Published: February 16, 2017

What Do We Mean By Facts?

Ken Angielczyk, MacArthur Curator of Paleomammalogy and Section Head, Negaunee Integrative Research Center

The use of the words “fact”, “hypothesis”, and “theory” in science can be confusing, especially if conducting research isn’t your everyday job! But these terms have specific meanings, and they’re part of an important process that scientists use to gather information about the world around us.

First, some quick definitions—here’s how scientists at the Field Museum (and around the world) use these terms:

  • A fact is an indisputable observation of a natural or social phenomenon. We can see it directly and show it to others.
  • ​A hypothesis is an idea that we can test with further observations. We set out to gather evidence to see if our hypothesis is supported.​
  • ​A theory is a carefully constructed possible explanation for what we observe, drawing together many facts and hypotheses. Theories become stronger as they explain more facts. If a theory explains facts conclusively, it becomes accepted as the most likely explanation for the observed facts.  A man wearing a hat, field clothing, and a backpack standing next to a tall hill of exposed rockSo, how do all of these concepts come together in real science?

Associate curator Ken Angielczyk explains how paleontologists approach the task of figuring out when in Earth’s history a particular species went extinct:

First, we look to evidence in the fossil record, which helps us determine how long a species existed on Earth. This duration of time can be seen in a species' stratigraphic range—the thickness of sedimentary rocks in which a species is found. We determine stratigraphic range by searching for fossils of the species in measured sections of rock.

For example, we might find specimens of a certain species 5, 15, 18, 20, and 23 meters from the base of a measured section of rock. So here, the stratigraphic range for that species spans the interval between 5 and 23 meters. Each occurrence of the fossil is a fact: we don’t need additional information to state that these fossils were found in these locations. So far, no specimens of our species have been found above 23 meters.

The more places we look without finding the species above the level of the last fossil occurrence, the more support we have for our hypothesis that the species went extinct at about that time. Of course, it only takes the discovery of one fossil above our previous highest location to falsify our hypothesis. A famous example is the discovery of a living coelacanth species in 1938, which falsified the otherwise well-supported hypothesis that coelacanths went extinct near the end of the Cretaceous Period of Earth history (about 66 million years ago).

To understand what we mean by “theory” in science, let’s look at the theory that the impact of a huge meteorite caused the mass extinction at the end of the Cretaceous Period. This draws on several facts and hypotheses that work together. One piece of evidence is the huge impact crater of the correct age on the Yucatan Peninsula. The fossil record also shows the last occurrence of many species at this time, including all dinosaurs other than birds. Just above this layer, we see a sharp increase in the spores of ferns and fungi, which appear to represent disturbed plant communities that existed in the immediate aftermath of the impact.

Our theory provides a unified framework that explains all of these facts and hypotheses. But like anything in science, the theory is open to challenge if enough convincing evidence against it is found.


Ken Angielczyk
MacArthur Curator of Paleomammalogy and Section Head

I am a paleobiologist interested in three main topics: 1) understanding the broad implications of the paleobiology and paleoecology of extinct terrestrial vertebrates, particularly in relation to large scale problems such as the evolution of herbivory and the nature of the end-Permian mass extinction; 2) using quantitative methods to document and interpret morphological evolution in fossil and extant vertebrates; and 3) tropic network-based approaches to paleoecology. To address these problems, I integrate data from a variety of biological and geological disciplines including biostratigraphy, anatomy, phylogenetic systematics and comparative methods, functional morphology, geometric morphometrics, and paleoecology.

A list of my publications can be found here.

More information on some of my research projects and other topics can be found on the fossil non-mammalian synapsid page.

Most of my research in vertebrate paleobiology focuses on anomodont therapsids, an extinct clade of non-mammalian synapsids ("mammal-like reptiles") that was one of the most diverse and successful groups of Permian and Triassic herbivores. Much of my dissertation research concentrated on reconstructing a detailed morphology-based phylogeny for Permian members of the clade, as well as using this as a framework for studying anomodont biogeography, the evolution of the group's distinctive feeding system, and anomodont-based biostratigraphic schemes. My more recent research on the group includes: species-level taxonomy of taxa such as Dicynodon, Dicynodontoides, Diictodon, Oudenodon, and Tropidostoma; development of a higher-level taxonomy for anomodonts; testing whether anomodonts show morphological changes consistent with the hypothesis that end-Permian terrestrial vertebrate extinctions were caused by a rapid decline in atmospheric oxygen levels; descriptions of new or poorly-known anomodonts from Antarctica, Tanzania, and South Africa; and examination of the implications of high growth rates in anomodonts. Fieldwork is an important part of my paleontological research, and recent field areas include the Parnaíba Basin of Brazil, the Karoo Basin of South Africa, the Ruhuhu Basin of Tanzania, and the Luangwa Basin of Zambia. My collaborators and I have made important discoveries in the course of these field projects, including the first remains of dinocephalian synapsids from Tanzania and a dinosaur relative that implies that the two main lineages of archosaurs (one including crocodiles and their relatives and the other including birds and dinosaurs) were diversifying in the early Middle Triassic, only a few million years after the end-Permian extinction. Finally, the experience I have gained while studying Permian and Triassic terrestrial vertebrates forms the foundation for work I am now involved in using models of food webs to investigate how different kinds of biotic and abiotic perturbations could have caused extinctions in ancient communities.

Geometric morphometrics is the basis of most of my quantitative research on evolutionary morphology, and I have been using this technique to address several biological and paleontological questions. For example, I conducted a simulation-based study of how tectonic deformation influences our ability to extract biologically-relevant shape information from fossil specimens, and the effectiveness of different retrodeformation techniques. I also used the method to address taxonomic questions in biostratigraphically-important anomodont taxa, and I served as a co-advisor for a Ph.D. student at the University of Bristol who used geometric morphometrics and finite element analysis to examine the functional significance of skull shape variation in fossil and extant crocodiles. Focusing on more biological questions, I am currently working on a large geometric morphometric study of plastron shape in extant emydine turtles. To date, I have compiled a data set of over 1600 specimens belonging to nine species, and I am using these data to address causes of variation at both the intra- and interspecific level. Some of the main goals of the work are to examine whether plastron morphology reflects a phylogeographic signal identified using molecular data in Emys marmorata, whether the "miniaturized" turtles Glyptemys muhlenbergiiand Clemmys guttata have ontogenies that differ from those of their larger relatives, and how habitat preference, phylogeny, and shell kinesis affect shell morphology.

A collaborative project that began during my time as a postdoctoral researcher at the California Academy of Sciences involves using using models of trophic networks to examine how disturbances can spread through communities and cause extinctions. Our model is based on ecological principles, and some of the main data that we are using are a series of Permian and Triassic communities from the Karoo Basin of South Africa. Our research has already shown that the latest Permian Karoo community was susceptible to collapse brought on by primary producer disruption, and that the earliest Triassic Karoo community was very unstable. Presently we are investigating the mechanics that underlie this instability, and we're planning to investigate how the perturbation resistance of communities as changed over time. We've also experimented with ways to use the model to estimate the magnitude and type of disruptions needed to cause observed extinction levels during the end-Permian extinction event in the Karoo. Then there's the research project I've been working on almost my whole life.

Morphology and the stratigraphic occurrences of fossil organisms provide distinct, but complementary information about evolutionary history. Therefore, it is important to consider both sources of information when reconstructing the phylogenetic relationships of organisms with a fossil record, and I am interested how these data sources can be used together in this process. In my empirical work on anomodont phylogeny, I have consistently examined the fit of my morphology-based phylogenetic hypotheses to the fossil record because simulation studies suggest that phylogenies which fit the record well are more likely to be correct. More theoretically, I developed a character-based approach to measuring the fit of phylogenies to the fossil record. I also have shown that measurements of the fit of phylogenetic hypotheses to the fossil record can provide insight into when the direct inclusion of stratigraphic data in the tree reconstruction process results in more accurate hypotheses. Most recently, I co-advised two masters students at the University of Bristol who are examined how our ability to accurately reconstruct a clade's phylogeny changes over the course of the clade's history.