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

More segnosaur stuff

I've been extremely busy with non-dinosaurian work these past several months,
with practically no time to respond at length to any posts. But I do save the
material for future responses, and every so often I crank one out. Here is
one for Herr Wagner, to a post dated way back in May. I understand he's flown
the coop for a while, but I'm sure this will eventually make its way to him
soon enough:

Subj:   Re: Segnosaurs again [was Oviraptors]
Date:   97-05-01 08:23:13 EDT
From:   znc14@ttacs1.ttu.edu (Jonathan R. Wagner)
To:     Dinogeorge@aol.com
CC:     dinosaur@usc.edu

<<Dinogeorge wrote (Don't it give you a warm fuzzy to be arguming again,

Arguming? Who's arguming?

<<>How swiftly we forget that a number of what are now considered
>were once classified as coelurians:
        How swiftly we forget that _Basilosaurus_ was named as a lizard,
tyrannosaurs were considered carnosaurs, birds were considered closer to
mammals than to crocs, dinosaurs were thought to have been killed by a great
flood six thousand years ago...>>

You miss the point, as usual. The point is that small prosauropods are
similar enough to small, primitive theropods to have once been classified as
coelurians, so one might expect that derived prosauropods, that is,
segnosaurs, could now also >mistakenly< be classified as coelurians. Please
do >not< try to tell me that cladistic analysis is infallible.

<<>The similarities of some features of _Alxasaurus_ and _Erlikosaurus_
>to features found in some coelurian genera can easily be explained as
        You could just as easily explain the similarities between humans and
chimpanzees as convergences as well. This is something you never seem to get
in this (interminable) argument: Just because you *can* explain them as
convergences doesn't mean you should. Look for the explanation which best
fits the available evidence, with the fewest unjustifiable assumptions and
the least number of untestable hypotheses.
        Therizinosaurs are coelurosaurs.>>

Unlike many cladists, I look for explanations and phylogenies that make
sense. I do not agree that the explanation that "best fits the available
evidence, with the fewest unjustifiable assumptions and the least number of
untestable hypotheses" is >ipso facto< the "best" explanation. What does
"best fits" mean? That a simple, unweighted numerical majority of characters
suggests a certain relationship? Who proclaimed this particular method of
weighing evidence to be the "best"? Has anybody checked this particular
assumption against reality? Until this is done, and convincing evidence is
found that Nature >always< works parsimoniously in these matters, this is
nothing more than a philosophical homily. Because nobody has ever tested
cladistic analysis and method against known or controlled phylogenies, or
checked cladistic analysis with anything other than more cladistic analysis,
there is no particular reason to think that cladistics is any better or worse
than any other method of doing phylogeny, or that cladistic phylogenies are
any better supported than any other phylogenies constructed by knowledgeable
paleontologists using other methods. Why, for example, do cladists fiddle
with their cladograms before publishing them?

I happen to think that cladistic analysis yields a reasonable family tree
about 90% of the time, based solely on my own view of how many cladograms
look reasonable versus how many do not. It's those other 10%, where cladistic
analysis has been misled by excessive convergence (for example) or incorrect
character interpretation, that I think I can identify. Unlike you, I do not
accept cladistic analyses as infallible unless disproved by more cladistic
analyses. And I certainly do not give the same weight to minor, variable
characters randomly scattered throughout a dinosaur's skeletal anatomy as I
do to characters in unified postcranial character complexes such as feet,
limbs, pelvis, and vertebral column.

When cladistic analysis yields an incorrect result, which I think it does in
the segnosaur case, then any similarities >must< be convergences. I consider
a smattering of random theropod-like characters in the skull of an obvious
non-theropod to be of little significance--just homoplasies picked up during
the course of segnosaur evolution. Skulls have an immense number of
distinguishable features and character loci, and theropod-like characters can
be found in practically any archosaur cranium from rauisuchians to crocs to
other kinds of dinosaurs. With segnosaurs, someone has made the entirely
unjustified a priori assumption that they're theropods, so they're thrown in
with theropods in cladistic analysis. Since the analysis must yield >some<
kind of answer, we suddenly find segnosaurs as a sister group to
ornithomimids (as in Sereno 1997) or in some other equally untenable position
on the theropod family tree. Why not, for a change, toss segnosaurs in with
prosauropods and sauropods and see what happens?

<<        Once again, what the begumbus is a "coelurian"?>>

The name Coeluria Marsh 1881 has priority over Coelurosauria von Huene 1920
(or thereabouts). They're identical groups.

<<>I've read that paper several times and I have yet to find >any< feature of
the >skull of _Erlikosaurus_ that bars a prosauropod ancestry for
        That's not relevant. It is the shared derived characters which
elucidate phylogeny, not whether or not prosauropods *could* be their
ancestors or not. Heck, if you wanna stretch it, all theropods *could* be
descended from prosauropods, especially as you describe them!>>

Of course it's relevant. When the cladistic analysis goes haywire, it's time
to look around for other plausible groups. I consider most of the known
prosauropods to form a sister group to segnosaurs plus ornithischians within
the larger clade of herbivorous dinosaurs.

<<>In _Alxasaurus_ the two carpal elements that become the so-called
"semilunate" >carpal in _Therizinosaurus_ are still unfused
        Ontogenetic? Bones have been known to fuse late in ontogeny. In any
case, the form of the bone is clearly there. Which prosauropods have a
semilunate carpal block?>>

The (usual) missed point here is that the semilunate carpal is a touted
apomorphy of the theropod group that supposedly includes segnosaurs.
(And--why should prosauropods have a semilunate carpal block??) _Alxasaurus_
is an early segnosaur that does >not< have the semilunate carpal block.
Therefore the semilunate carpal block developed independently in segnosaurs
and theropods, unless it just happened to be reversed in _Alxasaurus_. Which
alternative is likelier, considering that the carpus of _Alxasaurus_ bears
little other resemblance to theropod carpi with semilunate blocks? So I
assert that the semilunate carpus does not exist in Segnosauria, and instead
there is a convergent fusion of some carpal elements in the later, more
derived forms.

>on the other hand, I consider the structure of the segnosaur foot as making
>virtually impossible for coelurians or any other theropods above the level
>_Herrerasaurus_ to be segnosaur ancestors.
        Until you can come up with a convincing explanation for how this
works, I simply cannot believe it.

You must be kidding, right? Here the shoe >has< to be on the other foot, so
to speak: It's up to the theropod people to show how the highly derived
theropod foot reversed into the prosauropod-like segnosaur foot. Just saying
it's a "reversal" is not nearly enough; to a cladist, anything that doesn't
fit a pet phylogeny is a "reversal." I don't have to worry about this
problem, since in my phylogeny segnosaurs simply acquired their
prosauropod-like feet from their ancestral prosauropods.

>As soon as I find anything in the literature convincing of a theropod
>relationship for segnosaurs, I'll let everyone know.
        I believe you once quoted to me: "They cannot see who do not open
their eyes."

So open yours already and stop simply spouting old-hat cladistic dogma.