Why Vertebrate Paleontologists Don’t Believe Anything

How often have you heard or read the phrase, “Paleontologists believe …”? This can occur when a new dinosaur is unveiled to the public. There will be some news coverage that says something like, “the study’s scientists believe that this dinosaur was an herbivore.” Although this is how vertebrate paleontology and other natural sciences are often described in popular media, in reality none of us write scientific papers in which we describe our findings as supporting what we believe or even that we’ve proven this or that. This is because vertebrate paleontologists, like all natural scientists, speak the language of probability. The trouble is, this is not the typical way we speak in day to day conversations.

For vertebrate paleontologists, understanding and interpreting anatomy is a must. Many of us often teach human and animal anatomy, and we’ve learned to read the skeleton’s shape to best infer what was probably there when our fossil animals were alive. It is important to emphasize that when we infer something, it is not simply a guess or what we choose to believe. Instead, we are constructing a hypothesis. In common usage, a hypothesis is considered an “educated guess,” but scientifically a hypothesis is a testable prediction that can be falsified. For example, tooth shape is strongly correlated with diet. So, if I am fortunate enough to discover a new dinosaur and it has a skull, I can predict that, if it is an herbivore, it will have teeth that resemble those of herbivorous reptiles and mammals, not those of flesh eaters. I can test my hypothesis through careful anatomical comparison of my dinosaur’s teeth with known herbivores. The hypothesis would be falsified should I find tooth shapes similar to those of known carnivores. The reality is, if I am telling you as a vertebrate paleontologist that a particular animal is an herbivore, there is probably some data to back me up. I don’t believe the dinosaur is an herbivore, I have data that support my hypothesis that it ate plants.

The skeleton and the rocks associated with it are the vertebrate paleontologist’s dataset. In a perfect world, even with the loss of soft tissues, these data would be enough to make numerous confident inferences about our vertebrate predecessors. But it is not a perfect world. In vertebrate paleontology, the fossils that come to us from deep time represent but a tiny proportion of all the living things that have come and gone on Earth. Therefore, we are dealing with a sample, and typically a small sample. Given how rare it is for a vertebrate animal to be fossilized and be noticed in time to rescue it from weathering, we have to be content with what nature has relinquished to us. So if you find a single dinosaur skeleton, for example, you are stuck with one individual that may or may not represent the typical member of that long lost population. Even in a healthy population, you can get quite a bit of variation in the skeleton from individual to individual. What if the individual dinosaur you’ve had the fortune to unearth was the largest or smallest individual in the population, or what if the dinosaur was very old or quite young? Sometimes these are very difficult things to tell, especially from a single skeleton that is often incomplete to boot. So, as a vertebrate paleontologist, you have to be comfortable with a healthy dose of uncertainty and a willingness to wait for more or better data to arrive.

The public often wants to know what a particular dinosaur or other fossil vertebrate “did” when it was alive. Could it run? Did it live in groups or was it solitary? Could it climb trees or was it stranded on terra firma? How confident we can be in our inferences depends on how many specimens you have and what sort of skeletal traits are present. If you have a lone individual, it would be extremely difficult to infer group behavior, for example. Often, we make a testable prediction of certain behaviors based on living relatives of the animal in question. For example, the living relatives of non-avian dinosaurs are crocodylians and birds. Therefore, inferences about dinosaur movement or feeding often involve matching bone shape to function in to living archosaurs (crocodylians and birds), and applying what is known about these form-function relationships to non-avian dinosaurs. Such inferences are the most probable given that non-avian dinosaurs are archosaurs. But as we go further away from archosaurs into the anatomy of other reptiles and then into mammals, we can be less confident about our inferences. Without extraordinary evidence, we have to be more circumspect about what we infer.

Granted, there are some spectacular instances where behaviors are preserved in the fossil record. A key example of this is the inferred brooding behavior of Citipati, an oviraptor (Prum, 2008). Several individuals of this dinosaur have been found preserved squatting on their eggs with their arms outstretched over their clutch much as many modern birds do when incubating their young. Although we can’t know the precise behavioral repertoire of this dinosaur, the inference that Citipati was brooding its eggs is strongly supported. Several predatory dinosaurs have been shown to be associated with clutches of eggs (Varricchio et al., 2008), but does this mean that all predatory dinosaurs behaved this way? That is would be a weakly supported inference that would require more data to support or falsify. Until then, we have to be patient not knowing.

In the end, vertebrate paleontologists like all natural scientists deal in probability. We don’t believe something to be true, we infer the likelihood of a particular aspect of vertebrate biology from the fossils based on data. What I do believe is that the public is best served if we use the language of probability to describe our findings. Popular science programs and media stories on vertebrate paleontology would better convey the science by substituting “infer” in place of “belief.”

Submitted by Matthew F. Bonnan, Stockton University

References Cited

Prum, R. O. 2008. Evolution: Who’s Your Daddy? Science 322:1799–1800.
Varricchio, D. J., J. R. Moore, G. M. Erickson, M. A. Norell, F. D. Jackson, and J. J. Borkowski. 2008. Avian Paternal Care Had Dinosaur Origin. Science 322:1826–1828.
Posted: 10/24/2016 12:11:15 AM by matthewbonnanadmin | with 0 comments
Blog post currently doesn't have any comments.
 Security code