[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]

Re: ADV: Re: ADV: Re: Classification: A Definition



On Wed, May 23, 2007 at 09:24:15AM +0800, gerarus@westnet.com.au
scripsit:
> >>>  > > I like the convenient of TOKOTs.  I prefer the name
> >>>  > > "Brachiosaurus" to the specimen number FMNH P 25017.
> >>>  > 
> >>>  > How would defining "Brachiosaurus" *as* "FMNH P 25017" cause
> >>>  > difficulty or information loss?
> 
> As Anthony Docimo already implied, only in that names are _far_ more
> easily remembered than numbers.

Which has something to do with why everybody gives their specimens
nicknames, too, but the unique identifier remains the institutional
collection number.  It's generally a good thing if the unique identifier
is readily recoverable.

> >>If we're after mathematically supported phylogentic hypotheses by
> >>whatever means, the actual cladistic analysis has to work from
> >>specific specimens.
> 
> Well, yes and no. Ideally a cladistic analysis should not be run
> without taking variation within the population into account. 

There are certainly constraints on how specimens should be used, but
when you code characters to run the analysis, it ought to be from a
specific specimen, and the folks reading your eventual paper ought to be
able to identify that specimen, no?

> Hennig (IIRC) formalised this in the jaw-breakingly-named concept of
> "character-bearing semaphoronts" - essentially, males should only be
> compared to males and females to females in relation to characters
> that are sexually dimorphic; characters that vary ontogenetically
> should only be compared between individuals of the same ontogenetic
> stage, etc. etc. 

Sure.  Though identifying all of those things ahead of time in fossil
taxa where the specimen count is small is going to be really
challenging.

> A paper referring to just this issue arrived in my inbox today
> (http://www.pnas.org/cgi/content/abstract/104/21/8731) suggesting that
> many phylogenetically significant characters of Catarrhini (Primates)
> are only present in males. I know that recognising such variation in
> fossil taxa can be difficult, but it remains a complicating factor,
> and cladistic analyses are better run on populations than individuals.

That's going to give you a two-stage coding process, though, which --
while desirable -- isn't going to be possible for all but a very few
dinosaurian taxa, because there just aren't enough of them.

> And in reply to Mike Taylor and Mike Keesey - you're right, I wasn't
> disagreeing with phylogeny as an ideal in classification. My idea was
> that in cases where the ideal could not be met, that should not
> prevent all development.

I would certainly agree with that -- otherwise, nothing gets done!

-- Graydon