By 1940, serious analysts — not hysterics, serious people — believed they were watching the permanent end of liberal democratic governance in Europe. Hannah Arendt was fleeing Germany. Churchill was widely considered a crank for believing liberal institutions could survive the decade. Serious historians were writing serious books about why the democratic experiment had failed as a form of governance and what would replace it. The evidence for this conclusion was not ambiguous: the interwar period had produced a systematic collapse of democratic governments across the continent, one after another, replaced by strongmen who — and this was the uncomfortable part — appeared to have the genuine support of their populations.
They were wrong. Not just about the specific outcome — about the nature of what they were watching. They were watching a pendulum at its furthest point of swing, which is exactly when a pendulum looks most like it has stopped moving.
That’s the model. Everything that follows is a variation on that observation.
What The Pendulum Model Actually Is
The model is not complicated. Large-scale social, political and religious change moves in oscillations rather than straight lines. The arc bends — eventually, unevenly, with substantial backward motion — in ways that only become readable as a pattern from historical distance. But the backward motion is not noise in the signal. It is the signal.
A few things this model is not, because the clarifications matter.
It is not Whig history — the comfortable Victorian idea that everything is inevitably getting better, that civilization is a river always flowing toward the sea. The pendulum model makes no such promise. It describes oscillation. Whether that oscillation settles toward genuine improvement over time is a separate empirical question, not a guarantee baked into the framework.
It is not Hegelian dialectic in simpler clothing. Hegel’s thesis-antithesis-synthesis carries an implicit destination — a meaning, a direction, a telos. The pendulum model is agnostic about destination. Oscillations can happen around a stable mean, a declining one, or an improving one. The model describes the motion, not the endpoint.
What it is: a tool for resisting the two most common analytical errors about historical change.
The first error is permanence projection — concluding that the current direction of travel is permanent. This is what the serious analysts of 1940 were doing. It is also what cheerful analysts of 1990 were doing when they declared liberal democracy the permanent final form of human governance. Both were looking at a pendulum in motion and concluding it had stopped.
The second error is completion bias — assuming that because an arc has been bending in a particular direction for a few decades, the work is done. That the bending is finished. That we can exhale.
Both errors make the same underlying mistake: taking a snapshot of a pendulum in motion and concluding it has stopped.
The Historical Record
The examples accumulate quickly once you stop looking for straight lines and start looking for oscillations.
Europe’s religious wars provide the longest arc. The Reformation began in 1517. The religious violence it unleashed didn’t reach its civilizational settlement — the Peace of Westphalia — until 1648. That’s 130 years of war, massacre, displacement and political disintegration before the continent produced the first modern framework for managing religious difference through sovereignty rather than suppression. Anyone living in 1600 had no structural reason for optimism. The settlement that eventually emerged was genuinely unimaginable from inside the violence.
The French Revolution is almost too good an example. 1789 to the stable Third Republic is roughly a century of oscillation: revolution, Terror, Napoleon, Restoration, 1830, 1848, Second Empire, Commune, Third Republic. The French did not arrive at modern republican governance in a straight line. They arrived at it through roughly a century of violent back-and-forth that would have looked, at any individual moment, like the definitive failure of the experiment. The pattern was only visible from the outside and after the fact.
Northern Ireland operated on a shorter timescale with the same logic. The Troubles began in 1969. The Good Friday Agreement came in 1998. The settlement didn’t emerge despite the violence — it emerged because both communities had reached the point where the costs of continuation exceeded any conceivable benefit of continuance. The pendulum had to reach its full extension before it could reverse. The people who lived through the extension were not consoled by the pattern.
South Africa in 1990 looked, from almost any vantage point in 1985, like an impossible outcome. Then it happened — not because the pressure suddenly became effective, but because the accumulated cost of resistance finally exceeded the accumulated cost of settlement for enough people with enough power to make a different choice.
Why Fifty Years Is Nothing
These examples feel surprising in retrospect because of a calibration problem. Human beings are built to think in lifetimes. A pendulum that completes its swing across four generations feels, to any individual watching it, like a straight line — either a permanent regression or a permanent advance.
Civilizational change operates on a different scale. Europe’s religious settlement took 130 years. The transition from feudal to industrial economic organization took roughly two centuries across the continent. The global reckoning with colonial extraction — the redrawing of political, economic and psychological frameworks built around European dominance — is arguably still in its early middle swing.
Within these timescales, fifty years is approximately the distance from one side of the room to the other. It is not nothing — the people living through any fifty-year period of the pendulum’s motion experience it as everything, because for them it is everything. But it is not the whole story, and mistaking it for the whole story is the source of most bad long-range analysis.
The near-term window always feels like the permanent state. It never is. This has been true consistently enough, across enough different civilizational contexts, that treating it as a reliable pattern rather than an occasional coincidence seems justified.
The Caveats Worth Taking Seriously
The pendulum model has two honest limits that deserve direct acknowledgment rather than footnote treatment.
Nuclear capability represents something genuinely new in the causal chain. Historically, the pendulum could reach its worst point — the worst wars, the worst oppression, the most complete collapses of civilizational order — and still reverse, because the underlying civilization survived to do the reversing. Nuclear weapons introduce, for the first time in recorded history, the genuine possibility of stopping the pendulum by removing the infrastructure that makes pendulum motion possible. This is not a metaphorical risk. It is an engineering problem with a known solution set that remains, as of this writing, unimplemented.
Climate disruption operates on a similar logic with a longer fuse. The civilizational pressures that flow from significant climate shift — migration at unprecedented scale, resource competition, agricultural stress in vulnerable regions — could compress the pendulum’s swing into shorter and more violent intervals while the background conditions steadily deteriorate. A pendulum doesn’t swing the same way in thinner air.
Neither of these is an argument against the model. They are arguments for taking seriously the near-term window — the dangerous extension point that the model identifies as the moment everything looks permanent — with more urgency than historical pattern alone would require.
The pattern offers long-run reasons for cautious optimism. The caveats offer short-run reasons for clear-eyed concern. Holding both simultaneously is the point.
The Honest Ending
The pendulum model is not optimistic. It is not pessimistic. It is an attempt to read the actual shape of historical change without the distortions introduced by either ideological comfort or ideological despair.
The historical pattern is genuinely encouraging over long time horizons. The moments that looked permanent — 1940, 1648, 1789, 1965, 1985 — turned out to be pendulums at their furthest extension rather than the final state of affairs. This matters. It is worth knowing. It changes how you read the present.
The same pattern says nothing reassuring about the people who happened to live at the furthest point of extension. The Thirty Years War killed a substantial portion of the German-speaking population of Central Europe. The survivors who lived to see Westphalia were not meaningfully consoled by knowing they had witnessed the pivot point of a civilizational arc.
The honest position, then, is this: the long arc appears to bend. The bending is not guaranteed by anything in the model, and two structural features of the current moment — nuclear proliferation and climate instability — could interrupt the pattern in ways that have no genuine historical precedent. The near-term window is dangerous. The near-term window has always looked dangerous. This time, some of the danger has a different character than it has had before.
Holding cautious long-term optimism alongside clear-eyed short-term concern, without collapsing into either reflexive hope or performative despair, turns out to be the actual intellectual project. It is slightly more demanding than the alternative — which is to pick a side, stop thinking, and be confident.
The bar for the alternative was not especially high to begin with.
Discover more from Long Thought Arcs
Subscribe to get the latest posts sent to your email.