Are Birds Dinosaurs?

A Case Study in the Flexibility of Social Kinds

Before turning to the matter at hand, a few preliminary remarks for my subscribers and regular readers. As I promised some weeks ago, it is my intention to post original writing at least once every fortnight, and on “off” Mondays to share either extracts of work in progress, or drafts of stillborn pieces that may yet be reanimated by the alchemy of public exposure, or, as is the case today, specially curated pieces that I for one take to be the “best of” the work originally appearing on my old website, at some point between 2005 and mid-2020.

The essay below originally appeared in February 2020, and was the basis of a couple of public talks I gave shortly before lockdown. It is, I take it, representative of an approach I’ve often tried to adopt in my scholarly and perischolarly work, namely, to stake out a position on issues of current social importance, but to do so stealthily, moving far, far away from anything currently in the news or in our newsfeeds.

I am aware that there is a lot going on in the world. Trump has the virus, for example, and many people are convinced that it is our moral and civic duty to do nothing but weigh in on this fresh news item. Many are working hard to construe it as a Greek tragedy. My own view (I could be wrong; we all could be wrong) is that reality will never line up with our narrative will, and anyhow if an ancient tragedian had been “writing the script” he would have made Trump get the virus about three weeks later than he in fact did. The vastly most probable outcome of the actual-world scenario is that by election day Trump will be fully recovered, neither his supporters nor his opponents will have budged an inch as a result of this “plot twist”, and nobody will remember why a month prior they found it so fitting and urgent to invoke Sophocles. 

Anyhow, the prime directive of this new newsletter might best be summed up as: No news-hooksor, as they say on the internet: No. News. Hooks. The belief that such things must be a part of everything we make time to think about is degrading to thought and harmful to our civic life. Dinosaurs, for example are interesting in themselves, and it is only a society with its priorities all wrong that would try to convince you that this interest is one best shed around the time you hit puberty. They are interesting in themselves, and they are also, to riff on Claude Lévi-Strauss, “good to think with”. That is, you learn about a lot more besides, when you stop to think about how we think about dinosaurs. 

So then, are birds dinosaurs or not? 


How long have dinosaurs been around? There is one obvious sense in which they ceased to exist 66 million years ago. There is another sense in which they began to exist only around the middle of the nineteenth century, when Richard Owen identified a “distinct tribe... of Saurian Reptiles” in 1842. Most animals have a long history of social salience before science comes along to tell us exactly where they belong in the order of nature. Not so with dinosaurs: they didn’t have any place in society at all until science informed us of their past existence, and from that point on their salience has been entirely wrapped up in cultural representations. These representations are anchored in something real, in a way that those of, say, unicorns are not, but the fact that we have fossilised skulls and vertebrae to point to in the case of dinosaurs, while we do not have equine skulls with a horn in the middle to point to in the case of unicorns, only makes it more difficult, not less, to understand what we may expect the folk-categorical term “dinosaur” to do. 

At first glance it may seem surprising that there should be a folk-category filled by representations of a class of beings that we (the “folk”) only know to exist at all thanks to what science tells us. But the folk are particularly adept at taking the austere information science delivers, and filling it in with fantasy. This is why, for example, black holes figure so prominently in science-fiction scenarios about cosmic consciousness. Yet in the case of palaeontology the people making the discoveries and fleshing out the dry bones with their imaginations, are often much closer to the folk than is generally the case of, say, black-hole cosmologists. 

And so the original image we have of dinosaurs as “terrible lizards”, an image that never really fit all the available evidence, even if at least some of them had large teeth, is one that was produced by field scientists who were simultaneously making the discoveries and letting these discoveries fuel their imaginations. And from the starter dough of their early imaginings, cultural representations begin to ferment and grow on their own.


As Claudine Cohen has shown, these representations have been very different in different cultural settings. Nowhere did the dinosaur take on a more iconic status than in the United States, in the early twentieth century, and this had to do largely with the political-geographical project of incorporating western frontier territories into the economic life-blood of the country through the extraction of natural resources from them. From the 1910s or so no natural resource was more important than the stuff that was to fuel automobiles. 

Of course the vast majority of organic matter that goes into the slow subterranean generation of fossil fuels comes from algae, and not from dinosaurs, but this did not prevent the Sinclair Oil Co. from using the image of a brontosaurus on its corporate logo beginning in 1933, marking out gas stations across the Arizona desert for American families assaying the length of their nation’s territory in a station wagon. Keychains and other souvenirs were available for purchase.

Screen Shot 2020-02-15 at 23.07.42

In the generations that followed, the cartoon representation of the dinosaur, often with a caveman on its back, would come to be as familiar as the “exotic” megafauna of Africa. By the 1970s there emerged however a pedagogical current, with which I’m very familiar from first-hand experience, that encouraged elementary-school children to uphold a certain standard of correctness in the way they spoke of dinosaurs: to insist on not placing a prosauropod of the Triassic next to a Cretaceous triceratops in a diorama, for example, let alone next to a caveman. This new standard of correctness as a value emerged simultaneously in American history with a countercurrent that deviated radically from more or less everything the accumulating evidence was telling us about the history of life on earth: namely, so-called “creation science”, a movement that appears to exist not only in tension, but also in dialectical interdependence, with the pretence of science and its volunteers from among the folk to be getting the earth’s distant past definitively right.

It is against the background of these parallel developments —both the new premium on correctness as a value among the folk, as well as the rise of a prominent ideological countercurrent that flagrantly rejects this value— that we must begin to reflect not just on the recent claim that birds are dinosaurs, but also on the implicit normative force of that claim. That is, it is not just that birds are dinosaurs, but also that you, fellow members of the folk, must affirm that they are. 

Screen Shot 2020-02-15 at 23.12.23


Now, in phylogenetic terms, a dinosaur is: (i) either a modern bird, or (ii) a triceratops (which went extinct 66 million years ago), or, finally, (iii) any member of any species descended from the last common ancestor of both birds and triceratops. So this excludes crocodiles, because the last common ancestor of birds and crocodiles occurs earlier than the last common ancestor of birds and triceratops, even though crocodiles are the closest living relative of all members of class Aves

This definition has the virtue of being clear, and permitting us to avoid the difficulty of setting up an arbitrary boundary as we move forward in time, between the last dinosaur survivor of the Cretaceous-Paleogene extinction event that killed triceratops and most other dinosaur species, on the one hand, and on the other the first so to speak post-dinosaur, descended from the so-called feathered dinosaurs, but now just too morphologically distinct to be considered a dinosaur itself. But fear of an arbitrary boundary, at least at the closer end of the temporal range of the dinosaur, seems to force the basic distinction such a boundary would have permitted to make its entrance under a new guise, namely, the distinction within this temporal range between the “avian” and the “non-avian” dinosaurs. 

It is not hard to find other cases in which the expansion of a classificatory term forces new distinctions within the space opened up by the expansion. For example, from the moment the American census began to distinguish between “racial” groups such as “black” and “white”, on the one hand, and partially overlapping “non-racial” identity groups such as “Hispanic” on the other —here, emphatically, I am using US census terminology, and not my own—, we unsurprisingly began to hear the phrase “non-Hispanic white” with much greater frequency. “Non-avian dinosaur” seems to work the same way: there is a group we often find ourselves wanting to talk about that is defined more broadly than it was in the past. But in many contexts we still want to talk about the group as it was defined in the past. And so we construct a negative phrase that indicates we are subtracting a current subset of it.

“White” is a folk category par excellence, and no respectable person today looks to it in the expectation that there is an underlying metaphysics or science of natural kinds forcing it upon us. The same is not generally thought to be true of “dinosaur”, though, and so the ease with which the folk revert to the phrase “non-avian dinosaur” upon being told that birds are dinosaurs, signals that there is a strong tension between the folk category and the phylogenetic definition. 

What should we make of this? Are the folk simply mistaken? The claim that birds are dinosaurs is a claim among other things that phylogeny trumps morphology, that is to say that to be a dinosaur is to have a dinosaur’s lineage, no matter what you look like, or how much you deviate from the most paradigmatic representation of, say, a lumbering, quadrupedal, somewhat tortoise-like sauropod. And here it is curious, and worthy of pause, that we play this trump card in a case where the paradigmatic members of the kind are long extinct.

Compare the case of whales and fish. John Dupré has compellingly shown that “fish” is yet another folk category par excellence. When we try to delimit it precisely, we find that the very task of drawing precise boundaries generates new problems in its wake. One problem is that many species that most folk and most scientists would be willing to call “fish” do not meet the standard definition of “gill-breathing craniate animal”. Consider the tellingly named “lungfish”, class Dipnomorpha, which is both capable of breathing air, and is the currently extant species closest to the ancestor of the first land-dwelling tetrapods, which emerged around 400 million years ago, and from which dinosaurs and mammals, including ourselves, and indeed including whales, would eventually evolve. 

Screen Shot 2020-02-15 at 23.17.35

That is, there is a species of fish, or “fish” in scare quotes if you’re not ready to commit, that is more closely related to whales (and to human beings) than it is to ray-finned fishes (class Actinopterygii) such as tuna or cod. In other words there is no meaningful phylogenetic basis for putting lobe-finned fish such as Dipnomorpha or coelacanths together with tuna and not with whales, even though whales have the curious distinction of having evolved from land-dwelling animals that themselves evolved from sea-dwelling animals (they have this in common with numerous species of sea snakes as well). 

Phylogeny, I mean, seems to be invoked inconsistently, with no clear rule that could explain when paraphyletic classes might be legitimately admitted into our system of carving things up. Now in the case of lungfish, tuna, and whales, we are dealing with species all of which are today extant, even though of course part of our consideration of where they stand (or swim) in relation to one another has to do with other species long extinct. But when we turn our attention to avian and non-avian dinosaurs, by contrast, we are faced with two different classes of beings in very different ontological predicaments: the ones exist, and the others no longer do. 


One might suppose on first consideration that this shouldn’t matter, that the present has no special status in our taxonomical efforts. This would be true, except that the present happens to be when we are making these taxonomical efforts. We have no choice but to work outward from it, and we find in this work that we have a different sort of epistemic access, for living species than for extinct species, to the criteria for classification that interest us. 

This is a problem that has been perceived by palaeontologists at least since Georges Cuvier sought to classify the enormous “saurians” excavated from the gypsum mines at Montmartre in 1798. Having more or less passively adapted the Buffonian conception of species, as that which is held together by the “unity of reproduction”, Cuvier was struck, as one cannot fail to be, by the fact that fossils do not reproduce, and so if the species you are trying to classify only exist in fossil form, you are not ever going to be able to apply the criterion by which they are marked off from one another. It is in part for this reason that Cuvier’s program of comparative anatomy veers into a structuralist project of establishing anatomical similarities across species, without any regard, or at least explicit regard, for what these similarities might imply about kinship or shared lineage.

This difficulty would be addressed directly in George Gaylord Simpson’s important 1951 article, “The Species Concept”. What Simpson calls “typological definition” of a taxonomic group is the establishment of the group on the basis of its correspondence “with an abstract or ideal morphological pattern.” He notes that typological classification is “pre-evolutionary and non-evolutionary”. Yet prima facie it would seem that palaeontology leaves us no other choice than to revert to it. 

Simpson is interested in examining the notion of species in particular as deployed in palaeontology, whereas both “dinosaur” and “bird” are significantly higher taxa than this. But we may nonetheless learn something from his observations. He maintains that “[t]he palaeontologist... uses the designation ‘species’ for two sorts of entities which are radically and fundamentally incongruent.” One possible approach to palaeontological species classification, for Simpson, is, starting from morphological features, “to recognise central lines as species and to distinguish branches as other species, ... even though their delimitation is genetically arbitrary at the point of branching.” Thus:

Screen Shot 2020-02-15 at 20.51.11

“Another possible approach,” Simpson continues, “is to recognise each evolutionary lineage as a unitary species until it divides and then to consider the descendent branches as species distinct from each other and from the single ancestral line.” Thus:

Screen Shot 2020-02-15 at 20.51.50

As for the first possibility, as represented by B (we are leaving some out here), Simpson notes that “the only reasonable criterion of choice would be designation of certain terminal branches as more important, or somehow definitive, than others.” Intriguingly, he mentions a view that he attributes to “a philosopher,” a certain H. Miller (as we learn only from the bibliography), in the long-out-of-print book, The Community of Man, published in 1949. (I have been unable to determine with any further precision who this Miller was; perhaps someone else will know.) According to an extreme formulation of this view, as Simpson summarises it, one might “take Homo sapiens as the supreme species, and consider its ancestry, from the beginning of life (or even before) as the main line, not specifically separable from H. sapiens... Taxonomists will surely agree that this result and the whole procedure involved are impractical if not absurd.” 

Again with the caveat that Simpson is speaking of species while birds and non-avian dinosaurs are both enormous clades, could it be that the contention that birds aredinosaurs resembles the absurdity Simpson has identified here? 

Suppose the meteor had not hit the earth, and that triceratops and all of its Cretaceous contemporaries had survived and evolved into new forms, differing at least as much from the ceratopsians as, say, the sparrow differs from the archaeopteryx. Suppose some of these descendants of Cretaceous dinosaurs eventually returned to the sea, like the ancestors of whales did, and, like the whales, evolved an outer morphology similar to that of fish. It seems unlikely that any taxonomist trying to make sense of such a scene of biodiversity would think to insist much on the fact that those ceratopsian-descended sea animals, on the one hand, and the birds on the other (let’s stipulate that they evolved in the meteor-free world more or less in the same way as in this one, which is naturally impossible, of course, but not logically impossible), are members of the same category of beings, even if it were known that they have a common ancestor. Again, whales and lungfish, or if you prefer, humans and lungfish, also have a common ancestor. 

This thought-experiment suggests that the insistence that birds are dinosaurs is conditioned in part by a conviction among those who promote it that birds are somehow definitive of the branch that reaches back to the common ancestor of birds and triceratops. This would not be in the way that “the philosopher” seems to imagine Homo sapiens to be definitive of the lineage that led to it — that is, in a way that strongly implies teleologically guided evolution. Rather, birds perhaps present themselves as the suitable end of the lineage in part because there is some need at the present moment, some sense-making exigency, to argue that the dinosaurs survived after all, that in spite of what we had previously thought, they are still with us. 


Before we venture any suggestions as to what the source of this felt need might be, it is worth noting that in general efforts to shorten the gap between the folk conception of birds and the folk conception of dinosaurs has involved as it were the avianisation of the latter, rather than the dinosaurisation of the former. From particular discoveries of filament-like layers protecting the outer surface of the bodies of many dinosaur species, the conclusion is quickly reached that these filaments must have been feathers; new artist’s renditions are hastily drawn up, and on occasion they have to be retracted when it is learned that they went further in the avian direction than the evidence permitted. Overcorrection from the Godzilla template for the T. rex has on occasion made the tyrant king of the lizards into a sort of chicken. Thus (to return to the pitiful figure we already encountered last week): 

Screen Shot 2020-02-15 at 23.21.22

There will always be corrections, and overcorrections, and renegotiations of our representations of the past. That is normal, and it is good that over the past several decades T. rex has been made more gracile, less like an upright Gila monster. But again, what is at stake in the particular direction in which our most recent overcorrections are sending us? Even if we are prepared to admit that birds are a branch of dinosaurs, we may still ask: Why must dinosaurs in turn be birds? 

I am not going to venture an answer to this second question today, except to say that it seems to fit a general pattern of revisionism in science education over the past decades that compels us to revisit whatever had previously been marked as “terrible”, and to discover that it was actually rather gentle or even cultured: the same pattern has been evidenced in cases as seemingly far apart as whales (perhaps the first trailblazer in this transformation), bats, sharks, and Neanderthals.

As to the particular need that I’ve already identified however, to have the dinosaurs surviving the great extinction event and into the present day, a few tentative remarks may be in order. Contrary to what some people’s intuitions here might be, it does not seem to me that these newly discovered (or newly legislated) survivors of the Cretaceous-Paleogene die-off have something to do with the current climate of ecological crisis, in which, after all, the prevailing sentiment is that nobody gets to survive a major extinction event, let alone come out the other side even more gentle and cute than before.

It may have something more to do with the dialectical relationship, to which I’ve already alluded, in which the public presentation of evolution is constantly formed and deformed by the political forces of evolution denialism. In this context, creationists have for decades weaponised the rather healthy skepticism that philosophers of science have brought to empirical claims about a distant unrepeatable past, and transformed it into a sort of universal denial of the possibility of having a science of the past at all. If dinosaurs were entirely a thing of the past, such skepticism could perhaps be seen as having more of a foothold than in a world in which dinosaurs are in fact well represented in the present. 



That’s all I’ll say about that for now, in order by way of conclusion to return to our guiding question: are birds dinosaurs, or aren’t they? Even if we agree with Dupré, as I think I do, that whales were fish until the mid-nineteenth century, and then ceased to be so as a result of new taxonomical legislation (which, as D. Graham Burnett has nicely shown, was itself influenced by actual legislation concerning the importation of whale blubber into the early United States: should it be taxed as fish oil? Or is there perhaps a way to get around that?) — even if we agree with Dupré about this mid-nineteenth-century adjustment, I was saying, it is not at all clear that a similar adjustment has yet been successfully carried out in the early twenty-first century as concerns the status of birds. 

Folk categories are not determined by fiat, but by actual usage among the folk. In certain scientific matters (say, geocentrism), the folk can be shown to be wrong over time, and gradually correct themselves. But in taxonomy it’s generally the case that the folk do no worse than the scientists. How can they, when any classificatory scheme is relative to the initial concerns of the classifier? 

Whales became non-fish, and then they became the gentle wise elders of the sea, not just “because science”, but because of a pervasive trust in scientific authority to reveal to us the true natures of things, and a fairly active post-whaling propaganda campaign on the part of scientific and political bodies to convince people to revise their conception of whales.

Screen Shot 2020-02-15 at 23.30.32

The political landscape is very different today, and not surprisingly what we see in response to the question whether birds are dinosaurs, as we see almost everywhere else, is a rift and a stalemate. Some people adopt the new ordering with pride and with a supercilious commitment to the correction of their peers who are still getting it wrong, while others hold back, somewhat suspicious but often unable to articulate the nature of this suspicion, and if you show them a sparrow they’ll say: “That just doesn't look like a dinosaur to me.”

Whether in the end these people cross over, and the twenty-first century proves to be the one in which the status of birds as dinosaurs is consolidated, depends much more on the calibration between scientific authority, the propagandistic aims of science education, and popular sentiment, than on working out any further finer points of phylogenetics. There are lessons here for many other debates about social kinds that are currently on people’s minds, some of which touch on questions of biology too. But I’ll leave those for another day.