The Worst Decision We Ever Made

Measure the skeletons on either side of the transition to agriculture and the result is not what the story leads you to expect.

Average height falls, in most regions studied, by several centimetres. Dental caries, almost absent in foragers, become common; enamel hypoplasia, which records episodes of childhood malnutrition, appears in a large fraction of teeth. Markers of repetitive strain show up in spines, knees and shoulders. Epidemic disease (measles, smallpox, influenza, tuberculosis) becomes possible for the first time, because it requires dense populations of humans living next to dense populations of animals, and that is exactly what a farming village is.

Jared Diamond called agriculture the worst mistake in the history of the human race, which is an overstatement written to make a point, and the point holds. By almost every measure of how an individual life went, the first farmers did worse than the foragers they replaced.

So why did it happen, and why did it happen everywhere?

The answer is that selection does not operate on how good a life is. It operates on how many descendants a strategy leaves, and on this measure farming is not close. An acre of wheat feeds perhaps ten to a hundred times as many people as an acre of wild resources in the same place. Foraging women, who carry children while moving, space births several years apart; sedentary farming women, with soft weaning foods available, do not. Farming populations grow, spread, and surround.

The individual can be worse off while the strategy wins. This is one of the most useful ideas history has to offer, and it applies to a great deal more than the Neolithic.

The second surprise is that it happened repeatedly. Agriculture was invented independently at least seven times and probably more like eleven: wheat and barley in southwest Asia, rice and millet in China, maize and squash in Mesoamerica, potatoes and quinoa in the Andes, sorghum and yams in Africa, squash and goosefoot in eastern North America, and taro and banana in New Guinea. Entirely different crops, different techniques, no contact.

Independent invention on that scale points to a common cause, and the best candidate is climate. The Pleistocene was violently unstable: temperature swings of several degrees within decades, carbon dioxide levels too low for efficient plant growth. The Holocene, beginning about 11,700 years ago, is a remarkably steady warm plateau, and every one of these inventions falls inside it. Farming may not have been possible before, whatever anybody knew.

What it was not, in any of these places, is a decision. Consider what domestication actually involves. A wild cereal shatters: when the seeds ripen the ear breaks apart and scatters them, which is how the plant reproduces. Occasionally a mutant appears whose ear stays intact. In the wild that plant is a reproductive failure, its seeds going nowhere. But it is the only kind a person cutting with a sickle carries home, and therefore the only kind that gets sown next year.

Harvesting selects for non-shattering automatically. It also selects for larger seeds, thinner coats and simultaneous ripening, for the same reason. Archaeobotanists can watch the proportion of non-shattering ears rise in the record, and it takes a thousand years or more: far too slow for anyone alive to notice, let alone intend. People were not breeding plants. They were harvesting, and the plants were changing underneath them.

And then there is Göbekli Tepe, which arrived in the 1990s to complicate the sequence everyone had settled on.

The story had been that agriculture produces surplus, surplus produces settlement, settlement produces hierarchy, and hierarchy produces monumental religion. Göbekli Tepe is a complex of massive carved limestone pillars in southeastern Turkey, raised around 9500 BCE by people with no domesticated crops, no domesticated animals and no permanent settlement nearby. The monuments come first.

Klaus Schmidt, who excavated it, suggested the causation might run the other way: that gathering large numbers of people repeatedly to build and to feast created a demand for reliable food in one place, and that the intensive use of wild cereals in the region followed from it. The argument is contested and the dating is not. At minimum, the tidy staircase does not survive.

Which is the last thing worth saying about the Neolithic. It was not a staircase anywhere. The archaeology shows communities cultivating a little and foraging a lot, herding seasonally, settling and dispersing, adopting farming and abandoning it, holding mixed arrangements for thousands of years. Some groups saw agriculture in operation next door and declined. Describing all of this as the moment humanity settled down and began to progress makes the actual evidence look like noise, when the actual evidence is the shape of the thing.