In the year 376 CE, a Roman army watching the lower Danube saw something that had no precedent in living memory: an entire nation, hundreds of thousands of people, standing on the river's far bank and begging to be let in.

These were the Goths (pronounced with a hard "g," as in "goat"), a Germanic people the Romans had known and fought and traded with for over a century, and they were not raiding. They were fleeing. They had their wagons, their herds, their old people and their children, and they were asking the Roman Empire, the thing they had spent generations harassing, for asylum. Something had come up behind them out of the east, something that had broken the most powerful Gothic kingdom on the steppe with frightening speed (the sources do not tell us in how many campaigns, only that it was fast) and was still coming, and the Goths had decided that throwing themselves on the mercy of Rome was less terrifying than staying to meet it.

That something was the Huns. This book is about what happened next: the roughly two centuries in which the western half of the Roman Empire was overrun, dismantled, and replaced, not by one invasion but by a chain reaction of peoples in motion that historians, borrowing a nineteenth-century German word, call the Völkerwanderung (pronounced "FUR-kuh-VAN-deh-roong," literally "the wandering of the peoples"). It is the story of how the ancient Mediterranean world came apart and how medieval Europe was assembled out of the pieces.

It is not, mostly, the story of the Goths themselves. They get their own book, because their journey from the Danube to a kingdom in Spain is long enough to deserve one. This book is about the wider shattering they were the first crack in: the Huns who started it, the Vandals who took Rome's breadbasket and then Rome itself, the Franks who would become France, the collapse of the Western emperors, and the Eastern Empire's catastrophic attempt to put it all back together. The Goths walk through these pages constantly, but in the background. The foreground is the system, and the system breaking.

A word, once, about how we know any of this, because it shapes everything that follows and it is more honest to say it now than to keep apologizing for it later.

Almost every written source for these two centuries was produced by Romans, Greeks, or Christian churchmen writing within the surviving (Eastern) empire. The peoples doing the migrating, Huns, Vandals, most of the early Franks, wrote little or nothing that survives, and what little they did was usually written down generations later by their descendants once those descendants had Latin clergy to do it for them. So when this book says a battle happened "in 451," or that a king "decided" something, the reader should hear an implied caveat that does not need repeating every sentence: the dates, numbers, and especially the motives are reconstructed from a handful of sources who were often far away, often hostile, and often writing decades after the fact. Where a claim is genuinely contested, this book will say so. Where it is reasonably solid, it will say that too. The general rule is that the shape of events is well attested and the vivid details are the part to hold loosely, because vivid details are exactly what ancient writers liked to improve.

With that on the table, the place to start is with the people who knocked over the first domino, because nobody in the Roman world had any idea who they were.

The Huns were not Germani. This is the single most important thing to understand about them, and it is easy to get wrong, because the Romans lumped almost everyone beyond the frontier under the catch-all sneer "barbarian," and a modern reader can absorb that flattening without noticing.

The Germanic peoples, Goths, Vandals, Franks, Saxons and the rest, spoke languages from the Germanic branch of Indo-European, the same vast language family that produced Latin, Greek, Persian, and Sanskrit. They were the cold-country, forest-and-farm peoples Rome had been managing along the Rhine and Danube for four centuries (the Germanic Tribes tells that long story). The Huns were something else entirely. They were steppe nomads: horse-herding, tent-dwelling, cattle-driving people from the Eurasian steppe, the immense belt of open grassland that runs for thousands of miles from the plains of Hungary east across Ukraine and Central Asia to Mongolia. Their language is essentially unknown. We have a few personal names and almost nothing else, and scholars have argued for over a century about whether it was Turkic, related to the later Mongolic languages, something Iranian, or something with no surviving relatives at all. The honest answer is that we cannot classify it, but it was certainly not Germanic and almost certainly not Indo-European. To put the distance in terms a reader can feel: a Goth meeting a Roman shared a deep common linguistic ancestor with him, the way English and Russian share one, distantly. A Goth meeting a Hun shared nothing, the way English and Chinese share nothing. These were not cousins of the Germani. They were strangers to everyone.

Where did they come from? The honest answer is that we do not securely know. A long tradition connects them to the Xiongnu (pronounced "shyong-NOO"), a steppe confederation that had menaced Han-dynasty China centuries earlier, but the link is debated and the evidence is thin enough that careful historians treat it as a hypothesis rather than a fact. What is solid is the behavior. The Huns appeared on the western steppe, north of the Black Sea, in the 370s, and they were a kind of military machine the Roman and Germanic worlds had never had to fight. Their power was the horse and the bow. Every man rode, from childhood. Their weapon was the composite bow, a short, brutally powerful bow built up in layers of wood, horn, and sinew so that a weapon small enough to shoot from a galloping horse hit as hard as a much larger one, and they could shoot it accurately at the gallop, wheel away, and shoot again before an enemy could close. Against this a Roman infantry unit, trained to lock shields and grind down an enemy that came to meet it, had nothing to grind down: the Huns simply rode in a wide circle out of reach, shot the formation full of arrows until it broke or charged, and if it charged they fell back faster than it could follow and shot it some more. They did not hold ground; they had no cities to defend and nothing to besiege. They could not be brought to a decisive battle they did not want, and they could not be exhausted by burning farms they did not own. Roman armies were built to fight other armies that stood still and fought. The Huns refused to be that, and for a generation no one had an answer.

The other thing the Huns did, and it is the engine of this entire book, is that they absorbed. When the Huns broke a people, they did not necessarily destroy or replace them. Often they incorporated them, took their warriors, their tribute, and their submission, and rolled them into a growing confederation under Hunnic overlordship. So the thing pressing on the Danube frontier was not a fixed nation of a known size. It was a snowball, and everyone it touched either ran or was rolled into it.

Here is the mechanism, stated plainly, because the whole rest of the book is variations on it.

The Huns hit the easternmost Gothic confederation, the Greuthungi (pronounced "groy-THUNG-ee," later remembered as the ancestors of the Ostrogoths), on the open steppe north of the Black Sea around 375, and broke them. The shock rolled west into the next Gothic group, the Tervingi (pronounced "TER-ving-ee," later the core of the Visigoths), who did not wait to be broken and instead ran for the Danube and Rome. That is one domino knocking the next. The Goths' arrival inside the empire in 376, the catastrophic battle three years later at Adrianople in 378 where the Goths destroyed an entire Eastern field army and killed Valens (the emperor of the Eastern half of the empire, on the throne since 364; for an emperor to be killed in battle by barbarians was almost unheard of, a humiliation Rome had not suffered on this scale in centuries), and the uneasy settlement of the Goths inside the Balkans in 382, all of that is the Gothic story, and it is told in full elsewhere. What matters here is the principle it established and the precedent it set.

The principle: a people pushed hard enough from behind does not negotiate over whether to move. It moves, and the only question is who is in the way.

The precedent: faced with a whole armed nation it could not destroy, Rome did the thing it would now do over and over until it had given the empire away. It cut a deal. It admitted the Goths as foederati (pronounced "feh-deh-RAH-tee," from the Latin foedus, "treaty"), allied peoples settled on Roman soil who kept their own leaders and their own internal law and served in the Roman army in exchange for land, supplies, and gold. This was not invented in 382; Rome had been settling defeated barbarians under treaty for two centuries, planting groups of beaten Germani as farmer-soldiers on emptied frontier land along the Rhine and Danube where they could work the soil and be called up to fight. But it had always done that on Rome's terms, with broken and divided peoples. It had never before done it with a whole intact nation that arrived under its own kings, on its own terms, having just annihilated a Roman army. The difference between a settled prisoner population and an armed allied nation living inside your border with its own command structure is the difference, eventually, between a province and a kingdom.

This is, in fact, an old trap (Carthage had armed and settled foreign mercenaries it could not pay, and they turned on it in the savage Mercenary War and nearly destroyed the city from inside; Gaulish factions had hired Germanic Suebi who then refused to leave), seven hundred years old by the time Rome learned it the hard way. Rome learned it anyway, because at every step the immediate alternative looked worse than the deal in front of it. That sentence is the closest thing this book has to a thesis, and it will be true on almost every page.

The Huns had knocked over the first domino. For the next thirty years the line of falling tiles ran mostly through the Balkans and the Gothic story, far enough from the Western capital that the men running Italy could persuade themselves it was a containable eastern problem. It was not. On the freezing last night of the year 406, while the Roman world was looking the other way, the dominoes reached the Rhine, and the half of the empire that had told itself it was safe began to come apart in earnest.

The Migration Period

The Domino Falls

The Rhine Gives Way

Kingdoms Inside the Empire

The Vandals

Attila

The Fall of the West

The Successor Kingdoms

The Eastern Reconquest

The End of Antiquity in the West