What, if anything, is a Tyrannosaurus rex?

 Leonard Finkelman writes...

Could you share a kiss with a Tyrannosaurus rex? (Let's set aside the question of whether or not you should; ethics is not my purview.)

This is the most contentious scientific debate of our time, presuming that "our time" is limited to the last two weeks. University of Toronto paleontologist Robert Reisz argues that dinosaurs likely sported fleshy, scaly lips covering their teeth; his research has, predictably, sent science journalists into a tizzy over the potential for T. rex to pucker up. So: did T. rex have lips?

AMNH 5027, courtesy of Wikimedia Commons. At least you wouldn't have to worry about tongue on the first date.

AMNH 5027, courtesy of Wikimedia Commons. At least you wouldn't have to worry about tongue on the first date.

I'm sorry (i.e., not sorry) to give precisely the sort of answer that a philosophy professor gives, but it depends. In this case, however, it doesn't depend on fossil evidence or on the biology of lips (although it's worth noting that reptile "lips" are physiologically quite different from mammalian lips). It depends on what a T. rex really is.

Following up on my last post, I think we have good reason to say that "Tyrannosaurus rex" isn't really the name of an animal; at least, we have as much reason to say that as we have to say that footprints aren't animals. My argument will cover some of the same ground laid out last time; by the end, however, I intend to explore new and treacherous conceptual territory.

What's in a name?

The question is clichéd, yes, but it's important to remember that the difference between speaking truly and speaking falsely might lie in the use of a name. This is why logicians and philosophers of language are so interested in names and naming.

When I took science journalists to task last month for sensationalizing the discovery of a Siberian "unicorn" skull, it was because of their too-casual use of the term "unicorn." That term has a variety of referents, including cryptids and mythological creatures, across a number of cultures. Elasmotherium sibiricum might be one of those referents, but it might not, and we currently have no way to tell whether or not it's one of the things the original coiners of the term "unicorn" had in mind. To call E. sibiricum a "real unicorn," then, is to presume knowledge that we don't have.

My argument depended on the causal theory of reference normally attributed to Saul Kripke (1980). According to that theory, the correct use of a name derives from a causal connection between the speaker's use of the name and the name's original coinage. As an example, consider my assertion that Barnum Brown discovered my favorite fossil, AMNH 5027. A partial sequence of events could be rendered as follows:

My assertion is true because the event that I've described and the origin of the name I used in the description are both links in the same chain of causes and effects. By contrast, consider the sequence of events involved in the assertion that E. sibiricum was a unicorn:

The question mark indicates an ambiguous connection: the journalist's understanding of the term "unicorn" might or might not be part of the sequence of events that began with the death of an E. sibiricum. Since the two events might be connected by correlation rather than by causation, the journalist's use of the term must be incorrect and her claim is therefore false.

As developed by Putnam (1975), the causal theory of reference is useful for philosophers of science who seek to explain how we can maintain reference to the same theoretical objects despite changes in theory. In particular, both Kripke and Putnam thought that the causal theory described naming conventions in the life sciences. Even though Kripke and Putnam have fallen out of favor with many philosophers of biology, I think that paleontologists share the causal theorists' concerns in naming trace fossils.

Trace fossil nomenclature

A distinction between body fossils and trace fossils runs deep in biological taxonomy. Body fossils are the mineralized remains of an organism itself; trace fossils are the remains of the organism's behavior. This distinction serves as the foundation for separate naming systems known as parataxonomies. In his introductory text about trace fossils, Bromley asserts:

'Which animal produced that trace fossil?' remains typical of the first question put to an ichnological sample [i.e., a fossil track]. This is not the right way of addressing trace fossils; they tell us much more interesting things than that, while rarely divulging the nature of their architects. (2012, p. 155).

He cites (among others) four reasons for the paleontologist's obligatory disinterest in connecting traces fossils with the animals that made them: first, one organism may produce multiple different kinds of trace; second, the same trace may be preserved differently in different substrates; third, different organisms may produce identical traces; fourth, the body fossils of a trace maker might not be preserved themselves (pp. 155-161). To reiterate an example from my last post: the state fossil of Connecticut, Eubrontes giganteus, is a kind of fossil track commonly attributed to a theropod dinosaur "similar to" Dilophosaurus wetherilli, but D. wetherilli itself is a candidate for the tracemaker. Furthermore, either of those theropods might also be responsible for a similar kind of fossil track, Grallator. We can visualize the whole mishegoss as follows:

The information preserved by body fossils and by trace fossils is insufficient to demonstrate a necessary connection between any of the taxa shown in that diagram, and so we cannot say that D. wetherilli is the same animal as the maker of E. giganteus--for the same reasons that a causal theorist would insist that it must be false to say "The maker of E. giganteus tracks had two crests on its head." Taxonomists therefore keep the names of trace fossils separate from the names of body fossils: the names pick out different kinds of things (Sarjeant 1990).

body fossils and paleontological inference

BHI 6230, nicknamed "Wyrex," from www.worldfossilsociety.org, because we've gone too long without a picture of a dinosaur.

BHI 6230, nicknamed "Wyrex," from www.worldfossilsociety.org, because we've gone too long without a picture of a dinosaur.

Paleontologists have generally considered the difference between body and trace fossils to be "profound" and so try to keep a clear line drawn between the two (Pemberton & Frey 1982). Nevertheless, there are good reasons to blur the line and perhaps even to step over it entirely.

In their seminal paper outlining the need for and perils of trace fossil parataxonomy, Pemberton & Frey show how 'the two categories may approach or overlap one another in subtle ways' (p. 845). For example, the impression of a snail shell is considered a body fossil by the International Code of Zoological Nomenclature, even though the impression's causal history bears significant resemblance to body-sized burrows that are considered trace fossils. Furthermore, some animal taxa--bryozoans in particular--use their burrows as functional skeletons. As burrows, these structures are considered trace fossils; however, taxonomists may nevertheless specify the structures as traits in zoological classification. In short: the distinction between "organism" and "organism's behavior" is almost certainly a fuzzy one.

The problem is compounded when we consider what body fossils really are. The American Museum of Natural History tells visitors that 'the "dinosaur bones" you see on display at the Museum aren't really bones at all'; body fossils are literally bone-shaped rocks formed by the gradual replacement of organic molecules with inorganic minerals. In this sense, the difference between body fossils and trace fossils is one of degree: neither are really the remains of the organism itself, but body fossils are a more proximate effect of an organism's life activities.

All of the foregoing strikes me as reasonable enough, but it implies what might be a prima facie unreasonable conclusion: that "Tyrannosaurus rex" is not the name of an animal, and that the animal causally responsible for T. rex fossils has yet to be named.

We've already established that taxonomists have good reasons for keeping distinct the nomenclatures associated with body fossils and with trace fossils--it is a mistake to say, for example, "E. giganteus is a relative of D. wetherilli," because those taxon names refer to different kinds of things. We've also established that body fossils are not substantially different from trace fossils. By the very same logic, then, it should be a mistake to say that "Struthio camelus is more closely related to T. rex than either is to Heloderma suspectrum," because the name "T. rex" refers to rocks and the other two names refer to animals.

To put it another way, let's consider the sequence of events that led to the claim that T. rex had lips:

The question at hand is whether or not the yellow arrow represents a necessary connection between the top row and the bottom. Do we hold beliefs about the fossils because of the animal? Does the fossil evidence necessarily depend on the way the animal lived?

At least one example seems to indicate that the answer to both questions is "no." Fans of paleoart have gotten accustomed to feathered tyrannosaurs. But we actually have good reason to think T. rex didn't have feathers: paleontologists found the specimen nicknamed "Wyrex," discovered in 2002, preserved with scaly (i.e., non-feathered) skin impressions. The evidence for feathers on T. rex is circumstantial, based on its phylogenetic relation to dinosaurs like Yutyrannus huali. Which is it? Given the fossil evidence, paleontologists could justify belief that T. rex went naked--but the evidence in this case might point towards a false conclusion. In other words, fossil evidence doesn't necessarily depend on the way that the animal actually lived.

The name "T. rex" was bestowed upon a fossil that bears only an ambiguous connection--correlative or causative--with some animal in the past. The same is true of E. giganteus. In that latter case we keep the trace name distinct from the names of animals, for the same reasons that we should be careful not to conflate E. sibiricum with unicorns. It follows, then, that "T. rex" should be a parataxonomic name. 


One might deny the causal theory of reference, of course; many philosophers do. But it seems that taxonomists--and particularly those interested in trace fossils--embrace the theory. If so, then by taxonomists' own lights we should consider "T. rex" to be the name of a group of bone-shaped fossils rather than the name of an animal species. It follows that "T. rex had lips" would be necessarily false because bone-shaped rocks don't have lips. Some extinct animal that left behind the bone-shaped rocks might have had lips, but I have no idea what to call that thing. (Might I suggest Finkelsaurus rex?)

Whatever we call it, though, I maintain hugging that animal would have been a better idea than kissing it.


  • Bertling, M. (2007). What’s in a name? Nomenclature, systematics, ichnotaxonomy. In Miller, W. (ed.), Trace Fossils: Concepts, Problems, Prospects, pp. 81-91.
  • Bromley, R. G. (2012). Trace Fossils: Biology, Taphonomy, and Applications (2nd edition). New York: Taylor & Francis.
  • Kripke, S. (1980). Naming and Necessity. Cambridge, MA: Harvard University Press.
  • Pemberton, S. G. and Frey, R. W. (1982). Trace fossil nomenclature and the Planolites-Palaeophycus dilemma. Journal of Paleontology 56:4, pp. 843-881.
  • Putnam, H. (1975). Mind, Language, and Reality: Philosophical Papers, vol. 2. Cambridge: Cambridge University Press.
  • Sarjeant, W.A.S. (1990). A name for the trace of an act: approaches to the nomenclature and classification of fossil vertebrate footprints. In Carpenter, K. & Currie, P.J., Dinosaur Systematics: Approaches and Perspectives. New York, NY: Cambridge University Press, pp. 299-314.