[Taxacom] quote of the week

[Pierre Deleporte]

John Grehan wrote :
>  Sorry, you must have misunderstood me somewhere along the way. I thought
> I had mentioned on this list that I am currently working on a
> morphological analysis. 
maybe, but you are accusing molecules all the time (while not even 
trying to analyse them, and this is the problem)
> Maybe I did not specifically mention computer
> programs, but I am using PAUP - principally because that is what all the
> other primate morphologists have used. But in addition I did to a Bremer
> support analysis, which the others did not do. 
>   
so it seems that you have found a cladistic analysis program at last... 
so you will hurry and treat the molecular data the same way as you are 
treating morphological ones?...
> We are looking at all character states that may be
> found within humans and the great apes that support any relationships
> between any of them. In other words, character states unique to humans
> and orangutans, humans and chimps, humans and both African apes, chimps
> and gorillas. We have not focused so much on other combinations although
> some may be valid. 
programs check for all possible 'combinations', even those you or 
anybody else didn't ever thought to 'focuse on'... this is why people no 
more have to check in advance what will be the possible contribution of 
each and every character to any conceivable topology (what no human 
brain can tract beyond some degree of size and incoherence ot the data 
set... e.g., 'all primates')...
this is precisely why we use programs... surprising, isn't it?   ;-)

now, are you certain that you are using your program according to your 
self-claimed criteria for 'cladisic' analysis? I don't think that 
default options in PAUP will search for a clique of 'uniquely derived' 
characters only, i.e. weighting "zero" partially homoplastic characters, 
as your unpublished algorithm seems to require;  this would necessitate 
implementing specific options of ad hoc programs...

maybe this is why you are still checking a priori character by 
character, which seems oddly contradictory with using a program at all...


Pierre


-- 
Pierre Deleporte
Université de Rennes 1
CNRS UMR 6552 
Station Biologique de Paimpont
F-35380 Paimpont   FRANCE
Téléphone : 02 99 61 81 63
Télécopie : 02 99 61 81 88