Appearance
3.7 — The Farming Revolution
Farming made people shorter, sicker, and made them work harder. The skeletal evidence in Chapter 3.3 is not ambiguous about this: early farmers were smaller than the foragers who preceded them in the same regions, had worse teeth, more signs of nutritional deficiency, more infectious disease, and more evidence of repetitive physical labour.
And every settled society on Earth today descends from people who did it anyway.
That is the puzzle. This chapter answers it, and the answer explains a great deal about how human societies work.
Where and when, and the crucial fact about it
| Region | Began | Founding crops |
|---|---|---|
| Fertile Crescent | ~9500 BCE | Emmer and einkorn wheat, barley, lentil, pea, chickpea, flax |
| China (Yangtze) | ~7000 BCE | Rice |
| China (Yellow River) | ~7000 BCE | Foxtail and broomcorn millet |
| New Guinea | ~7000 BCE | Taro, banana, yam |
| Mesoamerica | ~7000 BCE onward | Maize, beans, squash, chilli |
| South America (Andes) | ~7000 BCE | Potato, quinoa, beans |
| Sub-Saharan Africa | ~4000 BCE | Sorghum, pearl millet, African rice, yam |
| Eastern North America | ~3000 BCE | Squash, sunflower, goosefoot |
| South Asia | ~7000 BCE at Mehrgarh; independent rice and millet later | Barley, wheat; later Indian rice, mung, millets |
At least eleven independent centres, separated by oceans and continents, all within a few thousand years of each other after two hundred thousand years of our species doing nothing of the sort. That timing is the strongest single clue, and it points at something global. The something is the climate: the Holocene began, and it was warm, wet, stable, and had higher carbon dioxide than the glacial world. Chapter 2.8 and Chapter 3.6 supplied that setting; here it does its work.
India's own case. Mehrgarh, on the Kachi plain in Balochistan at the foot of the Bolan Pass, has occupation from around 7000 BCE with barley, wheat, and later cattle, sheep and goats. Barley and wheat arrived from the west, but the cattle are zebu — humped Indian cattle — domesticated locally from Indian aurochs, which genetics confirms as a separate domestication. Rice in the Ganges valley, at sites such as Lahuradewa, shows early cultivation of Indian wild rice; the relationship between that and Chinese rice domestication is an active research question, with the current picture being that the fully domesticated form spread from China and hybridised with locally cultivated Indian rice.
Domestication: what actually changes
Domestication is not just growing a plant. It is changing the plant, through generations of unconscious selection, until it can no longer survive without you.
The single most important change in cereals is the loss of shattering. A wild grass, when its seeds ripen, breaks apart at the stem so the seeds scatter — that is how it reproduces. A plant that scatters its seed is useless to a harvester. When people cut ripe stands with sickles, the plants that happened to carry a mutation preventing shattering were the ones whose seed ended up in the basket, and therefore in next year's field.
Notice that nobody decided this. It requires no plan and no understanding of heredity. Harvest with a sickle for long enough and non-shattering becomes universal in your crop. It also makes the plant completely dependent on humans, because it can no longer disperse its own seed.
Experimental work suggests the change can occur in as little as a couple of centuries under sustained harvesting, and the archaeological record shows it taking roughly one to two thousand years, which suggests early cultivation was casual and intermittent rather than intensive.
The other changes follow the same logic. Larger seeds, because the biggest were sown. Simultaneous ripening, because whatever was ripe on harvest day got harvested. Loss of dormancy — wild seeds stagger their germination over years as insurance, which is a nuisance in a field. Thinner seed coats. Less bitterness, because people preferred the palatable ones.
In animals, the same process gives a recognisable syndrome: smaller body, smaller brain, smaller teeth, floppy ears, curly tails, patchy coat colours, and a longer juvenile period. The famous Russian experiment breeding silver foxes purely for tameness produced these traits as a package within a few dozen generations, which suggests they are linked developmentally rather than each being selected separately.
Why, when it made life worse
Four explanations, and they are complementary rather than competing.
1. Nobody chose it, because it happened too slowly to notice. This is the most important point. The transition from occasional cultivation to full dependence took roughly two thousand years — a hundred generations. No individual ever made a decision to become a farmer. Each generation did slightly more of what their parents did, because at each step it made local sense, and the cumulative destination was invisible.
2. It supports far more people per unit of land, even while making each one worse off. This is the key insight and it needs stating carefully. Farming produces more calories per hectare and fewer calories per hour of work. So a farming population grows — and once it has grown, it cannot go back, because the land will no longer support that many people by foraging.
It is a trap, and it locks behind you. A group of a hundred foragers who take up farming becomes a group of a thousand farmers in a few centuries. Those thousand people cannot return to foraging; nine hundred of them would have to die first. Each generation is therefore committed by the choices of the one before.
3. Birth spacing, which is the underrated mechanism. Mobile foragers cannot carry more than one small child, and prolonged breastfeeding suppresses fertility, so births are typically spaced three to four years apart. A settled population has neither constraint, and cereal gruel allows earlier weaning. Birth intervals fall, population rises. This alone can drive the whole transition without anyone intending anything.
4. Storage, and what storage does to a society. Grain can be kept for years. That is genuine insurance against a bad season, and it is a real benefit. It is also the first accumulable, countable, seizable form of wealth, and Chapter 3.9 is about what that did.
And one honest complication. The picture of miserable farmers is drawn largely from the transition period. Over the long run, settled agricultural societies developed the surpluses that paid for everything in the rest of this volume. The trade was: worse for the average individual in the short and medium term, more people, and eventually cities, writing, specialists, science and medicine. Whether that is a good bargain is a question this book returns to in Chapter 16.1, and it does not have a clean answer.
What agriculture requires, and why some regions had it easy
Not all wild plants and animals are domesticable, and the distribution of suitable species was not even.
For a plant to be worth domesticating it needs to be reasonably nutritious, storable, capable of being grown densely, and — the crucial one — capable of self-pollination or easy cultivation. Of the world's roughly 200,000 flowering plant species, only a few thousand are edible and only a couple of hundred have ever been domesticated, with about a dozen supplying most of the world's calories today.
For a large mammal to be domesticable it must eat something cheap, grow fast, breed in captivity, not panic in enclosures, and — most restrictive of all — live in herds with a dominance hierarchy that a human can insert themselves at the top of. Zebras fail on temperament, gazelles panic, and pandas and elephants breed poorly in captivity or grow too slowly. Elephants are tamed, not domesticated: working elephants are usually captured wild.
Only about fourteen large mammal species were ever domesticated, and their wild ancestors were concentrated in Eurasia. Sub-Saharan Africa, Australia and the Americas had almost none: the Americas got the llama and alpaca and the guinea pig, and Australia got nothing. This is not a statement about the people. It is a statement about which animals happened to be there, and it is the strongest part of Jared Diamond's argument in Guns, Germs and Steel.
And the shape of the continents mattered. Eurasia's main axis runs east-west, so a crop domesticated in the Fertile Crescent could spread across thousands of kilometres at roughly the same latitude — same day length, similar seasons, similar diseases. The Americas and Africa run north-south, so spreading a crop meant crossing climate zones and adapting it each time. Maize took thousands of years to move from Mexico to eastern North America.
This is a real and substantial effect, and it is not a complete explanation of anything, which is a distinction the popular version of the argument often loses. Geography sets the starting hand; it does not play it. Chapter 9.4 and Chapter 16.3 return to this at length, and Chapter 6.15 applies it to India specifically.
What farming did to the land
Deforestation begins here. Clearing land for fields is the largest human alteration of the Earth's surface, and it started 10,000 years ago, not in the industrial era. Pollen records across Europe, China and India show forest declining and cereal and weed pollen rising, region by region, as farming spread.
Irrigation begins here too, and with it the first large-scale organised labour and the first environmental failure. Salinisation is the recurring one: irrigation water carries dissolved salts, evaporation leaves them behind, and over centuries the soil becomes too salty to grow anything. Mesopotamian records show wheat, which is salt-sensitive, progressively replaced by more tolerant barley, and eventually by nothing. Southern Iraq contains land that has been agriculturally dead for three thousand years, and Chapter 4.1 tells that story.
Soil erosion begins here (Chapter 1.9), because a ploughed field is bare soil exposed to rain and wind for part of every year.
And methane and carbon dioxide begin to rise, thousands of years before industry — from rice paddies, livestock and land clearing. The palaeoclimatologist William Ruddiman has argued that this early human influence was large enough to delay the next glaciation. The hypothesis is contested and it is taken seriously.
Where this shows up in your life
Your diet. Around 60 percent of the world's calories come from three grasses: rice, wheat and maize. Every one was domesticated by people with no writing, no metal, and no idea what a gene was, and no plant breeder since has added a staple to that list. What modern breeding has done is improve yields of those same species, which Chapter 6.29 covers as the Green Revolution.
Your teeth. Cavities are a farming disease. Starchy staples feed the bacteria that produce acid, and the human mouth was not calibrated for them. So is crowding of the teeth — softer cooked and processed food means less chewing during childhood development and a smaller jaw, in a mouth with the same number of teeth to fit. Wisdom teeth are removed today largely because of a dietary change 10,000 years ago.
Your inequality. The distinction between people who own land and people who work it starts here, and Chapter 3.9 traces it.
And the trap is worth carrying with you as a general shape, because it recurs throughout this volume: a change that makes each individual slightly worse off, but which allows more people to exist, will spread and cannot be reversed. It applies to farming, to intensive irrigation, to fossil fuels, and to several things in Parts 9 and 14.
What the next page covers
The other half of domestication was animals, and it came with an invoice nobody could have anticipated. Chapter 3.8 covers what living beside livestock did to human health — where measles, smallpox, influenza, tuberculosis and most of the great killers actually came from, why they required dense populations to survive at all, why Eurasians ended up with resistance that Native Americans and Pacific Islanders did not, and how that difference decided the outcome of the conquests in Part 8 before a single battle was fought.