What Is Palaeoart? Drawing Animals Nobody Has Seen

A reconstruction begins with the skeleton, which fixes proportions and joint articulation, and everything beyond that is inference of varying strength. Muscle attachment sites on bone indicate where major muscles were and roughly how large, which is the best-constrained soft tissue. Comparison with living relatives supplies the rest, and the standard method brackets an extinct animal between its closest living groups, so a feature present in both birds and crocodilians is reasonably inferred in a dinosaur while one present in neither requires specific evidence. Skin impressions, feathers and preserved tissue supply direct evidence where they exist and are patchy. Trackways constrain posture and gait. Colour is almost entirely unknown except in the small number of specimens preserving pigment structures. Behaviour is the least constrained of all, and a picture showing animals interacting is making a claim that the fossils rarely support.

Critics within the field have identified a set of habits that recur and that are mostly conventions rather than conclusions:

• •Shrink-wrapping, meaning rendering an animal with skin drawn tight over the bones, which no living animal resembles once fat, muscle and integument are included
• •Omitting soft tissue that leaves no trace, including lips, cheeks, wattles, fat deposits and display structures, all of which living animals carry in quantity
• •Perpetual aggression, since animals are depicted fighting, roaring and hunting far more than any real animal spends doing those things
• •Copying previous reconstructions rather than the specimen, which propagates errors across decades and is documented repeatedly
• •Drab colouration, since living animals are frequently vivid and reconstructions default to brown and grey without evidence
• •Anatomically impossible postures inherited from older art, including pronated hands on theropods, which persist long after being corrected

Colour was regarded as unknowable until the discovery that melanosomes, the pigment-bearing structures inside cells, are preserved in some exceptional fossils and that their shape correlates with colour in living animals, with elongated structures associated with black and spherical ones with reddish brown. That allowed the first evidence-based colour reconstructions, and several feathered dinosaurs have been restored with patterns supported by measurement rather than invention. The method has limits and critics, since the correlation is imperfect, since structural colours produced by microscopic architecture rather than pigment are harder to infer, and since preservation is patchy across a body. It nevertheless moved colour from pure speculation into partial evidence for a small number of specimens. For everything else, the honest position is that colour is unknown, and the reasonable approach borrows from living animals in comparable habitats rather than defaulting to grey.

The relationship runs both ways and the influence of images on interpretation is substantial. Charles Knight's paintings from the late nineteenth century onward established the visual conventions for dinosaurs for decades and were produced in consultation with palaeontologists, so the science and the art developed together. The reappraisal of dinosaurs as active animals in the 1960s and 1970s was propagated enormously by Robert Bakker's illustrations, which showed running dynamic animals and made an argument visually before the field had settled it. Film and television reconstructions reach audiences that no paper does and fix images that take decades to revise. Museum mounts are themselves reconstructions and carry the same influence. Researchers report that reviewers and readers respond to a claim differently depending on how it is illustrated, and several have argued that reconstructions should be treated as hypotheses requiring justification rather than as decoration accompanying the real work.

Standards have risen and the field has become self-conscious about its methods. Skeletal reconstructions drawn to scale from measured specimens are published as a foundation for fleshed restorations, which allows errors to be checked against bone. Artists increasingly document which parts of a reconstruction are supported by evidence and which are speculation, which is a straightforward honesty measure borrowed from scientific illustration generally. Comparison with living animals is applied deliberately rather than by default. Digital tools allow muscle reconstruction and biomechanical testing to inform the image. And a community of practitioners and critics scrutinises published work, with a substantial online discussion that has visibly improved the standard of illustration in books and exhibitions over the past two decades. The remaining tension is commercial, since dramatic images sell and accurate ones are frequently less dramatic, and illustrators working for popular publishers describe being asked for the older conventions specifically.

The skeleton fixes proportions, muscle scars indicate the major muscles, and everything else is inferred by bracketing between the closest living relatives, with colour and behaviour least constrained of all. Shrink-wrapping, perpetual aggression and copying earlier art are recurring errors rather than conclusions. Illustrations propagated the reappraisal of dinosaurs as active animals before the field had settled it.