An Empire Is a Communications Problem

Before the electric telegraph, the maximum speed of government was the speed of a horse.

This one constraint shapes almost everything about how large states worked. A message from a distant province to the capital, and an order back, took weeks; an army took months. Any province more than a certain distance away could organise a revolt, consolidate it and arm itself before the centre even learned there was a problem. The radius of effective rule was set not by how strong the army was but by how fast information moved.

Which is why every large pre-modern empire independently invented the same thing. The Persian Royal Road had stations at intervals of a day's ride with fresh horses and riders waiting; Herodotus reports the whole fifteen-hundred-mile run from Susa to Sardis in about a week, against ninety days for an ordinary traveller. Rome had the cursus publicus. The Mongols had the yam. The Inca had relay runners on a road system four thousand kilometres long. In every case the trick is the same: no individual travels the whole distance, so the message moves continuously at a speed no rider or horse could sustain.

Relay is the invention. The road is only the substrate.

The second problem is what to do once you are there, and the Achaemenid Persians produced the template. Cyrus and Darius governed a territory running from the Aegean to the Indus, containing dozens of languages, religions and legal systems, and they did it by leaving nearly all of them alone.

A province kept its own laws, gods, elites and language. Above that sat a Persian layer: a satrap governing, a garrison commander holding the troops, and a royal secretary reporting to the king (three officials with separate chains of command, so that no one man controlled a province by himself). What the empire asked was tribute and soldiers. What it offered, beyond not destroying you, was the road, the peace and a court worth appealing to.

This is indirect rule, and its logic is about cost. Removing a local elite means importing every official, policing every district and facing a population with nobody in it who benefits from your continued existence. Leaving the elite in place co-opts people who already know the terrain, already command deference, and now have something to lose if you fall. It is not a policy of generosity. It is the only affordable way to hold a continent.

Rome's version went further than anyone else's, and it is the most underrated thing about Rome. Citizenship was extendable. A conquered provincial, or more usually his grandson, could become a Roman citizen, then an equestrian, then a senator. By the second century CE emperors were coming from Spain, from Africa, from Syria. In 212 the Constitutio Antoniniana granted citizenship to virtually every free inhabitant of the empire.

The motive was probably fiscal in part, since certain taxes fell on citizens. The effect was structural. Ambitious men across the Mediterranean had a route upward through Roman institutions rather than only against them, and an elite with that route is very expensive to recruit into a rebellion. Rome's most effective export was not the legion. It was a way in.

None of which was cheap. Military spending consumed well over half of Roman state revenue, and comparable proportions elsewhere, which means that an empire's fiscal crisis and its military crisis are always the same crisis. Debasing the coinage is the standard response because it is the one available without a functioning tax administration, and it works until it does not. When soldiers stop being reliably paid they begin to notice that they are the ones who decide who is emperor, which is the entire plot of the Roman third century.

It is worth closing with the case that shows how much of this was contingent. The Inca ran a state four thousand kilometres long, moved whole populations, stockpiled provisions in thousands of warehouses along the roads, and levied labour rather than goods: all without writing, without money, without wheeled vehicles and without draft animals larger than a llama. Their records were khipu, knotted cords whose numerical content we can read and whose narrative content, if any, we cannot.

Everything the Eurasian empires used (coinage, alphabets, carts, horses) turns out to have been one solution rather than a requirement. It is easy to read a list of inventions as a list of prerequisites. Tawantinsuyu is the standing argument that it is not.