The Peril of Cliodynamics

Can history predict the future? The Atlantic asks. Nope, says objective science — the attractive dream of Turchin’s predictive Cliodynamic theory has some good points, yet overall remains dangerously wrong.

Nashhinton’s reproduction of mathematical model output from Cliodynamics theory.

A true systems framework that incorporates historical processes, however, can offer a powerful vision of what futures are more or less likely than others. Cliodynamics lacks this, and so fails.

Every now and again a scholar comes up with an idea that is so close to being right yet is desperately wrong it just hurts.

Peter Turchin’s Cliodynamics is one of those.

Historians using mathematical models to describe the unfolding of history all face a serious problem: bad data.

Most historical data is junk on its own. And to do proper mathematical evaluation of any system, good data —and a granular, probably cellular agent-based model is essential.

Good objective data is exactly what few historians have. Especially ones who focus on Big History — the major global patterns shaping our lives.

Effective scientific modelers understand that the world is full of patterns produced by causal processes. The difficulty is understanding which causes are dominant in any given situation.

This is immensely difficult when trying to understand the past. And of the academic disciplines most focused on days of yore, history is ironically one of the least able to comprehend global-scale processes.

To Turchin’s credit, he tries — and I suspect he has suffered for it in the esteem of fellow historians. In that discipline it has become fashionable to view one’s particular area of study so unique it can’t be compared to others.

So I applaud him for trying to go big and global — but damn if it isn’t frustrating to see Turchin rely on shortcuts that might get his work past academic gatekeepers but are deeply flawed when applied to the real world.

Sorry historians, I respect and adore much of your work, but it just isn’t solid enough to rely on in a systematic way. And yes, your object of study is unique in certain ways — but in many others what you see is only the local realization of a global potential.

To do Big History correctly, you need a working model of how humans actually live at multiple scale levels — individual, household, community — and a way to connect all those individual units across space and time. Then you have to run simulations and find a way to validate them, ideally experimentally.

Only after that is done can you truly start incorporating real historical data to estimate how relationships between agents should change, if your validation is done right. If that then matches up reasonably well with the fragmentary and biased sample of hard evidence from archaeology, anthropology, and careful review of the extremely biased written sources available, you may be onto something.

Doing this at the global scale, incorporating billions of individuals and millions of communities across centuries? That’s a challenge on the scale of modeling global atmospheric patterns today and correctly predicting both the past and future weather everywhere.

Unless you have access to a supercomputer or can wait a really long time, and have the ability to unite the sum total of historical data available in a common framework, this ain’t happening. Which of course means that Cliodynamics is one of those paradigms that can never be fully evaluated and can’t offer specific predictions about the scary future it predicts.

One of the cleverest techniques professors hoping to advance a particular theory, paradigm, or discipline employ is to use complex mathematics. Economists absolutely adore doing this — upper division and graduate courses are full of calculus.

This gives these disciplines a sheen of hardness, of respectability, Economics does not in fact deserve. Because as any climate modeler will tell you, the quality of input determines the quality of the output. And human data, as anyone who works with surveys, interviews, or focus groups will attest, is always a bit wonky. There’s an art to getting quality information that is difficult to replicate.

Cliodynamics is essentially an over-fitted model bound to sketchy data. There simply aren’t any truly global datasets with the resolution required to model historical patterns beyond the last few hundred years. At least not to the level needed to prove why periods of dramatic collapse and chaos routinely occur.

Turchin is right, for the record, about the difficult coming decade. He is also right that what is happening in the world right now is the direct result of old historical processes converging to throw a global crisis.

But as a broken clock is right twice a day, and Newtonian mechanics is enough to get a rocket into orbit but not maintain a global positioning satellite system, so is Cliodynamics lacking something vital — the very thing needed to take control of the systemic challenges we face.

In systems theory there is an old saying that the first step to managing the system is comprehending it. And as Niklas Luhmann asserted in a deliberately ontological way, systems exist.

Humans live embedded in layers of systems, all structuring our lives in small but important ways. Language is a system, trade is a system, knowledge is a system too. And there are of course the physical, natural systems: the climate, oceans, biosphere, even the galaxy can be decribed as a system of interacting parts.

The reason Turchin’s Cliodyamics predicts a singular, mathematically-inescapable grim future is for the simple reason that it isn’t a true system. It is a backward-looking projection forward, using data skewed by the tendency for major conflicts to show up in the historical record while day to day processes that produced those conflicts go undiscovered, lost because they left little physical trace.

Humanity is barreling towards an apocalyptic future, but the secret is that this is always true. Many communities, societies, even nations have disappeared.

And rarely because they were outright destroyed — the Fall of Rome mythos looms too large in our historical memory, obscuring the simpler truth that people stopped believing a certain form of organization worked. They switched to a different one — often over a period spanning a few generations — and historians later went back and labelled this trajectory shift a grave crisis, pointing to outcome processes like wars and regime change rather than the subtler shifts taking place in everyday life.

White men are obsessed with the impending fall because, deep down, we recognize how messed up the past five centuries of European colonialism and white supremacy have really been. So many white men are desperate to find some objective reason for the impending end of their long raid on the rest of the world, something other than the simple truth that it was a great evil that made us the worst oppressors in planetary history and provoked justified resistance.

The reality is that life is driven by deep, community-level processes because of the limits of human cognition. We naturally divide into tribes because we can only handle so many social relationships at once. We personify other tribes in stereotypical ways that are usually wrong once we engage in close, sustained contact with them.

Society and social values build up from these granular relationships — and economies form alongside them. Political systems emerge to balance the other two and set global rules to keep people from fighting — boundaries then form between jurisdictions. And there exists everywhere the potential for inequality, which generates gradients of power and the incitement to form new tribes tied to claims of eliteness.

Conflict is always about resources, real or perceived. Conflicts can reorder the system, but in most cases are themselves driven by changes in the balance of power that aren’t obvious until expressed in a material form, like a battle or revolution.

A global systems framework will show a cyclic tendency for humans to fall into major conflicts, but Cliodynamics is simply wrong when it describes these as inevitable. Collapse and reorganization are essential phases of the adaptive cycle, allowing a network of agents to reboot their internal organization on a periodic basis. Bad conflicts emerge when these phases are put off for too long, leading to extreme resource shortages and rising stakes for all agents.

Inequality breeds conflict because the presence of extreme inequality in the form of authority, wealth, or status creates perverse incentives probably best understood by playing Grand Theft Auto V. It is possible to work essentially honest jobs in the game, but it’s boring.

Why? Because if there’s a jewelry store you can rob and walk away with millions of dollars, what fool would drive a taxi? The root of criminal behavior is natural desire coupled to an inability to correctly estimate the odds of getting away with breaking the law. People are simply bad at this — individuals are extremely poor judges of personal risk most of the time.

Cliodynamics does identify inequality as a major driver of conflict, what it fails to do is show why this is the case or how to mitigate the impacts. By predicting an inevitable collapse driven by iron laws of history, it renders itself ascientific and fundamentally useless.

It more or less puts management of social trajectories on the same level as trying to organize against the heat death of the universe — admirable, but impossible. This fatalistic approach is characteristic of a dying philosophical paradigm, and that’s the real reason The Atlantic pops this up on their front page whenever America feels especially dismal.

It goes right back the troubles with Liberalism I’m always going on about. When confronted with the harsh truth that not everyone can ever be educated to accept the same philosophical worldview, most priests of Liberalism more or less curl up into a ball and insist we’re all doomed.

There is only salvation through faith and hope, or damnation in the fires of populism.

Which, I’m sure, is exactly what wealthy Romans told their Christian slaves when the Germanic Tribes came migrating across the Rhine and Danube in too great of numbers to deter. There Is No Alternative — it’s the lash, or barbarism.

Most people, given this choice, choose barbarism — quite often they find it more civilized by far. What is now called Europe’s Dark Ages were in fact a period of dynamic change and growth after a major climate crisis that helped bring an end to the oppressive Western Roman regime.

Until, of course, it reformed itself and took over Christianity. But that’s for another essay :D

The truth is that the future is now wide open — the shape of what is to come is entirely up to those who will live it. The next decade will be rough because inequality is too extreme for it not to be.

But therein lies the hope for us all.

In every collapse, inequality is both a driver and, ultimately, dies. Once the higher-order system fails people band together at lower levels to take what they can.

The first step in building a better future is to reject all the pessimistic theories of the past. They were designed for a different phase in the eternal cycle — that is why the old white men can’t lead effectively anymore, everywhere accelerating their own demise.

A worse world is coming. A better world may emerge after. This is the Ragnarok of our time.

It’s always a bad sign when a scholar uses a figure from Greek mythology to describe their work. Especially when Clio and Cassandra are, in the brutal Southern European cultural tradition, ignored and hated.

My theory, were I to lay it out, would be called Mimirdynamics, after the Norse god who was beheaded in a spat but so good at prediction Odin preserved his head with magic and consults it often. The future is always constrained by the past, but the existence of a system does not mean you have to accept it — once understood, the door is opened to manage it.

Across the world, in countries ravaged by poverty and government neglect, community organizations are rising to fill the gap. Even as countries like Syria fall into civil war towns and districts self-organized as they can to keep life going.

The future is coming — and it will be built. What matters now is a laser-focus on resources. Who has them, who needs them, and how to bridge the gap — that is the path to a world that doesn’t suck.

As for history? Truth is, we don’t know a fraction of what really was. So while a useful cautionary tale, history is rarely a good guide.

Bottom line on Cliodynamics: interesting, but like so much western social science, fatally flawed.