Re: Sereno's (2005) new definitions

[David Marjanovic <david.marjanovic@gmx.at>]

> > This seems to be based on a lack of understanding of how characters are
> > used to generate a phylogenetic hypothesis...
>
> Oh dear.  Don't let George hear you say that.  He'll give you a serious
> ass-kicking.  (Though I can't say I disagree with you, David.)

I have already told him (on another list). Apart from getting angry, he 
didn't reply.

> > This is mainly good old parsimony-with-one-character (in this case the
> > number of phalanges in the 5th toe). This character is deemed to be
> > irreversible, no matter what the other characters say.
>
> A beauty contest, in other words.  Winner takes all.

That's a way to say it, yes.

(However -- in the case of this particular character I demonstrated on said 
mailing list that the interpretation that reversals happened is on the 
currently accepted phylogeny equally parsimonious as the interpretation that 
no reversals happened and that instead the 5th toe disappeared several times 
independently.)

> > Yep -- but it will still be a troodontid (and probably a quite derived
> > one). That should be unambiguous enough for not harming its function
> > as a specifier.
>
> Here's the problem...  If a taxon is deemed to be a nomen dubium it 
> becomes
> an invalid OTU, and so must be excluded from phylogenetic analyses.
> Therefore, its exact position in a phylogeny cannot be determined.

This is not at all true. The holotype of a nomen dubium is an organism, and 
there's nothing wrong with putting individual organisms into a data matrix. 
After all individual organisms have a phylogeny. In fact, in the many cases 
where we don't know whether a given variation is intraspecific, the _only_ 
secure way of doing phylogenetic analyses is to use individuals as OTUs.

The T in OTU does stand for "taxonomic". But this is a fossil from the 
phenetic era. It is entirely meaningless. :-) (Phenetics was, I'd say, a 
theory of taxonomy.)

We can be reasonably certain that the holotype of *Troodon formosus* belongs 
to the clade commonly called Troodontidae. So I can't find anything to say 
against anchoring Troodontidae on *Troodon*.

However, if (an entirely speculative example) someone would define, say, 
"Troodontinae" as "everything closer to *Troodon* than to 
*Saurornithoides*", then we _would_ have a problem because without the 
referred material it seems not to be possible (at the moment) to determine 
exactly enough where within Troodontidae *Troodon* belongs.