The Poorest People in Eurasia Set Its Terms

From roughly 1000 BCE to 1750 CE, the single most consequential military fact in Eurasia was that a pastoral society could destroy a settled one.

This should be surprising. The steppe had a fraction of the population of China, Persia or the Mediterranean, almost no cities, no significant metallurgy of its own, and no agricultural surplus to speak of. It was, by every economic measure, poor. And it repeatedly imposed terms on the richest states in the world.

The reason is that a pastoral economy trains its entire population in the skills of war as a by-product of ordinary life. A boy who herds from horseback rides all day; a household that hunts with the bow uses it constantly. There is no training cost, because the training is the job. A farming state, by contrast, has to take men off the land (losing their labour) and teach them to ride and shoot, which takes years and money.

Then add mobility. Nomads concentrate for a campaign and disperse afterwards, live off herds that move with them, and present no fixed target. A settled empire can win every engagement it manages to force and still lose the war, because there is nothing to besiege and nowhere to march to. The asymmetry is structural, and it is why the standard answer was not campaigns but walls.

What the nomads generally wanted, though, was not the empire. It was access to it.

A steppe economy produces meat, milk, hides, wool and horses, and it cannot produce grain in quantity, fine textiles, metalwork or luxuries. Those had to come from the settled world, and the cheapest route was a negotiated flow: trade rights, border markets, subsidies, marriage alliances. Raiding was the fallback for when the flow stopped.

Thomas Barfield's argument, and it is a good one, is that nomadic confederations formed and dissolved in step with the strength of their settled neighbour. A unified China is worth extracting from, and extracting from it at scale requires a unified confederation; a fragmented China offers nothing worth uniting for. The Xiongnu rise with the Han and fall with them. The relationship is parasitic in the technical sense; the parasite has no interest in killing the host.

The Han policy for managing this was called heqin, peace through kinship: princesses, silk, grain and wine sent north in exchange for quiet. Chinese sources describe it as a humiliation, and by any accounting it was vastly cheaper than campaigning, which Emperor Wu tried at enormous cost with mixed results. It also created a dependency running in both directions, because the Xiongnu leadership redistributed those goods to hold its own confederation together. Cut the subsidy and the confederation frays, which is a lever the Han eventually learned to pull.

Two things are worth correcting about the popular image of these societies.

The first is that they were egalitarian. They were not. Herds are heritable, accumulable capital that can be multiplied by raiding and wiped out by a bad winter, which is a recipe for steep and unstable inequality. Steppe confederations produced royal clans, aristocratic lineages and sworn retinues, and the kurgan burials contain gold by the kilogram. The horsemen of the popular imagination are largely a projection.

The second is that they were destroyers only. The steppe was also the conduit: Buddhism moved along it, and so did technologies, diseases, metalworking styles and administrative practices. The Mongol empire's most durable consequence was not the conquest but the century of movement it enabled across the whole of Eurasia, which we will come back to.

What ended it was gunpowder, and not in the obvious way. Firearms did eventually neutralise the cavalry archer, but the deeper change is that firearms are industrial. They require foundries, powder mills, standardised parts, magazines and supply trains, which require taxation, bureaucracy and fixed installations. A pastoral economy cannot build any of that, and a fiscal state can.

So the advantage inverted. By the eighteenth century Russia was pressing east and the Qing were pressing north and west, and the corridor closed from both ends. The Qing destruction of the Dzungars in the 1750s, which was close to annihilation, is usually taken as the end of the steppe as an independent force in world history.

Three thousand years of one arrangement, and then, within about a century, none.