Re: What is a Dinosaur?

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

> >There is a basal dinosaur: The Most Recent Common Ancestor. Recently it
was
> >debated on the PhyloCode mailing list whether such ancestors are an
entire
> >species or just one breeding pair.
>
>    The problem with phylogenetic-type definitions (that is, everything is
> node- or stem-based) is that they don't help you with anything other than
> finding order to evolution.  Note that this isn't a _bad_ thing, simply a
> limitation!  After all, if I go out and dig up a bone, I can say that the
> bone (for example) has this and that trochanter, a head oriented this way,
a
> crest here, a fossa there, etc., etc., etc.  Knowing that "Dinosauria" is
> "the most recent common ancestor of _Megalosaurus_ and _Iguanodon_" (or
> whatever taxa one prefers to put there) "and all its descendants" doesn't
do
> _squat_ for helping identify whether or not the animal I've got is a
> dinosaur.  In the long term, it is _unavoidable_ that everything is
defined
> (not just diagnosed!) by the physical traits.  If we _only_ had the
> aforementioned definition of "dinosaur," then we'd never be able to say
> whether or not this, that, or the other bone belonged to a dinosaur.  In
> other words, we _have_ to _also_ have those dreaded character-based
> descriptors, or else, as Nixon & Carpenter have pointed out, one could sit
> down and do phylogenetics without ever having to look at a specimen -- it
> would become pure theory and no data.

But *Megalosaurus* and *Iguanodon* have certain characters. They will share
some of those, and many of those with certain other animals which are
_thereby_ identified as dinosaurs, and none others. Diagnoses aren't
abolished, as Nixon & Carpenter seem to think (I don't know their papers),
they are still there, just that they aren't called definitions anymore.

>    Please note that I am _not_ advocating reverting solely to character-
> (apomorphy-) based clades.  Certainly, if dinosaur (and other taxonomic)
> studies in the last several decades has taught us anything is that mosaic
> evolution certainly does occur -- where one taxon may have character
states
> A, B, D, and E, another may have A, C, D, and E, and another one B, D, and
> E, so having a taxonomic group that is "defined" (diagnosed, if you
prefer)
> by "possession of A, B, C, D, and E" creates the sorts of chaotic messes
> that node- and stem-based taxa are intended to avoid.

Situations like this "simply" require more & better research. In the above
example it may turn out that only D and E are synapomorphies of all three
taxa, or that the third one has lost A while the second has lost B, or that
convergences have occurred, or whatever. Look at birds --
Confuciusornithidae, apparently certain Enantiornithes, and Neornithes are
toothless (convergence), *Rahonavis* and something like Ornithothoraces have
joints between scapulae and coracoids (convergence), while e. g.
*Archaeopteryx* and *Sinornis* have shorter 3rd than 2nd fingers
(symplesiomorphy, or more exactly the opposite is an autapomorphy of
Confuciusornithidae).

> I _am_, however,
> saying that we should quit pretending that the "diagnosis" is somehow
> inferior or subordinate to the "definition."

Definitions are the stable things for theory, while diagnoses are the
changing ones for practice. Some HP had in his signature in 1994 "In theory,
theory and practice are the same; in practice, they aren't" :-)

> We _need_ both, and I
> personally see absolutely nothing wrong with saying "Our _current_
> definition of Dinosaur is 'the common ancestor of XXXsaurus and YYYodon
and
> all it's descendants' _AND_ 'animals displaying characters G, H, I, J, and
> K.'"  As more data comes to light, the latter part is subject to change,
of
> course...but at least we can then deal with new discoveries (or
re-evaluate
> old ones) based on something that we can observe directly.

What about "'the common ancestor of [...], unless the almighty Commission
should ever rule otherwise ;-) '; and 'those animals happen to share, as far
as the scrappy fossils allow us to say at the moment, G, H, I, J, and K, so
we think that animals 1, 2, 3 and 4 belong into this group while 5 and 6
don't, unless we find convincing evidence to the contrary'?