Neutral Poses & Elmerfuddasaurus

Here's a quick look at another kind of pose: the neutral pose, which is more or less the equivalent of the standard anatomical poses used for extant animals.  I'm actually quite fond of the unassuming aesthetic of this type of pose, but alas I suspect it isn't a good candidate for a pose to standardize on.  To see why let's take a quick look at the strengths and weaknesses of the approach:

Strengths:

Better shows off the center of gravity (full-tilt running poses always look unstable...because of course they are).  A reduced burden for the author/illustrator to get a biomechanically plausible gait (although there would still be the need to get a realistic stance).  A pose like this is possibly better for some artists (e.g. 3D artists).  Certainly there's no distraction created by the pose.

Weaknesses:

 Some people might find this pose dull, although scientifically speaking that's not much of a criticism.  I suppose when used for education or popular media that could be a drawback, although arguably there's good reason for people to see dinosaurs portrayed as animals, rather than stylized carnage-machines, so the argument could be made both ways.

The biggest drawback to a pose like this is it shows off less of the anatomy, since the limbs from the far side are obscured.  It's easy-enough to fix that, by arbitrarily moving the limbs like this:

You can even show off some other pose (in this case the finders are lightly flexed), perhaps the other wrist could also demonstrate the degree of pronation that is possible (in the case of Archaeoceratops it more or less already is), which would then pack in a bit more information.  Of course if you move the other hind leg then you are pretty much back to putting the animal in a walk.

Of course you don't have to put the feet on the ground - if we're looking for poses that are more neutral about locomotion we could instead illustrate as if the animal were lying on a virtual dissection table:

Unfortunately, even if this were explicitly labeled, I suspect someone somewhere would take it as intending to show off biomechanical behavior.  Jumping.  Or dancing.  Or maybe sneaking along, saying "Be vewy vewy quiet, I'm hunting wabbits.":

So it seems to me that you probably can't have it both ways - to get the benefits of the neutral pose we'd have to sacrifice showing off as much morphological data.

Conclusions:

Neutral poses have a couple of advantages - they may benefit some types of artists, they reduce the amount of biomechanical inference that is required, and their lack of "visual excess" means they won't distract from the anatomy.  Unfortunately they also obscure more of the anatomy, and trying to correct this by moving the limbs around quickly sacrifices the very things that were an advantage to the neutral pose.

For this reason I think neutral poses may have limited appeal - and if that's the case, it probably isn't an ideal candidate for a new standard.

Skeletal Poses: Do they matter?

Ok, first stop chortling.  Then take a good look at the handstand allosaur up there.  In several respects it's scientifically accurate - the bone outlines reflect the actual morphology of the fossils, and the proportions are correct, so it's a "realistic" skeletal reconstruction.  The pose is certainly unusual, but none of the joints are disarticulated.  In these respects it's better than many of the skeletals that appear in peer-reviewed journals.  Yet I think it's safe to say that most researchers would consider that allosaur to be in a biologically implausible position.

Do skeletal poses matter?

 Is this pose just as good as any other, or are in fact some choices more useful?  After the break I'll try to make the case that choosing a pose is an important part of making a skeletal reconstruction, rather than a random after-thought.

I shouldn't have to say this, but just to be clear: I don't think Allosaurus could do a handstand.  Even attempting it would probably lead to a dramatic reduction in life expectancy. Yet if all a skeletal reconstruction is supposed to do is to show off the bones, then the only real complaint in the image above is that the left leg obscures the pelvis more than necessary.

So why not use this pose?  Certainly it would be easy to build up a "brand" around such a pose.  Yet I'd submit to you that skeletal reconstructions with inaccurate biomechanics undercut the value of a skeletal by virtue of the added theoretical "baggage". Mike Habib, clever gentleman that he is, anticipated this point in his comment on the previous article, which I'll quote below:

"...it is distracting from the point of the reconstruction if the viewer spends time trying to work out if the pose is realistic. Ideally, a "standard" pose should be a 'no-brainer' for most taxa, so that viewers can focus on, you know, the *skeleton*."

In addition to distraction, poses that are not feasible (or even just unlikely) create other problems; some authors will avoid such skeletals (perhaps even choosing a reconstruction that is otherwise less accurate).  There will inevitably be well-intentioned artists that introduce incorrect poses into their work.  And of course other scientific illustrators may be scared off of using the same pose, making comparisons between bodies of work more difficult.

If we only dealt with ludicrous poses, this may seem like a straw man argument.  So let's consider a less overt example:

That's Silesaurus, from the original description in JVP.  The shapes of the bones generally reflect the individual elements described in the manuscript, and the proportions are quite good; clearly it's intended as a realistic skeletal reconstruction. The pose is certainly not wrong in some over-the-top manner, yet there are several problems with it.  Some differences are due to different interpretations of rib orientation and pectoral girdle positioning (but that's another post...), while others are not so easily categorized.

The vertebral column in general is problematic; the flex in the base of the neck and the overly-straight back are positions that may be possible, but would not be terribly common for the animal.  The forearms are pronated to a degree that is unlikely in such a  primitive dinosauromorph.  Even more clear-cut is the position of the right forelimb.  The right humerus (the upper arm bone) is so far forward it would be completely dislocated from the shoulder socket.  Moreover, given the position of the visible part of the humerus the proximal part would be articulating with the center of the coracoid, rather than the glenoid fossa (the shoulder joint).

If the only thing you care about is the bones, then I admit that how distracting these issues are depends on how closely you pay attention to biomechanics.  But the pose isn't without repercussions; a quick image search shows that several derivative skeletal drawings have been produced that perpetuate the same errors, and a decent number of life reconstructions also exhibit those errors.

(Image from here; artist unattributed.)

To some degree this is where we get to the crux of disagreements - people are often quick to criticize as outlandish the problems that appear at the macroscopic level (Allosaurus can't do a handstand!) while ignoring the problems that are less obvious, or at least the ones that fall out of their area of expertise.  As a result I'd be willing to bet cold hard cash that the handstand allosaur at the top would not make it past the same reviewers that gave a pass to the Silesaurus paper, even though the skeletal in that paper is a less biologically plausible pose than the allosaur.

If people really want to present just the bones, and not make any statement about functional anatomy at all, perhaps researchers should consider exploded diagrams:

Exploded diagrams have a proud tradition in technical illustration, and can be done without making any statement what so ever on functional morphology.  I should note that the above diagram is a butchered version of my Styracosaurus skeletal; in a diagram prepared from the start to be an exploded diagram I would expect the limb bones and possibly even the vertebrae to not be connected as in life.  Providing all of the bones scaled (and revealing only the preserved portions) would accomplish the purely descriptive goals of a traditional skeletal (perhaps even be superior, since nothing is hidden by the limbs) and completely relieves authors/illustrators from making explicit claims about how the animal went together.

So in conclusion, the point I want to make is this:

People do not have to put realistic skeletal poses in their papers.  They can use schematic diagrams (which partially relieves the burden) or use exploded diagrams (which completely removes it).  The exploded diagram in particular conveys more morphological evidence then a traditional skeletal drawing, while being 100% agnostic about biomechanics.

If authors/illustrators do choose to do a realistic skeletal reconstruction, then they should accept the need to place them in biomechanically sound poses.  Inaccurate poses can distract from the other purposes of a skeletal diagram, and may mislead paleoartists.  Down the line if such diagrams get incorporated into educational diagrams they also play a role in confusing students and consumers of popular scientific media...but that, two, is another post.

In the mean time, remember: Poses are important!

The Great Skeletal Repose of 2011

As many of you are no doubt aware, earlier this year paleontologist and scientific illustrator Greg Paul made a fairly public hubbub when (among other demands) he requested that all other illustrators stop using the skeletal poses he popularized the last several decades.  There was quite a bit of consternation over the issues he raised, filled with both sympathy (it's hard to make a living from paleoart) and skepticism (most people don't believe Greg has any legal basis to try and lay claim to an anatomical pose - I suspect those people are correct).

None the less, on March 8th of this year I wrote:

Allowing Greg (Paul) to establish a branding around the poses he popularized is a request I'm inclined to grant; after corresponding briefly with Greg I've decided to embark on the process of reposing my 100+ skeletal reconstructions.

That lead to a lot of questions.  I will be examining in greater detail skeletal poses and how we can make them as useful as possible, but first I wanted to address some of the common questions that came up from this.  Namely...

What was I thinking???

This is the main question I get.  While it's been phrased several different ways, the crux of it is some people are concerned whether the (substantial) time investment in changing the pose in all of my skeletals is worth it.  Of which the most substantive question is:

Will they will be less useful in another pose?

Several workers wrote to me with this concern - that by altering the pose it would make my skeletals less useful, since they would be more difficult to compare directly with Greg's.  I am a strong supporter of standards in science, so I'm sympathetic to this claim.  That said, due to the aforementioned hubbub the utility of those poses as a standard is rapidly eroding as several artists are now altering their poses, or actively advocating for everyone to use their own unique pose.  Since artists are frequently somewhat conflict-averse, I expect this exodus to continue, regardless of legal standing.

Given this larger perspective, I feel that we'd be better served to find a new pose to standardize on, perhaps one that can still be compared effectively with Greg's body of work.  An open standard by design, so that other researchers/illustrators can feel free to adopt it without fear upsetting someone else who uses it.  And by starting again we have an opportunity to "reboot" the standard skeletal pose, perhaps producing something even more useful then the original.

One obvious example: the "Greg Paul" pose for theropods, though iconic, is held back by its theoretical baggage.  The pushing off the left foot while dashing around at a full sprint pose is not something that all researchers agree is possible in all theropods.  Several times while providing a skeletal for another researchers publication I've been asked to alter the pose for this reason.  I've done this a couple of times due to my own incredulity; for example I illustrated Majungasaurus in a walking pose, since I'm skeptical that it could sprint:

So by undertaking this project we can take advantage of hindsight to create a standard that is both open and potentially solves some of the largest criticisms of Greg Paul's poses.

So what, you're just going to pick the new "standard"?

I do need to pick a new pose.  Or rather several (for various groups of dinosaurs).  But it won't be much of a standard if I am the only one using it.  Instead, I'm hoping to crowd-source this discussion, involving any individuals who have a stake and wish to participate.  To that effect I'm working on a series of articles on such subjects as: Do skeletal poses even matter?  And if they do, what is the best way to go about creating a pose?  Who are we serving with these poses?  And how can we balance the sometimes conflicting needs of the "consumers" of skeletal reconstructions?

I'm actively communicating with some people, and hope to engage others to pick up the torch.  I hope to get a wide range of responses on the blog, and perhaps to inspire others to create articles on the subject.  The best results can only be achieved if we get generate a robust conversation on the subject.  I hope you'll participate!

By the way, if for some reason you'd rather share an opinion privately, feel free to email me and I can incorporate your concerns into a future discussions anonymously.

Schematic vs realistic skeletals: Follow up

I'm happy to say I've received some excellent feedback on the previous blog post on schematic skeletal diagrams.  Several comments in particular share a similar feeling, which I'll repost here:

I don't see any reason someone would make a reconstruction more schematic than necessary (due to perspective, converting bones to lines, incompleteness or deformation, lack of available material, etc. that you mention). In the strat column and cell examples, there are obvious reasons not to make them realistic, but what about skeletons? Surely the only reason to not draw bones correctly is to save time, but in that case I'd argue it's better to not to include a reconstruction than to make a half-assed one. -Mickey Mortimer

This is a very.. charitable interpretation of what's going on. I agree that there is a place for schematics, but I think they should be made to look schematic (you see this sometimes, where bones are reduced to oblongs, and laid out in a vary schematic way). Just labelling skeletals as schematic will do little to stop artists using them I'm afraid.  -John Conway

I would have to side with John on the matter of schematic representations. You write "When writing a professional paper, which one of these styles is "better" depends on the needs of the authors, the time, ability, and access to the data that the illustrator has, and a host of other practical concerns. Far be it from any of us to dictate that one type of skeletal diagram is suitable in all cases." But I can't see any way in which is schematic diagram is better than a realistic one. In short, surely the only reason to put up with the schematic is when the data just isn't there to do the job right? -Mike Taylor

As you can see, they all raise a similar question - why would anyone want to publish a a skeletal diagram that isn't realistic (or in some cases, "correct").  I suspect that they aren't the only ones with this question, so I thought it would be worth addressing the issue with its own post.  So let's start at the beginning....

Historical Perspective: 

As I showed in the earlier 3 part series on dinosaur skeletals, the reality is in the history of paleontology it simply has never been a standard requirement to invest the time and effort that goes into producing realistic skeletal reconstructions for publication.  Yes, for a period of time realistic skeletals were used by some paleontologists in Europe in the middle of the 19th century, but hardly all.  Concerns with improving the anatomical posture of mounts saw a brief return to publications in the U.S. in the 1920s and '30s, but neither of these periods saw anything close to a universal adoption of realistic skeletal reconstructions, nor did either period produce published guidelines on how to produce such skeletals - and don't forget that they make good examples by virtue of how unusual they are in the history of paleontology.  

So point 1: It may be be true in some objective sense that realistic skeletal drawings are preferable, but it's never been a standard in scientific publications.  Sure, several decades of skeletals by Greg Paul and others may have created an expectation in artists that realism should be the default, but that hasn't translated to professional publications.  In my opinion it's neither fair nor realistic to expect all researchers to start including realistic skeletals (especially given the issues discussed below) in published papers cold-turkey, but it is fair to ask them to label their diagrams more explicitly (indeed, better labeling is something we should always strive for).  Why isn't it fair?  Well, there's a host of...

Practical Concerns:


Writing a paper takes a while.  While the process isn't really the mysterious and inaccessible dark art that some assume, it does take time and effort.  And unfortunately producing realistic skeletal drawings largely has been a mysterious dark art, without explicit guidelines, and with only a couple of people that produce them (and there isn't any universal consensus on who those people are).  The result is that a paleontologist that wants to get a paper out on a new dinosaur could be looking at a really significant investment of time (and possibly money) to try and include a realistic skeletal reconstruction.

Also remember that many researchers don't have research specialties that lend themselves to supervising the creation of a realistic skeletal reconstruction; even at the best of times it can be hard for technical and artistic professionals to find a common language, but for an expert in stratigraphy or systematics it may be even more difficult to direct a staff artist or art student on how to produce a realistic skeletal drawing.  This wouldn't be so bad if scientific illustrators had a set of guidelines they could follow when producing realistic skeletal reconstructions, but hey, that brings me to the last point....

Skeletal reconstructions need to stop being a dark art!

Let's all have a moment of honesty here; how is a young scientific illustrator supposed to go about learning to produce a realistic skeletal reconstruction?  Greg Paul has written a single, traditionally hard to attain article on the subject, and has written several guides to how he interprets common debates on dinosaur functional morph.  That's about it; the result was that many young artists took the "make it look like Greg Paul's" strategy, but there are several problems with this.  First, it's not always clear to illustrators when something is a well-established anatomical consensus, or whether it's an interpretation largely unique to Greg.  Making this worse is that Greg has (somewhat notoriously) issued a mass cease and desist request for people to stop copying his look.  Since there aren't many sources to tease out which parts are science (and therefore not copyrightable) and which are stylistic (and therefor subject to his copyright), many artists are probably feeling like throwing their arms up in surrender.

It's my hope that through several projects (including a modest contribution from this blog) that the science behind this process can be better documented, become better subject to testing, and generally move from the realm of dark art to the evidence-driven process it should be.  But I also think we need to be realistic about what stage we're at.  Right now there aren't even a handful of people who regularly publish on the subject, and the publications that do exist often are often made in the gray literature and are not subject to further testing.

So what can we do?


I think first and foremost we have to be realistic about the size of the challenge in front of us.  Given the name of this site it shouldn't surprise the reader to learn that I think skeletal reconstructions can (and generally should) be done to realistic standards.  I also believe the creation of them should be a data-driven activity, with a methodology that is transparent and subject to testing by others.

But a lot of work has to be done before we get there.  In the mean time, properly labeling a published skeletal as either schematic or realistic will be useful to the process; by making it clear when a skeletal isn't intended as realistic it will increase the accuracy of paleoart (since artists won't use it), make it more obvious which taxa are still in need of a realistic skeletal, and improve the "signal to noise ratio" when people try to understand what are common anatomical assumptions.

Proper labeling is also easy to do, making it a reasonable request of anyone getting ready to submit a manuscript (or reviewing them), as it requires a minimal investment of time and improves the usefulness of the paper itself.  Yes, this may be the "low-hanging fruit" in a larger revamp of skeletal reconstructions, but that doesn't mean it isn't worth pursuing.