History as Science: How Complexity Thinking Is Transforming Foresight

Complexity science is poised to reshape the way we do foresight. For decades, futurists have relied mainly on qualitative tools such as scenario planning to help organisations think strategically and prepare for uncertain futures. But as we begin to understand how complex systems behave—how patterns emerge, collapse, and re-form—we may be standing on the edge of a profound shift in how we think about the future.

Photo Credit: Mariola Grobelska/Unsplash

Complexity science seeks to understand the behaviour of complex systems—physical, biological, and social. It has already revolutionised physics and biology, and in recent years has begun to make inroads into the social sciences. The economist Eric Beinhocker and the physicist Doyne Farmer have brought complexity thinking into economics, developing models that simulate how markets behave and respond to interventions. Their work, known as Complexity Economics, has grown influential enough that several central banks now use such models to test policy scenarios.

These advances already challenge the assumption—common among futurists—that the future cannot be "predicted" in any meaningful sense. Yet the work of Peter Turchin—a mathematical biologist turned historian—may ultimately push this challenge much further.

Turchin’s career pivot from ecology to history is remarkable in itself. Historians traditionally resist attempts to identify universal patterns in human affairs. History, they insist, is “one damn thing after another”—a succession of unique, context-dependent events. While history offers lessons in human nature and experience, the conventional view holds that it reveals no general laws from which the future could be inferred.

As we know from other fields, it is often outsiders who challenge prevailing dogmas—and Peter Turchin fits that mold. Drawing on his background in mathematical modelling, he developed cliodynamics—named after Clio, the muse of history, and “dynamics,” the study of change. Cliodynamics treats history as data: it collects vast records of past events, encodes them systematically, and looks for the deep forces that drive social change. As Turchin writes in End Times:

Pundits and politicians often invoke ‘lessons of history.’ The problem is that the historical record is so rich that anyone can cherry-pick examples to support whichever side of a policy debate they favour. Clearly, inference from such cherry-picked examples is not the way to go. Cliodynamics is different. It uses the methods of data science, treating the historical record—compiled by generations of historians—as Big Data. It employs mathematical models to trace the intricate web of interactions among the many ‘moving parts’ of complex social systems like our own societies. Most importantly, cliodynamics uses the scientific method, in which alternative theories are subject to empirical testing with data.

Using historical data on roughly one hundred crises spanning several millennia, Turchin and his collaborators identified a recurring pattern across complex societies: long stretches of internal stability and prosperity—what he calls integrative phases—followed by periods of growing inequality, unrest, and political violence—disintegrative phases. These cycles are driven not by external shocks but by internal dynamics, and their length varies with each society’s characteristics. For the disintegrative phase to end, the structural conditions that caused it must be reversed—either through bloody conflict or through deep, systemic reforms.

In End Times, Turchin applies this model to contemporary America. He argues that the United States has entered a disintegrative phase, driven by three interlocking forces—structural conditions that precede all major social crises:

1. Popular immiseration: declining real wages, falling social mobility, and worsening well-being.

2. Overproduction of elites: an expanding class of ambitious, educated individuals competing for too few elite positions, creating intra-elite conflict.

3. Fiscal distress of the state: the growing inability of government revenues to sustain expenditures, eroding institutional capacity.

Among these, Turchin sees elite overproduction as the most destabilising. When large numbers of ambitious young people—“elite aspirants”—find themselves locked out of the system, they often become counter-elites, seeking to overthrow the existing order. History shows that such moments often precede revolutions and civil strife.

In addition to structural-demographic factors, Turchin’s analysis highlights how ideology evolves during pre-crisis periods. The consensus that once held elites together breaks down. Competing belief systems rise—some radical, promising to build a new and better society; others traditionalist, promising to restore a lost golden age. In both cases, the result is polarisation. Divisive, sectarian ideologies eclipse unifying ones. The public comes to believe the nation is on the wrong path, and calls for “justice” or “restoration” intensify. These are the social signatures of what Turchin calls an age of discord.

Applying this framework, Turchin predicted as early as 2010 that the United States was approaching a period of rising polarisation and outbreaks of political violence—a forecast that proved strikingly accurate by 2020. Recent polling in the United States suggests that the radicalisation Turchin described continues to deepen, with growing numbers of citizens viewing political violence as a legitimate response to perceived national decline. Whether this will lead to open conflict depends less on ideology itself than on the structural conditions that determine how revolts spread.

Turchin notes that political violence behaves like other self-organising natural phenomena—wildfires, earthquakes, or epidemics. In each case, small triggers can ignite large events if conditions are right. The spread of revolt, like the spread of fire, depends on how much combustible material (radical energy) exists and how easily it connects (through networks). As in nature, these events follow power-law dynamics: small disturbances are common, but large, system-wide crises—though rare—are inevitable. As the historian John Lewis Gaddis (one of the few in his field to engage deeply with complexity science) points out, even though we can calculate the relationship between frequency and intensity of crises, we cannot anticipate when a particular situation will reach its point of maximum intensity—when tension turns to revolution. The intersecting variables can only be reconstructed in retrospect. What we can do, however, is infer possible trajectories, identify likely outcomes, and design interventions that reduce systemic stress.

Together, these insights point to a profound question: if we can now model the social forces that drive instability, what does that mean for how we think about the future? This is where Turchin’s work intersects with the practice of foresight. Futurists have long emphasised that the purpose of foresight is not prediction but preparation—to build mental maps that help decision-makers navigate uncertainty. Yet as complexity science advances, we now have tools to simulate the behaviour of large social systems, track early warning indicators, and test “what-if” interventions in silico. These tools don’t make the future predictable in a deterministic sense, but they do allow us to anticipate risks and test responses more rigorously than before.

Turchin himself cautions that “prediction is overrated.” When he speaks of prediction, he refers to scientific prediction—the use of modelling to test hypotheses and refine theory—rather than prophecy. In this sense, history as a science is less about forecasting the distant future and more about identifying how crises emerge and what kinds of interventions might alter their course. The goal, he insists, is not to foresee the future perfectly but to shape it.

“What we should really strive for is the ability to bring about desirable outcomes and avoid unwanted ones. What’s the point of predicting the future if it’s bleak and we can’t change it?”

That sentiment aligns closely with the ethos of foresight. Both fields seek not prediction, but agency—the capacity to understand the forces shaping our world and to act before crises erupt.

As futurists, we should pay close attention to these developments. Just as historians have largely resisted quantitative methods, foresight practitioners may be tempted to view models and simulations as overly deterministic. But complexity science doesn’t replace qualitative foresight—it enriches it. By integrating data-driven pattern recognition with imaginative scenario building, we can create richer, more grounded maps of possibility.

This merging of foresight and complexity science offers not prediction for its own sake, but understanding—the kind that makes informed, proactive action possible. The next evolution of foresight may therefore be complexity-informed foresight: blending sensemaking, narrative insight, and computational modelling to better anticipate—and perhaps avert—the crises that shape the course of civilisations.

---

In End Times, Peter Turchin brings his structural-demographic theory to bear on the present, tracing how the forces that once tore apart ancient empires now shape the United States. It’s not a book of doom but of diagnosis—a map of the pressures that drive societies toward crisis and the reforms that might pull them back. For anyone interested in how complexity thinking can illuminate history—and perhaps the future—it’s an essential read.

---