What did extinct dinosaurs look like? We’ve been wondering this since the time before the word “dinosaur” was even coined, strange bones inspiring thoughts of creatures unlike any we see around us today. That fascination still draws our imaginations and drives scientific discovery, although this is not cumulative exercise of simply acquiring more bones and data. It’s a messy process, what we think we know and what could possibly be constantly colliding with each other. Even something as seemingly simple as whether some dinosaurs had cheeks is fertile ground for debate.

Think of a dinosaur like Triceratops. This massive, three-horned herbivore had jaws like garden shears, only with teeth instead of blades. And given this animal’s plant-munching habits, some kind of cheek would have been advantageous to keeping those clumps and shreds inside the mouth as the dinosaur chewed. But how can we tell? No one has found a delicately-preserved Triceratops skull with pristine skin impressions solving the problem for us (if such a fossil even exists). The answers, so far as the current collection of evidence goes, relies on what we know about the relationships between muscle and bone. That’s what anatomist Ali Nabavizadeh drew upon to investigate the chewing muscles of some herbivorous dinosaurs.

Nabavizadeh focused on ornithischian dinosaurs, or the major dinosaur group that contains familiar Mesozoic stars like the armored dinosaurs, horned dinosaurs, hadrosaurs, and more. (The other herbivorous dinosaurs - the long-necked sauropodomorphs and the varied theropods who evolved a plant-eating diet independently - were not part of the study.) These are the dinosaurs that seemed to be the best candidates for having cheeks, particularly the later and more derived species that had evolved their own ways of chewing plant material. If we’re going to call dinosaurs like Edmontosaurus the “cows of the Cretaceous,” then it’s worth asking if they had cheeks like cows do.

The problem, Nabavizadeh points out, is that no birds, crocodylians, or reptiles have cheeks like mammals do. If dinosaurs had cheeks, the underlying musculature would be very different. So here’s where the bones come in. Muscle scars and other clues can help reveal the anatomy and extent of musculature in extinct animals. If dinosaurs had their own type of cheek, there should be anatomical signs of it.

What Nabavizadeh found differs from what’s been proposed before. There appears to have been a “rostrally-expanded muscular support system” from the lower jaw to the cranium of dinosaurs like Triceratops, Edmontosaurus, Ankylosaurus and some other ornithischian dinosaurs - that is, a muscular connection between lower jaw and cranium that reached further forward. This not only made biting and chewing more efficient from a biomechanical point of view, but would have helped contain food within the mouth without requiring a specialized and novel sort of cheek to evolve. Large, herbivorous ornithischians had their own, unique anatomical solution to chewing on plants all day, further evidence that life, uh, finds a way.