Cliodynamics: Treating History As Science and Why That’s a Bad Idea

An example

Let’s focus on one main graph, which even Turchin devotes much time to in his blog.

The US political violence database shows distinct peaks with a fifty year cycle. According to this, the next peak in the pattern of violence will occur in 2020. The short cycle is so very striking that one may tend to forget the longer secular cycle, marked at the top. The second cycle is just starting. Turchin says that the long secular cycles are much more interesting than the shorter 50-year cycles, primarily because these longer cycles are better understood and also because these are generic features of society and the period of instability in them.

But aren’t these cycles present just in older societies? Do they percolate into modern societies? The answer is not known, but Turchin is keeping his eyes open.

Turchin is careful to point out that “we do not mean strict cycles in the mathematical sense”. He explains how non-linear cycles are involved and factors from both within the society and outside it affect the cycle.


The greatest criticism for this field, which might grow or not depending on the nature of future data, is specificity, or rather the lack of it. When you try to make history into a scientific subject, you have to incorporate two important ingredients: explanatory and predictive power. It must be able to explain the nature of political instability as part of a greater underlying framework or concept and/or must be able to predict some quantities based on that. At the moment, Cliodynamics fails at both ends – miserably!

Also, if you’re looking for patterns, you can always find them. Take a look at the graph on the US political instability presented above. What about the not-so-small peaks that occur between the two large peaks? Further, look at the peak at 1970. The height of this peak is actually lower than the intermediate peaks that occur at about 1890. This seems pretty much like pick and choose in order to find. How do you know which events to take and which to ignore?

Crucially, this type of pattern-based model-building leaves out very important major events. Would you expect this kind of model to predict an event like 9/11? Given that this is so very important an event in world history, won’t it be a shame if this much of non-linear mathematics cannot predict it?

What I am trying to drive home at is simply this: Cliodynamics and the patterns that it looks at may be simple flukes. You’d definitely find patterns if you look at a long record hard enough and historians have always known that. The use of non-linear mathematics and a large database of historical events ought to give back something more.

History isn’t repetitive in cycles; it might be illusory to think so. With the human society constantly changing and modernizing in a monotonic fashion, are we to really expect concrete cycles embedded? Then what explains the monotonic behaviour? If the cycles are not that stringent, why call them cycles at all?

Ironically, Cliodynamics might itself be like the cycle of history which predicts that these kinds of theories crop up, beautiful enough they are to attract a few proponents, and eventually die out. I have no idea about the periodicity of this kind of events, but it might just be a case of déjà vu.

Published by

Debjyoti Bardhan

Is a science geek, currently pursuing some sort of a degree (called a PhD) in Physics at TIFR, Mumbai. An enthusiastic but useless amateur photographer, his most favourite activity is simply lazing around. He is interested in all things interesting and scientific.