Sauropod Vertebra Picture of the Week

Photo copyright Derek Bromhall, borrowed from ARKive.

Let’s say you want to paint an elephant. Where will you locate your elephant, and what will it be doing?

If you depict an elephant standing on a glacier at 14,000 feet, your depiction is accurate, because elephants have been caught doing that. Elephant, standing in a dunescape with no water or vegation in sight: accurate, for the same reason. Elephant, swimming in the ocean out of sight of land: accurate. Elephant, scraping salt out of the wall of a cave: accurate. Elephant, rearing to pull down otherwise unreachable vegetation: accurate. Elephant beating the hell out of a monitor lizard for no apparent reason: accurate. Depictions of elephants doing these things might not be familiar–at least to those of us who don’t live around elephants and therefore don’t get to see them doing all the wacky stuff that real animals do–but they are all accurate, in that elephants actually do these things. A lot, apparently, given that all of the above behaviors were documented in the space of just a few decades. Who knows what you might see if you could watch all the elephants, all the time, for a million years or so.

Is there any reason to think that extinct animals were any less versatile?

On the other hand, just because elephants occasionally go for strolls on glaciers or voluntarily rear up on their hind legs to reach higher does not mean that glaciers are their usual habitat or that rearing is a big part of their behavioral repertoire. So these things are accurate, in that they do happen, unfamiliar, in that they are not widely known by most laypeople*, and unusual, in that they are in the long tail of elephant behavior.

* Before you flood the comment section with, “I knew that about elephants!”, consider the implicit possibility that you are not most laypeople. Does your grandmother know that elephants do all this weird stuff?

So we’ve got three potentially orthogonal axes: accuracy, familiarity, usualness. If this was xkcd, at this point I’d draw a Venn diagram. But it’s not and I’m lazy, so I’m just going to pick three possibilities that illustrate an ascending scale of weirdness. First, the most vanilla (by behavioral weirdness, not artistic achievement) wildlife art depicts animals doing things that they actually do (accurate), frequently (usual), that are known to most people (familiar): giraffes eating out of trees, lions with bloody faces crowded around a dead zebra. Second, art that depicts animals doing things that they actually do (accurate), frequently (usual), that are not known to most people (unfamiliar): hummingbirds eating dirt, mud turtles (kinosternids) climbing trees. Third, art that depicts animals doing things that they actually do (accurate), infrequently (unusual), that are not known to most people (unfamiliar): mammals raising the adopted offspring of other species that are their typical predators or prey, grey whales in the Mediterranean Sea.

The question is, what expectations do we have for paleoart or wildlife art in general? Do paleoartists have a responsibility to only depict extinct animals doing things that are accurate, usual, and familiar? Maybe, if an art director for a book or documentary requested a vanilla dinosaur doing vanilla stuff, but outside of that situation?

Tree-climbing Protoceratops by John Conway, inspired by tree-climbing goats, borrowed from Tetrapod Zoology.

As will probably come as no surprise, I skew pretty hard in the other direction. Paleoartists are vastly more important to paleontology than wildlife artists are to zoology, because they have to do everything that artists of extant wildlife do–and one more crucial thing. If, say, a mammalogist needs to be reminded of the complexity and sheer otherness of her study animals, she can usually go out and observe them for a while, and see herbivores eating meat and carnivores eating plants and interspecies sex and all kinds of crazy stuff that real animals do. Paleontologists do not have the same luxury. It is all too easy to slip into the trap of thinking that we know what our animals were like in life. Consider, for example, the difference in temperament between black and white rhinos, or African and Asian elephants, and then consider Morrison sauropods or Two Medicine ceratopsians, and tell me you know anything about the behavioral differences between Apatosaurus and Diplodocus and their ecological ramifications. We need to be periodically shaken out of our comfortable assumptions and creeping anthropomorphizing (sensu Witton–not just attributing human traits to animals, but casting them in standard roles). We need to be confronted with the essential weirdness–and indeed unknowability–of our study animals. And we need paleoartists to do at least some of this shaking and confronting.

I’m not saying that paleoartists have a responsibility to deliver the unfamiliar or unusual in their art, any more than they have a responsibility to only draw vanilla stuff. I don’t think that paleoartists have a responsibility to anything but accuracy, and I mean accuracy in the inclusive, “not directly contradicted by the fossil record” sense* instead of the exclusive, “only what the evidence will support” sense. I’m saying that we–paleontologists, dino enthusiasts, science writers, museum docents, interested citizens–need the unfamiliar and unusual in paleoart as much or more than we need the comfortable and familiar, and we can only ask for it and be grateful when it appears.

* Hat tip to John Conway for this very useful turn of phrase.

Now, on the flip side, just because there is a huge amount that we will never know about extinct animals does not mean that we should give up trying, or that we should play down the reasonable inferences that we can make. Triceratops probably fought each other more than Centrosaurus, for example, or at least inflicted more damage on the squamosals of their conspecifics (evidence, discussion, link to paper). Would a painting showing two Centrosaurus beating the hell out of each other with their horns and doing all kinds of gnarly damage to each others’ heads therefore be inaccurate? Of course not–I am certain that at some point in the multi-million-year history of centrosaurs, two of them did in fact beat the hell out of each other in just that way. But neither would that painting show their usual mode of settling differences, so far as we can tell from our current interpretation of the available fossils (count the caveats there). That’s what the usualness axis is all about–getting comfortable with the  distinction between what animals occasionally do and what they commonly do.

Scavenging Styracosaurus by Mark Witton–go here for the full-size version and Mark’s thoughts on ceratopsian carnivory.

There is a lot that we simply won’t ever know. Which is why I advise caution in assessing accuracy. As long as whatever the animal is doing doesn’t violate the laws of physics, I think it’s hard to rule out that it could have happened, somewhere, at least once. So the interesting discussions will probably center not around accuracy but around usualness. It’s hard to argue that a styracosaur never scavenged a carcass, but do we think that scavenging and even predation were common behaviors for ceratopsians? Given that squirrels are notorious for killing and eating chipmunks, and that deer apparently eat the eggs and nestlings of ground-nesting birds as often as they can get them, the possibility that carnivory was a usual feature of ceratopsian behavior is worthy of serious consideration. At least, we can say that (1) it is consistent with the behavior of many extant herbivores, and (2) it is something that ceratopsians were  well-equipped to carry out. And given those antecedents, it is a difficult hypothesis to falsify. Then again, “difficult to falsify” does not mean “true”–so there is room for interesting discussions.

And that’s really what this post is all about: fostering productive conversations. I have seen and been part of many paleobiology conversations that went nowhere because accuracy, familiarity, and usualness were all scrambled up–often in my own mind. I’m not saying that this particular parsing of the issues is the best possible–indeed, I hope that it inspires someone else to come up with something better. But I also think that it is better than nothing, and that couching things in these terms might help us zero in on our points of genuine disagreement, and thereby make some progress, whether we’re talking about paleobiology, paleoart, or both.

What do you think?

UPDATE: Dave Hone has blogged on this sort of “what if” stuff, at least thrice: here, here, and here. That last post includes more of John Conway’s art from his “All Yesterdays” slideshow at the SVPCA 2011 icebreaker, which was awesome.