Re: Apomorphy-based definitions - who needs them?

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

> Two years ago the manuscript of a new high school biology
> textbook had a cladogram, and the synapomorphy for birds was "feathers".

A high school biology textbook with a cladogram!?! Hey! The future has
begun! :-)

> They seem to cause a lot more headaches than the apomorphy-based Mammalia
> (which we might have to refine slightly ever few decades).

Please tell which character you use for Mammalia.

> A
> paraphyletic group is just a truncated clade that has an upper boundary as
> well as a lower boundary.  Therefore, there is nothing unnatural or
> unscientific about a paraphyletic taxa.  Cladists keep saying they are,
but
> that will never make it true.

Hm... a singly paraphyletic group is exactly twice as arbitrary as a clade
:-) -- it has an arbitrary beginning as well as an arbitrary end. More
arbitrary ends for multiply paraphyletic taxa like traditional Reptilia,
Thecodontia etc.. That end is an arbitrary end as well as the arbitrary
start of another taxon. On the contrary, the arbitrary start of a clade need
not be the end of any taxon. Phylogenetic taxonomists need not worry "ow, if
I don't include this fossil in Mammalia/Aves, I must refer it to Reptilia...
no, I can't do that, I can't make that a reptile. But can it really fit in
Mammalia/Aves?" and so on. In phylogenetic taxonomy, Tritylodontidae belongs
to Mammaliamorpha but not Mammaliaformes or Mammalia; *Sinoconodon* belongs
to Mammaliamorpha and traditional Mammalia, but not to Mammaliaformes and
the crown-group; and so on. None is in any danger of falling into Reptilia
or suchlike. I think this way the transitional nature of every fossil is
much, much better represented.
        In another way, paraphyletic taxa are much more arbitrary than
clades: they are the inevitable consequence of Linnaean ranks. Putting
Mammalia, Reptilia and Aves next to each other on the same level (as
classes) is obviously arbitrary and not inevitable.

> Paraphyly is a real evolutionary
> phenomenon, while Hennig's convention (of a mother species giving rise to
> two "matricidal" sister species) is just that----a convention that makes
> cladistic analysis more useful.  Sister groups don't really exist in
nature,

Sister _species_ may be a rare phenomenon.* But who says the borders of
clades must coincide with the borders of species? All those ancestors can
(usually) easily be individuals or breeding pairs that can sit squarely in
the middle of a species.

* Certainly not a totally inexistent one -- if a population is split in half
by a barrier, and the 2 new populations evolve away from each other at the
same speed, Hennig's convention is indeed the best way to describe reality,
because it would be completely arbitrary to refer one of the daughter
species, but not the other, to the mother species. But I'm certain this case
is very rare.

>      Anyway, I obviously won't be adopting "pygostyles" for any
> apomorphy-based definitions, or anything else that is imprecise and likely
> to arise convergently.

I think pygostyles can be described pretty precisely -- the vertebrae either
fuse or don't. If necessary you could probably find a way like using fusion
of the centra or neural arches or whatever. Why do you think they are likely
to arise convergently?
        Anyway, I would not use them in your place -- we know far too few
tail ends of relevant groups IMHO.

> The "true" semilunate seems to have enlarged rather
> suddenly (geologically speaking) between ornithomimes and maniraptors.

Depends _heavily_ on your preferred phylogeny and on our knowledge of the
fossil record. At the moment, it does seem that the (usually) fused distal
carpal 1 + 2 is either small and has a rather flat proximal surface, or big
and has a semilunate proximal surface. Who knows what transitional state
will next come out of Liaoning or wherever else, maybe tomorrow, so you
can't even count on the "several decades" a definition might stay unchanged.