I just finished this book by James C. Scott, and it’s amazing. It’s one of those books that’ll unwind your mind and rethread your head. I’m tempted to rewrite the early history of Almea, and you may want to do the same with your worlds. Oh, hey, is it clear that the post title is the book title? It’s called Against the Grain: A Deep History of the Earliest States.


Actually a bad guy

Everything you know is wrong

At some point, maybe in sixth grade, you probably read some histories that suggested, more or less:

  • humanity progressed from foraging, to pastoralism, to agriculture, the last being true civilization
  • each of these steps was an advance in freedom and prosperity
  • nomads and foragers did not understand agriculture, otherwise they would have immediately adopted it
  • agriculture was necessary for large permanent settlements
  • since agriculture developed, the world has been dominated by large agricultural states

All of these statements are wrong. A truer set of statements would be:

  • People prefer foraging or pastoralism, but can be coerced into agriculture
  • Agriculture (and to a lesser extent pastoralism) is a step backwards in freedom and prosperity
  • Nomads and foragers understand how crops work and sometimes plant them, but prefer not to be tied down to a much more tedious and unhealthy lifestyle
  • Large permanent settlements preceded agriculture by a few thousand years
  • For most of history, the bulk of humanity has lived outside the effective control of states

There’s a simple reason the state dominates history, as opposed to humanity: because that’s what generated stone cities and writing that survived. It takes a lot more work to uncover what happened before states appeared, or in areas where the towns were built from perishable materials. Quite a lot of that work has been done in Mesopotamia, which is the focus of the book. (On the other hand, there’s a huge amount that we’ll just never know.)

Our secret weapon: Fire

A nice trendy argument is when the Anthropocene began: the geological era dominated by humans. Was it when we noticed global warming, or when the industrial revolution began? Scott makes a case that it began 400,000 years ago, when hominins mastered fire. Fire greatly changed our diet, and our own bodies and brains, because it allowed us to cook both meat and vegetables, unlocking a great deal more nutrients. Our huge brains are the product of fire: the other great apes can’t support equally sized brains with their diet of raw food. Fire has shortened our guts, which are about a third as long as those of chimps’, because we don’t need as much digestion. We can eat a wider range of things; that, and the warmth of fires, allowed us to greatly expand our habitat.

What’s less realized is that we also used fire to transform the landscape. Sometimes this was accidental; sometimes a purposeful hunting/foraging technique. Fire could be used to chase prey into a killing zone. More subtly, it encourages certain crops which we happen to find useful, and animals that grazed on those crops. Just about every landscape we consider “natural” has already been modified, thousands of years ago, by humanity, largely through fire.

A little fact which underlies the scale of this change: when Europeans reached the New World, their diseases killed off perhaps a majority of the natives– who for centuries had been using fire to clear the forest. The forests sprang back, incidentally absorbing so much CO2 that global climate cooled, from roughly 1500 to 1850.

Next there’s an extended discussion of what happened in lower Mesopotamia and when. The first oddity is that it looks like there were permanent settlements by 6500 BCE, about 1500 years before solid evidence for agrarian villages. The second is that it took another 2000 years before states developed. (Mesopotamia was not the pioneer in sedentism; there was year-long settlement at various sites around 10,000 BCE.)

What sort of area could support sedentism before agriculture? Not the arid desert that much of this region is today… but at that time it wasn’t desert, it was wetlands, if not actually under the sea: almost half of the current land from Baghdad to the Persian Gulf was then part of the gulf.  Ancient Ur was on the coast. The land was about 10 meters below the current level; the difference is due to 10,000 years of sediment from the Tigris and Euphrates.

It was very rich in resources, especially as it was a mixture of coastal and river environments. Frequent changes of the watercourse allowed planting on rich, naturally cleared silt without irrigation. Animals and birds abounded, and migratory gazelles and asses came through, and could be corralled into narrow areas for hunting. If an area is sufficiently rich, you don’t need cultivated fields to support villages.  (Another such area is the pre-Columbian Pacific Northwest.)

Agriculture isn’t an either-or proposition. The same people could hunt, forage, take care of animals. The same plants that were later cultivated grew wild, and foragers had long had the equipment to harvest it and prepare the seeds for cooking. The sort of opportunistic sowing just described (right after a flood) took little extra effort.

This ecological zone survived to modern times, but farther south, and the lifestyle did too, among the Marsh Arabs. Saddam Hussein drained the marshes in the 1990s, displacing half a million people and turning the marshes to desert. The dikes were breached after the US invasion; the marshes are partially restored but few of the people have moved back.

Why plant?

The big question is: if things were so good, why did Mesopotamia move to agriculture at all?

Ester Boserup posited that the change must be out of desperation, e.g. overcrowding, the loss of large game, climate change. For Mesopotamia, there doesn’t seem to be evidence for any of these. Scott can’t suggest anything better, so really we don’t know why the changeover happened.

Of course, once it has happened, it has a certain inertia. You can support a much higher population with agriculture– which means that though individuals can and do retreat from the lifestyle, entire populations can’t.

Co-evolution in the village

There’s a section on what Scott calls “late Neolithic multi-species resettlement camps”… that is, farming villages. The jargon is meant to underline that a bunch of co-evolution was going on, as crops, weeds, domestic animals, uninvited vermin, and people all adapted to living together.

A lot of this was driven by the humans, of course. In general we want crops with increased fruit or grain size, no toxins, no hard cases or spiky protrusions, and which are easy to harvest– e.g., heads that don’t shatter. For animals we want docility, increased fertility, tolerance for cramped conditions and a monotonous diet, and comfort around humans. We also get some unintended consequences: less genetic diversity and robustness; and among the animals, neoteny, reduced sexual dimorphism, and a certain stupidity. (This even affected our vermin: rats and mice who live among us, for instance, are smaller than their wild counterparts.) Many of our crops and domestic animals couldn’t survive without us.

Something that affected all the species was disease. Cramped and unsanitary conditions spread diseases not only within but between species. (Measles comes from sheep or goats; smallpox from camels; influenza from waterfowl.) And epidemics were one of the failure modes of this lifestyle: they could wipe out a settlement, a kingdom, or an army.

More subtly, living in villages affected us too. Evolution did not stop with the Cro-Magnons; we’ve become adapted not only to cooking but to grains and to large quantities of alcohol (historically healthier than the nearby water). In the West, we’re adapted to drinking milk in adulthood. We have some resistance to all those new diseases. Arguably we too are domesticated animals, subject to some of the same changes, including smaller size, duller teeth, neoteny, less sexual dimorphism, and tolerance for crowding and stress.

The bad guy enters

Cue the Imperial March, because now our villain enters: The State. States appear in Mesopotamia around 3100 BC, and everything goes to hell.

In brief: with the state, you get all the drudgery of agriculture, plus coercion and oppression. Someone evidently noticed that if 90% of the people were farmers, a quarter or half their produce could be taken from them, supporting an elite: kings, nobles, priests, soldiers, merchants, craftsmen. (To be precise: if left alone, the people wouldn’t produce this surplus; the state coerces them to produce more than they otherwise would.)

It’s a bad bargain for the farmer… which is why, to the extent of their power, the authorities kept them from leaving. And that’s if they were free to begin with: there was extensive use of slaves, and one of the main purposes of war was not to conquer territory, but to grab captives.

Scott’s particular insight is that states worldwide, up to at least 1800 CE, were based on grain, and that this was no accident. (For the purposes of this discussion that includes rice and maize.) Grain is a tax collector’s dream: it ripens all at the same time, so you can go right in and take a large part of the harvest. (To ensure this uniformity, states often mandated that fields be planted at a particular time.) Grain can be stored for years, and it’s one of the highest nutrient-per-weight foods, so it can be transported long distances.

Can you have a state based on tubers or manioc instead? Not nearly as easily. Tubers don’t have to be harvested all at once; indeed, the best place to store them is in the ground, till they’re needed. If the tax man wants a share, he has to go and dig them up, and if he does, he has a wagonload that’s heavy, easily spoiled, and barely worth transporting.

All the major empires, Scott asserts, are based on grain– and their effective area of control, as opposed to the lines they or we draw on maps, is the limit of grain cultivation. Beyond that are two major populations.

The misfits

One is the non-grain-growers: people who don’t fit, or don’t want to fit, into the tax man’s grain system. Scott has written another book, The Art of Not Being Governed, about the huge region that never quite fit into the East Asian states: southwestern China, northern Vietnam, Thailand, and Burma, and Assam. These are largely mountainous areas where it was hard to grow grain, and where the people grew other things, as well as raising animals, and if necessary melted away when the administrators and armies attempted to enforce control.

The other exception is the pastoralists, as well as mixed groups (like the ancient Germans and the Jurchens) who farmed or herded as circumstances warranted. Though his picture of states is grim, he presents the nomads as far healthier, happier, and more egalitarian.

One reason, it must be said, is that the nomads noticed that the surplus of the peasants could be skimmed off as easily by themselves as by their own elites. So the peasants endured not only the depradations of the taxman, but that of the horselord.

States naturally fought back, but it’s not easy to defeat nomads, who after all have no cities to loot, and can easily melt away into the steppes. But cooperation was often preferable to war. Nearer nomads could be bribed to fight farther ones, or be co-opted as cavalry.

Most of our sources come from states, and we should be skeptical when states claim that non-grain areas or nomads acknowledge their suzerainty. That was a way of saving face; the reality was often that effective control over either was impossible, and huge sums were spent to keep the nomads happy.


The whole structure of states was precarious. States could collapse due to defeat in war, or ecological change, or epidemics, or by peasant revolts, or by the increasing toll of deforestation and salinization. In early Mesopotamia, states were particularly prone to collapse– as Scott puts it, the interregna outnumbered the regna. One historian, Robert Adams, notes that the Third Dynasty of Ur was unusual in lasting a hundred years. Mesopotamia as a whole seems to have collapsed from 1800 BCE to 700 BCE; during this period urban settlements had 1/4 the area they’d had previously. The Greeks famously collapsed around 1100 BCE, losing their cities and literacy for hundreds of years.

Such times are called dark ages, but given the general misery under state control, they may well have been a relatively pleasant breathing space for the people. They were certainly more egalitarian, and cultural output was probably not less; it simply switched from written to oral modes. (The Iliad and the Odyssey are products of Greece’s dark age.)

If you put all this together, and try to look at humanity as a whole before 1500, it may well be that the majority of humans were outside state control, and all the better for it.


If there’s a takeaway for your understanding of history, or for your conworld, I’d suggest something like this:

  • The fluidity of people about foraging/herding/agriculture. It’s not a progression, and the same population, or individuals, might engage in all three.
  • How long it takes between sedentism and states. (I’m sure I didn’t leave enough time in Arcél…)
  • The importance of grain. Think hard about starting a state outside river valleys suitable for grain production.
  • The frequency of collapse in the first millennia.
  • The fact that states are bad news for much of the population.

These are not ironclad rules, especially in fantasy. It’s not that all cities were hellholes. (Just one detail: Chinese cities were probably healthier than European ones, simply because the manure was a valuable substance and removed from the city.)

There were also mitigations Scott doesn’t mention, such as debt jubilees. (See David Graeber… I think that’s the first time I’ve cited him as being more cheerful than another book.)

Some grains of salt

As ever, I have a few cavils. One is that Scott can be annoyingly low on details. You won’t get any explanations of how Sumerian city-states differed from the Assyrian or Babylonian empires. He gives population estimates without explaining where they came from or how reliable they are. He admits that slavery and war pre-existed states, which surely undercuts his major villain, but he provides no way to estimate how much.

More seriously, I’m not sure that his ideas apply so well to Africa, or the Americas, or India.

  • There were kingdoms in Africa, for instance, but so far as I know agriculture never depended mostly on grain, as it did in Egypt or Mesopotamia or China.
  • He mentions the Inka and Maya, but on his own admission maize is not as easy for the tax collector as wheat– it can be left in the field to dry.
  • As for India, at one point Scott says that only two large empires appeared in its history, the Guptas and the Mughals– a statement of colossal ignorance.

I’m inclined to think his ideas apply well enough to temperate areas, but he should have left tropical areas to another study.

Finally, I think he over-paints the picture of the state as tyranny and “barbarism” as pleasant and egalitarian. You could be captured and sold into slavery by nomads, or as a nomad. Or you could be forced to serve in the khan’s wars. And the state/nomad balance didn’t always favor the latter: e.g. Rome was not really bothered by the Germans until the 200s.

And the lot of peasants varied– e.g. it seems to have always been better to be a colonist, when your numbers were few and therefore you had to be treated fairly well. I’m inclined to think it’d be better to be a Chinese peasant in the 1C than the 18C, because game and trees were still available.  For that matter, you’d really want to be born in the beginning rather than the end of a dynasty: taxes were lower, the state was more organized, and bandits were held in check.

(Also, wasn’t 19C Ireland a potato state? Scott doesn’t even mention it.)