The Code of Hammurabi Was Not a Law Code

The black stone in the Louvre that everyone calls the Code of Hammurabi was not a code, and almost nobody in Babylon was ever tried under it. That is my reading, and it is not an eccentric one: it has been the working view of a good part of Assyriology for sixty years, even as museum labels and courthouse friezes go on presenting the stele as the first statute book of humankind. The object is real enough. It is a shaft of diorite a little over two metres tall, found in three pieces at Susa in south-western Iran in the winter of 1901–02 by Jacques de Morgan's French expedition and published within months by the Dominican scholar Jean-Vincent Scheil.¹ It was carved in Babylon around 1750 BC and carried off to Elam as war booty some six centuries later. It carries, in something like four thousand lines of Akkadian cuneiform, a long prologue, between 275 and 300 short provisions, and a long epilogue. What it does not carry is any sign that it was consulted in court.

The word "code" arrived with Scheil, who titled his 1902 edition Code des lois de Hammurabi, and it stuck because it flattered a story Europeans wanted: law begins with a written, numbered, comprehensive statute, and here at last was the ancestor of Justinian and Napoleon, older than Moses. Nearly everything in that sentence is wrong. It is not comprehensive, it was not numbered (the paragraph numbers are Scheil's), and it does not behave like a statute in any document that survives from the courts of Hammurabi's own kingdom. What it behaves like is two other things at once: a royal monument, and a work of scribal scholarship. Martha Roth, whose translation is the standard one in English, deliberately titles her edition Law Collections from Mesopotamia and Asia Minor, and the neutral word "collection" is doing exactly the work she intends.¹

What the stone says about itself

Start with the parts of the stele nobody quotes. Roughly a fifth of the text is prologue and epilogue, and neither has anything to do with the rules in the middle. The prologue opens in the heavens: the gods Anu and Enlil have raised Marduk of Babylon to supremacy among the gods, and have named Hammurabi, the pious prince, to make justice appear in the land, to destroy the wicked, and to see that the strong do not oppress the weak, rising over the people like the sun-god Shamash.¹ Then comes a roll-call of more than twenty cities and their temples, each with a pious epithet: Nippur, Eridu, Babylon, Ur, Sippar, Larsa, Uruk, Isin, and on out to Mari, Ashur, and Nineveh. Read as a list, it is a map of the empire Hammurabi assembled in the last decade or so of his reign, presented not as conquest but as caretaking. This is the same genre as every other Mesopotamian royal inscription: the king as shepherd, provider, and builder, addressed to the gods and to posterity.

The epilogue is more revealing still. It calls the provisions dīnāt mīšarim, from the Akkadian words for verdict and for justice: not laws in the abstract but the sort of judgment a righteous king renders.¹ In Roth's edition the king says, in my paraphrase, that he has set these judgments before his own image as king of justice, and that any wronged man should stand there, have the stele read to him, understand where he stands, and go away with his heart quieted and a blessing for the king. He then tells whatever ruler comes after him to heed the words and not alter them, and closes with a very long curse-list calling down the gods on anyone who erases the inscription or substitutes his own name. That is what a monument says. Statutes do not bless the reader.

The wronged man is invited to hear the stone read aloud and to calm his heart. That is what a monument says. Statutes do not bless the reader.

Notice too where such stones stood: in temple courtyards, in the presence of the god. The Louvre stele was almost certainly taken from Sippar, the city of Shamash, and fragments of at least one duplicate stele were found alongside it at Susa; further copies presumably stood in Babylon and other cities.¹ The relief at the top shows Hammurabi standing before the enthroned Shamash, who holds out the rod and ring, the emblems of ordered rule. The rules are displayed to the god who guarantees justice, and to the king who comes after, and only then to a public that, overwhelmingly, could not read them. In this the stele belongs beside the Egyptian idea of the ruler as the one who "does ma'at" that I take up in Ma'at: Egypt's Real Constitution: the king's first duty is to be seen establishing order, and the writing is part of the seeing.

"If a man…": the shape of a scholarly text

Now the middle. Every provision has the same grammar: šumma awīlum, if a man, followed by a condition, followed by a consequence. If a man accuses another of murder and cannot prove it, the accuser is put to death (§1). If a man puts out the eye of a member of the free class, they shall put out his eye (§196). If a builder builds a house that collapses and kills the owner, the builder is put to death (§229). Scholars call this casuistic form: not "murder is forbidden" but "if this case, then this outcome". It is not the natural grammar of a lawgiver. It is the natural grammar of a Babylonian scholar, because it is exactly the form of the great omen series and the medical diagnostic handbooks, whose entries run, in my own summary: if a city stands on a hill, things go badly for the people in it; if a woman bears a child with a lion's ear, a mighty king will arise; if a sick man's temples throb and his eyes redden, he will not recover. The lists that Babylonian scribes used to organise the world, the same habit of mind that let their successors turn nightly sky-watching into eclipse forecasts (see The Babylonians Invented Prediction, Not the Greeks), were lists of cases, each a protasis and an apodosis. The "laws" of Hammurabi are written in that idiom because they were composed by the people who wrote in that idiom.

The observation is not mine. The Dutch Assyriologist Fritz Rudolf Kraus made it the centre of a short, famous article in 1960 whose title asks the question plainly: what is the Codex Hammurabi? His answer was that it is not a law code at all but a piece of scholarly literature, a scribal composition in the tradition of the lists, presented in the king's name.³ Jean Bottéro developed the argument in an essay that is still the best single thing to read on the stone: the provisions, he showed, are not a set of general rules but a series of model verdicts, arranged the way a Babylonian treatise arranges its cases, and their purpose was to display, exhaustively enough to impress, the king's justice at work, not to bind a judge.² Bottéro's word for it was a treatise, and he set it beside the divination handbooks rather than beside Roman law.

Look at what is in the collection and what is not, and the scholarly character stands out. There is no general provision on homicide, only the false accusation of it. There is a great deal on the hire of oxen and boats, on the wages of a field labourer, a shepherd, a physician setting a bone, and on the price of a builder's work, all as fixed tariffs. There are long stretches on marriage, dowry, inheritance, and adoption. Whole areas of what any actual community must have needed are missing, and the sequence moves by association of ideas, from one case to a neighbouring variant, in the way a compiler extends a list rather than the way a legislator covers a field.²¹¹ A judge who opened this text hoping to find the answer to most disputes would have closed it again.

The collection had ancestors, and the ancestors had the same shape

Hammurabi's is also not the first such text, which matters, because it means the form was inherited rather than invented for use. The Laws of Ur-Namma, in Sumerian, come from the founder of the Third Dynasty of Ur around 2100 BC, some three and a half centuries earlier; the Laws of Lipit-Ishtar of Isin, also Sumerian, from around 1930 BC; and the Laws of Eshnunna, in Akkadian, from a rival kingdom up the Diyala only a generation or so before Hammurabi.¹ All are casuistic. Ur-Namma and Lipit-Ishtar both open with a prologue in the king's voice about establishing justice and freedom in the land, and Lipit-Ishtar has an epilogue and curses. Eshnunna opens with a tariff of prices and wages. Whole clauses recur across the series: the goring ox that kills a man appears in Eshnunna and again in Hammurabi (§§250–252) and again, later and elsewhere, in Exodus. What is being handed down is a genre with a stock of standard problems, edited and expanded in each king's name, the way an omen series or a lexical list was edited and expanded. Roth's edition prints them together for exactly this reason.

The comparison also punctures the stele's most famous line. "An eye for an eye" (§196) is often taught as the primitive baseline that later law softened. The reverse is closer to the truth: the earlier Ur-Namma collection settles injuries with silver payments, and Hammurabi's own text does the same when the victim is a commoner (a mina of silver, §198) or a slave (half his value, §199).¹ Bodily talion in Hammurabi is a status rule for the free class, and on the face of it an innovation rather than a survival. Whether it was ever carried out is another question, and it brings us to the courts.

What the courts actually did

The Old Babylonian period is one of the best-documented stretches of the ancient world, with tens of thousands of clay tablets: sale and loan contracts, marriage and adoption agreements, court records, and letters, including a large dossier of letters from Hammurabi himself to his officials in the south, above all to Šamaš-hāzir, his land administrator at Larsa, and to the governor Sîn-iddinam.⁶ If the stele were a statute in force, this is where its fingerprints would be. They are not. In this whole mass of practice there is, so far as I can find in the scholarship, not one trial record that cites a provision of the collection, no judge who grounds a ruling in what is written on the king's stele, no litigant who appeals to a paragraph. A handful of documents mention a stele, or a wage reckoned by the stele, and whether they mean this monument or another royal proclamation is itself disputed; even if every one of them refers to Hammurabi's stone, they concern tariffs, not the trial of cases.⁶

Hammurabi's own letters are the sharpest test. In them the king decides disputes daily: a soldier's field wrongfully seized, a bribe alleged against a judge, a claim over a house, a tax farmer's arrears. He never once refers his officials to a written law, his own or anyone's.⁶ Nor do the verdicts of the local courts, which are given by panels of judges and elders in the city gate or the temple, and recorded with the names of witnesses and the oath sworn, resemble the mechanical outcomes on the stele. Where the same subject appears in both, the practice often diverges: interest rates in real loan documents do not sit neatly on the stele's figures, and the death penalties the stele prescribes so freely are strikingly hard to find carried out in the records.⁸

In this whole mass of practice there is not one trial record that cites a provision, no judge who grounds a ruling in what is written on the king's stele.

Meanwhile the kings of Hammurabi's dynasty did legislate, and we can watch them do it. Their instrument was the mīšarum edict, a periodic royal proclamation, usually early in a reign, cancelling certain classes of debt, releasing debt-slaves, and correcting abuses; the best-preserved is the edict of Ammi-ṣaduqa, Hammurabi's fourth successor, and Kraus, again, was its editor.⁴ These edicts are cited in practice: contracts are dated by them, loans are voided on the stated ground that the king has issued his edict. That contrast is the whole argument in one frame. When a Babylonian king wanted a rule to change what happened in courts and counting-houses, he issued a decree, and the documents show it working. The stele is not that kind of text, and the documents show it not working, or rather not being asked to.

The case for the other side, put fairly

None of this is undisputed, and the strongest opponents are not naïve. Raymond Westbrook, who did more than anyone to bring the cuneiform collections into the mainstream of legal history, accepted that the collections were products of the scribal academy, cast in the omen-list mould, and not legislation in the modern sense. But he argued that they were nonetheless reference works of real law: distillations of a common legal tradition shared across the ancient Near East, which a judge could consult as a guide to how cases of a given type were customarily decided, much as a treatise summarises common law without itself being enacted.⁵ On that view the silence of the trial records is unsurprising: nobody cites a treatise in a verdict; they simply decide the case the way the treatise says such cases are decided. Westbrook's two-volume History of Ancient Near Eastern Law is built on the assumption that the collections and the practice documents can be read together as evidence of one working system.⁷

Dominique Charpin has pressed harder, and from the practice side. Working through the letters and archives, he has pointed to cases where the outcome in a real document tracks a rule in the collection closely enough to suggest the collection was known and applied, and he takes the epilogue's invitation to the wronged man at face value: the stele was meant to be read to the public, and copies of it circulated on tablets, so it was not sealed away in the world of scholarship.⁶ Sophie Démare-Lafont has offered a subtle version of the same conclusion: the collections were law, but subsidiary law, a fallback that filled the gaps and corrected the outcomes when local custom or a contract did not settle the matter.⁹ And Roth herself, whose word "collection" I have leaned on, does not treat the question as closed. In an essay on the Mesopotamian legal traditions she reads the stele as operating in several registers at once, royal ideology and scribal science, but with a live relation to the norms of the courts, and she has been careful not to reduce it to literature.¹⁰

I take these seriously, and I concede the part of them that survives contact with the evidence: the collection is not fiction. Its cases were drawn from, and drawn towards, the practice of Old Babylonian courts, and it is a fine source for how a Babylonian thought about a case. What I do not concede is that this makes it a law code, or that "reference work" is a description rather than a hope. A reference work is something someone refers to. The proposition that judges consulted the collection is not supported by any judge doing so, and the parallels Charpin finds are what one would expect if the scribes who compiled the collection and the scribes who drafted the court records were trained on the same exercises, which they were. That the same stock cases appear in Ur-Namma, Eshnunna, Hammurabi, and Exodus tells me the collection was fed by a tradition. It does not tell me the tradition was fed by the collection.

What it was, then

Here is the reading I end up with. Late in a long reign, having taken Larsa, Eshnunna, and Mari and made himself master of Mesopotamia, Hammurabi commissioned a monument to say what kind of king he was. The form his scribes gave it was the highest form they had, the treatise of cases, expanded from the collections of earlier kings and dressed in the language of the omen lists, because that was what learning looked like. He set copies before the gods in Sippar and elsewhere and told the wronged man to come and listen. Then the scribal schools took the text back. It was copied as an exercise and a classic for more than a thousand years, in the Old Babylonian schools, in the libraries of Assyria, in the Babylon of Nebuchadnezzar, long after any pretence that it governed anything; Roth notes that the last copies are separated from the original by over a millennium.¹ Nobody transmits a statute that way. Scholars transmit a canon that way.

Two things follow. The first is that the modern story of "the first law code" gets the history of law backwards. Real Mesopotamian law lived in contracts, oaths, ordeals, custom, the verdicts of local judges, and the edicts of kings; the stele is a portrait of that world made by people who knew it well, and it should be read as a portrait, with the king in the frame. It is the difference between a document that governs and a document that describes, and Babylon, which knew both, put them in different genres. The builder-king who took the same city apart and put it back together twelve centuries later, and whose bricks I have borrowed for the figure above, suffers from the same modern habit of reading a monument as a record of what happened (see Nebuchadnezzar II: The Builder We Remember as a Destroyer).

The second is stranger and, to me, more interesting. The Code of Hammurabi is the earliest long text we have that tries to make justice look like a system: a complete, ordered set of if-then rules, displayed as though the whole of right conduct could be written down and read out. It was not true, and the Babylonians who used courts knew it was not true. But the ambition survived the fact. It passed through the schools, into the Bible's own casuistic laws, and eventually into the codes that a museum label now claims the stele began. What the stone founded was not a legal system. It was an idea about what a legal system should look like, and that idea has outlived every court that ignored it.