Hephthalite Confederation

A name in six languages

Drafted from (eds), History of Civilizations of Central Asia, Volume III: The Crossroads of Civilizations, A.D. 250 to 750 (1996)formation

The Greeks called them Ephthalitai, the Persians Haytal, the Arabs Hayatila, the Indians Sveta Huna — white Huns — the Chinese Yeda, and their own coins carry a Bactrian form usually read as Ebodalo. Every one of those is somebody else's word, and the question of who they actually were has been argued for a century and a half without resolution.

The positions are roughly these. They were Huns, part of the same movement out of the Altai and the Central Asian steppe that produced the European Huns and the Kidarites before them. They were Iranian-speaking, and the Chinese sources that describe their language as unlike the Rouran and the Turks are describing an Iranian one. They were a Turkic group. Or — which is the direction most recent work goes — the question is badly posed, because a steppe confederation is a political thing rather than an ethnic one and absorbed whoever joined it.

What is not disputed: from the 440s or thereabouts a confederation with its centre in Bactria and Tokharistan was the dominant power between the Sasanian frontier and the Tarim basin, and for about a century it was strong enough to kill Persian kings and break Indian empires.

It is worth being blunt about why they are usually left out. The history of Sasanian Iran is normally written as six centuries of war with Rome, because that is the frontier both sides wrote books about. The eastern frontier produced no literature at all and produced most of the consequences: it killed a king of kings, extracted tribute from his successors for a generation, and forced the fiscal reforms that remade the empire.

What makes them hard to write about is that they left no writing of their own. Coins, overstruck on Sasanian silver; a few Bactrian documents; seals. Everything else is Chinese, Byzantine, Syriac, Armenian, Sasanian and Indian — six traditions writing about an enemy or a curiosity, in six languages, with six sets of assumptions and six versions of the name.

What killing a king of kings was worth

Drafted from (eds), History of Civilizations of Central Asia, Volume III: The Crossroads of Civilizations, A.D. 250 to 750 (1996)peak

Peroz I fought the Hephthalites three times. The first campaign, around 469, ended with the king himself captured; he was ransomed, left his son Kavad as a hostage, and undertook not to cross the frontier again. The second ended in a humiliating settlement. The third, in 484, ended with Peroz dead on the field with most of his army, in what the sources describe as a trap involving a concealed ditch.

The consequences inside Iran were a generation of tribute payments, a fiscal crisis, the Mazdakite convulsions and eventually the reforms of Kavad and Khusrau I that rebuilt the state. The consequence on the other side is the one worth dwelling on.

A steppe confederation with no cities had acquired a fixed annual income from the richest state in western Asia. Empires normally extract from confederations. This one paid.

It also acquired standing. In 498 Kavad — the same man who had been a hostage — was deposed by his own nobles and priests, escaped, went to the Hephthalites, married a woman of their ruling house, and returned with their troops to take his throne back. He kept it for another thirty-three years and paid for the loan.

A confederation without a bureaucracy, without a capital and without a written language was, for about twenty years around the turn of the sixth century, deciding who ruled Ctesiphon. That is the height of Hephthalite power, and there is no better measure of it available: nobody publishes a figure for how much territory the confederation held or how many people lived in it.

The coins say the same thing in metal. Hephthalite issues are Sasanian drachms overstruck — the defeated empire's own silver, with the fire altar and its attendants left standing on the reverse, and a new portrait and a tamgha hammered over the front. Reissuing your victim's money with your own head on it is a public statement, and it is not one that any of the usual vocabularies of royal legitimacy has a word for.

The trench

Drafted from Caesarea, History of the Wars, Books I-II: The Persian War (1914)peak

Procopius describes the Hephthalites at some length before he describes what they did to the Persian king, and the description is worth having first, because it is a Roman writer taking apart the category he has just put them in.

Not what the name says

They are Huns in fact as well as in name, he says — and then spends a paragraph explaining that almost nothing else about them fits.

They do not mingle with the Huns anyone in the Mediterranean knows, because their land is nowhere near. Their territory lies immediately north of Persia and their city, Gorgo, sits on the Persian frontier, which is why the boundary is perpetually disputed. They have never raided Roman territory except alongside a Persian army.

And then:

They are not nomads like the other Hunnic peoples, but for a long period have been established in a goodly land... they are ruled by one king, and since they possess a lawful constitution, they observe right and justice in their dealings both with one another and with their neighbours, in no degree less than the Romans and the Persians.

In no degree less than the Romans and the Persians. A sixth-century Constantinopolitan is saying that a steppe-descended people on the Oxus kept law as well as his own empire did. He also notes that they are the only Huns who are not, to his eye, ugly — which is the kind of remark that tells you what the default assumption was.

The companions

The one custom he reports at length is not a legal one.

Wealthy Hephthalites attach to themselves friends — twenty or more — who become permanent banquet-companions and share in all their property, holding a kind of common right in it.

Then, when the man who has gathered such a company together comes to die, it is the custom that all these men be borne alive into the tomb with him.

Procopius sets that down two sentences after praising their justice, without appearing to notice any tension. Both things were true, and he reports both.

The prostration

Peroz of Persia went to war with them over the frontier and was drawn into a defile shut in by precipitous mountains and screened by forest, with the Hephthalites closing the entrance behind him. The Persians saw clearly what straits they were in and thought the situation hopeless.

The Hephthalite king sent messengers. He upbraided Peroz at length for the senseless foolhardiness by which he had destroyed himself and his people — and then offered to let the whole army go, on two conditions: that Peroz prostrate himself before him as the proven master, and that he swear never to campaign against the Hephthalites again.

Peroz consulted the Magi with him. Their advice was a piece of theological engineering: as to the oath, settle it as you please — but Persians prostrate themselves before the rising sun every day, so meet the Hephthalite king at dawn, face east, and make the obeisance then.

He did exactly that, gave his pledges, and withdrew with the whole Median army intact, gladly.

What the ditch was for

Not long afterwards, disregarding the oath, he came back.

He brought every Persian and allied force he could raise, and of his roughly thirty sons he took all but one — leaving behind only Kavadh, just past boyhood.

The Hephthalites reproached their own king for having let Peroz go. He asked them, laughing, what of theirs he had abandoned — their land, their arms, anything they owned? They answered that he had abandoned nothing except the one opportunity on which everything else depended.

Then he told them to wait, on the grounds that the Persians were still inside their own borders. And in the plain where the invasion would come, he marked out a very large tract and dug a deep, wide trench across it — leaving one narrow strip of ground intact in the middle.

Peroz died on that field with most of his army and most of his sons. The son left at home became king, and the tribute the Hephthalites then levied on Iran ran long enough to help produce the fiscal crisis, the Mazdakite convulsion, and the reforms that rebuilt the Persian state.

The other frontier

Drafted from (eds), History of Civilizations of Central Asia, Volume III: The Crossroads of Civilizations, A.D. 250 to 750 (1996)peak

While the Hephthalites were taking tribute from Iran they were also going south, and the second campaign is the one with the larger long-term consequence.

Huna forces crossed into Gandhara and then into the Gangetic plain from the later fifth century. The Gupta empire ended in 550, and the standard account makes the Huna invasions a principal cause: Skandagupta beat them off in the 450s at considerable cost, they came back, and by the early sixth century Toramana and then his son Mihirakula were ruling from Sakala in the Punjab and raiding far into the interior. Yashodharman of Malwa defeated Mihirakula around 528 and left an inscription boasting about it, which is most of what is known.

Mihirakula has a bad reputation in two traditions. Indian sources and the Chinese Buddhist pilgrim Xuanzang, writing a century later, describe him as a destroyer of monasteries and a persecutor of Buddhism on a large scale. Whether that is what happened, or a memory of one episode generalised, or the standard treatment a hostile tradition gives an invader, is not settled — and the coinage shows Huna rulers patronising Shaivite and Buddhist establishments as well.

A caution belongs here that the older literature does not make. Modern scholarship increasingly separates the group in India — the Alchon, who struck their own coins and used their own tamghas — from the Hephthalites proper in Bactria, treating them as related but distinct successors of the same movement rather than as one empire operating on two fronts. What follows keeps the older, simpler grouping for convenience, but the distinction is real and should not be assumed away.

What is secure either way is the shape of the thing. The same chain of displacements out of the steppe that had ended Greek Bactria in the second century BC was still running six hundred years later, and inside a single century it killed a Sasanian king, put another one on his throne, and broke the last classical empire of northern India.

What a Chinese embassy saw

Drafted from (eds), History of Civilizations of Central Asia, Volume III: The Crossroads of Civilizations, A.D. 250 to 750 (1996)aside?What aside means

In 518 the dowager empress of the Northern Wei sent a mission west to collect Buddhist texts. Song Yun and the monk Huisheng travelled through Tokharistan and Gandhara and were back in 522, and their account survives in a later compilation. It is the only description of the Hephthalites by somebody who was there.

What they report is worth setting out because it is the whole of the ethnography. The king had no walled city and moved with the seasons, living in a large felt tent — they give its dimensions — with carpets on the floor. He sat on a golden couch or throne carried on four gold phoenixes. His principal wife wore a horn-shaped headdress several feet long, covered with pearls, and the other wives wore smaller ones by rank. The court did not observe Buddhism, and Song Yun's attempt to interest it in Buddhist doctrine went nowhere; the mission was received with formal courtesy and no particular deference to the Wei.

They also record that the Hephthalites had subjugated more than thirty kingdoms, took tribute from all of them, and had no writing.

That last item is the whole difficulty in five words, and it is worth spelling out what follows from it.

There is no capital, because the king moved with the seasons and no source names a seat. There is no reliable king-list: the names have to be assembled from coins, Chinese notices and Sasanian mentions, and they do not match each other. Nobody can say how one ruler followed another, or how much country the confederation held, or how many people were in it.

What can be said is how it was paid for and how it fought — tribute, tolls on what crossed, and levies from the tribes in it — and what it did to two of its neighbours. It is a power recorded almost entirely through its effects on other people.

Compare the alternatives. Saba' in South Arabia is known from tens of thousands of its own inscriptions and no narrative; Greek Bactria from its money; the Parthians from what their successors decided to leave in. The Hephthalites are the fourth case and the hardest: described only by the people they frightened, in six languages, none of which was theirs.

Two empires at once

Drafted from (eds), History of Civilizations of Central Asia, Volume III: The Crossroads of Civilizations, A.D. 250 to 750 (1996)end

In the 550s a new confederation came out of the eastern steppe. The Göktürk khaganate destroyed the Rouran, took the Mongolian plateau, and kept moving west along the northern edge of Hephthalite territory.

Khusrau I, who had spent his reign rebuilding the Sasanian state on the survey and the fixed assessment, saw the arrangement available and took it. A Sasanian alliance with the Türks was agreed, sealed by a marriage, and between roughly 557 and 561 the two powers attacked from opposite sides: the Türks from the north across the Oxus, the Sasanians from the south-west.

The confederation broke, and neither of the two who broke it took the whole of it. The territory was partitioned along the Oxus: Sogdiana and the north to the Türks, Tokharistan and the south to the Sasanians. The Hephthalite principalities did not all disappear — several continued as vassals of one side or the other for decades — but the tribute stopped and the confederation as a power did not recover.

It is worth noticing what it took to do it. A confederation with no cities, no writing and no bureaucracy, which had killed a Sasanian king, taken his successors' money for a generation and installed one of them on his own throne, was brought down by a coalition of the empire it had been extorting and a larger confederation from further east. One of them was not enough.

The partition did not hold either. Within forty years the Türks and the Sasanians were fighting each other over the same ground; within eighty the Sasanian empire was gone and Arab armies were on the Oxus. What followed in Tokharistan was a set of local dynasties — the Turk Shahis among them — governing the old Hephthalite country under whichever larger power was nearest, until the caliphate and then the Samanids absorbed it.

The confederation itself left almost nothing: overstruck coins, a handful of Bactrian documents, six versions of a name, and the reforms a Sasanian king had to carry out because it had beaten him.

Haytal

Drafted from Strange, The Lands of the Eastern Caliphate: Mesopotamia, Persia, and Central Asia, from the Moslem Conquest to the Time of Timur (1905)afterlife

A confederation that left almost nothing behind left one thing, and it is the item it had least control over: a word other people used for it.

The Arabic geographers of the ninth and tenth centuries knew the Haytal. They knew, correctly, who they had been — in the fifth century AD the chief enemies of the Sasanian monarchy, the people the Byzantine writers called Ephthalites and everyone else called the White Huns. That much was still common knowledge three hundred years after the confederation had been partitioned out of existence.

But the word had drifted. By the tenth century Haytal had stopped meaning a particular people and had come to mean, loosely, everything beyond the Oxus: all the Turanian peoples and all their lands. Muqaddasi uses it that way. The formal Arabic name for the region was Ma wara al-nahr, That Which Is Beyond the River, and Haytal was the alternative — an ethnic name turned into a direction.

Le Strange adds a further suggestion in a footnote, and it is worth having as a suggestion rather than a fact. On the upper Oxus, in the mountains, there was a district the geographers write variously as Khuttal, Khutlan and Khuttalan. Timur's historian calls it Khutlan when describing the campaigns there. It appears, Le Strange says, to be the same word as Haytal.

If he is right, then a people known to us only in six other languages' versions of their name ended up with the name itself doing two separate jobs: a vague word for the far side of a river, and a specific mountain district that kept it for another eight hundred years.

The country they held

Tokharistan, their centre, went on being a place with an identity long after they stopped.

It lay east of Balkh, along the south bank of the Oxus as far as Badakhshan, bounded southward by the ranges north of Bamiyan, and it was divided into an Upper and a Lower. Muqaddasi and the others name a great many of its towns and say almost nothing about most of them, so that the majority can no longer be placed on a map at all: Khulm two days east of Balkh, a small city with a good climate among large villages; Siminjan, larger, with a Friday mosque and good fruit, lying in a maze of valleys that had been settled by Arabs of the Tamim; Baghlan beyond it; and Andarab, with fine markets, in valleys under the Panjshir range whose recesses held silver mines.

That is what a Chinese envoy's thirty subjugated kingdoms had turned into by the tenth century: a list of market towns in an Islamic geography, most of which are now unidentifiable.

What that adds up to

The Hephthalites had no capital, no script, no chronicle and no king-list that survives intact. They beat the greatest power in western Asia, took its money, chose one of its kings, and broke an empire in India, and it took an alliance of two states to end them.

And what came through is a place-name, a district on the upper Oxus that may or may not be their name, and a vague Arabic word for everything on the wrong side of a river.