Half a Million Tablets, and Most Are Unread

Something like half a million cuneiform tablets have come out of the ground since excavation began in Mesopotamia in the nineteenth century, and no one has ever counted them properly. That figure is not a census; it is the rough consensus of the people who catalogue collections for a living, an order of magnitude rather than a number, and it is treated that way throughout — as an estimate, flagged as one, every time it appears.¹ What is not in dispute is the shape of the problem the estimate points to. This is, by any reasonable reckoning, the largest surviving body of writing from the ancient world, larger than the papyri of Egypt, larger than everything that survives in Greek and Latin from before the Roman empire, and it is simultaneously scattered across two dozen institutions on four continents, mostly unpublished, mostly unread even by specialists, and mostly, if the archaeologists are right about what is still under the mounds of southern Iraq, not yet excavated at all. The archive of Babylon is not a room you can walk into. It is a backlog.

This essay walks through that backlog in three pieces: where the excavated tablets actually sit, what a handful of the most famous ones actually say, and what happened to the collection that stayed in Iraq when a museum in Baghdad was left unguarded for a day and a half in April 2003. The three pieces connect. Two other essays on this site have leaned on this archive without saying much about the archive itself — the reading of Hammurabi's stele as a scribal treatise rather than a statute book rests on the absence of any citation to it in tens of thousands of Old Babylonian court documents, and the case that Babylon built a predictive science before Hipparchus rests on centuries of nightly astronomical diaries that had to survive, travel, and eventually be read by someone who could do the arithmetic. Both arguments live or die by tablets that happen to have been dug up, shipped somewhere, and deciphered. Most of the corpus has not had that luck.

Not all in England

The popular image of "the tablets" is the British Museum, and the popular image is not wrong so much as incomplete. The British Museum does hold the largest single collection in the world, something on the order of 130,000 tablets and fragments, and a large share of that number — on the order of 30,000 pieces — comes from one findspot: the library that the Assyrian king Ashurbanipal assembled at Nineveh, on the mound the Arabic-speaking villagers who lived over it called Kuyunjik.¹ Ashurbanipal's scribes copied and collected literary, religious, medical, and divinatory texts from across Mesopotamia in the seventh century BCE, and when the library was buried in the destruction of Nineveh in 612 BCE it was, in effect, sealed. Austen Henry Layard began digging at Kuyunjik in 1849; his assistant and successor Hormuzd Rassam, an Iraqi Christian from Mosul who ran his own excavations at the site through the 1850s, recovered a large share of what followed, including, in 1852, the room that held the library's core.² Almost the entire library ended up in London, under the arrangements the Ottoman administration allowed to nineteenth-century foreign excavators, arrangements that had nothing to do with the informed consent of anyone in Mosul.

But that is one collection, however large, and the rest of the half million is spread wide. Berlin's Vorderasiatisches Museum holds a major share, built substantially from German excavations at Babylon, Assur, and Uruk in the decades around 1900. The Louvre holds a significant collection, including the Hammurabi stele itself, taken from Susa in 1901–02, and a large run of tablets from French excavations at Mari and Tello. Istanbul's Museum of the Ancient Orient holds tablets that never left Ottoman, then Turkish, territory — a reminder that "dispersal" did not send everything west. The Iraq Museum in Baghdad holds the national collection built from a century of excavation on Iraqi soil, including tablets recovered under Iraqi-run digs after independence. In the United States, Yale's Babylonian Collection is the largest, on the order of 40,000 tablets, assembled mostly through the purchases and excavation shares of J. B. Nies and Albert T. Clay in the early twentieth century; the University of Pennsylvania and the Oriental Institute in Chicago hold substantial collections of their own, mostly from their own excavations at Nippur and elsewhere.¹

The mechanism behind that list is a nineteenth- and early-twentieth-century story, and it is not one story but several, run together by time. Some of it is straightforward removal: Layard and Rassam dug under British consular protection with essentially no local oversight and shipped what they found to the British Museum by river raft and sea. Some of it is the system called partage, formalised later and imperfectly, under which an excavating foreign mission and the host country's antiquities service divided a season's finds, a system that in principle gave something back and in practice still moved the majority of the best material abroad. Some of it, especially for the American collections, is outright purchase on the antiquities market, where a tablet's excavation context — the single most valuable piece of information a tablet can carry, more valuable in most cases than the text on it — was already lost by the time it reached a dealer. This essay does not referee it. The same century that scattered the archive also, in fact, preserved a great deal of it: tablets bought or divided out before 1914 mostly survived two world wars, an Iraqi monarchy, a republic, and a war whose damage comes up below; a share of what stayed in one place did not survive nearly as well. Dispersal was loss for the people whose history it was, and dispersal was, in a narrow and uncomfortable sense, also insurance. Both things are true, and the fact pattern speaks for itself well enough that it needs no verdict attached here.

What they actually say

It is worth saying plainly what fills those half a million tablets, because it is not what a museum case implies. The overwhelming majority are not literature, not law, not astronomy. They are receipts. They are loan contracts, sale contracts, ration lists, tax accounts, marriage settlements, court depositions about a stolen sheep, letters between a merchant and his agent about a shipment of tin, and above all school exercises — the same handful of lexical lists and proverbs, copied by apprentice scribes thousands of times over a thousand years, because that was how you learned to write cuneiform. This is the archive's real glory rather than its disappointment. No other ancient society left behind anything like this density of ordinary paperwork. We do not have Athens's grocery lists. We have Babylon's.

Three tablets, though, earn their fame, and each is given here in paraphrase rather than translation, because that is the only honest way to handle them on a site that does not pass off modern scholars' renderings as its own.

Three tablets, in paraphrase

The Flood Tablet

The eleventh tablet of the Epic of Gilgamesh, recovered from Ashurbanipal's library at Nineveh, tells Gilgamesh the story of a great flood sent by the gods, survived by one man — Utnapishtim in this version — who is warned in advance, builds a boat to exact measurements, loads it with his family, his craftsmen, and the seed of every living thing, rides out a storm that terrifies even the gods who sent it, and afterward sends out birds to find dry land, in rough summary of a text substantially longer and stranger than that sentence can carry.³ George Smith, a self-taught cuneiformist working at the British Museum, identified the flood narrative on the tablet in 1872 and read a paper on it to the Society of Biblical Archaeology that November, arguing that this Assyrian text told essentially the same story as Genesis, centuries older than any Hebrew manuscript then known.⁴ The most-repeated story about that moment — that Smith, on making the identification, leapt up and began taking his clothes off in excitement — comes down to us as anecdote, reported after the fact by a colleague; it is labelled exactly that here, an anecdote, because there is no verifying it was witnessed as told.⁴ What is not anecdote is the effect: Victorian London, mid-argument over Darwin and the historicity of the Bible, was electrified by a cuneiform tablet, and the Daily Telegraph funded Smith's return to Nineveh to look for the missing lines.

The complaint to Ea-nasir

From Ur, from around 1750 BCE, a tablet in the British Museum preserves a letter from a man named Nanni to a copper trader named Ea-nasir, and it reads, in paraphrase, as a complete customer-service complaint: Nanni says he sent his servant with silver to buy copper ingots of good quality, that what arrived instead was inferior metal, that when he sent the servant back to reject it Ea-nasir refused to take it back or refund the silver, that Nanni had trusted Ea-nasir's own house and been treated with contempt, that other merchants in the same trade would not dream of behaving this way, and that he wants his money or his copper, properly, and expects better than to be humiliated by an agent sent through hostile territory to collect what was owed him.⁵ It is not a legal filing and it is not literature. It is a furious letter from a customer who has been shorted on a shipment, nearly thirty-eight centuries before Yelp, and it has become, informally, the internet's favourite piece of cuneiform for exactly that reason. It is fair to call it the oldest surviving written customer complaint, and to leave the joke where it belongs, alongside the genuine point: Old Babylonian trade ran on a dense web of long-distance credit and reputation, and Nanni's letter shows what happened when that web frayed for one buyer.

Nanni's letter shows what happened when the web of long-distance credit and reputation frayed for one buyer, nearly thirty-eight centuries before anyone thought to leave a review.

Plimpton 322

Plimpton 322, held at Columbia University and dated to roughly the same Old Babylonian period, is a table of numbers arranged in four columns and fifteen rows, and what those numbers are has been argued over since Otto Neugebauer and Abraham Sachs published the standard edition in 1945.⁶ The columns give, in essence, sets of three whole numbers that satisfy the relationship we know from the Pythagorean theorem — the sums of two squared numbers equalling a third squared number — arranged in a sequence that changes in a fairly regular way from row to row. That much is not contested: the tablet is a list of what a modern reader would call Pythagorean triples, compiled roughly a millennium before Pythagoras is supposed to have lived. What the list was for is where scholars split. Neugebauer and Sachs read it as a table connected to solving problems about right triangles. A more recent argument, from Eleanor Robson among others, holds that it is better understood as a teacher's list of exercise problems, generated by a known scribal method for producing such triples, rather than evidence of a developed trigonometry; a separate and more publicised 2017 claim, from Daniel Mansfield and Norman Wildberger, argued for an actual exact trigonometry encoded in the ratios, a claim that has drawn serious pushback from specialists in Babylonian mathematics.⁷ That dispute is not adjudicated here; "trigonometry a thousand years early" is a live and contested reading, not a settled one, and that the tablet's plain content — the triples themselves, computed and tabulated with evident system — is remarkable with or without it.

Baghdad, April 2003

The clearest measure of how precarious this archive still is came on 10 and 11 April 2003, when the National Museum of Iraq in Baghdad was looted over roughly a day and a half after its staff evacuated and coalition forces, present nearby, did not secure the building. Reporting in the years since gives a range of roughly 15,000 objects taken in that window, a figure drawn from the museum's own subsequent inventory work.⁸ Recovery figures vary depending on who is counting and when: the museum's own statement some years after the looting put recovered items at around 4,300; other reporting closer to the fifteenth anniversary put the recovered total nearer 7,000, through a mix of Interpol-assisted seizures abroad, an Iraqi amnesty that saw looters and middlemen quietly return objects, and finds inside Iraq itself.⁹ On either figure, several thousand objects — by some counts more than eight thousand — remain unaccounted for. Two of the museum's signature pieces, the Warka Vase and the Mask of Warka, both Uruk-period objects roughly five thousand years old, were among the items recovered, the vase in pieces, handed over anonymously in a car, the mask found buried in a farmer's field north of Baghdad.⁹

Cuneiform tablets, small, dense, and comparatively easy to move, were and remain a particular casualty of this whole period, and not only inside the museum. Site looting across the archaeological mounds of southern Iraq accelerated sharply after 2003, as the state's ability to police remote sites collapsed; freshly dug tablets with no excavation record at all began surfacing on the international antiquities market through the 2000s and 2010s, a pipeline that is documented in general terms even where any single tablet's path through it is not.⁸ The Iraq Museum itself reopened to the public on 28 February 2015, after more than a decade of restoration and reorganisation, with UNESCO among the bodies marking the occasion.¹⁰

Several thousand objects remain unaccounted for, and on the mounds outside Baghdad, the looting did not stop when the museum's doors were finally guarded.

The reopening did not end the damage to the wider record. In 2015, ISIS militants filmed themselves smashing statues in the Mosul Museum and using explosives and bulldozers to level large parts of the ancient Assyrian site of Nimrud, an act of destruction documented in enough footage and subsequent site survey that it is not in serious dispute.¹¹ Against that loss, there has been a slow current of return: the British Museum handed a small group of looted tablets back to Iraq in 2019, part of a broader pattern of institutions and governments repatriating objects once their illicit removal after 2003 could be documented, and the United States has, through a series of legal actions, returned thousands of smuggled Iraqi antiquities over the past decade, including objects at the centre of the Hobby Lobby case, in which the company agreed to forfeit thousands of illegally imported artefacts it had acquired for a planned museum.¹² The exact figures in that last sentence are hedged deliberately; the specific tallies differ by report, and understating them is safer than repeating an unsupported number.

The archive is a backlog, not a room

Put the three pieces together and the shape of the argument is this. Half a million tablets, by rough consensus, have been excavated. They sit in something like eight or nine major collections and dozens of smaller ones, on a distribution that a colonial-era excavation regime drew and that a century of purchase, partage, war, and looting has since redrawn without making it any less scattered. The tablets that get famous — a flood story, a furious customer, a table of numbers — are famous precisely because someone read them, translated them, and told the story well enough that it left Assyriology and entered the wider world. That is a vanishingly small fraction of the half million. Assyriology is a small field: there are, worldwide, only so many people trained to read cuneiform at a research level, and the corpus has always outnumbered its readers by an order of magnitude no one bothers to compute, because the answer would only be discouraging. A large share of excavated tablets, sitting catalogued in museum storerooms in London, Berlin, Baghdad, and New Haven, have never been fully published, which in practice means never been fully read by anyone since the ancient scribe who wrote them.

And beneath all of that is the plainer fact: most cuneiform tablets ever written were never excavated at all. The unexcavated tells of southern Iraq — the mounds that mark buried cities, many surveyed but not dug, some not even surveyed — are understood by the archaeologists who work the region to hold far more written material than has ever come out of the ground, a point that is close to consensus among excavators rather than a specific count that can responsibly be given a number.⁸ The ground under Iraq is not a closed file. It is a set of first editions that no one has opened yet, and every year that site looting, war, or simple lack of funding for excavation continues, some of them are lost before they are ever read once.

This is why this archive keeps coming back in essay after essay on this site. The claim that no Old Babylonian court cited Hammurabi's stele rests on the tablets that survive from the courts, not on the tablets that do not; if a larger dossier of trial records turned up tomorrow from an unexcavated mound at, say, ancient Larsa, that argument could in principle be strengthened or overturned by evidence that has not yet seen daylight. The claim that Babylon's astronomers built a working predictive science before Hipparchus rests on Astronomical Diaries recovered from one findspot and painstakingly edited across the twentieth century; there is no guarantee the surviving run is representative of everything a Babylonian temple ever recorded, only that it is what happened to reach London intact. Neither argument is wrong for depending on a partial record — every historical argument depends on a partial record — but both are, quite literally, only as good as an archive that is scattered across a dozen cities, mostly unpublished, and mostly still underground. "The archive of Babylon" was never a room in London. It is a backlog the size of a civilisation, and the ground still holds the first editions.