Seeing Like A State is the best political book I have ever read. I have long had tendencies along these lines, but could never put them into words. Scott expresses himself beautifully and cites numerous examples.

The full text of the book is available here.

Here are some other good reviews of this book:

• Slate Star Codex
• Sam[ ]zdat
• Ribbonfarm
• Brad DeLong
• Crooked Timber

---

Prerequisites: None.

Originally Written: Spring 2021.

Confidence Level: Core political belief.

---

---

Personal Preface

Once upon a time, I was in middle school and was learning about the Industrial Revolution. We were each given a large sheet of paper and were shown some symbols we could use for different types of buildings. We were to draw a small town, which so many of each type of building. And be sure to draw small. The teacher then told us that our town had grown and needed to add these new buildings … again … and again … and again. Soon, everyone’s paper was a cluttered mess. The lesson was that rapid unplanned growth leads to problems.

Our homework was to take the same sized sheet of paper and the same final list of buildings and to plan a city. Every student came back with their streets in a grid and their buildings separated by type … except me. I first drew a river curving across the paper, then arranged the buildings, mixed together, across some imagined hills.^[1]Unfortunately, I don’t still have this assignment, or I have included my drawing here. I thought that looked like a pleasant place to live.

In hindsight, this project seems very ideological: the public school system teaching children to see like a state. But even then, although I did not know what I was doing, I refused.

---

Introduction

Most political texts have stronger conclusions than arguments. From a few simple assumptions, they claim to derive everything you need to know about how a society should be organized. Seeing Like A State is the opposite. It has much weaker conclusions then arguments. The main conclusion is: When a powerful centralized institution wants to radically remake society, it should check with local knowledge first. Along the way to this conclusion, Scott develops an entirely new way of viewing the political world.

This new perspective really doesn’t fit on any traditional political spectrum. Scott simultaneously has conservative, socialist, and anarchist tendencies. And yet, his ideas clearly form a coherent whole that can easily be recognized when other writers reference them.

The key idea in Scott’s work is legibility. How easily can a social system be understood by a centralized institution?

The social system could consist of trees in a forest, crops in a field, bees in a beehive, farmers in a village, residents in a city, or citizens in a country. Any type of interaction within any of these groups might be legible or illegible. The centralized institution is usually the state, although large corporations pursue similar goals.

In order for a centralized institution to do anything, it first needs legibility. States repeatedly reach into new aspects of people’s lives, make them legible, and then use this new power to achieve the state’s goals.

The growth in state capacity often combines with a blind trust in Order / Science / Rationality to create an ideology. Scott calls it High Modernism. I am not particularly fond of this name because, while the ideology is most prevalent in the Modern Era, it has existed since the first states. I will instead call the ideology Legibilism.

Combining growing state capacity and Legibilism with authoritarianism and a weak civil society is a recipe for disaster, even if the government has good intentions. The great utopian social engineering schemes of the twentieth century are the clearest examples.

Legibilism itself is neither inherently good or bad. It may or may not be aligned with what actually makes life good. Optimizing for legibility, which is necessary before a state can optimize for anything else, can be harmful. Scott picks examples where legibility is particularly poorly aligned with actual good things.

Seeing Like A State contains many wonderful examples. I classify Scott’s examples into three categories: earlier developments of state capacity, modern Legibilist control over nature, and modern Legibilist control over people.

---

Earlier Developments of State Capacity

These first few examples don’t follow all of Scott’s pattern. Instead, they provide background for the expansion of legibility and how deeply it infiltrates our lives.

Some of the categories that we most take for granted and with which we now routinely apprehend the social world had their origin in state projects of standardization and legibility.

[p. 64]

Names

Traditional naming schemes, and modern nicknames, have tremendous variety. Names might change based on age, relationship, or any number of factors. The Christianization of Europe ensured that every person had one permanent name. If some circumstance required further specification, a temporary qualifier could be added: the person’s occupation, residence, father, or a personal characteristic. For example, one the leaders of the Münster Anabaptists was named Jan van Leyden or John of Leiden or Jan Beukelsz or Jan Beukelszoon or John Bockold or John Bockelson.

The state would like every person to have a unique, permanent identifier. This makes any interactions it has with its subjects much easier – especially the collecting of taxes and military drafts.

So the state conducts a census. The first censuses involved more assigning names than recording them. The earliest census was in China, in the 300s BC, which also permanently established patriarchal family structures. The earliest attempts at census taking in Europe began in the 1400s, although it took several centuries to be successful. Civil society resisted what they accurately saw as an expansion of state power.

Colonial governments were more willing to override popular resistance and impose permanent surnames on the population. The Philippines acquired surnames in 1849. The governor created a catalog of acceptable Spanish surnames and distributed a copy to each region. Each town would take a few pages and register all their families with those surnames. Official documents were only valid with the proper name and children were required to exclusively use their Spanish name in school. Since the catalog was alphabetized, as you travel around an island, the surnames progress through the alphabet. This naming pattern can still be seen on some of the islands of the Philippines.

Systems of Measurement

We use standardized units for measurement. For most of the world, this means meters (and companions). Even in the US, the foot is unambiguously defined (as $30.48$ cm). Pre-modern societies had “local, interested, contextual, historically specific” measurements [p. 27]. A distance might be a stone’s throw or a rice-cooking time.

Not only are these units extremely diverse, they are also being continually renegotiated and reflect local power dynamics. Conflicts over the size of a bushel, and thus how much was owed in feudal dues, sometimes led to the town’s bushel being literally carved into the stone church.

Modern measurement systems eliminate this conflict:

The centuries old dream of the masses of only one just measure has come true: The Revolution has given the people the meter.

– French Revolutionary decree [p. 32]

Ever since, people have been expected to measure everything using this standard.

Land Ownership

Traditional land use practices are extremely illegible. Who is allowed to do what on what land varies from village to village, from season to season, depending on what kind of land it is, and with many other factors. You often could say that it was managed communally, although even that suggests too much systematization. These tend to be unwritten norms and are continually renegotiated. For a particular example, in many Russian villages, each family had the right farm a few rows in every field. This gave them access to all of the village’s microclimates and spread their risks.

Illegibility makes traditional land use hard to tax. Tax collectors want to estimate how productive each piece of land is and who is responsible for paying taxes on it. So they create a cadastral map, which shows precise property lines and who owns each plot. Unambiguous private property was created by the state to collect taxes.

Cadastral maps have the force of law and can override traditional land rights. The map itself has the power to transform society.

People who cooperate with and manipulate the development of cadastral maps can make themselves owners of land they had little right to before. While ‘the tragedy of the commons’ is a classic example in economics, what actually happened to the English commons is that lords declared themselves the owners and forced all of the peasants who had ‘trespassed’ there for generations to leave. While traditional land use certainly could be oppressive, the transition to private property was often at least as corrupt.

• 

• 

National Languages

You can probably apply the pattern now without guidance.

Traditional language consists of numerous local dialects. The state creates dictionaries, grammar books, and style guides. All official documents must be in the official language.

This massively devalues local knowledge in favor of centralized knowledge. Instead of state representatives needing local guides to navigate, locals now require guides to interact with the increasingly powerful state.

Standardization vs Simplification

You might be tempted to call these examples ‘state simplification of society’ instead of ‘state standardization of society’. These are similar, but not identical ideas. Standardization makes things easier for a centralized institution to understand. Simplification makes things easier for a random individual to understand. Standardization and simplification are often associated, but not always. Creolization of language tends to simplify, but not standardize. Handbooks tend to standardize, but not simplify.

Standardization involves using the same categories and measurements for everything, regardless of how relevant these categories are for a particular situation. Large institutions have “a ‘strong incentive to prefer precise and standardizable measures to highly accurate ones’, since accuracy is meaningless if the identical procedure cannot reliably be performed elsewhere” [p. 81]. To achieve accuracy, people are forced to fit the categories instead of adapting the categories to fit the people.

The contrast between standardization and simplification shows us the difference between legibility and transparency. Transparency is simple. A typical person can easily get an accurate view of a large institution. Legibility is standardized. A large institution can easily get a precise view of each individual.

Many modern reforms increase both legibility and transparency. If the reform is more transparent than legible, it can increase the power of civil society relative to the state. If the reform is more legible than transparent, it will increase the power of the state relative to civil society.

---

Modern Legibilist Control over Nature

‘Scientific’ Forestry

Seeing Like A State opens with the example of ‘scientific’ forestry. I think that Scott choose this example both because it clearly demonstrates his ideas and because most people don’t already have strong opinions on it. If his first example were city planning, people would respond with a story about how they got lost one time – and miss the broader pattern.

Here is how scientific forestry began in Prussia and Saxony:

• The state owns some forests and expects them to produce lumber. How much lumber can each forest produce?
• State officials survey the forest, counting the number of trees of each species and size.
• Subsequent studies determine that the Norway spruce is the most productive tree species.
• Whenever a forest is cut, instead of allowing it to grow back naturally, foresters plant exclusively Norway spruce.
• State officials become enamored with the image of trees growing in straight rows.
• More and more states adopt these ‘scientifically proven’ practices for their own forests.

In the short to medium term, this is wildly successful. Forests produce higher and more predictable yields of standardized lumber.

But problems are developing:

• While the Norway spruce is a hardy plant with few vulnerabilities, every tree in the forest is vulnerable in the same way. Damage from weather or pests occurs less frequently, but more catastrophically.
• In a natural forest, there is a complex web of numerous species that maintain the soil. Only a few of these species can survive when there is exclusively Norway spruce, so the soil degrades. You are effectively mining the nutrients of the previous forest without replenishing them. By the second or third generation, ‘scientific’ forests produce less lumber than natural forests.
• Peasants had many uses for the many species in a natural forest independent of their wood production. These uses are excluded from a ‘scientific’ forest.

Once we realize that the new forests produce less than the old, we can just go back, right?

Well, we could. But state foresters have grown to love their ‘rational’ ‘scientific’ forests. They continue to practice and export ‘scientific’ forestry long after it stops being viable.

This example shows a general pattern:

• We begin with a complex web of interactions: either an ecosystem or a society.
• A large centralized institution has a goal for this ecosystem or society.
• It first has to radically simplify things to make it legible for a central planner. This simplification ignores or eliminates everything not relevant to the goal.
• This undermines plant / human flourishing.
• The aesthetics of legibility, almost always realized as straight lines and grids, becomes a goal itself.

This process can occur regardless of whether the goal itself is good or bad.

Monoculture Farming

Forestry is a particular example of a much older and more important practice: agriculture. Tropical farming and farmers are Scott’s particular area of expertise.

Traditional Agriculture

All agriculture is a simplification of nature. Natural flora is conceptually divided into ‘crops’, which are useful for humanity, and ‘weeds’, which are not. An area of the landscape becomes a field, which has been heavily modified to promote crops and discourage or remove weeds. The crops are selectively bred, both to make them more useful for humans and to make them better suited for a field. Most crops “have been so adapted to their altered landscape that they have become ‘biological monsters’ which could not exist in the wild” [p. 265].

Nevertheless, traditional agriculture maintains a remarkable diversity. There are thousands of landraces: local varieties developed locally, adapted to the local environment and culture. Each individual farmer would be familiar with multiple landraces, allowing him greater flexibility for his farm. Diversity within a field is further increased through polycropping, especially in the tropics. Multiple crops are grown in a single field. In eastern North America, the most famous traditional combination is corn, beans, and squash. In the tropics, even more species are sometimes grown together. Diversity between farms was maintained by having many small cultivators who had different land, family/labor structure, and goals.

Legible Agriculture

New farming practices developed in the early to mid 1900s dramatically reduced this diversity. “Modern, industrial, scientific farming, which is characterized by monocropping, mechanization, hybrids, the use of fertilizers and pesticides, and capital intensiveness, has brought a level of standardization into agriculture that is without historical precedence” [p. 266]. Crops are grown in a large field of identical plants in evenly spaced rows. Machinery is used whenever possible. Crops like the ‘supermarket tomato’ were bred to be compatible with machinery, not for their taste. Most corn grown today are hybrids which have high performance but don’t breed true, so farmers have to buy new seeds every year. Chemical fertilizers and pesticides are used in abundance. In extreme cases, the entire farm would be planned on a different continent using a few standardized measurements of the average climate and soil type. Legilibism has triumphed over the diversity of traditional practices.

So does legible farming works better than traditional farming?

Scott says no. Legible farming usually claims that it is scientifically proven to have higher yields, but this is not true in the tropics. Unintended environmental consequences make it less sustainable in the long term. The main benefit of legible farming is making state control and exportation easier.

Let’s go through some of the problems that legible farming causes.

Plant Pandemics

Large fields of identical crops are extremely vulnerable to pandemics. Both the genetic variability of landraces and the presence of other crops in the field hinder the spread of fungi, insects, or other pests from plant to plant. Scott lists several examples, including the corn blight of 1970. It was cured by adding another landrace to the hybrid. Modern hybrids are not only the descendants of landraces. They continue to depend on them, even as they displace them.

Somehow, Scott failed to mention the Bananapocalypse. Commercial bananas do not have seeds and are reproduced through cuttings. If you live in the developed world, every banana you eat is genetically identical: the Cavendish banana. But the Cavendish banana was not the first choice for imports. Gros Michel (Big Mike) was slightly larger and had a significantly stronger flavor. During the 1920s-1950s, a fungal wilt called the Panama disease destroyed the Gros Michel plantations. Only an echo remains in our diet, through banana-flavored candy. The Cavendish banana was adopted primarily because it resisted this disease. A new variety of the blight that overcomes this resistance was discovered in Taiwan in 1989. It has since spread across Southeast Asia to Australia by 2015 and to Colombia in 2019. Another bananapocalypse is underway and we will have to find new bananas to replace it.

Plant pandemics are not unique to modern agriculture. Agriculture was not industrialized when the blight caused the Irish potato famine of 1850 or as the boll weevil devastated the cotton fields of the southern US from 1896 to the 1920s. Ireland did only have a few breeds of potato and cotton was grown in large fields on plantations, so Scott’s ideas partially apply here. An even better counterexample is the chestnut blight, which entered the US from East Asia in 1904. Over the next 40 years, it eliminated the most common and productive large tree in the mostly unmanaged forests of the eastern US. Even though plant pandemics exist outside of modern agriculture, Legibilism makes plant pandemics worse.

Pesticides and Fertilizers

The modern proscription for plant pandemics is pesticides. This works for a while, until the pests develop resistance. Pesticides can have broader environmental impacts, as was famously chronicled in Rachel Carson’s Silent Spring.

A similar attitude exists towards fertility. If the soil is barren, add chemical fertilizers until the crops grow. This neglects aspects of fertility such as the soil structure and the organisms that live in the soil. Scott does not mention the major environmental consequence here: the Dead Zone that forms in the Gulf Mexico each summer. Fertilizer coming down the Mississippi River causes massive algal blooms. As the algae die and decompose, it deprives the water underneath of oxygen. Fish leave and life on the bottom dies in an area of thousands of square miles. Similar dead zones exist elsewhere, where rivers wash fertilizer into the sea.

Thin Tropical Soil

The problems of legible agriculture are even more extensive in the tropics, which usually have thin soils and lots of rain. If an entire field is harvested at once, the soil will be washed away by the next rain. Polycropping that mixes multiple annual crops harvested at different times with perennial crops ensures that the ground always remains covered. Shifting ‘slash and burn’ agriculture which leaves the largest trees is often more sustainable than fixed field farming because it gives the soil time to recover – as long as the population density is not too high. Not only does traditional agriculture have fewer environmental consequences, it also can maintain higher yields.

Farmers’ Values

Even when monocropping does give a higher yield on average, a farmer might prefer more diversity. By planting both drought resistant and rain-loving crops, he can have a more reliable harvest. Would you rather use a strategy that produces 1000 days of food 90% of the time and 100 days of food 10% of the time or a strategy that produces 500 days of food 90% of the time and 400 days of food 10% of the time? If you are selling for the highest average profit, the first is better. If you are feeding your family and the food won’t store for years, the second is better.

And then there are the personal reasons of the individual farmers. They might value the self-reliance that comes from producing all of the different food that you eat. They might try to avoid labor bottlenecks by not planting and harvesting all of their fields at once. Particular varieties might be especially well suited for local dishes. Some crops might have special ritual significance. A farmer who doesn’t follow the best practices for getting the highest yield might not be ignorant – he might just have a different value system.

Legibilist Response

Why do agricultural experts reject local practices and values? Part of it is aesthetic. They know what a well run farm looks like and it’s not this. A monocropped field planted and harvested with machinery looks like the experimental plots designed to test one variable at a time. The experts can make average measurements of the soil and climate and make recommendations for the appropriate planting and harvesting times or fertilizer and pesticide use. Whether or not this advice is followed can then be monitored. Control is also important. The crops produced are easy to tax, redistribute, or sell to a large corporation. Ease of management and product accessibility are more important to the Legibilist than the yield.

Mētis

Scott introduces a new term to describe the knowledge of traditional farmers: mētis. “Broadly speaking, mētis represents a wide array of practical skills and acquired intelligence in responding to a constantly changing natural and human environment.” [p. 313] This is exactly the kind of knowledge which resists being made legible.

Why don’t we just use the term ‘traditional knowledge’? Because practical new technology is quickly adopted by ‘traditional peoples’, as long as it is on their terms. As an extreme example, farmers in West Africa who had only started growing maize a few generations earlier already had “elaborate rituals and myths about a maize goddess or spirit who had given them the first kernals.” [p. 331] Oral traditions are especially flexible in adopting new ideas. The term ‘traditional knowledge’ suggests recording and preserve what already exists. This is not what mētis is.

Mētis can only be learned through experience. Even if you have read all of the relevant guidebooks and watched all of the relevant Youtube videos, you still don’t completely known something until you’ve tried it yourself. This is true for carving sculptures – or for working math problems.

Mētis is local. Mētis developed over a career of fixing American cars does not translate into mētis for fixing Japanese cars. Farmers from the eastern US took their mētis (and deep plowing) to the Great Plains and created a dust bowl. You don’t just have mētis – you have to have mētis in something. This is different from the original Greek, where Odysseus was said to have mētis in general.

Mētis operates best in an environment with continual small changes. Each time is a little different, but close enough to still be familiar. If there are no changes, everything could be written down and codified. If the changes are extreme and unprecedented, then prior experience is no guide. Mētis is remembered through stories and rules of thumb. Knowing which one is currently relevant is an integral part of the knowledge.

Mētis is interested. The people who develop it have skin in the game. I don’t think this is as important as Scott does. You can develop mētis for video games, even if they have little impact on your life. I do not have this mētis because I did not play video games growing up, so when I do try to play, I do not know how to walk. Video games do have fast feedback mechanisms that make it easy to know if you are succeeding. Since mētis is usually important to people’s livelihood and unevenly distributed, it can lead to some nasty local power dynamics. But we should oppose the restrictions of the guilds without destroying the knowledge that led to their wealth.

I have been using my own examples here. Scott uses different ones, like riding a bike and piloting a ship into a particular harbor and team sports and many more. I will highly two of his from traditional agriculture:

• Squanto’s maxim: “plant corn when the oak leaves were the size of a squirrel’s ear” [p. 311]. Spring comes early some years and late others, so you should use plant growth to determine when it is time for plants to grow. You should be familiar with the entire sequence of new growth in spring so in case it is a weird year and the oaks are early, you can be more cautious. This maxim has the unusual property that it is more general than the corresponding standardized knowledge: a farmer’s almanac. I also like it because you can so easily imagine the fable of Corn, Oak, and Squirrel that could go with this maxim.
• This one is my favorite passage from Seeing Like A State, so I will quote it in full:

While doing fieldwork in a small village in Malaysia, I was constantly struck by the breadth of my neighbors’ skills and their casual knowledge of local ecology. One particular anecdote is representative. Growing in the compound of the house in which I lived was a locally famous mango tree. Relatives and acquaintances would visit when the fruit was ripe in the hope of being given a few fruits and, more important, the chance to save and plant the seeds next to their own house. Shortly before my arrival, however, the tree had become infested with large red ants, which destroyed most of the fruit before it could ripen. It seemed nothing could be done short of bagging each fruit. Several times I noticed the elderly head of household, Mat Isa, bringing dried nipah palm fronds to the base of the mango tree and checking them. When I finally got around to asking what he was up to, he explained it to me, albeit reluctantly, as for him this was pretty humdrum stuff compared to our usual gossip. He knew that small black ants, which had a number of colonies at the rear of the compound, were the enemies of large red ants. He also knew that the thin, lancelike leaves of the nipah palm curled into long, tight tubes when they fell from the tree and died. (In fact, the local people used the tubes to roll their cigarettes.) Such tubes would also, he knew, be ideal places for the queens of the black ant colonies to lay their eggs. Over several weeks he placed dried nipah fronds in strategic places until he had masses of black-ant eggs beginning to hatch. He then placed the egg-infested fronds against the mango tree and observed the ensuing week-long Armageddon. Several neighbors, many of them skeptical, and their children followed the fortunes of the ant war closely. Although smaller by half or more, the black ants finally had the weight of numbers to prevail against the red ants and gain possession of the ground at the base of the mango tree. As the black ants were not interested in the mango leaves or fruits while the fruits were still on the tree, the crop was saved.

[p. 333]

If you don’t know how to manage ant vs ant genocide, how could you consider yourself an expert in farming?

---

Modern Legibilist Control over People

The natural world is not the only place to be standardized. Human societies are also messy and irrational, from the perspective of a centralized institution.

Urban Planning

Unplanned cities look a lot like natural forests. There is a complicated array of forms, functions, and interactions, but buildings and people have replaced plants. Organically grown cities have “streets, lanes, and passages [that] intersect at varying angles with a density that resembles the intricate complexity of some organic processes” [p. 53]. Buildings of all different shapes and purposes are built next to each other.

The city looks disordered from above or to a stranger. The locals navigate it with ease. The complex network of streets follow the most common daily movements they and their ancestors have followed for generations.

Scott uses Bruges as his example of an illegible city. But there is no city less legible than Venice. Venice is located in a shallow lagoon, with only a few secret navigable channels to the city. Within the city, there are two mutually incompatible methods of transportation: alleys and canals. This did not make Venice less successful as a city. She was one of the dominant powers in the Mediterranean for about a thousand years. Napoleon was the first person to conquer Venice. When the Austrian army took over after the defeat of Napoleon, they insisted on two new things: all islands must be connected by bridges and all the streets must have names, which they painted on the walls of every intersection. Even today, maps and directions from locals are better than maps produced by outsiders.

In contrast, planned cities have grids. This is not just a modern phenomenon. Chang-an (now Xian), the medieval capital of China, and even Alexandria in Egypt, were built on a grid.

• 

• 

Unsurprisingly, advocates of ‘modern’, ‘rational’, ‘scientific’ values preferred planned cities. Descartes made this argument:

These ancient cities that were once mere straggling villages and have become in the course of time great cities are commonly quite poorly laid out compared to those well-ordered towns that an engineer lays out on a vacant plane as it suits his fancy. And although, upon considering one-by-one the buildings in the former class of towns, one finds as much art or more than one finds in the latter class of towns, still, upon seeing how the buildings are arranged – here a large one, there a small one – and how they make the streets crooked and uneven, one will say that it is chance more than the will of some men using their reason that has arranged them this.

– Discourse on Method [p. 55 in State]

The criticism here is that the city looks random. This is portrayed as an argument from reason, but it is really an aesthetic argument. I find this especially ironic because Descartes is aware that locally, the disorganized city is at least as beautiful as the organized city. The aesthetic he promotes is only viable from a distance or from above.

Sometimes, the state redesigns a city to make it more legible. The most notable example is Haussmann’s redesign of Paris from 1853 to 1869 under Napoleon III. The goal of this redesign was counterrevolutionary. Paris had revolted nine times since the end of Napoleon I’s reign in 1815, usually unsuccessfully. Haussmann carved broad boulevards through the poorest, least legible, and most revolutionary districts to provide easy access to all of Paris for the police in times of peace and for the army in times of insurrection. Imposing facades along the boulevards covered, but did not replace, the tangle of medieval buildings. The legible public works projects of Haussmann also allowed for a cleaner, healthier city. However, the people in the paths of construction were displaced to new suburbs that were as dense, poor, unhealthy, and revolutionary as the city had been before.

The sharpest critic of Legibilist Paris I have seen is Victor Hugo, in Les Miserables (1862):

Everywhere parallel rows of trees, buildings erected on a line, flat constructions, long, cold rows, and the melancholy sadness of right angles. Not an unevenness of the ground, not a caprice in the architecture, not a fold. The ensemble was glacial, regular, hideous. Nothing oppresses the heart like symmetry. It is because symmetry is ennui, and ennui is at the very foundation of grief. Despair yawns. Something more terrible than a hell where one suffers may be imagined, and that is a hell where one is bored. If such a hell existed, that bit of the Boulevard de l’Hopital might have formed the entrance to it.

I don’t know if Scott has read this critique, but I’m sure that he would agree.

Legibilist plans for Paris became even more extreme in the 1900s with the work of Le Corbusier. He wanted to completely rebuild the city and fill it with rectangular buildings and even wider streets, all arranged on a grid. The city should have clearly separated residential and business districts, each of which is aggressively optimized for a single function. A central monumental plaza is also important. When Paris refused to be completely rebuilt with all of its unique urban history, tradition, and aesthetic tastes erased, Le Corbusier took his plans and rebuilt Chandigarh, India instead. Local conditions should not influence the Plan.

Legibilist cities rarely survive on their own. Few people enjoy living in them. Usually, less legible suburbs grow around the city, where less planned, mixed use suburbs allow people to create a local culture for their community.

We can make the comparison to legibilist farming. Complicated, mixed use neighborhoods are replaced by grids and single use zoning, much like how irregular polycropped fields are replaced by monocropped fields with straight rows. The legibilist core depends on its illegible suburbs for its vitality, while the legibilist farms depend on the diversity of landraces. The legibilist cities and farms are parasitic on the illegible ones.

Is there mētis for city planning? There are many communities with long traditions of managing the complexities of farming. Their survival depends on their continual success, so we can trust that their knowledge is valuable. There aren’t communities that regularly design cities – and live in the cities that they design. Since there is no tradition, with strong feedback mechanisms, I don’t think that anyone today has mētis for city planning.

Scott suggests that Jane Jacobs is closer to city planning mētis than anyone else. Her emphasis is on the lived experience of a city, not on its abstract form. A successful city does not have to conform to a rational plan; it does have to improve the lives of its inhabitants.

I think an even better example is the Hanseatic League, an organization of German towns that monopolized trade in the North and Baltic Seas during the Late Middle Ages. At its height, it included hundreds of towns. We don’t know exactly how many towns were involved because the Hanse never made a list. There are two ways that towns joined the League: either an existing town sought the support of other towns in a conflict against their lord or citizens (burghers) from an existing town founded a new town. The diversity and regular experimentation of the Hanse should have created mētis for planning towns and legal systems (Magdeburg Law and Lübeck Law) in medieval Germany.

One reason that I am excited about charter cities is that they provide an opportunity to create mētis for city planning today – if the developed actually live in the cities they design. We are still far away from this now. But if we got to the point where there are dozens of charter cities and we’ve learned what does and does not work in particular circumstances, then the community would have gained mētis.

The Revolution

Scott began as a leftist, before moving off the political spectrum in an entirely new direction. To honor his past, Scott includes a chapter on better and worse ways to have a communist revolution.

Unsurprisingly, Scott despises the centralized, state-led revolution of Lenin. He prefers Rosa Luxemburg and Aleksandra Kollontay, who thought of revolution as a living process where workers continually recreate their own local conditions.

Communism in practice has always followed Lenin in massively expanding the power of the state.

Ujamaa Villages in Tanzania

We now arrive at Scott’s biggest example, the ujamaa^[2]The word means ‘extended family’ in Swahili. village campaign in Tanzania under the leadership of President Nyerere from 1973 to 1976. This example might seem arbitrary (unless you’re from East Africa), so we should justify why this gets to be the centerpiece of the work:

• Scale. At least 5 million Tanzanians were relocated to new villages. Nyerere claimed that “70% of our people have moved their homes in the space of about three years” [p. 245]. Similar programs existed elsewhere in Africa, but were not as large.
• Dramatic Changes. Tanzanian society extremely illegible before the reforms and extremely legible after the reforms. Such dramatic changes should have easily observable effects.
• Documentation. Tanzania was (relatively) politically open, so there is thorough publicly, available data. Many consequences of the reforms are known.
• Good intentions. We could also consider the Soviet collectivization of agriculture (which Scott does) or apartheid South Africa’s resettlement schemes, but those cases were more malicious. It is not clear the extent to which the tragedy was the goal. Tanzania had an idealist, post-colonial government that sincerely desired to improve their new country.

Prior to 1973, Tanzania was extremely illegible. The country was mostly rural and rural life “included shifting cultivation and pastoralism; polycropping; living well off the main roads; kinship and lineage authority; small, scattered settlements with houses built higgledy-piggledy; and production that was dispersed and opaque to the state” [p. 238]. There was attempts at plantation agriculture under the colonial regime and at villagization during Nyerere’s first decade in power, but they were largely unsuccessful and only impacted a small fraction of the population.

Nyerere wanted rural life to be organized into villages. He explained why as early as 1962:

Unless we do we shall not be able to provide ourselves with the things we need to develop our land and to raise our standard of living. We shall not be able to use tractors; we shall not be able to provide schools for our children; we shall not be able to build hospitals, or have clean drinking water; it will be quite impossible to start small village industries, and instead we shall have to go on depending on the towns for all our requirements; and if we had a plentiful supply of electric power we should never be able to connect it up to each isolated homestead.

[p. 230]

Nyerere wanted people to live in villages so the government could provide development for the people.

Nyerere’s plan to develop Tanzania involved dramatically increasing legibility at (at least) two levels. The complex interactions among plants in traditional agriculture would be standardized, as well as the complex interactions among people in traditional rural life.

Operation Planned Villages began in December 1973. Since voluntary villagization had not proved popular, they turned to compulsion. The militia and army were mobilized to transport people to the newly built villages. If people did not cooperate, their houses would be pulled down or burned. Famine relief would only be provided to those in villages. The few instances of earlier voluntary villagization, like the Ruvuma Development Association, were subsumed into the government plans or disbanded.

In the villages, people were expected to “grow cash crops (as specified by the agricultural experts) on communal fields with state-supplied machinery” [p. 239]. Each household also had an individual plot that they had more freedom with. The communal fields massively underperformed expectations. Farming productivity fell, so food imports were needed the next few years. The amount spent on food imports during those three years was enough to have bought a cow for every family.

The people were forced to live in houses along straight roads. The speed of the operation ripped the peasants from their traditional social networks so they (hopefully) would become modern producers guided by experts. The agricultural field officer also designed a labor plan for the communal fields, despite not knowing anything about the population other than the number of households. The Tanzanian peasantry responded with “flight, unofficial production and trade, smuggling, and foot-dragging” [247]. Scott does not go into detail here because he has written previous books on peasant resistance.

The failure of ujamaa villages was almost guaranteed by the [Legibilist] hubris of planners and specialists who believed that they alone knew how to organize a more satisfactory, rational, and productive life for their citizens. It should be noted that they did have something to contribute to what could have been a more fruitful development of the Tanzanian countryside. But their insistence that they had a monopoly on useful knowledge and that they impose this knowledge set the stage for disaster.

[p. 247]

After 1976, villagization efforts subsided. Most of the people who could be moved to villages had been. Villagization remained the policy of the Tanzanian government, and was supported by international development agencies like the World Bank and USAID, through the end of Nyerere’s reign in 1985.

---

Conclusion

After so many examples of the failures of Legibilism, you would expect a strong condemnation of the ideology. But Scott is not so certain:

Aided by hindsight as it is, this unsympathetic account of [Legibilist] audacity is, in one important respect, grossly unfair. If we put the development of [Legibilist] beliefs in their historical context, if we ask who the enemies of [Legibilism] actually were, a far more sympathetic picture emerges. Doctors and public-health engineers who did possess new knowledge that could save millions of lives were often thwarted by popular prejudices and entrenched political interests. Urban planners who could in fact redesign urban housing to be cheaper, more healthful, and more convenient were blocked by real estate interests and existing tastes. Inventors and engineers who had devised revolutionary new modes of power and transportation faced opposition from industrialists and laborers whose profits and jobs the new technology would almost certainly displace.

For nineteenth-century [Legibilists], the scientific domination of nature (including human nature) was emancipatory. … First, virtually every [Legibilist] intervention was undertaken in the name of and with the support of citizens seeking help and protection, and, second, we are all beneficiaries, in countless ways, of these various [Legibilist] schemes.

[p. 96-97]

This is why Seeing Like A State has such weak conclusions. As much as Scott doesn’t like Legibilism personally, he remains aware of the good that it has accomplished. The result feels almost as if Marx had developed the theory of the oppression of the proletariat, and then stopped to acknowledge the successes of capitalism.

On the one hand, I am impressed with Scott’s intellectual humility. Being able to recognize the limits of your theory’s applicability is extremely valuable – a lesson Scott himself tries to teach. If Marx had been more cautious about calling for revolutions and had reminded people not to destroy the good aspects of bourgeois culture, he might have saved tens of millions of lives.

On the other hand, it limits Scott’s influence. Scott does not call for an Anti-Legibilist Movement. It would be hard for someone else to start one when the key thinker is only conditionally supportive.

Even though the nuanced view is more accurate, I think that having some Anti-Legibilist activists would improve society. Currently, most opposition to Legibilism comes from people or groups that Scott also dislikes: anti-vaxxers or Luddites or NIMBYs. I would like to see Scott directly address how we can improve our institutions by emphasizing mētis and deemphasizing legibility:

• Is it possible to have democracy without a census?
• What illegible alternatives are there to the National Institute of Standards and Technology?
• How could utilities be decentralized?
• What is the best way to organize or privatize state owned enterprises?
• What is the best way to break up monopolies?

Scott’s work helps us to understand the past. I want him to help us plan for the future.