[Taxacom] base alignment

John Grehan jgrehan at sciencebuff.org
Tue Aug 29 08:28:09 CDT 2006

Pierre wrote:

> Sorry but this is exactly the definition of compatible characters: no
> homoplasy between the ingroup (taken as a whole) and any outgroup. It
> seems
> that I am correct.

OK - no problem.

> Right, but now you have to specify which procedure you will use for
> ingroup: compatibility analysis or otherwise?

I have no specific axe to grind as to which analysis to use. In the
orangutan case I am currently using PAUP since that is what others used
for the human-great ape relationship and I am happy to use the same
method in order to show that one gets a different result when
incorporating certain characters.

> Please note that using compatibility analysis outside (your prior
> selection
> of putative "cladistic characters"), and standard parsimony inside,
> be logically inconsistent: your results would differ according to the
> phylogenetic level of delineation of your ingroup versus outgroups. 

The characters for orangutans are pretty darned unique. Tho or three
might come under scrutiny - thick dental enamel occurs in two monkeys
(one OW, one NW), ischial callosites are absent in NW monkeys,
prosimians, and all other mammals).

In the
> extreme, analysing the data matrix of "Life" would leave you with no
> for prior selectioin of "cladistic (copmpatible) characters at all...
> logical inconsistence seems obvious enough.

It seems to me that systematists do not analyze the entirety of life all
at once. They select a putatively monophyletic group to start with.

> Some people perform(ed) compatibility analysis. I merely suggest that
> nobody calls compatible characters "cladistic characters" like you do.
> Please quote some paper using this terminology (just one)...

I can't be bothered. Maybe you are right and there is no one else say it
like I do. I don't really care. The real issue is whether there is more
cladistic evidence for the human-orangutan relationship than the
human-chimpanzee relationship. Everyone is free to analyze this problem
with whatever terminology and algorithm they prefer.

But for
> everybody I know "phenetic" means classifying on the basis of overall
> similarity of taxa, not selecting compatible characters.

I agree with this definition. And according to that definition cladistic
analysis of molecular characters appears to be an analysis of characters
established through overall similarity of base substitutions and gaps.

Others who are computationally advanced might also find my positions on
cladistic theory to be algorithmically naive. Perhaps so. In the end it
is the empirical data that will have to be addressed, and for the
orangutan question this has hardly begun. So far I have not had any
responses to the Mona Lisa paper saying that the analysis of evidence is
garbage or that my observations about questions in molecular analysis
are demonstrably wrong.

John Grehan

Taxacom mailing list
Taxacom at mailman.nhm.ku.edu

More information about the Taxacom mailing list