Our Deep History Our Deep History A museum of unexplained ancient sites and relics
Iraq rooms

Probably Sippar; copies in London and New York, Iraq

Atrahasis

Copied in the reign of Ammi-saduqa (a colophon, not a radiocarbon date)

The Atrahasis epic, a large tablet copied in the reign of Ammi-saduqa, probably at Sippar, now in the British Museum. The colophon, the scribe's note at the end, dates the copy, and the plot gives the gods a reason for the flood. SpeakingArch, CC BY-SA 4.0. Shown unedited.

Humans are too loud for the gods, so one god warns one man, and the flood is already a story in the Old Babylonian period. One published chronology places the copy near 1635 BC. A 2016 wood study reports a fit there or eight years later, and does not close the other chronologies.

Why this is extraordinary

Humans are made so the gods can stop working. They become numerous, and their noise keeps Enlil from sleep, so he sends plague, then famine, then a flood. Enki, the god of Eridu, warns Atrahasis to build a boat. The flood comes. The survivor's sacrifice draws the gods, and limits are set on human increase. The writing is cuneiform, wedge-shaped signs pressed into the clay. The best Old Babylonian copy carries a colophon of the reign of Ammi-saduqa, a note the scribe added to date his own work. Three chronologies are still argued. One of them, the middle, places that reign near 1635 BC. The high and the low move it by more than a century. A 2016 study of tree rings and radiocarbon at Kültepe and Acemhöyük reports that the timbers fit the middle chronology or a date eight years later, and argues that the high, the low, and the new chronologies do not fit the wood. The study says it does not decide between the two nearer dates. The low chronology is the step of several decades. The eight-year date is a different reading. The year remains a fit to the Venus observations, records of the planet Venus, not a carbon date of this clay.

The same flood speech, with another hero's name, was copied into Gilgamesh Tablet XI for Ashurbanipal's library at Nineveh. The Sumerian story of Ziusudra is older in language and is a different tablet. None of the three is a core from the end of the Ice Age. The Younger Dryas, a cold return about 12,900 to 11,700 years ago, sits thousands of years before any of these copies. A memory could in principle be that old. The tablets do not say so, and the silt layers at Ur, Kish and Shuruppak are local and are not one date.

The poem is evidence that a flood, a boat, and a warning were already a Babylonian story in the Old Babylonian period. It is also evidence of a plot, noise and punishment, which a river flood does not write by itself. The silt is weighed in the Shuruppak room and across the versions in the flood tablets room.

Silt, then the written stories. The Ice Age cold snap is older than the left edge of this drawing, and no tablet is placed there. Diagram: Our Deep History.

Why it matters

Atrahasis is the flood story Gilgamesh later borrows. Its date, its reason for the flood, and the silt under the cities can be set side by side without forcing them to be one event.

Enki, god of Eridu, is the warner. Enlil is the sender. The later retelling, with Utnapishtim as speaker, is Gilgamesh Tablet XI. The Sumerian telling, with Ziusudra and the founding of cities, is the Eridu Genesis. The yellow layer under the flood hero's city is Shuruppak. This room is the Old Babylonian poem.

Photo tour

5 photographs, unedited. On a phone, swipe the row. Choose one to look closely.

The Atrahasis epic, a large tablet copied in the reign of Ammi-saduqa, probably at Sippar, now in the British Museum. The colophon, the scribe's note at the end, dates the copy, and the plot gives the gods a reason for the flood. SpeakingArch, CC BY-SA 4.0
A fragment of Atrahasis in the Morgan Library, MLC 1889, exhibited with the reign of Ammi-saduqa. The same reign produced more than one copy, and this piece is not the great London tablet of Ku-Aya.Daderot, Public domain
A fragment the Metropolitan Museum files with the Atrahasis flood story. The flood survived as pieces in more than one collection, each one a copy, and none of them a silt layer.The Metropolitan Museum of Art, CC0
Comparison: Gilgamesh Tablet XI, from Nineveh, 7th century BC. The flood speech here is the later retelling. The hero's name and the library are not those of the Ku-Aya tablet.Photograph by Mike Peel (www.mikepeel.net)., CC BY-SA 4.0
Comparison: the Nippur tablet of the Sumerian flood story, Penn Museum CBS 10673. The hero is Ziusudra and the language is Sumerian. The museum files it as a flood tablet. It is not Atrahasis, and it is not Gilgamesh Tablet XI.Onceinawhile, CC BY-SA 4.0

See the evidence at the source

These pictures are not openly licensed, so they are not copied onto this page. Each link opens in a new tab.

What is still unexplained

These sit beside the consensus account, with the same visual weight. None of them is marked solved.

Observed A reason, and a warning The gods are tired of labour and create humans to carry it. Noise, not a moral catalogue, is what Enlil will not bear. Plague and famine fail. The flood is the third measure. Enki warns Atrahasis. A story that gives a reason and a leak inside the divine council is a composed plot. A flood can still sit inside a plot.
Observed The colophon The great British Museum tablet is the work of a named scribe, Ku-Aya, and its colophon places the copy in the reign of Ammi-saduqa. The reign is dated with the Venus tablet from the Nineveh library. The observations of Venus on that tablet can be fitted to a king's years in more than one way. Three chronologies are still argued, and the high and the low move the reign by more than a century. One published placement, the middle chronology, puts the reign near 1635 BC. The same 2016 study of Kültepe and Acemhöyük timbers reports a fit with that middle chronology, or with a low-middle date eight years later, and argues that the high, the low, and the new chronologies do not fit the wood. The study's text leaves the nearer pair undecided. A shift of chronology would move the copy. It would not move the poem into the Ice Age, and it would not close the reign's year.
Observed One plot, three heroes Ziusudra, Atrahasis and Utnapishtim survive a flood after a warning from the god of fresh water. The birds, the boat and the sacrifice are shared with Gilgamesh Tablet XI closely enough that the later poem is using this one. The Sumerian tablet from Nippur tells the flood after the founding of cities, Eridu among them. Shared plot is not a shared object. The clay in London, the clay in Philadelphia, and the clay from Nineveh are three things.
Open Dates that refuse one flood Silt at Ur lies in Ubaid levels, roughly the fifth to fourth millennium. The yellow band at Shuruppak sits at the Jemdet Nasr to Early Dynastic transition, often placed about 3100 to 2900 BC, and the report will not call it one blanket. Kish has another layer of another date. The Gulf floor was dry and then drowned across several thousand years after the Ice Age. A single world flood, or a single cold snap at 12,900 years ago, would need those layers to be one event. They are not.

Location and dating

Location
The principal Old Babylonian tablet was probably written at Sippar and is in the British Museum. Further fragments are in the Morgan Library and the Metropolitan Museum. Later copies come from Nineveh.
Period label
Current consensus — open to revision The working date is open to revision. The Ku-Aya copy belongs to the reign of Ammi-saduqa. One published placement, the middle chronology, puts that reign near 1635 BC. A 2016 study reports a fit with that placement or a date eight years later, and argues that the high, the low, and the new do not fit the wood. The room does not file one year. Nineveh copies are 7th century BC. The silt layers are older than both and do not share one century.
How this dating was done
The copy is dated by its colophon and by typology, which places a tablet from the style of its script. The reign is placed by the Venus observations, records of the planet Venus, which allow more than one chronology. The silt is dated by stratigraphy, reading age from the order of the layers, and by pottery, not by this poem.
What could change it
A radiocarbon result, a date from the carbon left in once-living material, on a sealed silt, and a second trench that maps the Shuruppak band, would show whether any layer can be the flood this poem tells. A colophon older than Ammi-saduqa would move the poem. The noise plot would still be literature.
Published ranges, drawn as bars. A bar is a claim about a span, not a point.
Shuruppak yellow layerContested
Ku-Aya copyContested
Nineveh copies
3100 BCE2000 BCE631 BCE
  • Shuruppak yellow layer. Ceramic window. Not a date of this tablet.
  • Ku-Aya copy. One published placement of Ammi-saduqa, the middle chronology, near 1635 BC. A 2016 wood study reports a fit here or eight years later, and argues that the high, low, and new chronologies do not. It does not decide the nearer pair.
  • Nineveh copies. Ashurbanipal. The flood passage reused in Gilgamesh.

What is actually there

Open questions

Question Status
Does the noise plot carry a memory of a real inundation, or only a story about the gods? Open
Which chronology does the reign of Ammi-saduqa belong to, once the Venus observations are fitted? Debated
Can any one silt layer be the flood of this poem? Open until the band is dated and mapped

The current narrative

Where publication stands today. It is the best-supported reading, and it is open to revision. It is not a verdict.

The current narrative — open to revision

A literary plot on a floodplain that really flooded

Atrahasis is a composed story, with noise, plague and a divine leak, written on a plain where rivers laid down silt. The memory, if it is a memory, is of local floods, enlarged into a world flood. It is not a date for the Younger Dryas.

What supports this

  • The colophon places a full copy in the Old Babylonian period.
  • Silt at several cities shows real inundation, and the cities are the ones the tradition names.
  • The plot has a reason and a structure that a deposit of clay does not contain.

What pushes back

  • The silt layers are not one date, so "the" flood is already too neat.
  • A purely literary reading has to say why the tradition keeps returning to Shuruppak, the last city before the flood in the king list.

What would settle it

This is the best reading so far. The poem is literature, the floods were real and local, and the step from one to the other stays open.

Emerging narratives

Other published readings. Each may be right. None is adopted here unless the evidence carries it.

An emerging narrative

One inundation, remembered

The warning, the boat and the landing preserve a single flood that ended a settlement the texts still name.

What supports this

  • More than one composition uses the same shape of story, which is what a shared memory would look like.
  • Shuruppak has a water-laid band on a living floor, and the king list stops the antediluvian cities there.

What pushes back

  • The band is not mapped as one blanket, and the tablets that name the city sit above it.
  • Ur and Kish record silt of other dates. One memory cannot be all of them unless the memory is of floods in general.
  • The copies are centuries to millennia after the layers.

What would settle it

One inundation is possible, and it is not the centre of the reading, until one layer is dated and the poem is tied to it by more than a shared city name.

An emerging narrative

A memory of the Younger Dryas or of the filling of the Gulf

The flood is the cold return at the end of the Ice Age, or the sea filling the Gulf from about 12,000 to 6,000 years ago, carried in oral telling until the scribes wrote it down.

What supports this

  • The Gulf floor was land and then sea. The Younger Dryas was a sharp cold return. Both are real.
  • A story about water covering the world could, in principle, keep a distant shoreline.

What pushes back

  • The Gulf rose over millennia, not in the seven days the poem uses.
  • The Younger Dryas ends about 11,700 years ago. The Ku-Aya copy is placed in the 17th century BC on the middle chronology, one published fit. Nothing on the tablet names either event.
  • Local silt of the fourth and third millennia is a nearer cause, and it is already more than one flood.

What would settle it

A distant background is not ruled out by silence, and it is not what the copies date. Collapsing every flood telling onto one cold snap asks the clay to say something it does not say.

Should the current view be reevaluated?

Consensus holds well

The poem is dated by a colophon, the flood section was reused later, and the ground shows repeated local floods. A single Ice Age event is not required by the evidence in hand.

What drives this

  • The Ku-Aya colophon
  • Silt of more than one date
  • A plot with a reason the silt does not contain

This is the museum’s own weighing of the evidence. It is not a visitor vote, and it is separate from the slider.

Weigh it yourself

Anonymous. Nothing here is stored except on this browser. Your lean does not change the assessment beside it.

A literary plot on a floodplain that really flooded Unsure One inundation, remembered

If the contested reading is right

What would follow if a reading that is not the consensus turned out to be true, and what could settle it. This explains the stakes. It does not adopt the reading.

The poem remembers one inundation at Shuruppak

The yellow band and the king-list break would be the same event, and the long reigns would remain a separate problem the silt does not solve.

Why it is still worth a test

Radiocarbon on the charcoal under the yellow band, and a section that maps the band beyond one pit.

Sources

  1. Atra-hasis: The Babylonian Story of the Flood W. G. Lambert and A. R. Millard, Eisenbrauns
  2. Noah's Flood Reconsidered Iraq, 1964
  3. Atrahasis Wikipedia
  4. Integrated tree-ring-radiocarbon high-resolution timeframe PLOS ONE, 2016. The same study cited in the Nineveh room. Reports a fit with the middle chronology or a date eight years later, and argues against the high, the low, and the new. It does not decide the nearer pair.

Image credits

← Iraq World map