david's really interesting pages | Just another WordPress site

Skeletal Reconstructions; discussion

Scott Hartman has launched a valuable discussion about skeletal reconstructions, to which this post is more or less an illustrated comment. So – if you haven’t already – you should check out his blog, and particularly the last posts about skeletal poses, reposes and style (parts one and two).


First off, there’s the discussion of what might be a standardized pose to replace the one claimed by Greg Paul. Above I’ve taken Scott’s Majungasaurus  and fit him to the implausible running pose. (Sorry Scott.) Everyone agrees that that’s not good. But if other theropods are shown in the extreme locomotion gait while  Majungasaurus alone is left to walk, that is information. namely, it communicates that Majungasaurus couldn’t run, ie. the walk gait was its extreme. I won’t get hung up on this, but assuming the goal of a skeletal reconstruction is to communicate maximum information, establishing an across-the-board walk pose seems to weaken this objective.

What IS the goal?

Scott’s posts are already leading up to this conclusion, but I’ll try to state it plainly. Skeletals should fulfill their function, and as there are numerous functions for them, there should be numerous ways of doing them. The exploded view seems ideal for those presenting an analysis of the bones themselves, while a posed reconstruction would better suit someone analysing biomechanical features. Problems occur when the presentation seemingly lays claim to information that isn’t in the model, so its great to see Scott point to this so directly. Really good stuff. It feels like those 3D skeletals on low-budget TV shows… they look like bones that have just been dug up and prettied. But they’re not at all like the bones that specialists know they should. Result: disconnection.

What is my goal?

I’ve been exploring the creation of just such 3D skeletals and have a couple of observations to contribute from the point of view of a 3D animator not all too far removed (approximately 1 1/2 years) from those clueless artists from the above-mentioned TV shows that you scientists, justifiably, criticize.

I want to create interactive, 3D skeletal/life reconstructions that are interlinked to biota profiles and phylogenetic timelines so that important concepts like deep time, evolutionary relationships and ecosystems are communicated passively, like a background tapestry while people satiate their fascination for Trex. Or Kentrosaurus.

Eye candy: here’s an Archeopteryx reconstruction as a walk-cycle animation consulted by Scott himself. It blends from stylized life-reconstruction to stylized skeletal. Or more exactly, an abstracted skeletal. I’m not trying to make something that looks like bone or stone, but rather something that conveys those features that makes these bones identifiable as belonging to Archeopteryx. I know that many viewers will find it odd, but I like the fact that it looks sculptural and not at all like bone. The look of bone is easy to achieve, and uninformed viewers immediately assume it to be bone, which lends it a superficial authority. Ah, that’s a bone! It must be the bone.

In this specific context of continuously progressing scientific results, it must be a good thing to communicate that this creature must have been so – not that it was so.

The thing is, I look at the scraps and shards that capable scientists puzzle together into what most people think is directly dug up out of the ground and I recognize next to nothing. I look at interspecies diversity and am grateful that I’m not the one determining where the line between species is to be drawn. The more I learn the more I understand how much interpretation is involved in each step along the way, from lump in the ground to a bone reconstruction, a skeletal reconstruction, a life reconstruction.

Instead of being put off by this, I find myself fascinated by it, and I want other people to be fascinated by it too. That requires an understanding of the scientific process… ie. an awareness that there are many interpretive steps involved in that visual.


In the Archeopteryx animation, the spine is a volume study without individual vertebrae. In this sketch (blatantly copying – once again – Scott’s lead) I’ve depicted Silesaurus as a sketch. It’s interesting to see how reactions vary and, unfortunately, I fear that it would be taken less seriously. The scientist interested in exact numeric counts of caudal vertebrae will be disappointed, of course, but to the interested layman, it otherwise conveys a large portion of the same information correctly. Just – it looks unfinished. It surrenders authority for the author’s signature, together with the implied fallibility that goes along with having been authored. Fallibility is an assumption that leads to independent confirmation of results… a key part of the scientific process?

Scott’s discussion has prompted me to more precisely wonder what style would best suit a skeletal reconstruction for popular, interactive publications.

There are 3 Comments to this article

Scott Hartman says:

Excellent discussion David. And I agree that educational needs are sometimes different, especially in 3d (and especially since 3d affords more possibilities to abstract or otherwise convey data).

This isn’t just a flaw of interactive dinosaur content – the standard models of atoms and molecules (i.e. the giant colored spheres that overlap) stem from the shorthand way the chemists illustrate complex chemistry (especially organic chemists), yet in many ways that oversimplification is actually more misleading to the average consumer of educational content, who doesn’t really know much about atomic structure, nor molecular bonding. The same could be said for the overly generalized illustrations of enzymes and their action (i.e. enzymes are not single shapes, and the “lock and key” fit is not always so much morphological as they are a match between intermolecular forces).

That said, I think that people who produce skeletal reconstructions that take maximal amount of the data available (which of course requires that we start doing a better job of publishing and peer-reviewing that data!) can make realistic 2d skeletals that aren’t misleading at all and contain a lot of data.

Scott Hartman says:

Oooh, to clarify my statement that “enzymes are not single shapes” I mean that showing them as a solid object, rather than that they are a molecule folded up on itself makes it harder to explain say conformation shifts, which means people end up just taking the narrator’s work for the fact that changes in temperature or ph alter an enzymes efficiency, whereas an object made up of a folded chain makes it easy to understand how it can move or change shape under different environmental circumstances.

d maas says:

Nice metaphor. When you say “people end up just taking the narrator’s work for the fact” I think of my current peeve: everyone (general audience) ‘knows’ that birds are dinosaurs, yet not one has been able to explain how we know that this is so, other than referring to the authority of ‘scientists’.

Archeopteryx to the rescue!

Write a Comment