Paleoart As Science

Adrian Currie writes...

A few months ago, my friend Isobel Ronai drew my attention to the following tweet (thanks Isobel!):

In light of the uses twitter is put to these days, tweet-based wish-fulfilment is not, generally speaking, a good idea. But in this case I’ll make an exception.

Naish is referring to a lovely paper he, Mark Witton and John Conway published in 2014. They both defend the importance and validity of ‘paleoart’, and engage in some paleoart activism. I’m going to extend one of Naish and his co-authors’ ideas.  I think that under certain conditions paleoart isn't just scientifically valid, but should count as science. That is, the best paleoart isn’t just educational, or rigorous, or useful for furthering scientific aims. It should be considered part of the main business of science itself.

So, what is paleoart?

If you're reading this blog, chances are that you have an affection for the denizens of the deep past. And if you like extinct critters, you probably like representations of them.

Representations like paintings:

Heinrich Harder's 1920 painting of a sauropod sporting a wonderfully lizardy stance...

Heinrich Harder's 1920 painting of a sauropod sporting a wonderfully lizardy stance...

 Representations like mounted skeletons:

hur hur 'mounted'. Another excuse to post a photo of the Jurassic Museum of Asturias' racy reconstruction.

hur hur 'mounted'. Another excuse to post a photo of the Jurassic Museum of Asturias' racy reconstruction.

Representations like toys and figurines:

Oh feathered  T. rex , will you ever get old?

Oh feathered T. rex, will you ever get old?

Representations like full-scale reconstructions

Get it? 'Scale'? The Crystal Palace's 1852 Megalosaurus.

Get it? 'Scale'? The Crystal Palace's 1852 Megalosaurus.

Representations like costumes

Just another case of dinosaur-on-dinosaur violence.

Just another case of dinosaur-on-dinosaur violence.

Representations using stop-motion pupperty

The mighty  T. rex  from the 1974 Doctor Who Serial  Invasion of the Dinosaurs.  Although, given the number of fingers, it could just be an overgown  Allosaurus ...

The mighty T. rex from the 1974 Doctor Who Serial Invasion of the Dinosaurs. Although, given the number of fingers, it could just be an overgown Allosaurus...

Representations using computer animation

And so on and so on. There are, then, plenty of ways of representing extinct animals; of making paleoart. And, um, it’s also clear that thre are plenty of ways in which they vary in their scientific validity.

As Jillian Noyes’ excellent guest post illustrates, there are also many reasons to be interested in how the denizens of extinct worlds are depicted: the artistry, how they reflect the ideas of the time (scientific or otherwise), the collaborative working relationships between scientists and artists, and so on. And undoubtedly, the popularity of art which draws upon and explores the extinct worlds which paleontology reveals to us - from the most commercial to the most high-brow - cannot be denied.

Despite this, according to Witton, Naish & Conway, paleoart is in a bad way.

Paleoart has long had a somewhat uncomfortable relationship with paleontology itself. On the one hand, such representations are incredibly important for popularizing paleontological work, but on the other hand, aesthetic and commercial interests (as well as just-plain-laziness) often lead these representations to, well, not exactly meet a rigorous scientific standard. 

The main target of Witton, Naish & Conway's discussion are the cultural and economic features of paleontology and its relationship with the public. In particlar, they're interested in how these can undermine the production and appreciation of good paleoart. For instance, they highlight a ‘culture of copying’ where, instead of developing reconstructions on the basis of recent scientific findings, paleoartists simply copy earlier images (see Naish's recent discussion). This leads to significant drag between developments in the state-of-the-art, and what we see in depictions. Not to mention the frustrating dissemination of dinosaur tropes. They call for action to combat these woes.

We are optimistic that increasing awareness and promotion for palaeoartists could ultimately see the current, often down-trodden palaeoart industry become a much more vital, interesting and economically sustainable one (9)

If, like me, you think that the social aspects of science are often where the epistemic action is, the kind of discussion that Witton, Naish & Conway have is fascinating and I might, in a later post, look at it more closely. Here, however, I want to think about the relationship between paleoartists and scientists in terms of knowledge-production.

So, is paleoart ever science?

We might say no. Although paleoart at its best draws upon the latest scientific ideas, it is not properly speaking science. Rather, it draws upon science. It is, perhaps, analogous to technology: although scientific knowledge is utilized in design and engineering, we shouldn’t properly speaking say that the development of new technology is science. No, technology and paleoart are simply applications of scientific knowledge.

I can imagine motivating such a view by pointing out that paleoart is necessarily speculative: it goes beyond what we know, what has been scientifically verified, and therefore into the realm of mere story-telling. And although scientists might tell stories sometimes, when doing so, they are not (to put things philosophically) doing so qua scientist. The scientist qua scientist only cares about what has been rigorously and empirically established, and so should have no truck with such fancy. Moreover, paleoartists are typically not scientists institutionally speaking.

However, in their discussion, Naish and friends briefly highlight how the production of paleoart has itself progressed palaeontological knowledge.

… it is unsurprising that palaeoartists both bring scientific concepts ‘to life’ and also advance ideas and hypotheses about the portrayal and even anatomy, behaviour and biology of extinct organisms. The production of rigorous reconstructions has shed light on the proportions, postures, gaits and body masses of extinct creatures. (3)

So, in some cases paleoart appears to have played a direct role in paleontological progress. In what remains, I’m going to suggest that it has so in virtue of making concrete otherwise abstract scientific ideas, and thus enabling us to identify new, hitherto irrelevant, evidence. That is, I’m going to expand upon what it is to bring scientific concepts ‘to life’, and explain why this is so crucial to progress in paleontology.

Karl Popper famously argued that science works best when it involves bold conjectures. Instead of the softly-softly approach of carefully and gradually building support, he encourages scientists to stick their necks out:

... there is no more rational procedure than the method of trial and error - of conjecture and refutation: of boldly proposing theories; of trying our best to show that these are erroneous; and of accepting them tentatively if our critical efforts are unsuccessful (from a lecture in 1953).

Now, I don’t think we should always be so bold, but I have argued that this Popperian sentiment is particularly important given a particular problem paleontologists often face: let's call this the problem of identifying evidence. Given some historical target, it is often tricky to work out what evidence will be relevent to it. Let me give you an example (drawing on the rather tired but dammit still fun tradition of philosophers going on about Sherlock Holmes).

In Conan-Doyle’s short story Silver Blaze Holmes and Watson are trying to piece together the disappearance of a famous racehorse and the murder of her trainer. In a famous moment, Holmes takes the lack of a dog barking as a crucial piece of evidence:

Detective: Is there any other point to which you would wish to draw my attention?

Holmes: To the curious incident of the dog in the night-time.

Detective: The dog did nothing in the night-time.

Holmes: That was the curious incident.

Basically, (spoilers!) Holmes ‘deduces’ (um, inducts) that had someone unfamiliar to the dog stolen Silver Blaze, the dog would, as dogs do, have caused a commotion. Makes sense (it is also a fun case of an absence counting as evidence, something Derek and I discussed recently).

Now, imagine that prior to investigating the case, Holmes was to consider what evidence will matter to it. Some sources of evidence are obvious: look for hoof-tracks, for instance, or try to find some witnesses to interview. But for all his cleverness, I doubt that even Holmes would guess that a lack of dog barking would be crucial for solving the mystery. This is because evidential reasoning often proceeds via scaffolding. It’s only once Holmes had started to peice things together, having followed some faint horse tracks out to the moor, and had some reason to think that she was taken during the night (and so on), that the dog’s behaviour became relevant to his investigations. Similarly, in paleontology it is often only from a scaffold of both a bunch of data and—crucially for my point here—a bunch of (bold!) hypotheses relating to that data, that we are in a position to identify evidence. My defence of paleoart as science turns on the idea that it enables us to identify evidence in this way (Kim Sterelny and I make a similar argument for storytelling in a forthcoming paper).

How does paleoart fit in here? Well, to put it bluntly: good examples of paleoart just are bold conjectures. They enable us to solve the problem ofidentifying evidence. In making paleoart, we must take a collection of fairly undetailed scientific hypotheses and fill in the details. Paleoart involves making the abstract concrete. The palaeontologist collects fossils and other traces from the past, she examines and measures them. She draws on what she knows about how fossils form, the anatomy and physiology of animals, and so forth, to support a bunch of hypotheses about the phenotype, evolutionary history and behaviour of the critters. This leaves a lot open: it only constrains the various ways in which the critters might have in fact been like. The shear amount of wiggle-room is beautifully demonstrated in Conway, Koseman & Naish’s brilliant All Yesterdays, where they present us with a bunch of dinosaur reconstructions which cohere with the scientific evidence, but diverge massively from what we’re used to. For instance, here’s a pretty bog-standard Parasaurolophus:

… and here’s Conway’s version:

If we take the authors of All Yesterdays at their word, the awkward, bovine-reconstruction is just as scientifically valid as the more traditional, dynamic one (by the way, Derek discusses All Yesterdays in this paper). The artist has to ‘concretise’ the scientific evidence by realizing it with more specificity then it contains. And this – certainly in principle, and occasionally in practice – can provide a scaffold from which more evidence can be identified. Indeed, the act of making such an image can form the basis of synthesizing further evidence, particularly about aspects like behaviour.

By bringing scientific concepts to life, paleoartists enable us to think about how to test their bold conjectures: they provide the scaffolding required for the identification of evidence.

And so, not only is paleoart at its best creative, beautiful, fun and educational. It should also be considered part of the epistemic business of science.