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

Re: Dinosaur Web Pages' Re-Opening [awfully long]

At 01:34 AM 9/4/97 -0400, Dinogeorge wrote (quoting me quoting him ad naseum):

>If Dinosauria is defined as the common ancestor of _Iguanodon_ and
>_Megalosaurus_ plus all its descendants (best definition I've seen, if I do
>say so myself), then Saurischia is congruent to it.
        Best, but lacks priority, see Padian and May.

>If you define Dinosauria
>as the stem group you describe, then Saurischia would by my phylogeny
>certainly NOT be synonymous with Theropoda, since it includes the
>phytodinosaur _Triceratops_.
        Saurichia is not affected by my definition of Dinosauria, and
certainly has never been defined by me to include _Triceratops_.
        In any case, defining:
        Saurischia = {+ "birds", -_Triceratops_} Gauthier 1986 and
        Dinosauria = {+Saurischia, +Ornithiscia} which reduces to 
                   = {+"birds", + _Triceratops_}
        Theropoda  = {+"birds", - Sauropodomorpha} which I ammended to
                   = {+"birds", - Sauropodomorpha, -_Triceratops_}
                        for the occasion.
Then, by your phylogeny [ ("theropods", ("sauropods", "ornithischians")) ],
Theropoda has the same common ancestor as Saurischia. Thus, by the terms of
phylogenetic taxonomy, the two taxa are synonymous.
        I goofed BTW, I meant node based taxon Dinosauria, as might have
been clear from the definition.

>If one can determine homoplasy within a phylogeny only by performing more
>cladistic analysis, then cladistic analysis is tautological and thus not
        No, because the idea is that the greatest number of overall shared
derived characteristics is more likely to tell you which groups are
phylogenetic. If we were to discard homoplaisies a priori, our analysis
would *certainly* mirror the prior conception of the phylogeny which led us
to discard the "homoplaisies" (since the only characters present in the
matrix would be characters we consider synapomorphic and non-homoplastic).
        Thus, a priori filtering of homoplaisies IS tautological. Allowing
the parsimony to determine which shared characters are homoplaisies is the
non-tautological approach, since all characters are treated as potentially
synapomorphic until demonstrated otherwise.
        So, George, how *do* you tell when a character is homoplastic? When
it doesn't fit *your* phylogeny? *That* is a tautology!

>If there is no way to check whether a cladogram models reality,
>cladistics is unscientific and is simply an exercise is making pretty
>patterns out of character matrices.
        This is simply soapboxing. We have covered this ground already.
Suffice it to say, lack of a priori knowledge about homoplaisy vs.
synapomorphy does not invalidate cladistic analysis. Indeed, if we had this
information already, we wouldn't need cladistics, the taxa would fall into
their groups naturally.
        The problem you don't seem to grasp is that in order to determine
which characters are homoplasies, you HAVE TO ALREADY KNOW THE PHYLOGENY.

><< >[...] important characters uniting the two groups into the clade
> >Phytodinosauria [i]nclude twin sternal plates,
>         This is likely to be plesiomorphic.>>
>Perhaps. And perhaps not.
        Uselss character until you demonstrate that it is synapomorphic.

><< >dermal armor
>         Plesiomorphic for dinosauria? I do not recall evidence for off-axial

>You might be right about this. Of course, it depends on what counts as dermal
>armor. Dermal armor is found in a number of phytodinosaurian groups but
>not--unless you count the small dorsal ossicles of _Ceratosaurus_--in
>Theropoda, as a rule.
        Of course, until you can demonstrate that dermal armor is primitive
for "phytodinosauria", the character is useless anyway. Just because
several; different groups developed it doesn't make it a valid synapomorphy.
To code it as such is to "code a trend", which IMHO is one of the most
subjective things you can do...

><< >a functional fifth pedal digit
>         Plesiomorphic for Ornithosuchia, and possibly more inclusive clades.
>[...]I do not recall there being any basal Ornithiscian with a functional
>fifth digit...>>
>The analyses are hardly "clear."
        But unclear analysis can still cast doubt on the value of this
character as a synapomorphy.

>I have yet to see any cladistic analysis
>that demonstrates beyond reasonable doubt that any terrestrial vertebrate has
>>regained< a functional digit that was previously lost.
        So digits are "special"? George, you have proven positively
intractable on this topic, despite numerous attempts by professionals,
semi-professionals, and myself to demonstrate the fallacy of your arguments.
I will not repeat the arguments here.

>Digital loss is a one-way street: down.
>Indeed, the occasional cladistic analyses that purport
>to demonstrate this phenomenon should be regarded with great skepticism.
        Now this *IS* tautological! You say a priori that it can't happen,
and any test which invalidates your statement is to be disregarded. No
wonder you haven't found an analysis which demonstrates this to you "beyond
reasonable doubt"...

><< >leaf-shaped herbivorous dentition.
>         As noted by Gauthier (1986), this morphology is a common adaption of
> reptilian tooth pattern, as it occurs in lacertillians as well as several
> archosaur groups.  While it is a potential synapomorphy, the diagnostic
> utility of this feature is doubtful.>>
>At the level of Phytodinosauria, it's very close to being a synapomorphy.
        I would not argue this, save for the fact that everyone else's
analyses regard it as a homoplaisy.

>It's certainly a derived state (primitive would be some kind of insectivorous
        One group's derived character is a more inclusive group's primitive

><< >(3) Heterodontosaurids are not ornithopods [...]
> and there is [not] an obturator process on the ischium. 
>        It is possible that many of common conceptions concerning the
> diagnostic utility of this element are in error. [...] >>
>In Ornithischia, the obturator process developed well along in the group's
>evolution. [...] it is a synapomorphy of Ornithopoda.
        This according to orthodox analysis. I would be interested to see
what happens if the definition of this character is clarified and the coding
of this character is explored (using _Lesothosaurus_ and _Eoraptor_ as
outgroups, perhaps).

> >theropods, the semilunate carpal evidently formed by fusion of the proximal
> >carpals (radiale and intermedium), whereas in segnosaurs it formed by the
>         Hasn't this particular homology been questioned?  I believe that
> Hinchcliffe's (?) work on avian ontogeny has some things to say about
>I'm going by Ostrom's description of the semilunate carpal in _Deinonychus_.
        While I have had my disputes with Hinchcliffe's interpretations, I
believe you should examine this paper before reiterating you argument. I
believe Ostrom's 1969 description predates Hinchcliffe's (1980?) work.

><< >carpal structure differs greatly from that of theropods and cannot be
>         "Always with you what cannot be done...">>
>Like I said before, a remarkable occurrence requiring remarkable proof.
        What sort of proof are you expecting? Do you want God to come down
and draw you dendrograms in the dirt? Cladistic analysis is accepted by many
many workers are an acceptable means of providing "proof", both "remarkable"
and mundane. If you choose not to accpet it, you will never be satisfied, as
it is pretty much all we have.
        Unless you can get on your direct pipeline to the Tree of Life and
request a 500 specimen transition series from a basal coelurosaur to a
therizinosaur, that is...

><< >from theirs. The forelimb and manus of segnosaurs are highly specialized,
> >so the fact that it the manus is tridactyl should not be used as a character
> >uniting segnosaurs with theropods.
>         This is patently ludicrous! Ornithomimids has a far more specialized
> manus, and no one argue the homology of their manus! Again you eliminate
> homology a priori of a phylogenetic analyses.>>
>Sorry, the ornithomimid manus is nothing like as specialized as a segnosaur
>manus. "Far more specialized" is going off the deep end.
        Perhaps, in retrospect, I should retract that statement. Substitute
in: "The ornithomimid manus is highly derived with respect to the
plesiomorphic theropod manus, and yet no one argues the homology of their
manus". In any case, my point holds.

><< >The ascending process on the astragalus differs in detail from that found
> in >theropods and probably developed independently.
>         So your argument then becomes "any feature which is significantly
> derived from the ancestral condition may no longer be considered
> homologous"?  Let's let the data speak for themselves: therizinosaurs and
> other theropods have large astragalar as.p.s.>>
>They certainly do, and in my phylogeny they developed them independently. I
>never said, "Any feature significantly derived from the ancestral condition
>may no longer be considered homologous." That's nonsense.
        Allow me to rephrase: "Any character which is significantly derived
from the ancestral condition, and for which homology would disrupt your
concept of the phylogeny, may not be considered homologous."
        Part of the problem with rejecting "homoplaisies" a priori is that
homologous elements may altered by evolution in such a way as to appear to
no longer be homologous, even though their function is similar. The flipper
of the plesiosaur (to my eye) does not always appear to be homologous to the
tetropod limb.  Yes, it has some bones in the right places and all that, but
it looks like it could have just as easily arisin convergently. Maybe the
bones just happened to arise there for functional reasons... How do I tell?
Simple answer: do a cladistic analysis [or find a series of increasingly
derived "transitional" forms...].

><<>The feet
>         The feet show clear signs of being secondarily quadradactyl (Russell
> and Dong 1993).>>
>Nonsense. The signs, if any, are not clear at all.
        Are Russell and Dong just pulling my leg then?

OBLIGATORY NOTE: I am NOT making fun of Dr.s Russell and Dong. I respect
them both highly, and am indebted to them for their contributions to
science. (Thanks to B. Blackwell for pointing out that one must do this from
time to time.)

><< >jaws
>         The lower jaw of _Erlicosaurus_ is similar to that of _Harpymimus_,
> as P. Buckholz [sic? sorry Pete] has pointed out.>>
>It's much more similar to the lower jaws of >any< prosauropod, as well as
>many sauropods and even stegosaurs.
        But you implied it was not similar to theropod jaws...

><< >teeth
>         These teeth are possibly just as similar to those of _Mononykus_ and
> _Pelecanomimus_ as they are to your "phytodinosauria".>>
>They are nearly identical (no pun intended) in conformation, relative size,
>to the teeth of most prosauropods. Including the interdental plates.
        I believe Pete has ripped this one apart. Although I did not have
time to read his post carefully, it sounded good enough to eat (no pun

>So are segnosaurs dromaeosaurids or birds? Those are the only theropods that
>display retroverted pubes.
        I was simply trying to demonstrate that retroverted pubes does not
automatically = ornithiscian. In this case it was almost certainly related
to diet, and possibly to shortening of the tail(?).

>Of course< retroverted pubes in segnosaurs are
>convergent with theropod retroverted pubes; but they could well be homologous
>with retroverted pubes in Ornithischia.
        Could be, but the rest of the evidence weighs out against this.

>         1.  Manus tridactyl>>
        An a priori assumption, with no supporting evidence.

><<         2.  Manual proportions similar to other ceolurosaurian groups>>
>Convergent. Too vague to be a real character.
        What ornithiscian has such long thin fingers with big claws?

><<         3.  Lip on manual ungual II (Archaey, Oviraptorsauria)>>
>So does this make segnosaurs archaeopterygids or oviraptorosaurs? Convergent.
        Sure, it could be, possibly it is a climbing adaptation. But
considering the evidence that they are most closely related oviraptorsaurs,
I'd say its homologous.

><<         4.  Enlarged preacetabular portion of Ilium (Neotheropoda)>>
>Also noted in all Ornithischia; probably a synapomorphy of segnosaurs and
>ornithischians, convergent with theropods.
        Excuse me? Oh, sorry, bad character transcription: "DORSO-VENTRALLY
enlarged bladedlike preacetabular portion of ilium" as in not that flat
twisted thing like ceratopsians, but big honking plate like T. rex.

><<         5.  Elongate ascending process of astragalus>>
>Just been through this one. Convergent.
        Me too. Your a priori assumption is rejected.

><<         6.  Leaf-shaped teeth (Ornithomimidae, Mononykus)>>
>So--are segnosaur ornithomimimds or mononykosaurs? Also noted in
>ornithischians and sauropodomorphs.
        Monony. may be in a clade w/ oviraptors and therizinosaurs, as might
be ornithomimids. The lack of teeth on oviraptorsaurs certainly does nothing
to discount this.
        In any case, your implicit criticism is valid. I should be listing
characters to support a counter hypothesis, not characters for several
different hypotheses. Touche.  :)

><<         7.  Preacetabular blade of ilium dorsovetrally elongate (ovir,
>Segnosaur apomorphy.
        Not an apomorphy in the sense I take you to mean, because it is
PRESENT in some ovirapotorsaurs and _Microvenator_. I believe you mean to
say "convergent".
        Homologous with oviraptorsaurs and _Microvenator_. Prove me wrong.

><<         8.      "       "       "       with hooked anterior margin
>(ornith, ov)>>
>Again, does this make them ornithomimosaurs or oviraptorosaurs? Convergent.
        Doesn't have to make them either, but may support an association
with either group or both groups.
        You seem to think I'm saying that if therizinosaurs share a
character with some group, they must be in that group. Does the shared
presence of an enlarged parasphenoid capsule make troodontids ornithomimids?
no, but it does suggest that they are more closely related to each other
than to any other known theropods.

><<         9.  Antero-dorsal expansion of lip of acetabulum (ov, microv)>>
        Come now, you've just a priori dismissed NINE characters as
convergences. You certainly must have a ton of evidence on your side. Yet I
have yet to hear anything I cannot counter except "they look like prosauropods".

><<         10. Obturator process triangular (ceolurosauria)>>
>Segnosaur apomorphy.
        Synapomorphic. I believe you again mean to say "convergent".

>The obturator process of segnosaurs is in a different
>location on the ischium.
        Yeah, towards the distal end, JUST like other coelurosaurs.

>Ornithopods have an obturator process; does this
>make coelurosaurs ornithopods or vice versa?
        Characters do not make taxa, and you know it. Yes, ornithopods have
obturator processes. What I am emphasizing is that the *triangular*
obturator process is a coelurosaurian character (which may be considered
convergent with that of some ornithopods).

><<         11. Pubic boot enlarged.>>
>Too vague. Enlarged which way? Anteriorly, posteriorly, both?
        Which way is it in prosauropods and ornithopods? Is it as big as in

><<         12. ?Ventral migration of obturator process (questionable)>>
>I'll say. See above.
        Actually, on later inspection, not questionable. Firmly established
on specimen with complete ischium. See illustration in _The Dinosauria_.

><<         13. Strap-shaped scapula (Theropoda)>>
>Too vague. Depending on what you consider "strap-shaped," sauropods,
>prosauropods, and ornithischians also have strap-shaped scapulae.
       Thin, unexpanded relative to the taxa you mention. I.e. "looks like a
theropod scapula".

><<         14. Tridactyl pes (reversed in Therizinosauroidea)>>
        And you accused ME of shooting down YOU with caustic little curt
answers?        ;)

><<         15. Semilunate carpal with transverse trochlea backs dig I and II
>         etc etc etc...>>
>Convergent; see above. The carpus of _Alxasaurus_ is much different from the
>derived carpus of maniraptorans.
        We have addressed this issue before, I believe you are incorrect.

>What you have here is a list of characters found randomly among theropods and
        Admittedly, yes, I should have developed a more cohesive theory.
Disregard the one reference above which does not relate directly to
Oviraptors, #6, and assume I was arguing therizinosaurs are the closest
relatives of _Microvenator_ and oviraptorsaurs amongst the Dinosauria. Now
what do you say?

>other dinosaurs that could easily have been acquired by convergence during
>the long interval of segnosaur evolution for which there is as yet no fossil
        But for which there is no evidence that they *were* acquired
convergently. Lacking this, you don't even have an explanation of *why* they
would be acquired convergently. What you have in support of your position is
a list of suspiciously convergent-looking characters which can be shown by
cladistic and non-cladistic analysis to be convergent, and (as a bonus) for
which perfectly good functional/ecological explanations of the convergences
could be made. So, which story sounds better established?

>There's >nothing< like a robust character suite here--just a few
>lumps and bumps
        In case you haven't been keeping up on current events, phylogenetic
reconstruction is based on "lumps and bumps", and the sum total of their
affinities. If we waited around for "robust character suites", we could
hardly expect to resolve most of the controversies in paleobiology.
        It was recognized a long time ago that gross similarities of the
sort that prosauropods and therizinosaurs share are not very useful in
reconstructing phylogeny. More recently, it was recognized that homoplaisy
does occur, and may be running rampant through the Dinosauria. Indeed, given
the number of convergences you invoked above, you of all people should
recognize this.
        Given these facts, it seems clear that a testable, minimally
subjective analysis which is minimally encumbered by a priori assumptions,
is most likely to produce the best hypothesis of the relationships in
question. THis means looking at "lumps and bumps". As my mom used to say,
"if you don't like it, lump it."  :)

><< >and limbs of segnosaurs all derive much more readily from
>prosauropod-like sauropodomorphs than from theropods
>         And snakes derive more readily from eels, but that doesn't make it
> right.>>
>Aw come on. Take almost any part of the segnosaur skeleton, from skull to
>tail, and compare it with the corresponding region in prosauropods. The
>similarities are too many to be lightly dismissed.
        And they have not been lightly dismissed. The theories of
knowledgeable and experienced people such as yourself, Gregory Paul, and
other dinosaur workers were considered heavily. However, recently, new data
have caused them to be dissmissed, not lightly, but with a heavy heart and a
heartfelt "well done" for doing the best that could be done with the
material then available.
        If you choose to cling to an older theory, this is you right, but it
does your arguments little service when you make statements like the one to
which I responded(in an admittedly disrespectful fashion, for which I

><< >(5) _Mononykus_ is not a bird, nor is it an alvarezsaurid. It's a
>         In the absence of a phylogenetic hypothesis to the contrary, we are
> left with this conclusion. Care to clade a better one?>>
>Since I don't accept cladistic analysis as anything more than simply one more
>way of constructing phylogenies, I don't see what good this would do.
        Since you have no means of demonstrating your argument without
constructing a testable phylogeny of your own, I suggest you either learn
cladistics, or develop an alternative which is just as testable, just as
relatively objective, and just as reproducable. Then develop an alternative
hypothesis. Then make statements like your original one above. Then we can
all compare the two, and accept or reject to our heart's content.

> <<I am interested to know what autopomorphies _Avimimus_
> and _Mononykus_ might share, which the latter taxon does not share with
> birds or arctomets. >>
>Very slender, distally tapering fibula, closely appressed to the tibia, for
        I was under the impression that this is pretty much halfway between
the primitive maniraptorforman condition and the derived bird condition. Am
I in error?

>one. _Mononykus_ can't be a bird because, like _Avimimus_, it doesn't have an
>avian metatarsus; it has an arctometatarsalian metatarsus.
        As do the non-arctomet elmisaurids. Indeed, there seems to be a
tendancy in all coelurosaurs towads a compressed MT III, and in at least two
groups, towards a more slender metatarsus.

>And a well developed tail with elongate chevrons.
        Quel horror, a theropod with a tail! As I recall, archaey has
chevrons too, does he not? Elongate chevrons may be apomorphic/reversed,
well developed tail ditto. Recall that it is only *shared*derived*
characters which elucidate phylogeny.

>So the slender fibula becomes a potential synapomorphy of _Mononykus_ and
>_Avimimus_. There may be others, particularly among the femoral
trochanters, >but I haven't finished looking at the literature.
        I'm still not sure about alvarezsaurids, and I haven't seen the
recent work. However, several workers I respect seem to believe they are
(all) birds, and no one has come up with an alternate explanation which I
can completely believe. As for _Avimimus_, well, I still like Bucholtz's
[sic?] idea, myself...

        Nice to talk to you again, George.  :)
    Jonathan R. Wagner, Dept. of Geosciences, TTU, Lubbock, TX 79409-1053
               "Not the One..." -- Zathras (not Zathras)