The Sumerians

← Encyclopedia

In January 1869, before the Société Ethnographique in Paris, a German-born French Assyriologist named Jules Oppert proposed that a people no one had ever heard of be given a name. The cuneiform tablets coming out of the mounds of Iraq recorded, under the Semitic Akkadian of Babylon and Assyria, an older and wholly different language — a tongue with no Semitic kin, plainly the speech of whoever had invented the wedges in the first place. The Assyrian kings styled themselves “King of Sumer and Akkad.” If Akkad was the Semitic half of that old realm, Oppert reasoned, then Sumer must be the other half, and its lost people the Sumerians. He had, in effect, named a civilization back into existence. It had been gone so completely that even the Hebrew scriptures, which remembered Babylon and Asshur and Nimrod, preserved no word for the people who had taught those nations to write.

The Sumerians are the inhabitants of the southern Mesopotamian alluvium — the flat, hot, river-laid plain at the head of the Persian Gulf, in what is now southern Iraq — from before 3000 BCE. They built the first things the world calls cities: Uruk, Eridu, Ur, Lagash, Nippur, Kish, Umma, Shuruppak. They pressed the first writing into clay. They divided the circle into three hundred and sixty parts and the hour, eventually, into sixty minutes, and the world still keeps their count. And then, as a spoken people, they vanished into the population that surrounded them, leaving their language to live on for two thousand years as the dead and sacred tongue of the temple and the school — until even that ended, and they were forgotten entirely, and had to be found again under the ground that buried them.

A people deduced, not remembered

Where the Sumerians came from is a question the evidence does not answer. Their language is an isolate: it is related to no other known language, living or dead, and a century and a half of attempts to marry it to Turkish, to Hungarian, to Dravidian, to Caucasian and beyond have failed to convince. It simply stands alone. Whether its speakers walked down into the plain from somewhere else, or were the descendants of the marsh and farming villages of the preceding Ubaid period (roughly 6500–4100 BCE) who became Sumerians where they stood, is genuinely unsettled. Some have noted that several of the oldest city-names — Eridu, Ur, even the word for the river-plain — do not parse as Sumerian, and have argued for a still earlier substrate people beneath them; others find the evidence too thin to bear the weight. The record opens with cities already built and writing already in use. Of any before, the plain keeps its silence.

They called themselves, in their own tongue, ùĝ saĝ gíg-ga — the “black-headed people” — and their land ki-en-gir, glossed as something like the place of the noble lords. The label Sumerian is not theirs at all but ours, back-formed by Oppert from a Semitic word for the same region. This is the strange fact at the root of the subject, and it has a name among scholars: the Sumerian problem. Two unrelated languages — Sumerian, the isolate, and Akkadian, an early Semitic tongue — grew up intertwined in the same small country, sharing one script, borrowing each other’s words, until a thousand years of contact had made them so alike in turns of phrase that telling the two peoples apart in the archaeological record is, in places, nearly impossible.

The recovery itself was a feat of pure inference. In the 1840s and 1850s, when Henry Rawlinson, Edward Hincks, and Oppert were prying open the cuneiform of Nineveh and Babylon, the script they deciphered wrote Akkadian — a Semitic language, cousin to Hebrew and Arabic. But the writing system fit that language badly. It was full of signs used for their sound alone, syllables that meant nothing in Semitic, and long lists in which a single line gave an Akkadian word beside an entirely foreign one. By 1850 Hincks had drawn the conclusion: the inventors of cuneiform had not spoken Akkadian. Some earlier, non-Semitic people had built the script to fit their own language, and the Semites had borrowed it, the way the Romans borrowed an alphabet not made for Latin. Rawlinson and Hincks called that ghost language “Akkadian” — exactly backward, as it turned out. It was Oppert, in 1869, who gave it and its speakers the name that stuck. A whole people, that is, was first detected as a discrepancy in somebody else’s spelling.

An administrative tablet in proto-cuneiform from the Uruk III phase, late fourth millennium BCE: accounting, the seed of writing. Louvre, AO 19936. Photo: Mbzt, CC BY 3.0.

What the black-headed people made

To live on the southern plain at all was an act of engineering. It has almost no stone, no timber, no metal ore; it gets too little rain to grow grain, and its two rivers flood in spring, after the crop is already in the ground, not when it is wanted. What the land has is silt and water that can be led. Farming there meant canals and dikes, and canals and dikes meant pooled labor, stored surplus, and someone to command both — and that, in the late fourth millennium, is the seed of the city. By around 3300 BCE one of those cities, Uruk, had swollen behind its wall to tens of thousands of people, the largest settlement the world had yet held. The Sumerians did not invent the village. They invented the dense, walled, administered, monumental thing that succeeds it.

And to run it they invented writing. Cuneiform began as bookkeeping. Its remote ancestors are small clay counting-tokens — a cone for a measure of grain, an ovoid for a jar of oil — sealed inside clay envelopes and then, more sensibly, just impressed on the outside. The oldest tablets, from Uruk around 3300–3200 BCE, are inventories: commodities drawn as little pictures, tallied with numerals, with no grammar and no sentences, a technology for remembering how much barley went to whom. Over the following centuries the pictures stylized into clusters of wedges struck with the cut end of a reed; the signs took on sound-values as well as meanings; and an accounting habit grew into a script that could carry a hymn, a lawsuit, a lament, an omen, a love song. It proved astonishingly portable. Because the wedges could be assigned to sounds, they were borrowed to write Akkadian, then Elamite, Hittite, Hurrian, and more — languages utterly unlike Sumerian — so that a southern Mesopotamian way of keeping store-accounts became the writing of half the ancient world for two thousand years.

The “war” side of the Standard of Ur, c. 2600 BCE, from the Royal Cemetery of Ur: ranked infantry, the four-wheeled battle-cart, the captive and the king. British Museum. Photograph: public domain.

Their mathematics was as durable as their script. The Sumerians counted in base sixty — the sexagesimal system — and they passed it to the Babylonians, and through them it has never left us. Sixty is a generous number: it divides evenly twelve ways, by two and three and four and five and six, which makes fractions behave. The hour of sixty minutes, the minute of sixty seconds, the circle of three hundred and sixty degrees, the degree of sixty arcminutes — all of it is Sumerian counting, still running under the surface of every clock and every navigator’s chart. It is the longest-lived invention of a vanished people: a way of dividing time and the sky that outlasted the people, the language, and the very memory that there had been such a people.

The city they built was, in their own understanding, the property of a god. Each town was the household and estate of its particular deity — Enki at Eridu, the moon-god Nanna at Ur, Enlil of the air at Nippur, the volatile Inanna of love and war at Uruk, whose Akkadian continuation is Ishtar — housed in a temple whose stepped tower, the ziggurat, lifted the sanctuary toward the sky so the god might come down to it. The temple was not only a shrine but a great landlord, holding fields and herds and the labor to work them; worship and economy were one operation, and the god’s steward administered both. Sumerian gave this order a vocabulary of its own. The me were the fixed decrees that made civilization run — kingship and priesthood, truth and falsehood, weaving and metalwork and music, even the forms of strife — each imagined as a real possession the gods had set in place, a thing that could be held, granted, or carried off. In the local frame the order of the world was a set of offices, and to possess them was power.

Kingship lowered from heaven

Over all of it stood the king, and kingship itself the Sumerians traced to the gods. The document that records this — the Sumerian King List, copied and recopied for centuries and best preserved on a clay prism brought from Larsa to Oxford in the 1920s — opens with a line that states the whole theory of Sumerian sovereignty in a breath: “After the kingship descended from heaven, the kingship was in Eridug.” Kingship is not seized; it is lowered, a divine office set down first at Eridu, then passing — as cities rise and fall — from one to the next. The List then runs through rulers with reigns of tens of thousands of years, a flood that sweeps over the land, and a resumption of kingship at Kish afterward, sliding only gradually from myth into datable history. It is a political claim dressed as a chronicle: whatever city holds the upper hand at any moment holds the one kingship that heaven let down, and the others merely wait their turn.

The Weld-Blundell prism, c. 1800 BCE: the fullest copy of the Sumerian King List, opening “After the kingship descended from heaven…” Ashmolean Museum, Oxford. Photo: M.atkinsonross, CC BY-SA 4.0.

Because the southern plain was never one kingdom but a dozen rival city-states, this theory of a single migrating kingship was the nearest thing the Sumerians had to a national idea — a way of imagining all those quarreling towns as successive holders of one sacred office. The ruler of each, styled en or énsi or, in the fuller claim, lugal (the “big man,” the king), governed as the god’s tenant and steward: he led the army, judged the lawsuits, dug the canals, built the temple, and answered for the city’s fortune to the deity who owned it. The oldest surviving law collections come from this world, two centuries before Hammurabi’s famous Babylonian stele — the laws of Ur-Nammu of Ur set out the redress for a broken bone or a runaway slave in the same casuistic “if a man… then” that legal codes have used ever since.

How completely the king belonged to that sacred order, the ground itself records. In the mid-third-millennium Royal Cemetery at Ur, excavated in the 1920s and 30s, a handful of deep shaft-graves held their occupants — among them a woman named on her seal as Puabi — laid out in gold and lapis with lyres, headdresses, and chariots, and, in the richest pits, with rows of attendants who had gone down into the earth with their lord. The Sumerian dead did not pass into nothing. They descended to a dim land below, where the shade depended on the living to feed and name it, or it would come back hungry. The cups and bowls that fill these graves, outnumbering every other kind of object, are that duty made monumental: the sovereign departs as a whole household departs, provisioned for the road.

The afterlife of a dead language

The Sumerians did not fall so much as dissolve. Around 2334 BCE Sargon, a Semitic-speaking ruler from the city of Akkad to the north, conquered the southern cities and forged the region’s first true empire, ruling in Akkadian from the Gulf to the upper rivers. After Akkad collapsed there was one last, brilliant Sumerian revival — the Third Dynasty of Ur, roughly 2112 to 2004 BCE, which raised the great ziggurat of Ur, generated a bureaucratic archive of staggering size, and saw much of the classic Sumerian literature written down and ordered. When Ur fell, to Elamite attack from the east and Amorite pressure from the west, power passed for good to Akkadian-speaking dynasties — Isin, Larsa, at last Babylon. By around 2000 BCE, and certainly within a century or two after, Sumerian had ceased to be anyone’s mother tongue. In the street and the field it had been displaced by Akkadian. The black-headed people were now simply part of the Akkadian-speaking population of the south, indistinguishable in the record from their neighbors. As a living people, they were finished.

As a language, Sumerian had two thousand more years to run. Long after the last child grew up speaking it, the scribes went on learning it — parsing it, copying it, reciting it in the temple — exactly as medieval Europe went on with a Latin no one spoke at home. It became the prestige tongue of ritual and scholarship, the language a Babylonian priest chanted a lament in, the language a schoolboy ruled his clay tablet to practice under the Persians and even the Greeks. The school that taught it, the é-dub-ba-a or “tablet house,” kept the old literature alive by sheer repetition: the disputes and debate-poems, the city-laments, the cycle of Inanna’s descent to the land of no return, and the Sumerian tales of the hero-king of Uruk that were gathered and recast in Akkadian as the Epic of Gilgamesh.

Tablet XI of the Akkadian Gilgamesh, the “Flood Tablet” from Ashurbanipal’s library at Nineveh, seventh century BCE — the Babylonian heir of a Sumerian flood story. British Museum, K.3375. Photo: BabelStone, CC0.

Through that long afterlife the Sumerian inheritance passed downstream more or less whole. The gods kept their names under Akkadian translation; the myths were retold; the king-list theory of sovereignty, the temple-economy, the law code, the school all carried forward into Babylon and Assyria and, through them, far beyond. The flood that drowns the world and the one righteous man who builds a boat — a story with notable resemblances to the flood of Genesis, resemblances that scholars have debated since the Assyriologist George Smith first read a Babylonian version of it aloud in 1872 — was told in Sumerian first, in the tale of the pious king Ziusudra. The conviction that the night sky was a text the gods had written, and that a trained reader could parse it, began among these southern cities and grew, over the following two millennia, into the celestial divination that the Greeks and Romans would call the wisdom of the “Chaldeans” — a label more Greek invention than ethnic fact, but resting on a science that was Mesopotamian to its root.

That root, by the time anyone in the West admired it, had no name. The later ages that traced their astrology and their flood-story and their towers to a fabled Babylon did not know there had been a people before the Babylonians who made the writing those Babylonians used. The Sumerians had slipped beneath the record of their own heirs. It took the cuneiform itself to give them back — the very thing they had invented, read at last after two thousand years of silence, confessing in its own awkward signs that someone else had made it. A civilization that taught the world to write left, as its last word, the evidence that it had once existed at all.

→ Related: Sumer · Eridu · Ur · Nippur · Uruk · Gilgamesh · The Assyrians · The Elamites · Mesopotamia · Babylon · Babylonia · Akkadian Empire · Mesopotamian Religion Sumerian Akkadian · Ishtar · Marduk · Anu · Berossus · Oannes · Divination · Astrology · Comparative Flood Myth · Chaldeans