How do paleontologists and artists bring dinosaurs back to life? Museums are filled with dinosaur skeletons and books are packed with restorations of what dinosaurs may have looked like, but how does a pile a bones get turned into a reconstruction of the living animal?
It all starts with the bones. While some dinosaurs, like Allosaurus, are known from multiple specimens that have allowed paleontologists to study their complete skeletons, other dinosaurs are known from a handful of scraps. For every dinosaur known from nearly complete remains there are scores more that are only represented by a bit of leg, a fragment of jaw, or a few vertebrae. How can scientists and artists restore these animals based upon such paltry evidence?
The science of paleontology has deep roots in related sciences like zoology and geology, disciplines where comparisons are important. Indeed, some of the first people to study dinosaur bones compared them to the remains of other animals to determine what sort of animal groups they might belong to. This was a very difficult task because nothing quite like the dinosaurs had been seen before.
Today paleontologists have an ever-growing collection of skeletons among which they can make comparisons. A piece of jaw by itself might be an enigma, but if it is compared with similar sections of jaw from known dinosaurs, scientists can get a better idea of what it may or may not be.
This works because dinosaurs, like all other organisms, evolved and share some traits in common with close relatives but not with other distant relatives. Albertosaurus is more closely related to Tyrannosaurus than either is to Velociraptor, for instance, but all three are coelurosaurs and more closely related to each other than any of them are to a sauropod like Apatosaurus. These relationships can be very important when dealing with an animal that is known from an incomplete skeleton.
Take Pachycephalosaurus, for example. Even though it was named in 1943, very few fossils of it have been found and certainly no complete skeletons. Even so, the fragments hold important clues if you know where to look. Bits of short forearm and long leg bones indicated a bipedal animal, and the skull, especially the teeth and jaws, revealed that it was an ornithischian dinosaur. This made it a relative of hadrosaurs and horned dinosaurs, and this is important because these dinosaurs had fairly conservative body plans. This meant that even though different genera may have sported different sorts of ornamentation on their heads, the rest of their bodies (from the neck down to the tail) were very similar across closely related forms. This allows paleontologists to look at close relatives of Pachycephalosaurus like Stegoceras for hints as to what the missing bones of Pachycephalosaurus might have looked like.
The word "hypothesis" is key here. Any restoration or reconstruction of a dinosaur is being continually subjected to our growing understanding of the fossil evidence. If a restoration of a dinosaur is created based upon fragmentary remains and comparative anatomy but a more complete skeleton shows that the dinosaur was really different in some ways, then the hypothesis will have to be revised. In this way scientific illustration closely mirrors the scientific process itself.
The alternative is to simply leave out parts of the skeleton that are unknown, which means that our dinosaur books would be inhabited by many creatures missing parts of their legs, tails, skulls, spinal columns, etc. We know they must have possessed such body parts, and rather than leave them out it is better to look at related dinosaurs for clues.
In an exchange posted a year ago at the blog "When Pigs Fly Returns," the paleo-artist Michael Skrepnick explained how artists navigate these difficulties in restoration. Artists like Skrepnick are not just guessing or just putting in whatever fits, but engaging in the scientific process to make a visual hypothesis of what these creatures were like when alive.