Why did we start farming?

Steven Mithen

  • Against the Grain: A Deep History of the Earliest States by James C. Scott
    Yale, 336 pp, £20.00, September, ISBN 978 0 300 18291 0

When our ancestors began to control fire, most likely somewhere in Africa around 400,000 years ago, the planet was set on a new course. We have little idea and even less evidence of how early humans made fire; perhaps they carried around smouldering bundles of leaves from forest fires, or captured the sparks thrown off when chipping stone or rubbing sticks together. However it happened, the human control of fire made an indelible mark on the earth’s ecosystems, and marked the beginning of the Anthropocene – the epoch in which humans have had a significant impact on the planet.

In Against the Grain James Scott describes these early stages as a ‘“thin” Anthropocene’, but ever since, the Anthropocene has been getting thicker. New layers of human impact were added by the adoption of farming about ten thousand years ago, the invention of the steam engine around 1780, and the dropping of the atomic bomb in 1945. Today the Anthropocene is so dense that we have virtually lost sight of anything that could be called ‘the natural world’.

Fire changed humans as well as the world. Eating cooked food transformed our bodies; we developed a much shorter digestive tract, meaning that more metabolic energy was available to grow our brains. At the same time, Homo sapiens became domesticated by its dependence on fire for warmth, protection and fuel. If this was the start of human progress towards ‘civilisation’, then – according to the conventional narrative – the next step was the invention of agriculture around ten thousand years ago. Farming, it is said, saved us from a dreary nomadic Stone Age hunter-gatherer existence by allowing us to settle down, build towns and develop the city-states that were the centres of early civilisations. People flocked to them for the security, leisure and economic opportunities gained from living within thick city walls. The story continues with the collapse of the city-states and barbarian insurgency, plunging civilised worlds – ancient Mesopotamia, China, Mesoamerica – into their dark ages. Thus civilisations rise and fall. Or so we are told.

The perfectly formed city-state is the ideal, deeply ingrained in the Western psyche, on which our notion of the nation-state is founded, ultimately inspiring Donald Trump’s notion of a ‘city’ wall to keep out the barbarian Mexican horde, and Brexiters’ desire to ‘take back control’ from insurgent European bureaucrats. But what if the conventional narrative is entirely wrong? What if ancient ruins testify to an aberration in the normal state of human affairs rather than a glorious and ancient past to whose achievements we should once again aspire? What if the origin of farming wasn’t a moment of liberation but of entrapment? Scott offers an alternative to the conventional narrative that is altogether more fascinating, not least in the way it omits any self-congratulation about human achievement. His account of the deep past doesn’t purport to be definitive, but it is surely more accurate than the one we’re used to, and it implicitly exposes the flaws in contemporary political ideas that ultimately rest on a narrative of human progress and on the ideal of the city/nation-state.

Why did people start farming? At the ‘Man the Hunter’ symposium in Chicago in 1966, Marshall Sahlins drew on research from the likes of Richard B. Lee among the !Kung of the Kalahari to argue that hunter-gatherers enjoyed the ‘original affluent society’. Even in the most marginal environments, he said, hunter-gatherers weren’t engaged in a constant struggle for survival, but had a leisurely lifestyle. Sahlins and his sources may have pushed the argument a little too far, neglecting to consider, for instance, the time spent preparing food (lots of mongongo nuts to crack). But their case was strong enough to deal a severe blow to the idea that farming was salvation for hunter-gatherers: however you cut it, farming involves much higher workloads and incurs more physical ailments than relying on the wild. And the more we discover, as Scott points out, the better a hunter-gatherer diet, health and work-life balance look.

This is especially true of the hunter-gatherers who dwelled in the wetlands where the first farming communities developed, in the Fertile Crescent, the arc of South-West Asia now covered by Jordan, the Occupied Palestinian Territories, Israel, Lebanon, Syria, southern Turkey, Iran and Iraq. Scott’s book focuses on Mesopotamia – the land between the Tigris and the Euphrates where the first city-states also appeared – though it takes many diversions into ancient China, Mesoamerica, and the Roman and Greek ancient worlds. Until about ten thousand years ago, Mesopotamia had been a world of hunter-gatherers with access to a huge range of resources: reeds and sedges for building and food, a great variety of edible plants (clubrush, cat’s-tails, water lily, bulrush), tortoises, fish, molluscs, crustaceans, birds, waterfowl, small mammals and migrating gazelles, which were the chief source of protein. The wild larder was routinely replenished by the annual cycle of the ripening of fruits and wild vegetables, and the seasonal changes that brought the arrival of migratory species.

Wetland environments were available to hunter-gatherers elsewhere in the world too. In China’s Hangzhou Bay, phenomenally well-preserved waterlogged sites show that hunter-gatherers became sedentary amid a bounteous range of wild resources. I do my own fieldwork in Wadi Faynan, in southern Jordan, which is now an arid, largely treeless landscape, but 12,000 years ago a perennial river flowed there. Where it joined with the river of Wadi Dana, an oasis-like niche was created. That is where the early Neolithic site of WF16 is now located (we call it Neolithic even though there is no trace of domesticated crops and animals). A dense cluster of about thirty semi-subterranean dwellings was constructed there between 12,500 and 10,500 years ago by hunter-gatherers who were clearly enjoying a diverse and resilient set of local resources: hunting wild goats, trapping birds, collecting figs, wild grass, nuts and so forth. I suspect they were also practising some form of environmental management, setting fires to promote young shoots, building small dams to retain and divert water, and undertaking selective culls among wild herds to sustain animal populations.

The key to food security was diversity: if, by chance, a particular foodstuff gave out, there were always more to choose from. And so hunter-gatherers could become sedentary if they wished, without having to grow crops or rear livestock. The first cultivation of barley and wheat came from the slight modification of wild stands – weeding, removing pests, transplanting, sowing seeds into alluvial soils. This would have provided hunter-gatherers with a new source of food at the cost of little additional effort. The mystery is why cereal-farming came to be so dominant. Why hunter-gatherers passed up their affluent lifestyle in favour of far more onerous and risky existences growing a narrow range of crops and managing livestock is a fundamental question to which we have no good answer. Was it by choice, or was that first sowing of seed a trap, locking people into a seasonal cycle of planting and harvesting from which we have been unable to escape?

Scott considers whether the Younger Dryas, a period of markedly colder and drier conditions between 12,900 and 11,700 years ago, forced hunter-gatherers into farming. But while the change in climate may have inspired more experimentation with cultivation and herding, the Younger Dryas is too early: communities committed to cereals and livestock didn’t arise until about ten thousand years ago. Scott overlooks another possible factor: religious belief. The discovery of the Neolithic hill-top sanctuary of Göbekli Tepe in southern Turkey in 1994 went against the grain of conventional archaeological understanding of the Neolithic. Here, around 11,000 years ago, hunter-gatherers had constructed a vast complex of massive decorated stone pillars in exactly the same place that domesticated strains of wheat had evolved.

The quantities of food needed to feed the workforce and those who gathered for rituals at Göbekli must have been huge: if the Neolithic gods could persuade people to invest so much effort in construction, and to suffer the physical injuries, ailments and deaths that came along with it, then perhaps expending those extra calories in the fields would have seemed quite trivial. Even then, Göbekli doesn’t help us explain why cereal farming and goat herding took such a hold elsewhere. Personally I find it difficult to resist the theory of unintended self-entrapment into the farming lifestyle, which was then legitimated by Neolithic ideology. We find evidence of burial rituals and skull cults throughout the Fertile Crescent.

I see the consequence in Wadi Faynan. A little less than 10,500 years ago the hunter-gatherer settlement of WF16 was abandoned and replaced by the Neolithic farming village of Ghuwayr 1. This had rectangular stone houses and storerooms. Piles of excavated grindstones testify to the back and knee-breaking work of grinding barley grain; meanwhile, domesticated goats had begun to eat up the local vegetation – the first step to today’s barren landscape.

It isn’t just the workload that makes the move to farming so surprising. Scott argues that these early Neolithic villages – his preferred term is ‘multi-species resettlement camps’ – were serious health hazards for all who stepped or slept in them. The villages were overrun by people and animals, and the waste they produced; mice, rats and crows took up residence, bringing micro-parasites of their own to join the burgeoning community of fleas, lice and mites. Until about 10,800 years ago domestic waste at WF16 was left on the floors of rooms or dumped between houses; after that date a centralised, communal waste dump appears to have been set up, but it would nevertheless have been a festering mess immediately adjacent to peoples’ homes. At the early Neolithic site of Boncuklu in Turkey, excavated by Douglas Baird of Liverpool University, a human toilet has been found in the middle of the village. Human susceptibility to illness would have been boosted by the increasing narrowness of the diet.

Outbreaks of disease, or chronic ill-health, may well provide a better explanation than ritual, ideology and conflict for the dense accumulations of human remains often found packed below floors or into special buildings in the Neolithic. The remains of around seventy individuals have been found in the Skull Building at Çayönü in Turkey. A mass of bodies, primarily infants, are currently being excavated by Roger and Wendy Matthews of the University of Reading from a single building at the early Neolithic site of Bestansur in northern Iraq. Signs of multiple burials have been found below house floors in late Neolithic sites throughout the region. Many Neolithic villages appear to have been abandoned quite suddenly: might that have been because of endemic disease?

It wasn’t just people: goats, sheep and cattle would also have suffered from overcrowding. A combination of selective culling and domestication reduced their sexual dimorphism, caused neotany (preservation of juvenile characteristics into adulthood) and changed their brains, both in size and function. The brains of domesticated sheep are 24 per cent smaller than those of their wild counterparts, and they are less timid, which probably relates to changes in the limbic systems relating to emotional response. The rodents in the settlements underwent similar changes, indicating that some of these were a consequence of village life rather than human selection. The implication is that we should assume the same process in people: how did the brains, personalities and patterns of thought of Neolithic farmers differ from those of their hunter-gatherer ancestors? Was this a critical step towards the complete domestication of Homo?


According to the conventional narrative, once the transition to farming had taken place, further progress would have been swift, culminating in the city-states with their glorious architecture, art and secure economies. This did happen eventually: by 3200 BCE Uruk in Mesopotamia was the largest city in the world, with between 25,000 and 50,000 inhabitants, soon joined by the city-states of Kish, Nippur, Isin, Lagash, Eridu and Ur. As well as monumental architecture, they all had thick city walls; a class of administrators and priests; centralised production of crafts such as ceramics and textiles; social hierarchies topped by a king; and – surely the key index of human ‘progress’ – tax collectors. But why did it take so long – about four thousand years – for the city-states to appear? The reason is probably the disease, pestilence and economic fragility of those Neolithic villages. How did they survive and grow at all? Well, although farming would have significantly increased mortality rates in both infants and adults, sedentism would have increased fertility. Mobile hunter-gatherers were effectively limited by the demands of travel to having one child every four years. An increase in fertility that just about outpaced the increase in mortality would account for the slow, steady increase in population in the villages. By 3500 BCE the economic and demographic conditions were in place for a power-grab by would-be leaders.

As Scott explains, the city-states were dependent on grain: wheat and barley in Mesopotamia, millet in China, maize in Mesoamerica. The reason is that cereals are easy to tax: they ripen at predictable times, the size of the harvest can easily be assessed, and the grain can be divided, transported and distributed in precisely measured rations by weight and volume. It is much more difficult to tax merchants who smuggle their goods, or to tax crops such as tubers that are hidden underground and can be dispersed throughout woodlands, or chickpeas and lentils, which have an extended ripening season. If the cereal farming takes place close to a river that can be used for bulk transportation, a potent power base can be established. That is what happened among the river and canal systems of Mesopotamia and Ancient China.

Record-keeping naturally follows. Writing was used for this purpose in Mesopotamia for five hundred years before it was used for storytelling and poetry. Huge numbers of lists on clay tablets recording crops, yields and taxes have been recovered. It remains unclear whether these tell us about projections and aspirations or are records of what was actually harvested. In China the earliest writing is found on oracle bones from 1600-1050 BCE, but record-keeping was adopted during the Warring States period, 476-221 BCE. The Qin verged on the obsessive, intent on imposing standardised methods of measurement throughout their empire, eliminating all local practices.

The third ingredient needed to form city-states, after grain and record-keeping, was people. Lots of people. The state with the most people was generally the richest in its region and prevailed militarily. Warfare between neighbouring states was endemic, not to gain territory but to seize captives. The aim wasn’t only to acquire slaves for unskilled labour but also people with technical expertise that could be used in boat-building, weaving and metalworking. In some cases whole communities were relocated closer to the centre of the state so that their crop production could be controlled. The key was to generate the agricultural surplus needed to feed the priests and administrators, the craftsmen, and the workers digging the canals and building the walls. But the peasants had to be compelled to do so, which incurred the risk of mass flight. The primary purpose of the walls around city-states may have been to keep people in, not to keep the barbarians out. (‘Barbarian’ and its variants are terms used by state centres to stigmatise those who live outside their control; the Huns, Amorites, Goths and other barbarian tribes were no more than loose confederations of disparate peoples brought together temporarily for military purposes and characterised as a ‘people’ only by the threatened state itself. Scott uses ‘barbarian’ simply as a shorthand for non-state people, whether they lived by hunting and gathering, farming or nomadic pastoralism.)

In describing the early city-states of Mesopotamia, Scott projects backwards from the historical records of the great slave societies of Greece and Rome. His account of the slaves and the way they were controlled seems strangely familiar. Much like migrant labourers and refugees in Europe today, they came from scattered locations and were separated from their families, demobilised and atomised and hence easier to control. Slaves, like today’s migrants, were used for tasks that were vital to the needs of the elites but were shunned by free men. And slaves, like refugee workers, were gradually integrated into the local population, which reduced the chance of insurrection and was necessary to keep a slave-taking society going. In some early states human domestication took a further step: written records from Uruk use the same age and sex categories to describe labourers and the state-controlled herds of animals. Female slaves were kept for breeding as much as for manual labour.

The city-states were fragile. The Old Kingdom of Egypt collapsed around 2100 BCE, Ur III around 2000 BCE and the Minoan Palaces around 1450 BCE – just three examples among many. Collapse could mean nothing more than the abandonment of the centre and the redistribution of the population into independent settlements, to be followed by the next cycle of annexation. The forces of destabilisation were such that it is remarkable any state was sustained at all. Many had short lives: the Qin lasted a mere 15 years and the third dynasty of Ur fewer than a hundred years.

Disease remained a signal problem. Virtually every infectious disease caused by micro-organisms and specifically adapted to Homo sapiens has arisen in the last ten thousand years, many of them in the last five thousand years as an effect of ‘civilisation’: cholera, smallpox, measles, influenza, chickenpox, and perhaps malaria. Germs travelled with trade and people, accompanying armies, slaves, or annexed peasants as they moved from state to state. Disease could destroy a harvest or a population of livestock, as could floods or drought. The frequency of flooding increased as a result of the deforestation caused by an insatiable demand for timber and fuel, and excessive grazing by goats. In the absence of trees, less rainwater penetrated the soil and reached the aquifers beneath, instead flowing destructively across the landscape. With deforestation came soil erosion, and with excessive irrigation came salinisation and a decrease in fertility. To this we should add continual warfare, which drew people away from farming and caused cities to be located with defensibility in mind rather than economic growth. There were also internal conflicts: battles for succession, civil wars and insurrections. And when an elite felt threatened it might become self-destructive, perhaps by taxing too heavily and leaving farmers to starve.

Scott writes about the normalising effects of state collapse. Often it was the best thing possible for a people now emancipated from disease, taxes and labour. In the subsequent ‘dark ages’ – a propaganda term used by the elite – democracy and culture could flourish. Homer’s Iliad and Odyssey date from the dark age of Greece. This is in marked contrast to the consequences of state collapse today, now that there is no longer an external barbarian world to escape into. When Syria collapsed its refugees had no choice but cross the border to another state, whether Lebanon, Jordan or Turkey.

According to Scott, the period of early states was the Golden Age for the barbarians. They could prey on a state as if it were just another resource for hunting or harvesting; the Berbers of North Africa described raiding as their version of agriculture. Or they could take part in trade. Hunters, foragers and marine collectors supplied sought-after goods from mountains and forests such as ivory, rhino horn, spices, feathers, beeswax and furs; others controlled the trade routes, and along with them the traffic of slaves. When it was in their interest, the barbarians could become mercenaries for a state, or even conquer it and themselves become the new ruling class.

In Scott’s picture, the barbarians and the city-states were entirely dependent on each other for their existence. They rose and fell together: the Huns and the Romans; the ‘Sea People’ and the Egyptians. And for the vast part of recorded history the majority of people lived in the barbarian world. Scott’s view is that the barbarian Golden Age ended as recently as four hundred years ago, when the power of the state finally became overwhelming, partly due to the invention of durable gunpowder. Which is, of course, a means to make fire sparked by flint – a return to the ‘moment’ 400,000 years earlier which marked the beginning not of the steady rise of civilisation, but rather the muddled and messy affair that is the human past.