Against the Grain is another work by Scott on the relationship between legibility and oppression. The societies described in each of his books are different: Seeing Like A State (1998) focuses on modern, utopian state projects, The Art of Not Being Governed (2009) focuses on Early Modern Southeast Asia, and Against the Grain focuses on the earliest states, especially in Mesopotamia.

Grains – wheat, barley, millet, rice, and corn – are the dominant food in all of these earliest states because they are the easiest to collect taxes on, not because they are the best for the people who farmed them.

To be frank, Against the Grain is not as good as the other books I’ve read by Scott. The reason is simple: Scott is far from his area of expertise. He does not have mētis for this material. His perspective is more limited and he doesn’t have the hundred exemplary anecdotes that make his other books so enjoyable.

I would be remiss if I did not include the best short review of this book: “basically 300 pages calling wheat a fascist”. I do not have the original source of this quote, but saw in the book review by Slate Star Codex.

---

Prerequisites: Familiarity with James C. Scott’s ideas. If you are unfamiliar with them, see my review of Seeing Like A State. Some knowledge of Biblical history would also be helpful.

The key idea in Scott’s work is legibility. Organically grown human societies are enormously complex. There are many local variations, grown from the preferences of many individuals. Although this makes sense to the people who live in the society, it does not to outsiders. Large institutions, especially the state (meaning government – not the US meaning), cannot function in this environment. In order to do anything, they first have to make society legible to a central administrator. So they bulldoze the organically grown society. Many aspects of our lives were created by the state forcibly replacing the organic society with something more amenable to the state – cities planned on a grid, monoculture agriculture, unambiguous property rights, standardized measurement systems, national languages with dictionaries & grammar rules, and even unique unchanging names. But, in order for the state to thrive, the individuals also have to thrive. When administrators push to “rationalize” society too much or too quickly, the results are catastrophic.

Originally Written: June 2021.

Confidence Level: Not my ideas.

---

---

I will begin with a summary of the book, then turn to my own response to it.

The Traditional Narrative

Much of the book is a counterargument against the traditional narrative of the foundations of civilization. We should state what this is first before arguing against it.

Agriculture and permanent settlements (‘sedentism’) were first developed in ancient Mesopotamia. These first settlements were small towns which grew as people moved in. These towns expanded into city-states, then larger and larger empires. Civilization progressed: more and grander architecture was built, trade networks expanded, learning spread, food production increased, and people’s lives became better. Occasionally, invasions of barbarians toppled civilization, causing a Dark Age. Monuments were destroyed, literacy faltered or vanished, and people were forced to live a more primitive lifestyle. Ultimately, over the millennia, civilization has triumphed over barbarism.

Scott attacks almost every point of this argument.

Sedentism, Agriculture, and City-States

The first link to break is the connection between agriculture, sedentism, and the first city-states.

Sedentism can exist without agriculture. Towns can exist in hunter-gatherer^[1]I am using ‘hunter-gatherer’ and ‘forager’ interchangeably. societies in particularly productive locations. Some of them are inhabited only seasonally to take advantage of ripening fruits and seeds or migrations of waterfowl, gazelles, or salmon. Some of them are occupied permanently, especially in wetlands. Wetlands are extremely productive. Regularly fluctuating water levels, either tidally or seasonally, can bring multiple ecosystems to a fixed settlement on a bit of raised land.

The oldest known towns are places like Göbekli Tepe, Turkey and Tell Abu Hureyra, Syria, and date from about 11,500 – 9,000 BC. They were probably inhabited seasonally to harvest wild wheat and barley. The technology for harvesting grains, from sickles to grinding stone, clearly predate domesticated. Much of Mesopotamia at the time was wetland. Since then, the Persian Gulf has retreated as rivers add sediments to their delta and state have systematically drained the wetlands. The last of the wetlands of Iraq were not drained until Saddam Hussein’s genocide of the Swamp Arabs in the 1990s. Early permanent settlements also developed in the wetlands around other major rivers like the Nile, Indus, and Yellow River.

Agriculture can exist without sedentism. Shifting fields (‘swiddening’) or planting crops and leaving them until harvest are common ancient practices and have persisted until today. This is sometimes derogatorily called ‘slash and burn’, although at low enough population densities, it is less environmentally detrimental than fixed field agriculture. You can think of a continuum of altering the landscape to increase food production, from using fire to clear undergrowth (which is even older than Homo sapiens) to intensive agriculture. Many people have also practiced combinations of farming, herding, and foraging.

The earliest evidence of agriculture dates from about 9000 BC, with the major early crops (wheat, barley, lentils, peas, chickpeas, vetch, and flax) and the early livestock (sheep, goats, and cattle) domesticated between 8000 BC and 6000 BC The first farming was probably ‘flood retreat agriculture’. When the flood waters retreated, seeds were thrown by the handful (‘broadcast’) across the new soil. This involves a minimum amount of drudgery and is easy to include with the hunting and gathering of the towns in the Mesopotamian wetlands.

The first towns which relied primarily on domesticated crops and livestock do not appear until about 5000 BC, after a globally cooler period that made Mesopotamia significantly drier. These towns were built on what we now think of as the best agricultural land: flat, dry ground with fertile soil brought in by rivers or (in China) wind. Local irrigation provided water to the fields.

Cities are not just larger towns. Cities have multiple significantly different types of buildings and are usually centered on a monumental palace or temple complex. Towns have little specialization between buildings. This suggests that cities have more social stratification and are led by a king or aristocracy, while towns are more egalitarian. The first writing, developed for administrative purposes, confirms the social hierarchy of these city-states.^[2]In China, writing was first used for divination, but was rapidly adopted to help administer early states. Cities are more likely to have walls than towns. and tend to be larger than towns, but do not have to be. At around 3500 BC, a few towns in Western Ukraine briefly were larger than the cities of Mesopotamia.^[3]I discussed them in my review of The Horse, the Wheel, and Language by David W. Anthony (2007).

Until the invention of the railroad, long distance transport of food (or any other bulk good) over land was impossible. The draft animals and driver have to eat. The farther it travels, the less food will actually make it to the destination. Water transportation is much more efficient. A small crew can move a boatload of grain, especially if you can take advantage of currents and wind. Early states had a small grain growing core and their power could only be extended along waterways.

The first city-states did not appear until shortly before 3000 BC. Sedentism, agriculture, and city-states were not closely linked. They appeared thousands of years apart.

The Life of a Typical Person

The next question Scott addresses is whether the typical person is better off in an agricultural state or as a hunter-gatherer.

Archaeology and history traditionally focused on monumental architecture and written texts, which only show the perspective of the elite. Only recently have there been extensive efforts to understand the lives of the rest of the population.

The elites were clearly better off in civilization. But was everyone else?

The clearest evidence comes from comparing skeletons of contemporary farmers and foragers. The first farmers were about five inches shorter on average and had more bone deformities than their neighboring hunter-gatherers. Foragers had less malnutrition and less repetitive work than farmers. Farming and states caused an increase in drudgery and a decrease in food availability or reliability.

Foraging involves taking advantage of multiple different food sources which vary from season to season and year to year. You might eat fish in the winter, migrating waterfowl and gazelles in the spring and fall, and wild fruit and grains in the summer. This is not done by haphazardly looking for food. For example, the gazelle migration was funneled through narrowing lanes to a small enclosure, where as many gazelles as could be preserved were killed. Because they drew from multiple food sources and even ecosystems, a bad harvest of one thing could be compensated by a good harvest of something else.

Farming involves a lot more work in planting, weeding, and harvesting – and then food preparation. Grains, in particular, must be all planted and harvested at the same time, causing major labor bottlenecks. Afterwards, grain grinding was so laborious that it left its imprint on almost all farming women’s bones in early Mesopotamia. Crops that can be harvested gradually, like legumes (beans, peas, lentils, etc.), root crops (potatoes, yams, cassava, etc.), and many other fruits and vegetables, do not have have this labor bottleneck. Planting only one type of crop (‘monocropping’) makes this problem even worse. Monocropping also creates a massive vulnerability. One bad harvest of one crop will cause your family to go hungry.

Yet grains provided a majority of the food for all ancient civilizations.^[4]I’ll talk about the Andes later. Why?

Let’s take the perspective of the tax (or rent) collector instead of the perspective of the farmer. Once the fields were planted and started to sprout, the tax collector could easily travel around the countryside and assess the expected yields based on each field’s prior performance. To make this assessment with the limited knowledge the tax collector had, the fields had to be fixed (each could be either planted or fallow) and planted with only one type of crop. After the harvest, the tax collector could return (with armed guards) to confiscate some percentage of the grain. If the crop were harvested gradually, the tax collector would have to come back frequently and the crops would be much easier to hide. Fixed fields planted entirely with grains are the most legible to the tax collector and therefore the easiest to appropriate.

All early civilizations were based on grains, not because grains were the best crops for farmers, but because they were the best crops for the state.

States also imposed costs on their subjects directly. Taxes or rents on the farmers were necessary to feed the non-farming elite. This was especially onerous during bad years when the one main crop that everyone grew underperformed. States wage war at a larger scale and with more expansive ambitions. To wage these wars, they conscript farmers to fill all the roles in the army not reserved for the elite. Farmers were also expected to provide corvée labor for the state. Some of this improved the lives of the farmers by building and maintaining irrigation ditches, but a lot of it was dedicated to building the monumental architecture that ancient kings and modern tourists are so fond of. A pyramid is unlikely to provide a net benefit for the people who actually built it.

Civilization also posed a much greater disease risk than foraging. Sedentism, agriculture (especially livestock), concentrated populations, and large trade or war networks each increase the risk of infectious disease. Sedentism puts people in closer contact with human and animal waste. Most human diseases and parasites originated in domesticated animals. Living in close quarters with animals increases the probability that a disease will jump species. A large concentrated population makes it easier for new diseases to spread. Epidemic diseases can only persist in large populations. In a small population, everyone will get infected quickly and either die or recover and be immune. The classic example of this is the measles epidemic of 1846 on the Faroe islands. Prior to civilization, the only persistent human infectious diseases were chronic diseases that took years or decades to kill people. In larger populations, there are enough new people, either through immigration or childbirth, for epidemic diseases to persist. Long distance travel by merchants or armies can bring in diseases from other places and increase the size of the network that a disease can infect. Although the germ theory of disease dates only to the late 1800s, practices like isolating the sick, quarantining travelers before they can enter, and fleeing the city when a disease appears are ancient.

Why Live There?

If early states made life worse for the farmers, why did people live in them?

The simplest answer is that most people did not want to. Although there is only scattered evidence, it seems as though there was a continual flow of people leaving the state for neighboring non-state societies. During crises, this trickle turned into a flood. The grain growing core depopulated as people fled to the hills.

Early states were continually trying to find more people to replenish or increase the population growing grain near their city. They mostly did this using slavery and war.

While slavery predates states, the first states oversaw a massive increase in unfree labor. Since slavery takes different forms, it can be hard to directly compare slavery in different societies. There was often not a significant difference between a slave and a subject who was required to grow grain and provide corvée labor to build monuments. Chattel slavery, when slaves can be bought as sold by individuals, is widely attested. In some societies, slaves are owned collectively.^[5]For example, Spartan peasants were collectively owned by the aristocracy. To read about how this form of slavery was significantly worse than other ancient slavery, see ACOUP. We should also not forget all of the women kept as concubines.

By far the most recorded form of slavery for the earliest states was war captives. Ancient monuments are covered with descriptions and depictions of how many captives were taken in war. Entire populations were seized and relocated to the grain growing core. There, they were the lowest of the low, forced to do the most arduous and dangerous tasks. The purpose of warfare for the first states was to seize population, not to seize land.

The first states were extremely fragile. The presence of cities, monumental architecture, and writing frequently vanished into a ‘dark age’. It was uncommon for any state to survive more than three generations of kings. The might Qing dynasty that first unified China (‘Qin-a’) lasted only 15 years. Only Egypt, which was surrounded by inhospitable deserts that prevented people from escaping, could maintain a state for long. States did not outcompete and expand out of river valleys until much later.

Many things could cause the collapse of a state. Wars, both with other states and with nonstate peoples, can scatter, kill, or enslave the population. Alternatively, they might try to leave the state intact and establish themselves as a new elite. New epidemics, either originating locally or accidentally imported, can kill many people and cause the rest to flee, possibly further spreading the disease. Environmental degradation could also cause collapse. The biggest environmental problem was salinization. Repeated irrigation gradually adds salt to the soil, which eventually turns the field into desert. Upstream deforestation was also common. Increased erosion upstream silted in canals and rivers. Floods became more devastating and could either destroy the city itself or reroute the river, leaving the city cut off from its irrigation and trade.

Figure 3: Historical routes of the Yellow River across northern China. Source.

Scott suggests that, in many cases, the collapse of civilization left people better off than they were before. They were less oppressed by taxes, corvée labor, or military conscription. Organizations and networks became smaller and more sustainable. Slavery and disease became less common. Food sources became more diverse and so more reliable.

The march of civilization looked like progress for the elites, but oppression for most of the population.

---

I now turn from summarizing Against the Grain to responding to it.

The Bible Against Tradition

Whenever someone argues against a ‘traditional narrative’, it is worth asking whether the narrative is actually traditional. I think in this case, it is mostly a fair assessment. The ancient writings we have are almost entirely from the perspective of the elite in the most civilized societies. The most influential text in this tradition is Julius Caesar’s description of Gaul. The word ‘civilization’ itself comes from the Latin word ‘civitas’, meaning city.

There is one ancient text that rejects this narrative and is so influential that it makes the narrative seem not traditional: the Bible.

Scott is aware of this and cites the Bible on several occasions. I don’t think he realizes the extent to which the Bible has challenged the tradition. If you are sufficiently familiar with the Bible, especially the Old Testament, most of Scott’s argument seems unoriginal.

The children of Israel are continually less civilized than their neighbors. They frequently criticize what the traditional narrative sees as progress, although there are a few tendencies in the traditional direction as well.

This starts at the very beginning. Adam is given grain based agriculture as a curse.

In the sweat of thy face shalt thou eat bread.

– Genesis 3:19

Abraham and his family flee from the great city of Ur to live in tents as a pastoralist (Genesis 11:31-12:5). This is a clear example of people choosing to leave civilization to find a better life.

In contrast, Joseph becomes an elite in Pharaoh’s court. He pursues a significant expansion of state power, gathering and storing grain for seven years so they can survive the following seven years of famine.

Then we have the Exodus. A group of people in a major civilization are kept as slaves and forced to build monumental architecture.^[6]Jewish tradition says that they built the pyramids, but the Bible only explicitly says they were making bricks. They chose to leave en masse and pursue a life of mixed agriculture and pastoralism elsewhere. Pharaoh tries to stop them but cannot because of multiple calamities that cripple the state. Most of Scott’s arguments are told and retold every year on Passover.

After the Exodus, the children of Israel live in a stateless society for generations. Although they occasionally acted together under the leadership of a judge, this was a much freer society than any of the surrounding kingdoms.^[7]Although it should be noted that the narrator is not convinced that freedom is a good thing.

Every man did that which was right in his own eyes.

– Judges 21:25

Eventually, the children of Israel ask for a king so they can be like other nations. Samuel the prophet warns them about all the problems a king will cause but eventually relents (1 Samuel 8).

The traditional narrative makes an appearance again with the magnificent reigns of David and Solomon. However, if we read these against the grain, we see that there is significantly more violence under the kings (except maybe Solomon) than the judges. Israel had more civil wars during David’s lifetime than under all of the judges. I think that the most concerning thing in the story of David and Bathsheba is the opening phrase, “at the time kings go forth to battle”, because it implies that war was something that happens every year (2 Samuel 11:1). In contrast, while we may not want to take “and the land had rest for forty years” completely literally, it does imply that times of peace existed (Judges 3:11, 3:30, 5:31, 8:28).

After Solomon, the kingdom devolved into a civil war that lasted hundreds of years.

Both kingdoms were eventually conquered by rapidly expanding empires. The bulk of their population was forcibly relocated to provide manpower for the empire’s core territory. The Bible does not present this as the progress of civilization. These empires are portrayed as the literal embodiment of evil.

The Bible’s counter to the traditional narrative has not only been important for religious scholars. Early Modern political philosophers, include Spinoza and Locke, used these passages to argue against the Divine Right of Kings and to establish the foundation of Liberal Democracy.^[8]See Theologico-Political Treatise by Baruch Spinoza (1670) and Two Treatises of Government by John Locke (1689).

Demography

There is another possible explanation for why grains are preferred. Grains have the highest productivity per land area of any crop. As the population grew, grains became necessary to feed everyone.

Scott addresses this in The Art of Not Being Governed, but not in Against the Grain. Agricultural productivity could be limited by available land or it could be limited by available manpower. For the historical societies we are most familiar with, there is more labor than land, so agricultural productivity is limited by land. For early states, whether in Bronze Age Mesopotamia or in Early Modern Southeast Asia, there was more land than labor, so agricultural productivity was limited by available manpower.

This difference can most clearly be seen in the spoils of war. If there is more land than labor, then societies go to war to seize captives and resettle them in the grain growing core. It is assumed that there is enough land there for them to work. If there is more labor than land, then societies go to war to seize land. It is assumed that the land comes with farmers who will work it. We can use the spoils of war to determine which was the limiting factor. It is much easier to look at monuments celebrating victories is than to determine ancient demographics.

We can now see why Scott does not emphasize the high productivity of grains. Land was not the limiting factor for production. The same population could have survived by spreading out and growing chickpeas.

Legitimacy Revolution

This got me wondering when the transition from labor limited to land limited societies occurred. This is both a demographic change and an ideological change. Empires might go to war for captives like their ancestors did, even though their core grain lands are already full of farmers.

To my surprise, I already knew the answer. Cyrus, king of Persia, allowed the Jews to return to Jerusalem and rebuild the temple. Babylon had taken the Jews captive and moved them to Babylon. It was still ruling people. Persia decided to rule land instead. Each nation could live in their own land and worship their own god/gods as long as they accepted the rule of the Persian King of Kings. This was widely regarded as a good decision and everyone loved him ever since. Except for the Greeks.

The Persians were not the only ones experimenting with establishing the legitimacy of their state on something other than pure violence. In Greece, the first democracies were being established. In China, Confucius was building an ethical system to govern the relationship between rulers and subjects. I will refer to this as the ‘Legitimacy Revolution’. From then on, agricultural states ruled territory instead of people and civilization gradually and intermittently caused less misery to a decreasing fraction of the population.

I’m not sure if Scott would agree with this last claim because, well, he is kind of an anarchist.

Scott and Malthus

An Essay on the Principles of Population by Thomas Malthus (1798) is a landmark book on demography which argued that the growth of the human population inevitably overcomes improvements in farm productivity. The consequences of progress are famines at an ever increasing scale.

Malthus’s predictions were completely wrong. Agricultural productivity grew faster than expected and the human population grew slower than expected. Famines had become increasingly rare and are mostly political rather than economic problems. Overeating is a bigger problem than undereating.

Malthus would have fared better if he had stuck to describing the past instead of predicting the future. For at least the previous 500 years, Europe had been close to its carrying capacity. Agricultural production had increased, but the population had kept up. Everyone had lived through multiple famines. Only terrible plagues increased the incomes of farmers, at least for those who survived. Ironically, Malthus wrote at the very end of what might be called the Mathusian era.

Scott seems to describe a pre-Malthusian era. In order for there to be more land than labor, the population must be well below its carrying capacity. This persisted from thousands of years. How?

Hunter-gatherer societies tend to have much lower fertility rates than farmer societies.^[9]This data is from modern hunter-gatherers, so we are assuming the same would be true historically. Lots of exercise, a more meaty diet, and late weaning meant that women in these societies averaged one child every four years. If this doesn’t work, abortifacients and infanticide were often also considered acceptable. Lower fertility allows hunter-gatherers to keep their populations low.

Women in farming societies average one child every two years. This is definitely fast enough to rapidly reach carrying capacity. So why didn’t they fall in Mathus’s trap?

Scott’s answer seems to be that everything was terrible. War and slavery killed more people and destroyed more families under the first states. The epidemics described above were important, and also the increased infant and maternal mortality rates caused by more crowded and less sanitary living quarters. Higher birth rates were countered by higher death rates.

The population did grow, but slowly. It took 5,000 years for the global population to double during the time agriculture was developed and spread across Eurasia.

Scott does think that faster demographic growth is what allowed states to outcompete nonstate societies. I am not entirely convinced because it’s not clear to me that states were more competitive until after the Legitimacy Revolution. Before then, most states lasted only a few generations and rarely expanded out of the river valleys ideal for grain agriculture. Afterwards, multiple large empires formed that persisted for centuries and controlled much of Eurasia.

Andean Civilization

Scott’s theory of early state formation works well across Eurasia and probably also in Mesoamerica. But it completely fails in the Andes.

Scott seems to be unaware of this. He mentions the Inka a few times, saying that corn was their main tax crops and that, while they did not have writing, they did keep records using knots (quipu). But the Inkas are no more the first state in the Andes than the Romans are the first state in the Mediterranean. The first cities of Norte Chico predate the Inkas by almost 5,000 years and form the third oldest cradle of civilization, after Mesopotamia and Egypt.

The Andes do not have large, productive wetlands for sedentism and agriculture to develop, or large river valleys for intensive agriculture and the first states. Instead, Norte Chico is a stretch of desert along the coast with fast streams dropping steeply from the glaciers at the tops of the Andes. It did not have grains. Corn had not yet been domesticated in Mesoamerica. Instead, most of the food was from the unusually productive sea. Agriculture did exist, but it was focused on less legible crops like potatoes and cotton. Norte Chico not only lacked writing, it lacked any carved stone artwork or even pottery. Norte Chico somehow managed to build cities and monumental architecture without most of the usual trappings of civilization.

Andean civilization breaks most theories of civilization, so it’s not surprising that Scott’s theory doesn’t work here. I would have liked for Scott to have directly addressed the Andes and considered how legibility may or may not have played a role here.

Conclusion

I am fully aware of the irony of reminding Scott about the importance of mētis and cautioning him against making strong conclusions without it. But I think it needs to be done. Against the Grain lacks both frequent exemplars and some of the broader context for his arguments. I was able to see this most easily for the Biblical counter to the traditional narrative and for Andean civilization, but it probably exists elsewhere too. Too fill this gap, Scott should have added a coauthor: either a historian or an archaeologist who has as much mētis as possible about societies so far in the past.

Nevertheless, this book does make some important points. The triumphs of civilization should not be measured entirely by how impressive their monuments are, especially if these monuments were built by slaves. The pursuit of legibility at the expense of individuals’ well being has existed as long as there have been centralized institutions, and it extended into the smallest and most frequent aspects of our lives. The food we eat is mostly grain, not because grains are the most nutritious or easiest to farm, but because they are the most legible.