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

Re: Torosaurus NOT Triceratops



> > Yes. It's not who analyses the data, but whether enough
> > methods are used.
>
> Um, actually, no. The data in this paper can be used to 
> argue that they're separate taxa OR to argue that Toro. is 
> Triceratops, just like you can pretty much show anything 
> you want using many published character matrices. All you 
> have to do is miscode one or two key characters for one 
> or two key specimens.

Then it's technically not the same dataset anymore...

> Note the asterisks in Longrich and
> Field's table. Statistically, i think that table/data
> shows that Torosaurus *is* Triceratops just as much as
> it shows the opposite. The more immature specimens are
> skewed toward Triceratops while the most mature are skewed
> toward Toro. I know that doesn't prove anything, but it's
> statistical.

Methods: is this skew statistically significant?

Never mind proving -- just quantify.

> My hunch is that Torosaurus is valid, at least *some*
> Toros. In the end they may all prove to be synonymous,
> but what about those Toros in UT? Just a sampling issue?
> And what about immature chars in some Toro skulls? John
> Scannella and the MOR group have tried to address that
> using histology, but the dataset is still incomplete.

>From a (too) incomplete dataset you can't say both things, you can't say 
>anything.

> All dinosaurs are morphotaxa (er, all *extinct* dinos),
> so sample size is crucial. There is a lot of circularity
> in trying to separate individual variation from ontogeny
> from taxonomy from dimorphism in similar taxa. You can
> apply the tools, but in the end, it's largely statistical.

Then let it _be_ statistical and do appropriate statistical analyses. I don't 
know how else to deal with individual etc. etc. etc. variation.

> I think in a year or two you will see someone arguing
> that Toro is the mature form only of *male* Trikes.
> Despite Horner and Padian on dimorphism, that seems to
> me a more likely

Wouldn't surprise me either. But I haven't even read the new paper.