Chapter 2 of 14

The First Harvest: Sumer and Egypt

About 28 minutes to read

The inspector's diary

In 2013 a French–Egyptian team led by the Egyptologist Pierre Tallet was excavating an ancient harbour at Wadi al-Jarf, on the Red Sea coast of Egypt. In a set of storage galleries cut into the rock, they found something that no one had expected: rolls and fragments of papyrus, the oldest inscribed papyri ever discovered. They dated from the final years of the reign of King Khufu, the pharaoh for whom the Great Pyramid of Giza was built, around 2560 BCE.

The most complete of these documents is a logbook kept by a man named Merer. He held the rank of inspector, and he commanded a team of roughly forty men. For several months his team ferried blocks of fine white limestone from the quarries at Tura, on the east bank of the Nile, across the river and along canals to the building site at Giza, which the papyrus calls Akhet-Khufu, "the Horizon of Khufu". The diary is laid out in neat columns, a few days at a time: where the boat set off, where it spent the night, when it arrived, when it went back for more stone. Merer also notes the name of his superior, Ankhhaf, a half-brother of the king, who was in charge of the whole operation.

Merer's diary is not a hymn to the god-king, and it is not a complaint. It is a work schedule. It is the record of a middle manager, accountable upwards for the movement of materials, responsible downwards for the men who moved them. Like the Uruk tablet with which this book began, the oldest papyrus we have is an administrative record. And like that tablet, it shows us not the people at the top or the bottom, but the person in between: the shepherd, keeping count.

This chapter is about the world that produced Merer's diary and Kushim's ledger — the first two great farming civilisations of human beings, in the river valleys of Mesopotamia and Egypt. It is here that all five parts of the farm first appear together: a fence, a feed, a body of shepherds, a story and, almost immediately, a breakout. To see how, we have to start with the most ordinary object imaginable: a bowl.

The bowls of Uruk

Anyone who has dug at a site in Mesopotamia from the second half of the fourth millennium BCE — the "Uruk period", named after the great city of that name — knows the bevelled-rim bowl. It is a crude, thick-walled vessel, roughly the size of a cereal bowl, made by pressing coarse clay into a mould rather than throwing it on a wheel. It is ugly, cheap and astonishingly common. Bevelled-rim bowls have been found in their hundreds of thousands, from southern Iraq to Iran, Syria and Turkey; at many Uruk-period sites they are the single most common kind of pottery. Excavators find them in heaps, often whole, as if they had been used once and thrown away.

What were they for? The most influential answer, proposed by archaeologists such as Hans Nissen and Gregory Johnson, is that they were ration bowls: standard measures in which institutions handed out daily allocations of barley, or of bread, to the people who worked for them. The theory fits the bowls' uniformity, their sheer numbers and the fact that they were disposable. It is supported by a striking detail in the earliest writing. In the proto-cuneiform script of Uruk, the sign for a ration or for bread looks remarkably like a bevelled-rim bowl, and the sign for "to eat" is a human head with that same bowl held to its mouth.

The interpretation is not universally accepted. Other scholars have suggested that the bowls were moulds for baking bread, containers for salt or curd, or vessels for temple offerings, and the truth may be that they served several purposes. But if the ration theory is even partly right, the bowls are a perfect symbol of the first farm. They represent a society in which large numbers of people no longer fed themselves from their own fields, but received a measured portion of food from an institution, in exchange for work that the institution directed. Each heap of discarded bowls is a heap of receipts for labour paid in grain.

Why grain?

In the last chapter I introduced the argument of the political scientist James C. Scott that the first states were built on cereals. It is worth spelling that argument out, because it explains why the first farms looked the way they did.

The domestication of wheat and barley in the Fertile Crescent was a long process, largely completed by around 8000 BCE. Settled villages, and in some places settled farming, are older still. Yet the first states — societies with kings, cities, taxes, writing and standing institutions of force — did not appear in southern Mesopotamia until around 3300 BCE. For some four thousand years, people farmed without being farmed. Something else had to happen first.

Scott's answer is that grain makes possible a particular kind of control. Cereal crops ripen all at once, at a predictable time, which means that the harvest can be watched and the share owed to the authorities collected on the threshing floor before the farmer can hide it. They grow above ground, in fields that can be surveyed and measured, so that the expected yield can be estimated in advance. Dried grain keeps for years and can be stored in granaries, transported in boats and counted out in standard measures. That makes it an almost ideal medium of tax, of rations and of wealth. Lentils, tubers and other crops have some of these qualities, but no other crop has all of them together. A state can be built on a granary. It is much harder to build one on a field of turnips that the peasants can dig up at night.

The flat, irrigated plain of southern Mesopotamia, the land between the Tigris and the Euphrates, suited this kind of farming extraordinarily well, and it also made escape difficult. Irrigation agriculture in a near-desert ties people to the canals: away from the water there is nothing to eat. By the end of the fourth millennium, Uruk had become what is often described as the first true city in the world. Population estimates are uncertain and vary widely, but most fall in the range of several tens of thousands. In the Early Dynastic period its wall, which later tradition credited to the legendary king Gilgamesh, ran for about nine kilometres and enclosed some five and a half square kilometres.

Walls have two sides. Scott points out that ancient laws, treaties and letters are full of concern about people running away — runaway slaves, runaway debtors, runaway peasants and conscripts — and that walls and gates controlled who left as well as who arrived. Grain made the farm possible. Geography, walls and the absence of alternatives formed its first fence.

The ledger

If grain was the first feed, writing was the first technology of management.

The archaeologist Denise Schmandt-Besserat has traced a plausible path by which Mesopotamian writing developed from small clay tokens of different shapes, used for thousands of years to represent quantities of goods such as measures of grain or heads of livestock. By the fourth millennium, tokens were being sealed inside hollow clay balls (bullae) to serve as tamper-proof records of transactions, with impressions of the tokens on the outside. Eventually the impressions alone sufficed, and the clay ball flattened into a tablet. Not every scholar accepts every step of her account. What is not disputed is the result: the earliest writing in Mesopotamia, the archaic tablets from the Eanna temple precinct at Uruk, is overwhelmingly administrative. According to the specialists who have deciphered them, chief among them Hans Nissen, Peter Damerow and Robert Englund, the great majority of these texts record goods, rations, fields, animals and workers. Most of the rest are lexical lists — standardised lists of words, including lists of official titles and professions — used to train the scribes who kept the books.

Writing, in other words, was not invented for literature. Stories, hymns and laws came later. The original purpose of writing was to allow an institution to keep track of what it was owed, what it held and what it had paid out, over periods longer than human memory and across more people than one official could know personally. A scribe with a tablet could hold an entire village to account.

The archaic accounts already show labour being measured in a way that a modern management consultant would recognise. Workers are counted, assigned to teams and given rations according to category. Over the following centuries, this accounting reached a level of precision that has rarely been equalled before the computer age.

The high point was the Third Dynasty of Ur, the so-called Ur III state, which ruled most of southern Mesopotamia from about 2112 to 2004 BCE. Its administrators left behind one of the largest bodies of written records from the ancient world: on the order of a hundred thousand tablets have been published, from archives such as those of Umma, Girsu and Puzrish-Dagan (modern Drehem), and many more lie unpublished in museum storerooms. The Ur III bureaucracy standardised weights and measures, organised the provinces into a rotating system of contributions to the centre (known as bala), and ran huge state livestock and textile operations. Its accountants measured labour in "worker-days", calculated how much work a given team should complete — so much earth moved, so much barley harvested, so much grain ground — and recorded the shortfall as a debt owed by the team's foreman.

In the province of Lagash, texts from the Ur III period record thousands of women and children working as weavers and grain-grinders in institutional workshops. Their rations were laid down in standard amounts by category. In the system reconstructed by the Assyriologist Ignace Gelb, adult male workers commonly received about sixty litres of barley a month, adult women about half that, and children less again, supplemented by oil, wool and occasionally other foods. That is enough to live on, but not much more. It is the ration of a herd that is kept, not one that is starved.

The scribes who managed all this were trained in schools known in Sumerian as the edubba, "the tablet house". The literary texts used in those schools, many of them copied in the early second millennium, show the scribes' sense of themselves as an elite: arduous training, harsh discipline, and at the end of it a career far above the toil of the fields. These were the first professional shepherds of the human farm. They did not own the storehouses. They made the storehouses work.

The great households

Who did own the storehouses?

For much of the twentieth century, scholars described early Sumerian cities as "temple-states" in which the temple of the city god owned all, or almost all, of the land, and every citizen was in effect a temple employee. This picture was built largely on the archive of a single institution in the city-state of Lagash, around 2400 BCE. Later research, by Igor Diakonoff, Ignace Gelb and others, showed that it was too simple. Alongside the temples there were palace estates, land held by extended families and communities, and land owned by individuals and bought and sold. Large institutions were enormously important, but they were not the whole economy.

The more accurate image is of a society organised around a few "great households". In Sumerian these were called é (house), the same word for the household of a god (a temple), a ruler (a palace) or a wealthy family. A great household owned land, herds, workshops and boats, and employed or supported hundreds or thousands of dependants. Some of these dependants worked the household's own land in return for rations. Others received plots of land to farm for themselves in return for service. Still others were tenants who paid a share of the crop. At the bottom were slaves, often war captives, and people bound to the household by debt.

At the top of each great household stood a very small group: the ruler and his family, the chief priests and the highest officials, who often came from the same families. This is the earliest form of the 0.01%. Below them was a larger group of scribes, overseers, stewards, foremen and specialist craftsmen: the 1%. And below them were the people with the ration bowls.

The archive from Lagash shows how closely the top of the system and the gods' property could be intertwined. In the generations before about 2350 BCE, the household of the ruler's wife administered extensive lands and personnel; under the ruler Uru-inimgina (also read as Urukagina), it was renamed the household of the goddess Bau. Whether this was a reform returning divine property from private hands, as his own inscriptions claim, or a change of title for the same ruling family's estate, as some scholars suspect, is one of the most debated questions in early Mesopotamian history. Either way, it shows that the boundary between "the god's property" and "the ruler's property" was thin, and that the owners of the farm found it useful to describe their wealth as the gods'.

The god's cattle

That brings us to the story.

Every farm needs a reason for the herd to accept its place, and the reason the first farms gave was that the herd was not really theirs to begin with. In the Mesopotamian creation story known as Atrahasis, preserved most fully in Old Babylonian copies from the eighteenth and seventeenth centuries BCE, the world before humans was a world of divine labour. The senior gods ruled; the junior gods, the Igigi, did the work, digging the rivers and canals. After thousands of years of this, the Igigi rebelled. They burned their tools, surrounded the house of the chief god Enlil at night and refused to go on. The senior gods' solution was not to share the work more fairly. It was to create human beings, moulded from clay mixed with the flesh and blood of a slain god, to bear the burden of labour so that the gods could rest.

It is a remarkable story to find at the very beginning of written literature. The first recorded labour dispute in history is a strike by gods, and its resolution is the invention of humanity as a workforce. The later Babylonian epic known as Enuma Elish repeats the idea: humans are made so that the gods may be at ease. The theology had a concrete economic meaning. The temple was literally the god's house; its land was the god's estate; the offerings placed before the god's statue were the god's meals, which were afterwards distributed to the temple's priests and staff. To work the temple's land, to pay the temple's dues and to accept its rations was not to be exploited by other people. It was to fulfil the purpose for which humans had been made.

In both Mesopotamia and Egypt, the ruler was described in the language of herding. Sipa, "shepherd", was a standard epithet of Sumerian kings, and later Mesopotamian rulers continued to use it. Egypt's rulers were also called shepherds, and an Egyptian wisdom text known as the Instruction for King Merikare, probably composed in the early second millennium BCE, describes humanity in words that could serve as an epigraph for this book. In Miriam Lichtheim's translation: "Well tended is mankind — god's cattle." The text goes on to say that the god made heaven and earth for their sake and gave them air and food. It is meant as an expression of divine care, and in its own terms it is a tender passage. But it says plainly what the herd is, who owns it, and who tends it on the owner's behalf.

There was also a more down-to-earth version of the story, written for the shepherds themselves. Egyptian scribal schools taught with texts that ridiculed every other occupation. The so-called Satire of the Trades, a Middle Kingdom composition attributed to a man named Khety, has a father taking his son to school and listing, one by one, the miseries of the smith, the potter, the builder, the reed-cutter, the fisherman and the farmer, in order to persuade him that the life of the scribe is the only one worth having. A New Kingdom schoolbook preserved in the Papyrus Lansing paints an even sharper picture of the farmer at harvest time: worms and pests have taken half the crop, hippopotamuses have eaten the rest, and then the scribe lands on the riverbank to collect the tax. There is no grain. The farmer is beaten, tied up and thrown in the well; his wife and children are bound. Be a scribe, the text concludes in effect, and you will be the one who collects, not the one who pays.

These texts were not propaganda for the farmers. They were training for the managers. They taught young scribes to see themselves as separate from, and superior to, the people whose grain they counted. That habit of mind — the conviction that one's own privileged position is the natural reward of learning — will reappear in every era of this book, and in Chapter 12 we will meet its modern descendant.

The interest is a young goat

If the story explained why people should work for the gods' households, debt explained why they could not stop.

Lending at interest is attested in Mesopotamia by the middle of the third millennium BCE, and it may well be older. The temples and palaces lent silver and grain; so did officials, merchants and wealthy families, very often the same people who managed the institutions. By the Old Babylonian period, customary maximum interest rates were 20% a year on silver and 33⅓% a year on barley.

The Sumerian word for interest, máš, is the same as the word for a young goat. The economic historian Michael Hudson has argued that the image came from herding: a loan was expected to give birth to new value, as a flock gives birth to kids. (Greek tokos and Latin fenus, both words for interest, carry similar associations of offspring and produce.) It is an apt image in a book about farms, and it also hides a cruelty. Flocks do not grow without limit; debts at compound interest, in principle, do. A creditor whose loan "breeds" at a third a year will find it has doubled in about three years. A farmer whose harvest fails cannot keep up.

The process that followed was grimly predictable. A farmer borrowed grain to survive the lean months before the harvest, or to pay a tax, or to cover a festival or a funeral. If the harvest was poor, the debt could not be repaid. The farmer pledged his land, or his labour, or — very commonly — the labour of his children or his wife, who went to work in the creditor's household until the debt was cleared. If it was never cleared, the family could lose its land altogether and fall into permanent dependency. Debt bondage of this kind was one of the main routes by which free farmers in ancient Mesopotamia became the dependants of the great households.

Here the farm metaphor needs one important refinement. Debt did not simply transfer wealth from the 99% to the 0.01%. It also shifted power within the elite. The creditors were often local officials, temple administrators and merchants — shepherds who had begun to accumulate herds of their own. Every free farmer who became a creditor's bondsman was one fewer free farmer paying taxes to the ruler, or serving in the ruler's army. A king whose peasants were all in debt to his officials was a king whose officials were getting dangerously strong.

The first breakouts: freedom as a return to the mother

That conflict at the top produced the first recorded reforms in history.

Around 2400 BCE the ruler of Lagash, Enmetena, recorded in an inscription that he had "established freedom" in his city, returning children to their mothers and mothers to their children, and remitting debts. The Sumerian word was amargi (often written ama-ar-gi₄), which literally means "return to the mother". It is the oldest known word for freedom, and it meant, quite concretely, the release of family members who had been taken into bondage for debt.

About two generations later, Uru-inimgina of Lagash left a longer and more famous account. His inscriptions describe a city in which officials and the ruler's household had seized everything: inspectors collected dues from boatmen, shepherds and fishermen; fees were charged for divorces and burials; the property of the temples had been taken into the ruler's estate; the powerful bought up the houses and donkeys of the poor at prices of their own choosing. Uru-inimgina claimed to have swept all this away, to have removed the inspectors, cancelled debts and established amargi. And he declared that he had made a solemn agreement with the god Ningirsu that the powerful would not do wrong to the orphan and the widow.

Historians read these texts with caution. Uru-inimgina had probably come to power irregularly, and denouncing the abuses of his predecessors was an excellent way of establishing his legitimacy. His reign was short: within a few years Lagash was conquered by Lugalzagesi of Umma. But the pattern his inscriptions describe — a ruler presenting himself as the protector of the weak against the predations of the rich, and cancelling debts to restore the free farming population — became a fixture of Mesopotamian kingship for the next thousand years.

In the Old Babylonian period kings regularly issued edicts of mīšarum, "justice" or "righteousness", often in the first year of their reigns and sometimes later. The most complete surviving text, the edict of King Ammi-saduqa of Babylon from the mid-seventeenth century BCE, is precise about what was and was not included. Arrears owed to the palace by farmers and tenants were cancelled. Consumer debts in barley and silver owed by ordinary people were annulled, and members of a debtor's family who had been taken into bondage were to be released. But commercial loans between merchants were not cancelled, and people who had been sold as chattel slaves were not freed. Severe penalties were laid down for creditors who tried to evade the edict by disguising a loan as a sale.

How often such edicts were issued, and how effective they were, is debated. Hudson and the anthropologist David Graeber have argued that "clean slates" were a regular and essential feature of Bronze Age economies, preventing debt from destroying the free population. Other Assyriologists are more cautious, noting that we have the full text of only a few edicts, that creditors developed contracts designed to circumvent them, and that private debt bondage clearly continued. Both things are probably true. The edicts were real, and they did return people to their families. They were also limited, temporary and designed to protect the ruler's tax base as much as the debtor's household. They were a release valve, not a demolition of the fence.

Seen through the lens of this book, the debt cancellations reveal something important about how the farm survives. The 0.01% does not always side with the 1%. When the shepherds grow too greedy — when they start to fleece the herd for themselves, threatening the owners' own harvest — the owners can step in on the herd's side, and present themselves as its saviours. Some of the most effective reforms in history were made by elites protecting the farm from its own managers.

There were other breakouts too, less visible in the records because they were made by people who did not write. Scott and other scholars emphasise that the most common form of resistance in early states was simply leaving: abandoning fields, moving out of reach of the tax collector, joining pastoral groups beyond the irrigated zone. The early states of Mesopotamia were remarkably fragile. The Ur III state, the most elaborately managed farm of the third millennium, collapsed around 2004 BCE amid crop failure, invasion and the defection of its provinces, after barely a century. When the centre failed, the people who had been counted on its tablets dispersed. It is possible, and some archaeologists argue it, that for many of them this was not a catastrophe but a relief.

Egypt: one farm, one owner

If Mesopotamia was a landscape of competing city-states and great households, Egypt was something even more ambitious: a single farm, nearly a thousand kilometres long, with a single owner.

The Nile valley was unified under one king around 3100 BCE. The geography was even more favourable to a grain state than Mesopotamia's. Every year the river flooded, laying down fresh silt on the fields along its banks, then retreated in time for planting. The valley was a narrow green strip between deserts; beyond it, survival was nearly impossible. A farmer could not easily leave. And the king, who in principle owned the entire land, could assess the harvest with unusual precision.

Egyptian administration did exactly that. The surviving fragments of the royal annals known as the Palermo Stone, carved in the Fifth Dynasty, record for each year of the early kings the height of the Nile flood, because the height of the flood predicted the size of the harvest, and therefore the tax that could be collected. In the Old Kingdom, years were often dated by the "count" — a periodic census of the country's cattle and other property, carried out at first every two years and later more often. Time itself was measured by the counting of the herd. Merer's diary is dated in exactly this way, by the count, to a year late in Khufu's reign.

The tax system that rested on these counts was enforced on the threshing floor. Tomb paintings and reliefs of the Old Kingdom officials who ran it, such as the vizier Mereruka at Saqqara, show the process with a frankness that is startling to modern eyes: village headmen are brought before seated scribes to account for their harvests, and those who have failed to deliver are held down and beaten. These scenes were not hidden. They were carved on the walls of the officials' own tombs, intended for eternity, as a proud record of the deceased's service.

Alongside the grain tax was the second great feed of Egypt: labour. The Egyptian state could call up its subjects for compulsory work on royal projects — canals, dykes, quarries, expeditions and monuments — for periods of the year. This system of corvée labour was one of the defining features of Egyptian society for thousands of years. Exemptions from it were a valuable privilege. Old Kingdom royal decrees, such as one issued by King Pepi I for the pyramid town of his predecessor Sneferu at Dahshur, solemnly freed particular temples and their personnel from being conscripted for royal labour. Such decrees show who counted as a shepherd: those whose service to the farm's sacred institutions excused them from carrying its stones.

The pyramid as a labour regime

That brings us back to Giza, and to the blocks of white limestone in Merer's boats.

For most of modern history, the popular image of pyramid-building was of armies of slaves toiling under the whip. The image owes something to the Bible's account of the Israelites in Egypt and something to Hollywood, but it is often traced to the Greek historian Herodotus, who visited Egypt in the fifth century BCE — two thousand years after the Great Pyramid was built. Herodotus wrote that the pyramid of "Cheops" (Khufu) took twenty years to build and that a hundred thousand men worked on it at a time in three-month shifts. He did not, in fact, call them slaves; he described a king who had closed the temples and forced all the Egyptians to labour for him. His figures are not reliable, and his story is the tale of a tyrant, told by priests a very long time afterwards.

Archaeology has replaced it with something more interesting. From the late 1980s onwards, excavations at Giza led by Mark Lehner and Zahi Hawass uncovered the settlement where the pyramid builders lived, now known as Heit el-Ghurab, and the cemetery where many of them were buried. The settlement included bakeries, breweries, storehouses, barracks-like galleries and administrative buildings. Animal bones show that the workers were supplied with meat in quantities that implied large herds of cattle, sheep and goats being driven in from estates elsewhere in the country. The skeletons in the cemetery show signs of heavy physical labour — arthritis, compressed vertebrae, injuries — but also of medical care, including healed fractures and, in a few cases, what appear to be amputations the patients survived. Many of the tombs were built for overseers and craftsmen, who were buried with honour close to the king's monument.

Graffiti left by work gangs inside the monuments show how the labour was organised. Teams bore names such as "Friends of Khufu" and "Drunkards of Menkaure", and were subdivided into smaller units. Lehner's estimates of the workforce at any one time are much lower than Herodotus's — in the range of perhaps twenty to thirty thousand people, including permanent skilled workers and a much larger rotating body of conscripted labourers. Estimates vary, and the true number is not known.

This picture has sometimes been summarised in the press as "the pyramids were built by free, well-fed workers, not slaves". That is an overcorrection. The rotating labourers were not volunteers. They were conscripts, drafted under a system of obligation that ordinary Egyptians could not refuse, probably for spells of a few months, perhaps during the flood season when fields were under water. Their food was not a wage freely bargained for, but rations issued by the state, from grain that the state had collected from them and their neighbours as tax. The rations were generous because a well-fed worker hauls more stone. The medical care was real because injured workers are wasted labour.

That is exactly how a well-run farm behaves. The Great Pyramid, made of an estimated 2.3 million blocks, is the most spectacular monument the human farm has ever produced: an enormous quantity of surplus labour, taken from the people of the Nile valley by tax and corvée, and converted into a tomb for one man. It was not built by slaves in chains. It was built by a whole society organised as a single estate, in which feeding the workers well was part of the efficiency of taking their labour.

And it was built for a reason that the builders themselves may well have believed in. The Egyptian king was the son of the sun god, the guarantor of maat — the order of the universe that kept the sun rising and the Nile flooding. If the king's eternal life was secured, the cosmos itself remained in balance. To haul stones for his tomb was to contribute to the survival of the world. That is the story at its most powerful: not a lie imposed by force, but a belief that turned the extraction of labour into an act of devotion.

The story had a further, telling development. By the Middle Kingdom, wealthier Egyptians were being buried with small figurines known as shabtis. Their inscriptions, later standardised as Spell 6 of the Book of the Dead, instructed the figure to answer "Here I am!" if the deceased was ever called to do labour in the afterlife — cultivating fields, filling canals with water, carrying sand. Even in the next world, the farm went on, and the corvée was still due. But those who could afford it sent a substitute.

The workmen's strike

The breakout in Egypt came late and in a small place, but it is one of the most vividly documented events in the ancient world.

At Deir el-Medina, on the west bank of the Nile opposite Thebes, lived the community of skilled craftsmen who cut and decorated the royal tombs in the Valley of the Kings. They were the aristocracy of Egyptian labour: literate in many cases, relatively well housed, and paid by the state in monthly rations of grain. According to the reconstruction of their pay by the Egyptologists Jaroslav Černý and Jac Janssen, an ordinary workman received about four sacks of emmer wheat and one and a half of barley each month, with more for foremen and scribes — enough to support a family and to trade with. They depended entirely on those deliveries.

In the twenty-ninth year of the reign of Ramesses III, around 1157 BCE, the deliveries were late. The state was under strain: the long reign had been burdened by wars, administrative corruption and economic difficulties. When the rations failed to arrive, the workmen did something that, as far as we know, had never been recorded before. They put down their tools and walked off the job. They marched out of the workplace, passed the guard posts and sat down behind the mortuary temples on the west bank, refusing to go back. The episode was written down by the scribe Amennakht, in a document now known as the Turin Strike Papyrus. In his account the men insist that they have come because of hunger and thirst, that they lack clothing, oil, fish and vegetables, and they ask for their complaint to be sent to Pharaoh and to the vizier so that they can be provided for.

The officials responded with negotiation, promises and partial deliveries. The work stoppages recurred several times over the following months before rations were restored. Later papyri from the reign of Ramesses IX tell of a darker breakdown: investigations into the robbery of royal tombs, in which workmen and local officials were accused of plundering the very treasures they had been employed to protect. When the feed dried up, some members of the herd began to raid the owner's storehouse.

There had been earlier and greater upheavals. At the end of the Old Kingdom, around 2200 BCE, central authority in Egypt collapsed into a period of regional rule and recurrent famine that historians call the First Intermediate Period. A later literary text, the Admonitions of Ipuwer, describes a world turned upside down: the poor have become rich, servants have taken their masters' goods, the offices have been broken open and their records taken away. How much of this reflects real events is uncertain — the text is literature, its date is disputed, and it was almost certainly written by and for the elite, as a warning of what disorder looks like. But the detail of the records being seized is telling. When the herd broke out, one of the first things it went for was the ledger.

What the first harvest teaches

Put the evidence from Mesopotamia and Egypt together, and the five parts of the farm are already in place by 2500 BCE.

The fence was geography, water and the walls of the city: the irrigated land on which survival depended, the desert beyond, and the law that pursued those who ran away. Above all it was debt, which bound families to creditors more effectively than any wall.

The feed was grain and labour: the harvest share collected on the threshing floor, the rations issued in return for work in the great households, and the corvée that built canals, temples and pyramids.

The shepherds were the scribes, overseers, stewards and priests — Kushim with his ledger, Merer with his diary, the officials in Mereruka's tomb watching the headmen being beaten, the schoolboys learning to despise farmers. They lived well, but they did not own the storehouses, and they were constantly tempted to start building herds of their own.

The story was that humanity had been created to work for the gods, that the land and its produce belonged to the gods and their earthly representative, and that the king was a shepherd whose care for the herd was a sacred duty.

The breakout took several forms: flight from the state; collapses in which the records themselves were destroyed; strikes by those who could not be easily replaced; and, most strikingly, reforms imposed from above, in which rulers cancelled debts and returned bondsmen to their mothers in order to protect the farm from its own managers.

There is a final lesson in these first farms, and it is perhaps the most important. Nothing in them required a conspiracy. The king of Ur did not need to plot against his subjects; he needed accountants, standard measures and a story about the gods. The creditor of Lagash did not need to hate the farmer whose children he took as pledges; he needed only the interest rate and a bad harvest. The Egyptian peasant who hauled stone for Khufu may have believed, sincerely, that he was keeping the sun in the sky. The arrangement worked because each person in it was doing what made sense in his own position — and because the arrangement had been made to look like the order of the world.

Bridge: law written in stone

The Ur III state fell around 2004 BCE. In the centuries that followed, new dynasties, many of them founded by Amorite chiefs from the edges of the old Mesopotamian world, competed for control of the river cities. One of them, at the previously modest town of Babylon, would produce a king who did something that no ruler before him had done on such a scale. Around 1750 BCE, Hammurabi had his laws carved on tall stone stelae and set up in public, crowned with an image of the king receiving his authority from the sun god of justice.

The first farms had run on accounts and on myths. The next stage would make the arrangement permanent in a new way: by writing down who owed what to whom, and what would happen to those who refused, and by declaring that the gods themselves had chosen the king to enforce it. In the next chapter, we turn to the Bronze Age empires and to the invention of legitimacy — and to the priesthoods who made the farmer's god into the landlord's partner.