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

RE: Stegosaur volume of Swiss Journal of Geosciences

Tim Williams wrote-

> > Well the postorbital horns were quite distinctive for the
> > time, and even now they're almost good enough to be
> > diagnostic for Ceratopsidae.  It's only through historical
> > accident that Ceratopsidae became a node-stem triplet clade
> > that doesn't encompass things like Zuniceratops.
> But how do we know that Ceratopsidae encompasses _Ceratops_?
> How many phylogenetic analyses have been done recently that include 
> _Ceratops_?  I'm guessing the answer is zero, or very close to it.

Long postorbital horns are only known in ceratopsids and their close sister 
taxa (Zuniceratops, Turanoceratops, etc.).  If you think you have evidence for 
a different placement of Ceratops, feel free to share.

I was countering your statement that Ceratopsidae should have never been named 
by Marsh because Ceratops was so fragmentary.  If Ceratopsidae happened to be 
defined as "everything closer to Triceratops than to Protoceratops", Ceratops 
would still belong to the clade with no question.  Even with the current 
"Chasmosaurus + Centrosaurus" definition, Ceratops was still most 
parsimoniously a chasmosaurine ceratopsid because of its long postorbital horns 
until the very recent discovery of basal centrosaurines and ceratopsoids with 
such horns.  Marsh's family lasted over a century without incident as a 
monophyletic taxon including Ceratops, so I think his erection of Ceratopsidae 
was justified.

Instead of ditching Ceratopsidae, why don't we define it in a way that will 
include Ceratops?  "Ceratops montanus <- Protoceratops andrewsi, Leptoceratops 
gracilis" or something?  The only difference from the current content would be 
the addition of Zuniceratops, Turanoceratops and such.  The Phylocode isn't 
even enacted yet so there are no official definitions of clades, whereas the 
ICZN has been going for decades and only cares that Ceratopsidae contains 
Ceratops.  Seems like the smart thing to do is to work within the ruleset that 
exists in a way that will also function under the Phylocode once it begins, 
instead of using an unofficial definition for a code that has not been enacted 
yet as an excuse for subverting a basic rule of the established code.

> > I'd rather have rule-based taxonomy than judgement calls
> > and "experience" any day.  Didn't we have enough
> > subjective, authority-based taxonomy in the past?
> Yes we sure did.  But that's not what I'm saying here.  I'm simply talking 
> about the naming of clades.

The naming of clades is part of taxonomy.

> Phylogenetic taxonomy is phylogeny-based, not authority-based.  For 
> phylogenetic taxonomy it's my argument that family-level clades should be 
> named after one of the better-known (scientifically speaking) members of the 
> clade.  Most families already are: Tyrannosauridae, Allosauridae, 
> Neovenatoridae, Coelophysidae, Brachiosauridae, Camarasauridae, Diplodocidae, 
> Stegosauridae, Ankylosauridae, Iguanodontidae, Pachycephalosauridae, 
> Herrerasauridae, and so on.

Allosaurus fragilis is probably indeterminate within Allosauridae based on its 
holotype. Neovenator should be a carcharodontosaurid under that family's 
definition, and Neovenatoridae was based on the results of a single dataset.  
Coelophysidae was also based on probably indeterminate material that had to be 
exchanged for a neotype.  Iguanodon and now it seems Stegosaurus were based on 
material indeterminate at some level, and the new Iguanodon is probably not the 
same as the original Iguanodon (likely to be Kukufeldia).  The "Brachiosaurus" 
that is well known is Giraffatitan.  Pachycephalosaurus' type is just a skull 
roof that could prove indeterminate once we have more complete 
pachycephalosaurines.  As far as I know, there hasn't been an in depth analysis 
of Diplodocus or Camarasaurus species in two decades, so who knows how longus 
(several caudals and chevrons) and supremus (8 vertebrae) will hold up.  These 
kind of examples are why we need strict priority-based taxonomy which does not 
fall when its eponym is found to be an undiagnostic member, as opposed to some 
subjective, authority-based system where things get renamed or lose their type 
status whenever someone thinks they're undiagnostic.

> I don't see why the same process can't be used to decide the names of 
> family-level clades, and have them all based on 'model' genera - again for 
> the sake of nomenclatural stability.  To give one example, at the moment 
> there is a carnosaur clade that includes both _Metriacanthosaurus_ and 
> _Sinraptor_.  The ICZN would say that the name of this family should be 
> Metriacanthosauridae rather than Sinraptoridae, because the former (as 
> Metriacanthosaurinae) was named first.  But _Sinraptor_ is based on an almost 
> complete skeleton, whereas _Metriacanthosaurus_ is known only from 
> fragmentary material.  So I would say Sinraptoridae is a better name for this 
> clade, because it's founded upon a well-known genus.  Sinraptoridae also has 
> the benefit of having a phylogenetic definition, which includes _Sinraptor_ 
> as a specifier. For the purposes of phylogenetic taxonomy, I can see no 
> advantage in having Sinraptoridae being trumped by Metriacanthosauridae.

As the above examples show, what seems like a justifiably valid taxon now may 
not be judged that way in the future.  It's not like our views on these matters 
will be any more valid in 2100 than Marsh's views are now.  Sure Sinraptor 
dongi's holotype is rather complete, but there are questions as to its validity 
compared to the earlier named hepingensis and Yangchuanosaurus.  
Yangchuanosaurus was named over a decade before Sinraptor, and has a more 
complete sacrum and tail.  Under your system, why not use "Yangchuanosauridae" 
as the proper name for the family?  It's all subjective anyway, so you can't 
say I'm wrong that the 15 years of genus priority, possibility of 
Yangchuanosaurus being the correct genus name, and more complete sacrum and 
tail trump Sinraptoridae being named and defined first and being based on a 
specimen with preserved manus and pes.  

> Robotically applying ICZN rules to phylogenetic nomenclature is often 
> downright unhelpful. It's the workers in the field of paleontology who are 
> best placed to decide which genera should give their names to higher clades, 
> not the ICZN.

"Robotic" application of rules is meant to sound bad, but what it really means 
is not privileging certain names subjectively.  Much like science, we should 
follow the taxonomic rules objectively, wherever they lead.  If Sinraptor turns 
out not to be a metriacanthosaurid, that's a scientific discovery, not a 
disaster.  And paleontological workers HAVE decided which genera should give 
their names to higher clades.  They're who name the clades in the first place, 
after all.

Mickey Mortimer