In the year four hundred and seventy six, a commander of the barbarian troops in Italy deposed a teenage emperor named Romulus Augustulus, declined to appoint a replacement, and sent the imperial regalia east to Constantinople with a note saying one emperor was enough.
That is the date in the textbooks. What is striking about it is how little anyone at the time seems to have thought it mattered. Italy went on being governed from Ravenna, by a man who used Roman titles, collected Roman taxes, and left the Senate sitting. The eastern emperor was still emperor. Consuls were still appointed. For most people the arrangement was a change of employer.
The date became important much later, when historians looking back wanted a moment. What actually happened was a transformation spread over three centuries, and it went at very different speeds in different places. The point of this chapter is to separate the parts that were slow from the parts that were sudden, and the parts that were a change of management from the parts that were a genuine collapse.
The empire nearly came apart in the middle of the third century, in about fifty years of civil war, plague, invasion and currency collapse. It was put back together, and the thing that came out was heavier than the thing that went in.
The emperor Diocletian, from the ranks of the army, split the imperial office among four rulers so that there was always someone near each threatened frontier. He doubled the number of provinces, which made each governor smaller and easier to control, and separated civil from military command so that no one man held both. He expanded the army and the bureaucracy, and he paid for all of it by taxing harder and by assessing tax in kind when the coinage would not hold its value.
He also tried to fix prices across the empire by decree, listing maximum prices for over a thousand goods and services with death as the penalty. Fragments of the edict survive on stone in a dozen places. It did not work, and its failure is one of the earliest well documented demonstrations of what happens when you legislate against arithmetic.
His successor Constantine did two things with permanent consequences. He legalised and then patronised Christianity, which had been a persecuted minority religion and within a century would be the official one. And he founded a new capital on the Bosphorus, positioned between the Danube and the eastern frontier, in the rich half of the empire.
That second decision quietly settled the west's fate. From then on the wealth and the attention were in the east, and the western provinces were the expensive end of the business.
The peoples who ended up ruling the western provinces were not strangers. They had been trading with the empire, serving in its armies and living along its frontiers for generations, and many of their leaders held Roman commissions.
What changed in the late fourth century was pressure from further east, probably from the movement of the Huns, which pushed whole populations across the Danube. In 376 a large body of Goths was admitted into the empire, mishandled and starved by the officials meant to settle them, and revolted. Two years later they destroyed a Roman field army and killed the emperor at Adrianople.
After that the pattern is less invasion than negotiation conducted with armies. Groups were settled inside the empire in exchange for military service, given land or a share of tax revenue, and gradually became the effective government of the areas they occupied. Rome was sacked in 410 by a Gothic king who had spent years trying to obtain a regular command and a settlement, and who sacked it partly because the negotiations kept failing.
By the middle of the fifth century the western emperor governed little outside Italy, and the armies he relied on were commanded by men from those same peoples. The deposition of 476 removed a formality.
How to describe all this is one of the live arguments in the field. One school emphasises accommodation and continuity, and prefers the term transformation. Another insists that whatever the process, the outcome for the people living through it was violent and impoverishing, and that calling it a transformation is a way of not looking. The evidence in the next section is why the second group has a strong case.
Forget the emperors. Consider a household on a moderate farm in Britain, Gaul or Spain, and ask what actually changed for them.
Before, they paid tax, in coin or kind, which travelled to pay soldiers and officials. They sold their surplus into a market that reached a long way, because a road and a sea lane connected them to towns that wanted grain and wool. They bought pottery that had been made in an industrial quantity somewhere else and shipped in, and it was good, thin walled, standardised, cheap. Their roof was tiled. Somebody in the district could write.
Afterwards, the tax still came, but it went to somebody local who had soldiers. The market shrank to what could be reached in a day or two, because long distance shipping depends on a security and a currency that were no longer there. The good pottery stopped arriving, and what replaced it was made locally, thicker, by hand, in fewer shapes. The tiles were not renewed. Writing contracted to the church.
This is not speculation. It is what the archaeology shows, and it shows it consistently across the western provinces, though at different dates and depths. Coin circulation falls off. Pottery becomes cruder and its distribution radius collapses. Building in stone and tile stops in some regions almost entirely. Cattle get smaller. In Britain, which was the extreme case, the material culture drops below the level it had been at before the Romans arrived, and stays there for generations.
That is a real fall, and it happened to people who had no vote in any of it. The eastern provinces show nothing comparable at this date, which is a useful control: this was not a general ancient decline, it was something that happened to one half of a state.
Against that, the things that did not break.
The church did not break. It had a structure of bishops mapped onto the Roman provincial system, and when the civil administration failed the bishop was often the last official standing, negotiating with besiegers and organising food. That is how a religious office acquired secular authority in the west, and the shape of the later medieval church follows directly from it.
The language did not break. Latin went on being spoken and slowly turned into the Romance languages, which is why a Portuguese speaker and a Romanian speaker are both speaking, at a great distance, the language of the Roman army and the tax collector.
The idea did not break. Every serious power in western Europe for the next thousand years claimed some relationship to Rome. The kings who took over the provinces issued law codes in Latin, used Roman titles, and struck coins with imperial imagery. Nobody said the empire was finished. They said they were continuing it.
And in the middle of the sixth century the eastern emperor Justinian very nearly put it back together by force, reconquering North Africa, Italy and part of Spain. The reconquest of Italy took twenty years of war and did more damage to the peninsula than the previous century of barbarian rule. Then plague arrived, and the recovered territories were mostly lost again within a generation.
It is worth naming the states that replaced the western empire, because the map of modern Europe is faintly visible in them.
The Visigoths ended up in Spain and southern Gaul, and ruled from Toledo until an Arab and Berber army destroyed the kingdom in a single campaign in 711. The Vandals took North Africa, which was the western empire’s grain supply, and taking it was probably the single most damaging blow of the century, because it removed the revenue that paid for everything else. The Ostrogoths held Italy, and their king Theoderic ran it for over thirty years with a Roman civil service, Roman law for Romans, and a policy of keeping the two populations legally separate.
The Franks took northern Gaul and are the ones who mattered longest, because they converted directly to the Catholicism of their subjects rather than to the Arian Christianity most of the other groups had adopted. That removed a religious barrier between rulers and ruled, and it gave them the bishops as allies rather than as opponents. Everything in chapter eleven grows out of that decision.
The pattern in most of these kingdoms is the same and it is not what the word barbarian suggests. A small armed elite sat on top of a much larger provincial population, kept the Roman tax registers going as long as they could read them, issued written law codes in Latin, and married into the local landowning families. Within a few generations you cannot reliably tell the two groups apart in the evidence.
What they could not do was maintain the expensive things: the standing army, the long distance shipping, the coinage of reliable silver, the professional bureaucracy. Those needed a tax base larger than any one kingdom had, and once they were gone the economy described in the previous section was the result.
Firm: the material record. Pottery, coins, building, animal bones and shipwrecks give us a picture of economic complexity that does not depend on any writer's opinion, and it is the strongest evidence in this chapter.
Soft: the narrative. Our written sources for the fifth century in the west are thin, fragmentary and mostly written by churchmen with a theological reading of events. Whole decades are known from a few lines of chronicle.
Soft: numbers and identities. How many Goths crossed the Danube, how large the settled groups were, and whether the peoples named in the sources were stable ethnic groups at all are heavily contested. Current thinking treats these identities as things that were assembled around successful war leaders rather than as nations on the move.
Contested, and worth knowing about: the whole framing. Since the 1970s a large body of scholarship has preferred to speak of late antiquity as a period of cultural transformation rather than decline, emphasising religious creativity and continuity. Since about 2005 there has been a strong pushback arguing that this reading is a scholarly aesthetic preference that fits badly with the archaeology of ordinary living standards. This chapter has sided, on the material evidence, with the second view about the west, while accepting the first about the east.
One thing to keep hold of: whatever we call it, the eastern half did not fall. It is about to be the subject of its own chapter, and it lasted another thousand years.