A General Theory of History – A bleg

Doctor Labyrinth, like most people who read a great deal and who have too much time on their hands, had become convinced that our civilization was going the way of Rome. He saw, I think , the same cracks forming that had sundered the ancient world, the world of Greece and Rome; and it was his conviction that presently our world, our society, would pass away as theirs did, and a period of darkness would follow.

Phillip K. Dick, The Preserving Machine

Imagine creating a scientific model of human society. A science to explain and predict the great movements of humanity through history. Tools that model societies, civilisations and empires rising, declining and falling. A light to shine on the paths taken through history and into the future . A way to uncover the mechanics of the most complex system we know of in the universe.

This idea, even the sniff of the idea, is amazingly enticing. I, like Paul Krugman, was brought into economics in a large part by Asimov’s Foundation trilogy– a series that revolves around the application of such a science. Economics seemed the only discipline that attempted anything remotely close to the dream. There are hints of that dream in my own honours research and first posts here.

But it is just a dream, and intellectual history is littered with failed attempts at such a model or general theory. These failures are still fascinating nonetheless. It’s a minor hobby of mine looking at these and having thought experiments about if such a discipline could be done, how would be done. To begin with I’m cataloguing attempts, which is what this bleg is about. I’m looking for suggestions about past attempts I’m not aware of. I’m trying to categorise the assumptions in each so we can consider their strengths and flaws. I have a short list of past attempts and a draft taxonomy of them, presented as a horrible Venn diagram over the fold.  

First up though I’ll just clarify what I am talking about here. It’s not just historicism. Things need not be deterministic, though they can be. They can also be probabilistic or allow for free will. It’s also not just world histories explained by a few factors. There needs to be some predictive capacity to a prospective general theory so that it could be applied to future societies, past societies on which the theory was not based or even historical counterfactuals. This excludes figures like Jared Diamond (Guns, Germs and Steel) or Gregory Clarke (A Farewell to Alms).

Here’s my ugly diagram.

The taxonomy:

The first two categories have different assumptions about long run behaviour.

Cyclical – There are repeated patterns in history that will keep repeating indefinitely. This is a very ancient concept with a long pedigree. The Mandate of Heaven in China has been very powerful in real political situations for thousands of years. The Islamic scholar Ibn Khaldun also described the fortunes of dynasties in a cyclical fashion. For the past century the idea that Western Society is echoing Roman decadence has itself been echoed so much as to be a punchline

Linear – Often universalistically, we see human society moving from one stage to another, from a distinct beginning to a distinct end. This includes Marxism, which is the second most influential attempt here after the Mandate of Heaven, and the most influential that envisioned itself as science, as well as other Hegelian descendants down to Fukuyama. I have also included Apocalyptism. This includes both the Abrahamic religions (especially Dispensationalism) with a series of stages between genesis and the end of the divine plan, and also visions of a distinct end outside this tradition, such as the Norse Ragnorok.

I considered adding a category called Equilibrium which would model society as constantly returning to a certain state after periodic shocks. This category would only have Malthus in it, and he did not seem to have requisite vastness of scope.

The next categories deal with the unit that is being studied in each instance.

States/Empires – These are focused almost entirely on the state and formal political institutions such as kingdoms, dynasties and empires. It’s tempting to see this see this as the result of an appreciation for Gibbon unmitigated by Pirenne’s broader conception of what constituted Rome, but Khaldun was doing it as well. It’s probably just that empires leave records (written and physical) and are easily identifiable.

Societies/Civilisations. – The unit of study captures a broader swathe of institutions than those of the formal state, including the entire civil society/art/culture etc. It also includes William Hardy Wilson, whom saw only aesthetics as important. They still attempt to create distinct units of these cultures though. Spengler sits between these two because whilst he professes to be describing the latter, I feel his own taxonomy of civilisations betrays a great reliance on the existence of states to define a civilisation.

Horrendously Large Populations – OK, this hasn’t actually been tried, but in the Foundation books Asimov makes it clear that the science is only applicable to very very large populations, far greater than the current number of humans. Populations in aggregate can be predicted just like the movements of fluids despite the unpredictability of individual people or molecules. I include this because it’s a perspective that must be considered.

Agent based modelling – Largely the opposite of Asimov. The behaviour of representative individuals is modelled, but this incorporates their reactions to other agents whom are reacting to them. This is all computed to discover the relevant macro behaviour. Essentially Schelling brought to the nth degree. I’ve named this Santa Feism after the Santa Fe Institute.

Lastly I’ve included the category “Religious” for those models that rely on a cosmic element.

I considered including materialist (Marx)/conceptual (Hegel, Dispensationalism) categories but the distinction does not interest me too much at this stage.

I can also see the potential for a three way split between chaotic determinism, path dependence and the law of the very large numbers. The first would stress the importance of initial conditions, the second divergent paths and chance and the third the relative unimportance of individual events and people compared to long run outcomes. These debates are present in historiography, but aren’t much use classifying the general theories I’ve found.

I’m also considering the addition of New Economic Geography, since it’s providence from Asimov via Krugman was relatively clear.

I’m eager to hear of other attempted methods though (maybe there’s an evolutionary approach more developed than crude social Darwinism), if anyone has reached the end of this rather perculiar post that is.