The Same 250 Years, Whatever the Weapons

Assyria ran on bronze and lasted 247 years. Britain ran on steam and telegraph and got about 250. Whatever imposes the limit does not care what you have invented.

The Turning of Days, Chapter 4: Glubb's Fate of Empires

In 1978 a retired British general published twenty-six pages that should have rearranged how the West understood itself, and did not, because almost nobody who needed to read it ever did.

Sir John Glubb was not an academic, which is probably why he noticed. He had spent thirty-six years in the Middle East, commanded the Arab Legion, spoke fluent Arabic and had lived among the Bedouin with a depth no visiting scholar accumulates. He read history widely, and not only his own side's.

He took eleven great empires and measured how long each lasted from rise to the end of dominance. Assyria. Persia. Alexander's successors. The Roman Republic and then the Empire. The Arab Empire. The Mamluks. The Ottomans. Spain. Romanov Russia. Britain.

The average came out at roughly 250 years. Around ten generations.


Now hold the thing that should stop you.

Those empires span three thousand years. They sit on different continents. They worship different gods, speak unrelated languages, and fight with entirely different technology. Assyria ran on bronze. The Ottomans ran on gunpowder. Britain ran on steam, steel and telegraph cable.

There is no material reason why bronze and steam should produce the same number.

If the limiting factor on an empire were technological, the numbers would diverge wildly, and the direction would be obvious: better logistics, better communications, better weapons, longer life. A state that can move an order from London to Delhi in minutes ought to outlast one where the fastest message travels at the speed of a horse.

It does not. Britain got about 250 years. Assyria got about 247.

Whatever is imposing the limit does not care what you have invented. Which means it is not in the equipment. It is in us.


Glubb was careful about the number and I want to be too. The dates are arguable. Empires linger in name after their power has gone. Some were struck down early by something external. He said as much himself, and called the pattern remarkable rather than exact.

The structure matters more than the arithmetic anyway, and the structure is where it gets uncomfortable, because he found the same seven phases in the same order every time.

Pioneers, erupting out of poverty and hard conditions with more courage than equipment. Conquest, where the army is the central institution and a talented nobody can rise on ability alone. Commerce, where energy shifts from taking territory to working it, trade networks spread and cities swell. Affluence, the age of splendour, grand buildings and the highest standard of living the system will ever deliver. Intellect, universities, scholarship, argument, the life of the mind at full stretch. Decadence. Then decline.

What runs underneath is the same thing Ibn Khaldun described six centuries earlier, and Glubb had read him: the qualities that build the thing are dissolved by the thing being built.


Two of his markers are worth separating out, because they are checkable rather than atmospheric.

The first is what happens to military service.

Early on, fighting is the most honoured thing a person can do, and the people with the most to lose do it. By the Age of Affluence that has inverted. Service becomes something the wealthy arrange not to do. The army fills with professionals, then mercenaries, then foreigners, and the defence of the empire is quietly contracted out to people with no stake in it beyond the wage.

The United States ended conscription in 1973. Its military is entirely volunteer, drawn disproportionately from the rural poor. The children of the class that decides where the wars happen essentially never serve in them. That is not an accusation of cowardice, it is a description of a structure, and it is Glubb's marker almost to the letter, running now for half a century.

The second is stranger, because it looks like an achievement.

The Age of Intellect is real. Universities multiply, scholarship flourishes, argument becomes the defining activity of the educated class. Glubb valued all of it and did not sneer. But he observed that it tends to produce not wisdom but fragmentation: pluralism first, which is genuinely valuable, and then relativism, which dissolves.

The failure mode is specific. It arrives when the educated class can no longer agree on what is true, what is good, or what is worth defending. Not because they are stupid. Because the machinery of endless questioning, pointed at everything without exception, eventually gets pointed at the foundations, and foundations do not survive being treated as opinions.

A society that cannot say what it values cannot defend what it values, and will not notice it has stopped trying.


The obvious objection is that 250 years is numerology. Pick your start dates carefully and you can make any set of numbers agree.

There is something to that, and the sequence is the stronger claim anyway. But the objection has a cost that is rarely paid. If the number is coincidence, you still have to account for the order: why pioneers always precede merchants, why merchants always precede the affluent, why the intellectual flowering always arrives late rather than early, across cultures with no contact and no shared assumptions.

Nobody who waves the number away has explained the sequence. They have just stopped looking at it.


Glubb's own conclusion was not that decline can be averted by better management. He had watched enough administration from the inside to know what it can and cannot do.

His conclusion was that the decisive variables are moral rather than technical. Duty and service giving way to selfishness. Money becoming the measure of everything, so that the rich are admired for having rather than for doing. Vast resources poured into amusement, and the heroes of the culture becoming its entertainers.

Those are not policy failures. No department fixes them, no election reverses them, and there is no technology that substitutes for them, which is precisely why 250 years buys you the same span whether you have bronze or satellites.

It is worth asking what year you think it is. Britain's run ends around 1950. The American ascendancy is usually dated from 1776, or from 1945 if you count only dominance.

Do the arithmetic yourself. Glubb declined to, and said it was not his place, and then wrote twenty-six pages that make it impossible not to.