Third Dynasty of Ur

After the Gutians

Drafted from (ed.), Routledge Worlds (2013)formation

Akkad fell around 2154 BC, and what came next is conventionally a dark century: the Gutians, a people out of the Zagros, held some part of Babylonia, and the later Sumerian tradition treated them as a catastrophe visited on the land for Naram-Sin's impiety. How much of that is true and how much is a literary construction written three hundred years afterwards is very much open.

What is not open is what happened at the end of it. A governor of Uruk named Utu-hegal defeated the Gutian king; a governor of Ur named Ur-Nammu, who may have been his brother, then made himself king of Ur, and within a few years held southern Mesopotamia from the Gulf to the latitude of modern Baghdad. The line he founded lasted a hundred and eight years and five kings of one family, with one throne passing to a brother rather than a son and not one disputed accession on record.

What they claimed to be

Ur-Nammu did not present himself as a new thing. He took Akkadian titles — king of Sumer and Akkad, and later king of the four quarters, which was Naram-Sin's — and he kept Akkadian in the administration alongside Sumerian.

It is worth being precise about what that claim was and was not. It was not a succession. A century separates the two states, the Gutians are in between, and nothing passed from Akkad to Ur in the ordinary sense — no institution, no family, no army. What passed was a vocabulary of kingship, publicly asserted by a man in a position to assert it, three hundred years after the last king who had used it.

Sumerian, one last time

The dynasty is called a Sumerian revival, and the label needs handling. Sumerian was the language of the royal inscriptions, the hymns, the law collection and the overwhelming bulk of the administrative record. It was almost certainly no longer the language most people in southern Mesopotamia spoke; Akkadian had been spreading for centuries and would be the only one left within a few generations of the dynasty's fall.

So this state wrote in Sumerian on a scale nobody had before and nobody would again, and one of the things it was doing was preserving. The scribal schools it founded are where the Sumerian literary corpus was fixed — the hymns, the laments, the debate poems, the king list — and everything later Mesopotamia knew about Sumer, it knew because this dynasty's schools copied it — including the lament for the fall of Ur itself, composed by scribes trained in the institutions that fell.

What Ur-Nammu built

Drafted from King, A History of Sumer and Akkad (1910)formation

The founder of the dynasty is best known through masonry. Ur-Nammu — read Ur-Engur by the excavators of King's generation — left his name stamped into kiln-baked bricks across southern Mesopotamia, and the largest thing he left is still standing.

The tower

The great temple-tower at Ur is his. He faced it with baked brick carrying his own name and inscription, and it is the form of ziggurat that everything later copies.

But the excavation showed something more interesting than a new building. Inside the structure were the cores of earlier and smaller towers — one of them raised by Naram-Sin, and others from before the Akkadians altogether. Ur-Nammu did not invent the ziggurat. He built a larger one around the ones already there, which is a different act and a more revealing one.

He reworked the whole precinct to match. He raised the level of the inner court above Naram-Sin's pavement. He straightened the line of the outer wall, using Naram-Sin's wall as a foundation wherever the two crossed, and extended the circuit to take in the mounds where much of the temple archive was later found.

What building on top of somebody meant

A ruler who encloses his predecessor's monument inside his own is making a claim that no inscription has to state.

The Akkadian empire had ended less than a lifetime before, and the Gutian interval between was the thing the Sumerian tradition would spend centuries explaining as divine punishment. Ur-Nammu's programme answers it physically. The old pavement is not destroyed, it is buried under a higher one. Naram-Sin's wall is not demolished, it is used as footing. The Akkadian tower is not replaced, it is inside the new one.

The message is continuity rather than conquest: this is the same holy place, run properly again, by someone who belongs in the sequence.

The long view from the same trench

The site keeps a record of everyone who did this after him.

Above Ur-Nammu's level, Kassite-period buildings across the water held the temple archives in their turn. Samsu-iluna — Hammurabi's son, four centuries later — left a clay cone in the debris of the temple court recording that he had dammed the Euphrates and cut it a new channel, probably to bring water to the part of the city that had grown up on the far side.

Ur-Nammu's ziggurat outlasted his dynasty by about sixteen centuries as a working building, and it is the reason a state that governed for barely a hundred years is visible on the ground at all. The hundred thousand tablets tell you how it was run. The tower is what it was for.

A hundred thousand receipts

Drafted from (ed.), Routledge Worlds (2013)expansion

There are more published administrative tablets from this hundred and eight years than from any other comparable stretch of the ancient world. The number usually given for the Ur III corpus is somewhere well over a hundred thousand, and it is still rising, and none of it is literature. It is receipts.

A dead sheep is recorded, with the date, the animal's condition, the official responsible for it and the official who received the carcass. A day's labour by a named woman in a milling establishment is recorded, with her ration. Reed bundles, fish, silver, barley, hides, the movement of oxen between provinces — all of it entered, all of it dated to the day by a system in which every year has an official name, the names are listed in order on tablets kept for the purpose, and the sequence has no gaps.

The bala

What the receipts show is a fiscal system unlike anything else attested for two thousand years. The Ur III state ran an arrangement called the bala: each province delivered into a central pool during its assigned months of the year and drew from that pool during the others. It is a redistributive rotation rather than a straightforward tax, and it is reconstructed entirely from the delivery records themselves — nobody in antiquity described it. Beside it, the gun mada, a levy in cattle and sheep owed by the militarised zone on the eastern frontier.

It is the earliest tax system anywhere that can be described in operational detail, by more than a thousand years, and the next one that can be described at this resolution is Neo-Babylonian, at the far end of Mesopotamian history.

What the receipts do not supply is a figure for how much ground the dynasty held. The standard tabulations of ancient empires give extents for Akkad, for Mitanni, for Hatti, for Egypt's New Kingdom, for both Assyrias and for Neo-Babylon, and they do not carry Ur III at all.

So the state whose provincial deliveries can be counted sheep by sheep is one of the few in the region with no published size. Documentation is not size, and the thing this dynasty is remarkable for is not the thing empires are usually measured by.

Shulgi

The reign that built the apparatus is the second and longest, forty-seven years. Standardised weights and measures, a single state calendar imposed across the provinces, a reformed accounting system with a standard year of three hundred and sixty days for the bookkeeping, the scribal schools, a road system with staging posts, a reorganisation of the army, and the conversion of the eastern frontier into a belt of garrisoned provinces.

None of that can be dated to a year; the reforms are spread across decades of one reign. What can be dated is the thing that came with them: in his twenty-first year Shulgi was made a god.

The king becomes a god

Drafted from (ed.), Routledge Worlds (2013)peak

From his twenty-first year Shulgi's name is written with the cuneiform sign that marks a divine name. Hymns are composed to him in the first person in which he runs from Nippur to Ur and back in a day, plays every instrument, reads any tablet put in front of him and knows five languages. Offerings are made to him in temples. His statues receive a cult. He is not a king favoured by gods; he is listed among them.

Naram-Sin of Akkad had done it first, three centuries earlier. What is different here is that it stuck: every king of this dynasty after Shulgi was divine, and the practice continued under the kings of Isin for another century.

It is worth keeping the distinction sharp. Rulers claiming divine sanction are everywhere in history, and what they are claiming is that a god approves of them. This is not that. Shulgi's name is written with the sign that marks a deity, his statues receive offerings, and hymns are addressed to him among the gods.

Why

The honest answer is that nobody knows, and the good arguments run in two directions. One is that the deification is a response to a governing problem: this state was assembling a bureaucracy across provinces that had been independent city-states with their own gods and their own loyalties, and a divine king is a focus of allegiance that outranks a local temple. The other is that it is a continuation of much older Sumerian ideas about the king's ritual marriage to the goddess Inanna, formalised.

What can be observed is the timing. The deification arrives with the administrative reorganisation, not before it and not after, and the hymns that celebrate the king as a god are also the texts in which the scribal schools that the reorganisation created taught boys to write.

The wall

Two reigns later, Shu-Sin built a wall. It is called muriq-Tidnim, "the one that keeps Tidnum at a distance", it ran for something like two hundred and seventy kilometres between the Tigris and the Euphrates, and it was directed against the Amorites — pastoralists moving into Mesopotamia from the west, whom the Ur III texts describe in a set formula as people who do not know grain, who do not bury their dead, who eat raw meat.

Nothing about the state's trajectory turned on the year the wall was finished. What it is worth having is the state's own reading of its position. A hundred years after Ur-Nammu, the most centrally administered polity in the world was building a fortified line against people who had no state at all, and within thirty years an Amorite would be a king in Mesopotamia, and within three hundred the dynasty this region's Babylon record begins with would be Amorite too.

The wall did not work. Neither does the explanation that the Amorites brought this dynasty down: the last chapter is about what the tablets actually show, which is more complicated and considerably worse.

The grain

Drafted from Michalowski, Mesopotamian Civilizations 15 (2011)contraction

Ibbi-Sin's reign is twenty-four years long and the archives fail during it, which means the collapse of the best-documented state in the third millennium is documented in the act of failing.

The provincial deliveries stop, one province at a time, and the sequence can be followed. Eshnunna's last dated tablet is in Ibbi-Sin's second year; Susa's in his third; Umma's in his fifth; Girsu's in his sixth or so; and by his eighth the bala system has effectively ceased and the year-names have stopped mentioning anything but Ur itself. And in his sixth year the price of grain at Ur, which the ration records let us track, goes up by something like a factor of sixty.

What caused it

This is genuinely argued about and will not be settled here. The candidates: Amorite movement cutting the roads; a failure of the irrigation system in the south, possibly from salinisation, possibly from a shift in the Euphrates channel; a climatic downturn; and the possibility that the bala itself was the fragility — a rotation in which every province depended on every other one is efficient until one link fails, at which point the failure propagates.

What the tablets show without ambiguity is the shape rather than the cause: a central redistribution failing outward from the periphery, and a capital paying sixty times the normal price for barley while its own accounting system is still issuing correctly formatted receipts.

Ishbi-Erra

The one narrative source for this is a set of letters between Ibbi-Sin and his officials, and it is at once the most vivid material about the dynasty and the least trustworthy. Michalowski's edition is the reason both halves of that sentence can be said.

The story they tell: Ibbi-Sin sent an officer named Ishbi-Erra north with a large quantity of silver to buy grain. Ishbi-Erra bought it — twenty gur of barley, the letter says — and then wrote to say that Amorites had entered the country, he could not get the grain through, and would the king make him governor of Isin and Nippur so that he could hold the region. The king's reply is furious. Within a few years Ishbi-Erra was a king at Isin, holding the north, and Ibbi-Sin held Ur.

Michalowski's warning is that these letters survive only as school copies made centuries later, in multiple recensions that disagree with each other. They are literary compositions about the fall of Ur before they are correspondence, and the events in them may have been shaped by scribes who already knew how the story ended. Something in the outline is almost certainly true — Ishbi-Erra was real, he did take Isin, the dates fit — and no individual sentence in the letters can be relied on.

That is an ordinary condition for ancient history and an unfamiliar one for this dynasty. Almost everything else known about Ur III comes out of administrative documents that had no motive to lie about anything. The central episode of its collapse comes out of a school text, and the difference is worth feeling.

2004

Drafted from (ed.), Routledge Worlds (2013)end

An Elamite coalition took Ur, sacked it, and carried Ibbi-Sin away east to Anshan, where he disappears. A garrison was left in the ruins and stayed for several years until Ishbi-Erra of Isin drove it out — which allowed the kings of Isin to present themselves as the restorers of Ur.

It is worth being exact about who took the city, because "the Elamites" covers three different things. The Elam of the fourteenth and twelfth centuries and the Elam of the eighth are later kingdoms with their own histories. The coalition that took Ur in 2004 belongs to the Elam of Awan and Shimashki, six hundred years earlier, a federation with its own dynasties and a succession that passed from brother to brother and then to a sister's son, which is a rule almost nothing else in the ancient Near East used.

And they did not stay. Nobody ruled Babylonia from Elam after 2004. The conquerors took the king, took the statue of the moon god, sacked the city and went home.

The lament

What survives from the moment is not a chronicle. It is a poem — the Lament for Ur, and there are several others for Sumer, for Nippur, for Uruk, for Eridu — in which the goddess Ningal pleads with the assembly of the gods for her city and is refused, and the city is destroyed because the gods had decided it would be.

The people, like broken shards, littered the streets.

The laments were composed, or given their surviving form, in the scribal schools of the kingdom of Isin, by the successors of the men this dynasty had trained. They were copied for centuries afterwards as school texts and as liturgy, which is how they survive. What that authorship actually meant for the poem, and for how much its account of the fall can be trusted, is worth taking slowly on its own.

Put that beside the archives. The administrative record of receipts, deliveries, rations and prices is contemporary, disinterested, and stops abruptly. The literary record is retrospective, theological, and never stops. They answer different questions and cannot check each other.

What the dynasty left

The Sumerian language stopped being a spoken language of government within a few generations, and after this every regime in Mesopotamia is Akkadian-speaking and most of them are Amorite. This is the last time the political centre of the Near East is in the far south.

What did not stop was the writing. Ur-Nammu's law collection — the earliest surviving one in the world, two centuries before Hammurabi's, with the same structure of prologue, conditional provisions and epilogue — was copied in schools long after the dynasty; Roth prints it alongside Hammurabi's and the Hittite laws in a single volume, which is a fair picture of how this region's legal writing actually travelled. The hymns and the laments and the king list went the same way.

And nobody wrote down how the throne passed. Five kings of one family, a hundred and eight years, not one disputed accession, and nothing in the whole enormous archive states the rule; Old Babylon is the same across eleven generations. Mesopotamia legislated the price of hiring an ox, the penalty for a badly built house and the compensation for a slapped face. It did not legislate this, or did not write it anywhere that has since been found.

The lamentation

Drafted from Kramer, Lamentation over the Destruction of Ur (1940)aside?What aside means

Sumerian literature produced a form for exactly one purpose: mourning a destroyed city. The lament for Ur is the fullest of them, and it is a poem about the fall of this state, written afterwards, in the language of the people it happened to.

The storm

The catastrophe is not narrated as an invasion. It is narrated as weather.

Enlil, having taken the good storm away from Sumer, calls down another: an evil storm, a storm that annihilates the land, the great storm of heaven, an afflicting storm, a destructive storm. It is helped by evil winds, by fire, by darkness and by unbearable heat. Kramer notes that of the thirty-two lines of one song, only two mention any destructive agent that is not an element.

The enemies are there — Sutians and Elamites, breaking up the Ekishnugal with pickaxes — but they are almost incidental to the way the poem understands what happened. What struck Ur was decided in a divine assembly, and the men with the axes were the instrument.

What was in the streets

Then the poem becomes specific, and stops being figurative at all.

The people of Sumer lay on the walls of their cities and their dead filled the gates. Those killed by weapons lay unburied and untended; those who escaped the weapons were brought down by the storm. Weak and strong alike died of famine. Parents who stayed in their houses were caught by the fire. Suckling infants were carried off by the water. The breasts of nursing mothers were pried open.

Judgment and counsel perished in the land. Parents abandoned their children and husbands their wives, and everything they owned lay scattered.

The goddess who has nowhere to go

The centre of the poem is Ningal, Ur's own goddess, and what makes it unbearable is that she is treated as a displaced person rather than as a deity.

Her city has been cursed by Anu; Enlil has turned hostile to her house. Dust has gathered in her rivers and the sparkling water does not flow. There is no grain in her fields and the field worker is gone. Her palm groves and vineyards have brought forth mountain thorn. Her possessions have been carried off upriver and down; her gold and stone and lapis lazuli lie scattered, and her ornaments are now on the bodies of people who do not know what precious stone is. Her sons and daughters have been taken away captive. She is no longer queen of Ur.

Where, the poem asks on her behalf, shall she sit down, and where shall she stand up? She goes out from the city and finds no rest; she goes out from the house and finds nowhere to live. She is a stranger in a strange city, and curses and abuse are heaped on her there.

She goes to her husband to weep about the house and city that have been destroyed. And then:

Like a fallen ox she will lie down beside the ruins of her city and her house and will not rise up.

Kramer's rendering of the poem's own verdict on the destruction is that it was attacked without cause.

What it was for

This was not a private grief. Laments of this kind were liturgical, composed and performed for the rebuilding of temples, and the copies that survive were made by scribes in the kingdom of Isin — the state whose kings had taken Ur back from the Elamite garrison and presented themselves as its restorers.

So the poem is doing two things at once. It records, in detail nobody had any motive to invent, what happens to a city when the state that fed it stops working. And it is also an instrument of the successor regime, because a city mourned properly is a city whose restoration means something.

The most complete account of the end of this polity was written by the people who profited from it, for use in a ceremony.

The kingdom continues

Drafted from King, A History of Sumer and Akkad (1910)afterlife

Elam took Ur in 2004, carried off the last king and the statue of the moon god, and left a garrison in the ruins. Then, as far as anyone can tell, they went home.

That absence of any Elamite administration afterward is the more surprising once it is set beside what did fill the gap. Leonard King, surveying the period in 1910, is the source for how completely: the people who destroyed the most thoroughly administered state in the world did not take over its administration, and made no attempt to.

What Isin did instead

What followed in southern Mesopotamia was continuity of a kind that makes the word "fall" awkward.

The kings of Isin, beginning with the officer who had walked away from Ibbi-Sin with an army and a grain contract, called themselves king of Sumer and Akkad. They used Ur's titles, kept the divine determinative in front of their names for another century, dated by year-names in the same style, and went on issuing law in the same forms — the collection of Lipit-Ishtar of Isin survives, and it sits in the same tradition as Ur-Nammu's before it and Hammurabi's after.

They also rebuilt Ur, restored the temple of the moon god and installed their own daughters as his priestesses, which is what the kings of Ur had done.

King's judgement is that the kingdom of Sumer and Akkad was continuous across 2004. A dynasty ended. The state did not obviously stop.

Why it fell, on the evidence available

The candidates already on the table are Amorite disruption, irrigation failure, climate, and the bala system's own structural fragility. King's survey offers a fifth, distinct from all of them, and it is the one an administrator would reach.

He attributes the decline of Ur to what happens when a highly centralised administration is combined with a complete delegation of authority by the king: provincial governors accumulating several offices at once, and misgoverning the provinces they held. Not invasion, and not climate — the failure mode of an over-organised state whose organisation depended on the centre paying attention.

That is a reading from 1910 and much has been added since. It has the merit of being drawn from the same documents the dynasty is known by. The state that recorded every sheep is the state whose collapse can be read in the recording.

What actually survived

Not the territory, which broke up; not the dynasty, which ended in captivity in Elam.

What survived was everything the scribal schools had written down. The Sumerian literary corpus was fixed under Ur III and copied for the next fifteen hundred years by people who did not speak the language. The lament for Ur's own destruction became a literary form, and was reused for other cities. The year formulas, the weight standards Shulgi had reformed, the calendar, the shape of a law collection and the practice of dating documents by an event all continued without interruption into Isin, into Larsa, and into Babylon.

Ur III is the earliest state whose bureaucracy is fully legible, and the reason it is legible is that its own successors kept the system running and its own schools kept teaching the texts. A hundred thousand receipts survive because nobody ever decided to stop writing them that way.