julia_clarke's picture
John A. Wilson Professor and HHMI Professor, Jackson School of Geosciences, University of Texas at Austin


I would like to put to rest the notion that evolution as a process should conform to words and concepts that we find familiar, comfortable, and perhaps even universal. More immediately, I'd like to stop having to always explain whether or not each new feathered dinosaur specimen we discover was a bird.

In many ways it's an understandable question. Most scientists have accepted for years that living birds are one lineage of dinosaurs. The idea that dinosaurs live on in birds even crept into popular consciousness through Jurassic Park. So perhaps it's not surprising that when scientists discover a new feathered dinosaur, people—including scientists and science journalists—often want to know, "Did it fly?" Consider the first-discovered feathered dinosaur, the so-called "Urvogel" Archaeopteryx. Debate continues in the scientific literature: Was it a bird?

As a paleontologist working on the evolution of living birds, I find myself having this exchange over and over again. For example, I describe a small feathery species newly uncovered from the fossil record. After detailing its known features, I might note that it may have had some form of aerial locomotion. There is inevitably a pause. Then the question, "Ok, but was it a bird?" Impatient with scientists and their endless modifiers and complex phrasing, the asker wants to get this story clear, "Ok, but did it fly?" Tell it to me straight.

The questions may sound innocent enough and perhaps are extremely intuitive to ask. However, although these seem like scientific questions, they mostly aren't. They concern primarily what we want to identify as part of the class of entities "birds" and what is part of the class "flighted".  We might think we have these straight in the present day, but try looking back through a dirty lens at life more than 100 million years ago.

Paleontologists must use the shape and form of bone as well as, in rare cases, feathery impressions to track the ecologies of the long dead. To do this, they must use data on form-function relationships in the living. This task in and of itself is difficult and ongoing. But what is more difficult is to translate combinations of structures that are not present in any living species into an understanding of how it moved. For example, flighted living birds have a joint between the scapula and coracoid where the upper arm bone, humerus, meets the pectoral girdle. Yet we have species in the fossil record with feathered forelimbs of impressive span (shall we call them wings?) but lacking this kind of joint articulation. Subtle features of the feathers and their relative proportions may differ from any living bird. Is this creature a bird?

How did it move? Did it have a form of sustained flapping flight but unlike that in any extant species? If we could time travel back to a Cretaceous forest, would we call this movement flight? What if a species beat its wings only briefly to move from branch to branch? What if it utilized these "wing" beats to climb trees or jump? What if it was only volant as a juvenile, but as a large bodied adult it maintained feathery forelimbs for signaling to a mate but flew no longer?

All of these hypotheses have been forwarded, and all may have been true for different denizens of Jurassic and Cretaceous environments. We can debate whether these creatures flew and whether or not they were birds, by our contemporary definitions, but doing so, we risk losing sight of the bigger scientific questions. All too quickly, we can fall into a rabbit hole of defining (and defending) terms, when we'd do better to seek a more precise understanding of the emergence, the relative evolutionary first appearances, of the many features comprising the flight apparatus in living birds.

Feathers make their first appearance in taxa that could not have been volant as adults. Precursors to feathers, simple filaments, are found in tyrannosaurids and an array of other relatives of living birds. While hundreds of characteristics of bone and feather have revealed these deep genealogical relationships within dinosaurs, we still seem to be searching to pin "bird" and "flight" to single characters.

I am not the first to remark that the debate over what to call a bird and what to term flight is not useful and actually at odds with evolutionary thinking. But, I have been surprised by the persistence of this debate even among specialists. For example, exchanges over how to apply the formal taxonomic name "Aves" are ongoing. While events unfolding in deep time via evolutionary processes are arguably the least likely candidates for dichotomous or categorical thinking, this mode of thought runs rampant and engenders false controversies that obscure interesting questions. It is tracking the more complex pattern of asynchronous change in many novel traits that will inform generalities about how the evolution of shape and form may work.

Arguably the hypotheses we investigate should be arrayed relative to one another in relationships other than opposition. However, often the categories we are comfortable talking about will artificially organize them into this apparent relationship. Indeed, across science I would argue we in fact have many "urvogels" lingering evidence of similarly strong collective cognitive investment in the existence of classes of entities we consider intuitive and natural. These can hold us back.