Urartu

A state built against Assyria

Drafted from Zimansky, Studies in Ancient Oriental Civilization 41 (1985)formation

The Armenian highlands are not a place that produces empires. High, cold, mountainous, cut into valleys that do not communicate easily, with a short growing season and no great river plain — the kind of country that produces chiefdoms and keeps producing them. From the middle of the ninth century BC it produced instead a centralised state with a royal dynasty, a chancery, a territorial administration and a network of vast fortresses, which held together for two hundred and seventy years and then vanished.

Zimansky's account is that the anomaly has an external explanation. Urartu is not what that landscape generates on its own. It is what that landscape generates when the Assyrian army starts coming north every few years and there is no alternative to organising.

The fortresses

The state is legible chiefly as architecture. Along the valleys the kings built citadels on rock outcrops — Tushpa on the crag at Van, Erebuni above what is now Yerevan, Argishtihinili, Ayanis, Bastam, Karmir Blur — each with cyclopean walls, each with an inscription naming the king who built it, and each with storage on a scale that is the point of the whole thing. Rows of magazines holding grain and wine in jars sunk to the shoulder in the floor, capacities recorded in the inscriptions, running to hundreds of thousands of litres in a single fortress.

Around them, state-built infrastructure: irrigation canals, orchards, vineyards established by royal order and recorded as such. The canal Menua cut to bring water to Tushpa runs for over fifty kilometres and parts of it still carry water.

What is thin, by comparison, is everything not royal. The countryside between the fortresses has produced remarkably little; there is not much in the way of Urartian villages, ordinary cemeteries or private inscriptions. This is a state that appears in the archaeological record almost entirely as the works of its kings, which is why Zimansky calls it a superstructure — an apparatus laid over a landscape rather than grown out of one.

Where the wealth came from

The kingdom lived on its fields and on raiding, and the second of those in a strict and structural sense rather than as an occasional windfall.

The royal inscriptions are relentlessly quantitative about campaigning. King after king records marching north and west into Transcaucasia and eastern Anatolia, burning towns, and bringing back the count: so many thousand cattle, so many thousand sheep, so many thousand horses, so many thousand men and women. The territories raided were mostly not annexed. The people were resettled.

That is what makes the raiding a revenue system rather than a habit. The deported population is what worked the fortress estates, and the herds are what stocked them. The raiding is not incidental to the agricultural base; on this reading it is the mechanism by which a fortress in a high valley had anything to store. Assyria financed itself the same way, which is the joke: the state built to survive Assyria paid for itself by doing to its own neighbours what Assyria was doing to it.

Biainas

Drafted from Sayce, The Hittites: The Story of a Forgotten Empire (1888)expansion

The kingdom did not call itself Urartu. That was the Assyrian name, and the Hebrew scriptures took it from the Assyrians as Ararat.

Its own name for itself was Biainas, and that name has survived in the one place it could: the lake. Van is what is left of Biainas.

Reading the kings

The kingdom left inscriptions all over eastern Anatolia, and Sayce, who was the first person to read them systematically, describes exactly what they are.

They are written in cuneiform borrowed from Nineveh — taken over, on his reckoning, in the time of Ashurnasirpal or his son Shalmaneser — and in a language that bears no resemblance whatever to Assyrian. That is the whole kingdom in one observation: an enemy's writing system used to record a language the enemy could not read.

What the inscriptions record is buildings, offerings and campaigns. Temples and palaces put up; gifts made to the gods; and the marching, in the terms the kings themselves used.

Menuas at Palu

Menuas reigned in the ninth century and went west. He overran the land of Alzi and found himself among the Hittites — the successor kingdoms of northern Syria and the Euphrates bend — where he plundered two cities belonging to a Hittite prince and captured a number of soldiers, whom he dedicated to the service of his god Khaldi.

On another occasion he marched as far as Malatya, and on the way back he had an inscription commemorating his conquests cut into the cliffs at Palu, on the north bank of the Euphrates, about halfway between Malatya and Van.

His son Argistis made war on the same people again, and one of his inscriptions shows Malatya itself counted among their fortresses.

What the reach was for

Set that beside what the Assyrians were doing at the same moment and the shape of the kingdom becomes clear.

Urartu's expansion runs north and west, into Transcaucasia and up the Euphrates — away from Assyria, along the routes that carried metal, horses and timber. It takes prisoners and dedicates them to the state god. It records the taking of towns and does not annex most of them.

This is a state doing to its own neighbours precisely what Assyria was doing to it, on a smaller scale and for the same reason: a fortress kingdom in high country cannot feed itself out of its valleys alone, and everything it needed had to be brought in.

The cliffs at Palu are as far west as the writ ran. A king who could not stop the Assyrian army coming north every few years could still put his name on a rock four hundred kilometres from his capital, in a script he had taken from the people he was hiding from.

The mare

Drafted from Luckenbill, Ancient Records of Assyria and Babylonia, Volume I (1926)peak

By the middle of the eighth century BC Urartu was not a mountain kingdom holding its valleys. It was contesting northern Syria with Assyria, and it had allies there.

That is the measure of the peak, and the evidence for it is Assyrian, because Tiglath-Pileser III had to go and undo it.

How far it had come

Sarduri of Urartu, the annals say, revolted against me and made common cause with Mati'-ilu — Mati'-ilu being the king of Arpad, in what is now northern Syria, six hundred kilometres from Lake Van.

The word revolted is Assyrian bookkeeping and should be read as such: Urartu had never been an Assyrian possession, and its king owed Assyria nothing. What the phrasing conceals is the real situation, which was that a highland kingdom had built a coalition across the top of the Fertile Crescent and was in a position to decide who held the north Syrian trade routes.

The battle happened at Kishtan and Halpi in Commagene, on the Euphrates — far from both capitals, which is where the two states had come to meet.

The flight

Tiglath-Pileser's account survives in more than one version, and the differences are the interesting part.

Both agree that he broke Sarduri and took his whole camp. One version says Sarduri became frightened at the fury of Assyrian arms and ran away, alone, to save his life. The other is more specific, and better: to save his life he mounted a mare and escaped to Mount Sibak, a steep mountain, at night, and ascended it.

A mare rather than a stallion, at night, up a mountain. It is the kind of detail that survives because it was humiliating and because somebody saw it, and it has outlasted every inscription Sarduri put up about himself.

He was then shut inside Tushpa, his own capital on the lake, with large numbers of his men killed in front of the gate. Tiglath-Pileser did not take the city. Instead he had his royal image made and set up in front of it, which is what you do when you cannot get in and want the failure to look like a decision.

What he did take

The annexation list is the part that shows what a great state losing a battle actually costs.

He marched, he says, through the whole width of Urartu from its upper border to its lower — sixty double-hours in one version, eighty in the other — and found none to oppose. Then the strongholds: Ulluba entire, and Kashtirru, Parisu, Tashuha, Galananal, Sikibsa, Asurdai, Babutta, Lusia, Bisia; and Kuta, Urra, Arana, Taba, Uallia as far as the Euphrates; and Kilissa, Ezieda, Abbissa, Harbisinna, Tasa, Enzi, Anganu, Benzu.

They were not sacked and left. They were brought within the Assyrian border, assigned to the provinces of the commander-in-chief and of Nairi. A new city was built in Ulluba, named Ashur-ikisha, populated with deportees, and given an Assyrian governor.

Note also, in the same list, a stronghold called Sardaurri. Urartian kings named fortresses after themselves; this one now belonged to Assyria under the name of the king who had lost it.

The high point, read backwards

Nothing in Urartian inscriptions describes a defeat, because royal inscriptions in this world never do. Almost everything we can date about the kingdom's extent comes from the people who fought it.

So the peak has to be read out of an enemy's annals, and it reads clearly: to be worth a campaign of this size, to hold allies in Syria, to require a march across the entire kingdom to subdue, Urartu had to have been the second power in the region. Its capital held. Its frontier provinces did not, and it never got them back.

Thirty years later Sargon II would come up the same valleys and write the most detailed military narrative in the ancient Near East about it.

The eighth campaign

Drafted from Zimansky, Anatolian and Caucasian Studies (1998)contraction

In 714 BC Sargon II went into the highlands, and afterwards had an account of it written as a letter to the god Ashur. It runs to something over four hundred lines and it is the most detailed military narrative surviving from the ancient Near East.

It is detailed in ways royal inscriptions usually are not. The terrain is in it — gorges the chariots had to be taken apart to cross, mountains described as standing like the blade of a dagger. The exhaustion is in it: the troops complaining, the water problem, the king on foot. The intelligence is in it, including reports from informers. It reads, in places, like a campaign diary that somebody afterwards addressed to a god.

The campaign avoided the Urartian fortresses, which Sargon evidently did not propose to besiege, met Rusa I in open country in the mountains and beat him, marched through the country burning, and then turned aside on the way home to sack Musasir.

Musasir

Musasir was the principal sanctuary of Haldi, the god in whose name every Urartian king says he does everything. It lay outside the kingdom's core, in the hills on the Assyrian side, which is why it was reachable. Sargon took the temple treasury, and the letter itemises it at extraordinary length — the weights of gold and silver, the vessels, the furniture, the statue of the god himself, carried off to Assyria. A relief in Sargon's palace at Khorsabad showed Assyrian soldiers stripping the temple, and it is one of the few images of an Urartian building made by anyone.

Rusa, the letter says, heard what had happened, threw himself on the ground, tore his clothes and took his own life with his own iron dagger.

What actually happened to Urartu

It survived. That is the part worth holding onto, because the document is overwhelming and the outcome was not.

Urartu was not annexed, not made tributary, and not destroyed. Its kings went on building fortresses for another century — some of the largest, at Ayanis and Bastam, are later than this campaign. What it stopped doing was contesting Assyria. After 714 the two states are largely at peace, and by the seventh century they appear to have had a working relationship against the nomadic pressure coming down from the steppe, which had become the more serious problem for both.

So the two great events in Urartian history are both Assyrian victories, and neither of them ended the kingdom. It was built to survive the Assyrian army and it did exactly that, at the cost of never being anything more than a state that survived.

There is no way of scoring that. Empires are measured by what they took and what they spread, and a kingdom whose achievement was continuing to exist for two and a half centuries against the most efficient army in the world registers on neither.

Nobody knows

Drafted from Zimansky, Anatolian and Caucasian Studies (1998)end

Urartu is one of the very few kingdoms of the ancient world whose ending cannot be described at all.

Whether it was conquered is open. Whether the destruction levels at the great fortresses represent one episode or several spread over decades is open. Who the agent was — Medes, Scythians, Cimmerians, some combination, or nobody — is open, and each candidate rests on a late Greek or biblical source rather than on anything contemporary. Even the date is a convention: 590 is a round number standing in for a process somewhere in the first third of the sixth century.

This is a stronger kind of ignorance than usual. For most vanished states the event is known and the cause is argued — everyone agrees Hattusa was emptied and burned around 1180, and the dispute is about why. Here there is no cause to dispute, because there is no event to attach one to.

Nineteen years

The one comparison worth making is with the neighbour.

Assyria ended in 609, comprehensively: the great cities destroyed and never reoccupied, the language of the court no longer written, and within two centuries Greeks passing the mounds were being told they had been raised by somebody else entirely.

Urartu was a seventh of Assyria's size at its peak and was built specifically to survive Assyria. It did — by about nineteen years on the conventional dating, which is close enough to be a coincidence and satisfying anyway. What it did not survive was whatever came out of the steppe afterwards, and the same disappearance of the written record that makes its ending unknowable is why nobody can say what that was.

What did not carry over

Armenia occupies this ground later, and the connection is weaker than the geography suggests. Armenian is Indo-European; Urartian is not, being related to Hurrian and to very little else. No source shows an Armenian dynasty claiming Urartian legitimacy. The Assyrians' name for the country, Urartu, survives as the biblical Ararat and as a mountain's name, and a name on a map is not an inheritance.

What did carry over went to the enemy. Urartian cuneiform is Neo-Assyrian cuneiform adapted to write a language unrelated to Akkadian — so the one thing this kingdom's writing passed on, it passed to the state it had been built to keep out.

The rule nobody wrote

There is a pattern across this whole region worth ending on, and Urartu completes it.

Look at the dynasties whose successions are undocumented: Ur III, Old Babylonian, the Egyptian Old Kingdom, and this one. Four dynasties across three writing systems and two thousand years, each of which passed the throne down for centuries without recorded difficulty, and not one of which wrote down how it was supposed to work.

Now look at the two that did write it down. Hatti produced Telipinu's Edict around 1500, after a century in which nearly every accession had been settled by murder. Assyria produced Esarhaddon's succession oath in 672, drafted by a king who had taken the throne in a civil war caused by his father's assassination.

So the ancient Near East does not divide into states that had succession rules and states that did not. It divides into states that had been badly hurt by not having one and states that had not yet been. A written constitution for the throne is a scar, and the dynasties whose successions worked never saw the need to acquire one.

The mountain moved

Drafted from Sayce, The Hittites: The Story of a Forgotten Empire (1888)afterlife

The kingdom disappeared around 590 BC without anybody recording how, and what outlived it was a word in somebody else's book — attached, eventually, to the wrong mountain.

Where Ararat was

Ararat was what the Assyrians called the country around Lake Van and the district south of it. That is where the name sits in the Hebrew scriptures, where the ark comes to rest on the mountains of Ararat and where the sons of Sennacherib flee after murdering their father.

It was not, Sayce notes, extended northward to the peak now called Mount Ararat until post-biblical times. The title originally belonged to the Kurdish range in the south. A country's name became a mountain's name, and then moved about a hundred and fifty kilometres to a different mountain, and stayed there.

So the most famous thing about Urartu is a misattribution of its own name, and the state that carried the name for two and a half centuries had nothing to do with a flood story.

What survived to be read

The inscriptions did, scattered across eastern Anatolia in a borrowed script and an unrelated language, and they were unreadable for a very long time.

The script could be sounded out once Assyrian cuneiform was deciphered, which is why nineteenth-century scholars could copy the texts and pronounce them without knowing what they said. Sayce, writing in 1888, was among the first to work through them and to establish what they contained: the buildings, the offerings, the campaigns, the names of Menuas and Argistis and the god Khaldi.

Everything now known about the kingdom's own view of itself comes from that material, and the material was standing on cliff faces and fortress walls in plain sight the whole time.

A state with no successors, and a country with one

Nobody claimed to continue Urartu. No later kingdom took its titles and no dynasty traced itself to its kings, and the god Khaldi disappears from the record altogether — which is the strangest part, because gods normally outlast the states that endow them.

The country carried on under another name. When Darius had his account of his accession carved at Behistun in three languages, the province the Old Persian version calls Armina is called Urashtu — Urartu — in the Babylonian one. The Persians had a new name for the place and the Armenians were becoming its people, and Babylonian scribes went on writing the old name for a century after there was no kingdom left to attach it to.

The kingdom was built for one purpose — to survive the Assyrian army — and it achieved that for two hundred and seventy years, which is longer than the empire it was built against lasted. Then Assyria fell, the reason for the fortresses went with it, and within a generation or two Urartu was gone as well, with no enemy left to explain it.

What is left is a lake with the kingdom's own name in it, a mountain in the wrong place carrying the name its enemies used, and several hundred inscriptions in a language that has no relatives.