolution"
that relationships of  taxa are established without any evidence, I can't
see any way that one's best assessment of the data can be worse (even when
wrong, as we are so often) than accepting no information.


kip will




More information about the Taxacom mailing list