Cliodynamics explains history scientifically

Author: Daniel Mullins

Seshat overall coordinator Peter Turchin explains that Seshat: Global History Databank is an integral part of an exciting new scientific approach to historical research—Cliodynamics.  The job of Cliodynamics, according to Peter, is to use the scientific method to produce the data necessary to empirically test competing theories. This is a two-step process:

Step 1: Translate rival verbal theories about historical processes into mathematical models and extract quantitative predictions from these models.

Step 2: Test these quantitative predictions using rich datasets like Seshat: Global History Databank.

For more videos, see our Vimeo page.

To explore relevant publications, see Cliodynamics: The Journal of Quantitative History and Cultural Evolution.


Peter Turchin: History is a wonderful discipline. I love reading history books. That’s why I got into history. Traditional historians have done a great job describing past human societies and how they changed. What they haven’t done so well is explaining why human societies changed in some particular ways and not others. Well, actually, historians are proposing explanations for why things happened in history. The problem is that there are not too few explanations. In some sense, there are too many. Let’s take the example of the fall of the Roman Empire. One German historian actually counted how many explanations different historians proposed for the fall of the Roman Empire. He found there were 210. So, clearly, not all of them can be correct. What do we need? We need some kind of mechanism which will decide which explanations are better and which are worse. And we do that by empirically testing theories.

Cliodynamics is history as science as opposed to traditional history as humanities. Cliodynamic’s job is to reject bad explanations in favour of good explanations. When Cliodynamics is done, we should see a cemetery of bad theories and only better theories will be still standing. How do we go about it? What we do is we through the scientific method in its all full majesty at questions in history.

So, how do we actually test theories empirically against each other? Its not quite as simple as you might think. History is complex and human societies are dynamical systems with many interacting parts. A naked human brain is not very good at extracting predictions from theories when they deal with such complicated objects. So the first step we have to do is we have to translate the verbal theories into mathematical models. Mathematical models ensure that any predictions that come from them will really fall from the assumptions that you make to start with. We take mathematical models and we ran them forward and we extract quantitative predictions about them about some aspect of history. We do it for rival theories and then we see where the predictions from rival theories disagree. This is the place we need to look to be able to determine which of the theories is better and which is worse. So, to recap. The process involves two steps. In the first step we translate verbal theories into mathematical models, then we extract predictions from the models and in the second step we test these predictions with rich data sets like Seshat: The Global History Databank.


Notes for Editors: 

  • For further information contact Jill Levine via email or see
  • Seshat: Global History Databank is a large, international, multidisciplinary team of evolutionary scientists, historians, anthropologists, archaeologists, economists, and other social scientists. Our team includes scholars from various backgrounds, policy makers, and enthusiastic volunteers. Seshat is governed by an editorial board, who oversee work done by postdoctoral researchers, collaborators and consultants, and research assistants all over the world.
  • Cite this page: “Peter Turchin explains how historical processes can be explained using Cliodynamics”

Name *
Email *