[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]
Fwd: Re: Dinomorph Strikes Back!... or does it?
I'm forwarding Kent Stevens' reply to Scott Hartman's post of yesterday. DV:
I appreciate Scott's comments very much. The following is protracted
(= long-winded) and of interest perhaps to only those who are worried
about sauropod neck intrinsic curvature.
First of all, I want to say that I'd be delighted if we find a
sauropod that in life had an upturn at the base of the neck. It
would be emotionally uplifting, if not merely a pun, to find that not
ALL sauropods went around looking like Eeyore (from Winnie the Pooh),
all dejected and droopy.
So please find wonderfully uncrushed, undistorted, sauropod
cervicodorsal vertebrae, put those vertebrae into articulation (and I
mean CLOSE articulation) with tight articulation at the centra and
zygapophyses superimposed, and let's see what the resultant curvature
of the neck is. I'm glad to see the Wyoming Dinosaur Center is
getting the bones closer to articulation, but the vertebrae Scott
shows us are still way out of articulation at the centra, and one has
to do some additional visualization to try to determine the
orientation that would arise between successive vertebrae if they
were as closely separated as they would have been in life. (Also,
while this specimen is gorgeous, there is some obvious post-
depositional distortion in the cervicals in question that prevents
the pre- and post-zygaophyses from aligning as they would have in
life.) How close is close between condyle and cotyle? Look at the
intervertebral separation of articulated sauropod cervical series as
found in situ (like on the wall at DNM) and there is but a centimeter
or less between condyle and cotyle, as would be measured deep within
the cotyle... and careful, that tight spacing within the ball-and-
socket joint is not apparent when one looks casually at the gap
between the lateral margins of the cotyle and condyle, for obviously
there must be a wide spacing so that the more cranial vertebra can
articulate by rotating over the surface of the associated condyle of
its caudal neighbor. In modern vertebrae with opisthocoely, the cup-
shaped socket rides closely upon the ball-shaped condyle with little
Here are some comments on Scott's contribution to the DML:
> This may all be part of the "Dinomorph Project", but let's not
> confuse this with computer reconstructions done using the Dinomorph
> This is Photoshop manipulation of illustrations; frankly, it's no
> less "artistic" than the other reconstructions that are lamented on
> the same page.
I respectfully disagree, and it is not merely a matter of personal
taste regarding what is "art" and what is not (or what the definition
of "is" is :).
Composting of images is NOT "artistic". Granted it's something that
can be done with paper and scissors, like in a grammar school art
class, but it's NOT art.
In creating the digital composites, I used scans of the original
illustrations, at MOST modifying the scale of the images
(isotropically, to preserve proportions of the lateral views) so they
were iso-scale. Then I placed them in articulation with neutral
position simultaneously at centra and zygapophyses. I introduce none
of the subtle manipulations of the shape of the centra as one can
find in some side view silhouette illustrations. Please look at the
silhouette illustrations for Brachiosaurus (on my website under
Brachiosaurus, and also available in PDF form elsewhere on the
site). Check for yourself the induced wedge shapes in some of the
illustrations. To quote Dave Barry, "I'm not making this up".
> Photoshop composites depend strongly on the accuracy of the
> illustration (sort of hard to quantify off hand...), and any errors
> by the original artists will be faithfully passed on in the
> Photoshop composites.
That is correct. In fact, some original artwork, such as in Hatcher,
are not consistently scaled (i.e. not actually 1/10 even though so
labeled, as is apparent if one cross-checks with the actual
dimensions), and that can get misled when trying to composite two
vertebrae side views that are of different scales. Greg Paul got
some snickers from the audience at the Norman SVP showing some
pictures of diplodocids seemingly hunting for truffles. The problem
was he did not account for scale inconsistencies within the
illustrations, as I showed in my talk (which followed immediately
after) based on the same original illustrations (I presume, at
least: I used Gilmore and Hatcher's images). You view the slides
from the SVP talk by going to the "selected presentations" pulldown
menu on my site, or by looking at the illustrations Mike Parrish and
I published in the two recent sauropod volumes.
When one is careful to correct any scale differences, the Apatosaurus
CM 3018 presacral series comes out beautifully arched (see my
Apatosaurus page plus the sauropod book chapters), and Diplodocus CM
84 comes out similarly, but something funny happens at mid-neck (a
truffle-hunting specialization? no, probably post-depositional
distortion as Mike Parrish and I confirmed at Jim Madsen's studio by
hand manipulating into neutral articulation the plaster casts of
those CM 84 vertebrae!). Every vertebral series I have put together
by photo-compositing) comes out with a smooth, straight cervicodorsal
transition when articulated in neutral position, even Cetiosaurus
(thanks to illustrations from John Martin which I have cross-checked
with the original specimen in the UK) and now Mamenchisaurus).
But I hasten to agree with Scott re: "any errors by the original
artists will be faithfully passed on in the Photoshop composites".
I've therefore been careful to cross-check those illustrations with
the photographs, when available. And in lateral view, the steel
engravings generally match up well with the photographs. By the way,
the posterior and anterior views one finds in these old monographs
usually suffer considerable perspective distortion, for the camera
was close to the subject and the nearer end appears relatively too
large compared to the farther end, an issue that does not present
itself overly when considering the lateral view.
But compare the monograph illustrations (which are generally quite
accurate geometrically) with some of the whole-skeleton silhouette
drawings where subtle distortions can indeed create the appearance of
neck curvature in neutral pose, such as in the artistic renditions of
Brachiosaurus I have analyzed quantitatively (look at the table in
the presentation slides from the Bozeman SVP talk).
> Another, more intractable problem is that the search for "wedge
> shaped" centra in procoelous vertebrae cannot be done simply by
> looking at lateral views of the centra.
You probably mean opisthocoelous, not procoelous, but anyway: Of
course not, and I don't trivialize the notion of "wedge shape" by
only looking at centra, and moreover, certainly not to merely looking
at their posterior (cotyle) margins. Again, look at our book chapter
illustrations, and better yet, please read what Mike Parrish and I
carefully wrote on the subject of determining neutral pose
curvature. Those chapters were subject to careful peer review, and
no small amount of a priori skepticism on the part of some of the
reviewers, something that only helped the text. Oh, and again, if we
got something wrong in our methods or conclusions, let's fix it; we
are not trying to prevail, just trying to get it straight (sorry
about the pun).
> Mechanically speaking, the "wedge" or "keystone" shape that causes
> an upward (or downward) arch arises from the articulation of the
> centra. Since sauropod cervical articulations are convex/concave,
> this cannot be ascertained simply by looking at the posterior rim
> of the centra.
And again, I don't just look at the posterior rim, as I just
mentioned. In doing this photo-compositing, I place multiple
vertebrae into articulation (with condyle inserted within associated
cotyle). Look at the lateral view of Diplodocus on that page at my
site, or look at the illustrations in our two sauropod book chapters.
It takes considerable effort and persistence to communicate some
concepts, and the notion of vertebral articulation and neutral
deflection seems to be one that requires practical experience, and
not just reading words. Take a real turkey neck, prep it down to the
bones, and play with the vertebrae. If the fact that the centra are
heterocoelous is distracting, get some giraffe vertebrae and go at
it! Giraffe necks go up because they are wedge shaped. Really.
For the AMNH exhibition "Dinosaurs: Ancient Fossils, New
Discoveries" (see AMNH under "products" menu on the DinoMorph site)
the AMNH very kindly CT-scanned an entire giraffe neck and I put
those vertebrae into articulation both in DinoMorph (watch the movie
showing them in articulation) and, in our book chapter in the Currie
Rogers and Wilson volume look carefully at the lateral views).
> The Dinomorph Camarasaurus page credits the Wyoming Dinosaur Center
> as having mounted a Camarasaurus with a neck in the horizontal
> posture "very much in agreement with the osteological evidence",
> the skeleton was in fact mounted that way because of the media
> exposure the Dinomorph project had received. To do so resulted in a
> gap between the zygopopheses of the anterior dorsals that you could
> shove a man's thigh through, see here:
> In light of the unacceptability of this, we recently remounted it
> in a closer approximation of articulation. This resulted in an
> upright neck, albeit not one as upright as Greg Paul used to
> restore. A photomanipulation of the actual bones that was used to
> figure out how we had to changes the armature can bee seen here:
> Unfortunately we were constrained to modifying the existing
> armature, so the bones are perfectly articulated.
I think you mean "are NOT perfectly articulated", but continuing ...
> That said, it is much closer now than it was in the neck-straight-
> out position. We are in the process of scanning all the bones into
> 3D computer space, but manipulation of the bones in a sand box
> confirmed that there was no way to keep the anterior dorsals in
> articulation and have the neck coming straight out at the shoulders.
Sigh... I'd suggest looking at other specimens of Camarasaurus, and
visit the AMNH big bone room (AMNH specimen 5761 has two vertebrae at
the cervicodorsal transition that are actually fused into a straight
line, pretty clear indication of how they were held in life). Then
again, just a click or two away on the internet, just look at the
mount "Annabelle" (I provide a link on my Camarasaurus page), as well
as pictures of the CM 11069 specimen which I provide, courtesy Mike
Parrish, along with others specimens of this sauropod. But maybe the
WDC specimen is indeed different than all the others, and maybe it
really does indicate that this specimen, at least, bent up at the
shoulders when it was alive. That would be exciting, albeit a bit at
odds with the other fossil evidence for this taxon.
> Kent is doing some exciting work, but manipulating drawings in
> Photoshop has its own share of error sources. It's worth pointing
> out that it isn't simply "artists" who have argued for an upright
> neck posture in some sauropods. Per Christiansen, among others, has
> published biomechanics arguments that support this interpretation.
There are indeed some physical arguments for a vertical neck is over
a horizontal neck (basically similar to why it is easier to build a
structure that stacks vertically rather than one that is
cantilevered). I agree with the argument. The only problem that
seems to rear its head (pun again, sorry) is that the neck bones
don't permit the desired upright neck postures. The devil is in the
> Don't get me wrong, the jury is still out
Remembering the jury's verdict on the O.J. Simpson trial and the
Michael Jackson trial, do you really want to use a jury as a basis
for determining the truth?
> on how many (if any) sauropods held their necks upright, and if so,
> how high. But Photoshop manipulation of drawings is no stronger a
> form of evidence than the skeletal reconstructions that proceeded
> the technique. What we really need are good 3D data sets from well
> preserved and complete specimens.
I respectfully disagree. Painstaking Photoshop(tm) compositing of
high-resolution scans of scale-calibrated monograph illustrations is
an independently verifiable, and hence more scientifically reliable,
form of evidence than an artist's simplified, and often stylized,
silhouette drawing. I rest my case.