This summary of Seeing Like A State by James C Scott explores how states simplify and standardize complex societies to make them “legible” to distant rulers. In doing so, the state ends up reshaping reality to fit its narrow view.
[Estimated reading time: 33 mins]
Buy Seeing Like A State at Amazon (affiliate link)
Table of Contents
Key Takeaways from Seeing Like A State
- States try hard to make their subjects legible:
- State power is usually exercised from a centralised position, by officials far removed from their subjects. These officials need to understand and “see” their subjects to exercise many forms of state power, such as taxation.
- Legibility relies on facts that are standardised, static and written. This allows information to be aggregated, but it necessarily ends up “flattening” human complexity.
- Legibility is also selective: the simplifications capture what authorities care about rather than what communities consider meaningful.
- High modernism is the belief that society can be designed from first principles using science and reason:
- High modernism often views history and local knowledge as inconveniences that need to be cleared away, instead prioritising abstract, technical expertise.
- Visual order is a key feature of high modernist ambitions. But the order that emerges organically can be more functional, even if it looks “messy”.
- High modernism has led to disaster in the 20th century when paired with coercive state power.
- High modernism prioritises technical, universal knowledge (techne) over local, practical knowledge (mētis):
- Techne is more accessible to the state, as it can be written down as a set of rules and principles and applied to many different places.
- In contrast, many societies have developed mētis that is well-adapted to the real world, but involve too many variables to replicate in a lab.
- Both forms of knowledge are valuable, but states have generally devalued mētis, leading to some grave mistakes.
- Over time, the state’s “view” can end up shaping reality:
- Because officials govern according to their simplified legible facts, people end up adjusting their lives to fit the categories.
- When states devalue local knowledge, they erode local power as well as the skills and adaptability of populations. The state can therefore end up creating the very dependent, unskilled populations that require the state.
Detailed Summary of Seeing Like A State
States try to make their subjects legible
In the past, state power was exercised indirectly through local elites who had direct understanding of their communities. But such elites had their own interests, and could withhold resources and knowledge from the state.
The shift to direct rule and taxation required an enormous leap in state capacity. Because state power is usually exercised at a distance, the state needed to make its subjects “legible” to a central, distant viewer. That is, the state “sees” its subjects through a host of abstractions: taxes paid, land records, average incomes, mortality rates, and so on. Without legibility, effective state intervention is impossible.
Legibility is an ongoing project
The state’s project of increasing legibility is an ongoing one. The degree of legibility required depends on how ambitious the state is. A more ambitious state may want to extract more taxes, raise a stronger army, or get everyone to speak the same language. Such a state is more intrusive, and so requires a higher degree of legibility.
Example: Making populations sedentary in Southeast Asia
Before Southeast Asia was colonised, land wasn’t that valuable because there weren’t enough people to work it. Populations were sparse and mobile, which made them hard to tax. “Voting with your feet” had a very literal meaning.
States therefore tried to create permanent, fixed settlements to gain control over their populations. Some of this was very coercive. For example, in Thailand, the state devised a system of tattoos to literally mark who “belonged” to whom. Bounty hunters made money capturing runaways to return to their owners.
Scott thinks that the desire to increase legibility is also what drove colonialists in Southeast Asia to prefer large plantations over smallholder farms. Even though smallholder farming is much more productive for almost all crops, it’s much easier for the state to control and tax a few large plantations than a lot of small farms. [Joe Studwell offers an alternative explanation — he chalks it up to a mistaken belief that the economies of scale that apply to manufacturing also apply to agriculture.]
Legible facts are standardised, static and written
The state narrows its focus by relying on facts that are:
- Standardised and aggregated. Subjects need to be grouped in ways that allow for collective assessments—e.g. totals and averages. Standardised subjects are considered to have uniform needs and are treated as basically interchangeable.
- Static. Even when facts look dynamic, this is usually just the result of multiple static observations over time using “connect the dots”. The state remains blind to what happens between the dots.
- Written. Written records can be stored, copied, and transmitted across time and space. Oral and unwritten traditions by contrast are invisible to distant officials.
Example: Employment figures
Many people have working lives that are exceptionally complex and may change on a day-to-day basis. But whether someone is “employed” or not is a stylised fact that can be aggregated and used in centralised decision-making. [Gig workers are a good example of a type of work that has challenged some of the traditional categories.]
People who gather and interpret aggregate data usually understand that their categories are simplifications that hide a wealth of variation. Averages are statistical fictions that may not describe anyone’s experience. However, once the categories are set and data is recorded, the “facts” can take on a life of their own.
Legibility is selective
State simplifications capture only those aspects of social life that are of official interest.
Narrowing your vision does come with real advantages. When you focus on certain limited aspects of a complex reality, it becomes possible to “zoom out” and see the big picture. A narrower focus can also confer analytical power—if you focus solely on, say, crop yields, you can carefully tease out which factors affect yields.
But tunnel vision inevitably has blind spots. Second- or third-order consequences are hard to predict, and many events (e.g. droughts, wars, epidemics) fall outside of planners’ models. For example, agroeconomic analysis tools are very useful in examining the microeconomics of an individual firm, but often fail to account for externalities like waste disposal, pollution and increased fragility.
Example: Crop yields are not everything
You can treat all rice, all corn, and all millet as “equal” if you only consider the amount of money gained from selling it on the market.
However, there is a lot that can be harvested from a plant other than its seed grains. For example, you can use stalks as trellises or temporary fencing, and husks as wrappers or fodder. Families might also prefer more reliable crops even if they have lower average yields, so that they have food even when markets fail.
Moreover, while the drive to maximise profits has certainly increased yields a lot, there are also costs:
- Taste and nutritional quality have been deprioritised (for example, the “supermarket tomato” developed after WWII was selected for yield and durability for transportation rather than taste); and
- The cultivars we use today have far less genetic diversity than ever before, so are more disease-prone. So far we’ve coped by using pesticides. But long-term pesticide use is starting to produce pesticide-resistant strains of pests and weeds, and some pathogens have developed cross-resistance to whole classes of pesticides. We also don’t fully understand all the ways in which pesticides impact soil and groundwater quality or human health.
The state’s selective focus means that it will ignore things that fall outside its field of focus, even if they are important to the people being governed.
Example: Rice variants in Sierra Leone
Farmers in Sierra Leone often chose rice variants that textbooks claimed were lower-yielding, because they added more chaff that would have to be removed. However, that “chaff” was precisely what stopped birds from eating their rice before harvest time.
What is high modernism?
High modernism was not limited to any political ideology, and was held by those on both the left (e.g. Lenin and Trotsky) and the right (e.g. Nazism). Scott uses the term “high modernism” to describe the belief in scientific and technical progress associated with Western industrialisation between roughly 1830 and WWI.
Redesigning society to improve welfare
The idea that one of the central purposes of the state is to improve the lives of all members of society is actually quite novel. Although states had long cared about improving their subjects’ productivity or health for tax or military purposes, it was only in the 19th century that the welfare of a population began to be seen as an end in itself.
At [high modernism’s] centre was a supreme self-confidence about continued linear progress, the development of scientific and technical knowledge, the expansion of production, the rational design of social order, the growing satisfaction of human needs, and, not least, an increasing control over nature.
— James Scott in Seeing Like A State
Perhaps it’s understandable that people who had witnessed stunning scientific advances in medicine, engineering and management would assume that the same scientific mindset could “perfect” the social order, too. But they were wrong.
Example: Soviet collective farms
The state and collective farms that the Soviet state implemented in the 1930s failed to deliver on any of the goals envisioned by their socialist leaders. The sowing plans imposed from above were often made by people who knew little about the crops they were mandating or local soil conditions. Planners clearly favoured monoculture and specialisation, with entire regions producing only wheat, livestock, or potatoes.
Collective farms have done well with some crops, like the major grains. Wheat in particular is a sturdy crop, well-suited to large-scale mechanisation. But many other agricultural products—including fruit, vegetables, livestock, eggs and dairy—are not. Raspberries, for example, are very delicate. They are virtually impossible to pick by machine, and last no more than a few days once packed.
Even though the state farms and collective farms absorbed 10% of the labour force, they produced merely 2.2% of gross farm production. For the next 50 years or so, yields for many crops stagnated or even fell.
Sweeping aside history
High modernism is almost exclusively focused on the future. Many high modernists dreamt of wiping the slate clean so they could start from scratch. The Pol Pot regime, for example, wanted to create a whole new Cambodian nation starting at “year zero”. Colonialists similarly wanted to break with history and tradition, building new capitals instead of compromising with what already existed.
Scott acknowledges that “breaking with the past” can be very attractive when the past was truly terrible. For example, even though the initial land reforms in Bolshevik Russia and post-revolutionary China involved vast simplifications, they effectively enfranchised millions who had lived in virtual serfdom.
But new cities tend to be “thin” cities, which makes them easier to control from outside. While a “thick” city is built by countless people contributing to it over time, a “thin” city contains only the vaguest outlines of the complex activities that take place in “thick” cities. It doesn’t have the diverse sources of social cohesion and coordination that develop over time.
It is not a coincidence that many of the high-modernist cities actually built—Brasilia, Canberra, Saint Petersburg, Islamabad, Chandigarh, Abuja, Dodoma, Ciudad Guayana—have been administrative capitals. Here at the center of state power, in a completely new setting, with a population consisting largely of state employees who have to reside there, the state can virtually stipulate the success of its planning grid.
— James Scott in Seeing Like A State
Visual order vs functional order
High modernist planners are drawn to visual order: straight lines, geometric layouts, and functional separation (e.g. separating pedestrians from cars, and having separate zones for work, residences, entertainment and shopping). Functional separation helps planners optimise for efficiency, because trade-offs inevitably arise when you try to optimise for two or more functions at once.
But visual order is not the same as functional order. Straight lines are rarely practical in reality. If the topography is irregular, it’s much easier and more cost-effective to work with the terrain than to level everything for a straight road. A city that looks “messy” from above may nevertheless perfectly meet the needs of the people living in it. Most complex systems do not display a surface regularity, but may still contain a deeper order. You might be able to create a city from the top-down, but not a community.
Example: Le Corbusier
The European architect Charles-Edouard Jeanneret, better known as “Le Corbusier”, was the embodiment of high-modernist urban design. He loved simple, straight lines and was offended by disarray and complexity. Above all, he wanted to optimise productivity. He saw shopping and meal preparation as nuisances that would best be discharged by centralised services.
Le Corbusier’s drawings for Paris, Buenos Aires, and Rio de Janeiro were all designed to make a powerful visual impact from a distance, but were completely divorced from local context. His drawings could’ve been for any city, anywhere. For this reason, none of his designs were ever adopted—even Stalin found his plans for Moscow too radical. But he did influence the design of Brasilia (see below).
[Le Corbusier’s Moscow was] a city of nowhere,… [a city] that is neither capitalist, nor proletarian, nor socialist, … a city on paper, extraneous to living nature, located in a desert through which not even a river must be allowed to pass (since a curve would contradict the style).
— Soviet architect El Lissitzky, as quoted in Seeing Like A State
One of the most influential critics of high modernist urban planning was Jane Jacobs, author of The Death and Life of Great American Cities (1961). [Jane’s Walks are available in many cities — worth checking out!] Instead of looking at cities from the top-down, Jacobs took a “bottom-up” approach starting from the street level.
Jacobs understood that many activities serve multiple purposes. Most people don’t just go out to do a single task; they’ll often do several errands at once, and maybe stop to chat with a friend along the way. So, unlike high modernists planners who favoured single-use districts, long streets and architectural uniformity, Jacobs favoured mixed-use, short streets, and buildings of varying age and condition (to allow for a variety of rental terms and uses). Jacobs saw the city as a “social organism” that is constantly changing. For such organisms, diversity is the key to resilience and economic success.
Example: Jane Jacobs on “eyes on the street”
Jacobs explains how, when a friend needed to use her apartment, she just left the key with the nearby deli owner who had a special drawer for such keys. In fact, every nearby mixed-use street had someone who played that role—someone on “sidewalk terms” with his customers.
Similarly, the public peace is kept by a large number of unpaid observers. These observers can quickly intervene if, for example, they see a suspicious man hanging around young children. When a street is animated and busy, you’ll naturally get a lot of “eyes on the street”. Formal institutions like the police only work when they are undergirded by this rich, informal public life.
Problems with high modernism
Scott is careful not to dismiss high modernism or state simplifications entirely. Many of its techniques did greatly improve efficiency. Moreover, some high modernist techniques have allowed us to discover new valuable new knowledge. Centralised data collection makes it possible for officials to intervene early in epidemics, understand economic trends, and make all sorts of other policy decisions.
Scott further acknowledges that some public works—constructing dams, irrigation systems, and public transport—necessarily require a high degree of state capacity. There are many tasks for which a centralised, high-modernist approach can be the most efficient and equitable solution.
High modernism has led to disaster when paired with coercive state power
Scott’s main concern is when the arrogance of high modernism is combined with with coercive state power. High-modernists tend to see politics as obstacles to their visions. Under their views, only those with the scientific knowledge to discern and create a superior social order are fit to rule. Those who refuse to yield to the scientific plan need to be educated to its benefits—or swept aside.
This view is deeply authoritarian. The worst disasters have typically occurred in the states of the former socialist bloc and in revolutionary Third World settings, where authoritarian state power could run roughshod over any resistance. But the ideas behind these schemes were Western in origin.
Example: Tanzania village resettlement
Between 1973 and 1976, Tanzania undertook the largest forced resettlement scheme in independent Africa, relocating over 5 million people into planned villages. The campaign was started by Julius Nyerere, Tanzania’s head of state, who wanted to promote agriculture that would yield larger surpluses for export, deliver services more efficiently, and encourage communal, socialist cooperation. Nyerere originally insisted that no one should be forced to move.
Unfortunately, the campaign ended up being coercive and violent. Since officials were judged on metrics like the number of villages created or people resettled, they traded off values like autonomy that weren’t being measured. Authorities forcibly pulled or burned houses down to prevent people from returning.
The campaign was also an economic failure. Villages had been planned by central officials without consultation. They often picked sites without enough arable land, or far from fuelwood and water. Some villages were laid out as a long street of houses stretching for miles along a major road—convenient for administrators to monitor, but terrible for the villagers. Agricultural production fell sharply, necessitating large food imports. Eventually, many Tanzanians drifted back to sites more suitable for grazing and cultivation.
State plans to resettle people have rarely gone as anticipated. Yet they have almost always disrupted prior communities whose cohesion derived mostly from nonstate sources.
Things only worked because of informal, unplanned practices
Political elites typically begin with changes in the formal structure and rules, because these are the easiest to rearrange. But everyone knows that the work actually carried out in an organisation isn’t fully described by what is written down in handbooks and guidelines. That is why “work-to-rule” strikes are so effective.
It’s also why completely top-down attempts to plan a village or city inevitably fail. There is always another, far more “disorderly” and complex city that makes the official planned city work.
Example: Brasilia
Brasilia was designed by Lucio Costa, an architect heavily influenced by Le Corbusier. It was built as a new city from the ground up, with no references to the traditions or practices of Brazil’s past. Brasilia was intended to be an exemplary “city of the future”.
In most Brazilian cities, family dwellings were cramped. The public square and crowded “corridor” streets had therefore been important sites where children would play, adults might run into friends, and ceremonial processions could take place. But Brasilia was not designed to facilitate this. There was a square—the Plaza of the Three Powers—but it was way too large for a public gathering space. Instead, Brasilia was designed for vehicles. It is a city without crowds, lacking the “bustle” of street life.
Of course, things don’t always go as planned. The manual labourers who built Brasilia set up their own unplanned settlements on the periphery. By 1980, 75% of Brasilia’s population lived in these unplanned settlements.
State officials have often come to tolerate a range of informal practices that helped the official “plan” survive. Formal order is always, to some considerable degree, parasitic on informal processes—processes which the formal scheme does not recognise, but without which it could not exist.
Is this an argument for free markets?
This analysis of high modernism may sound like an argument for the ‘invisible hand’ of markets over top-down, centralised economies.
However, Scott cautions against this interpretation for two reasons:
- First, the market is itself an instituted, formal system of coordination. It is therefore similarly dependent on a larger system of social relations, such as social trust, community, and cooperation.
- Second, but perhaps more importantly, the economy is a subsystem of a finite and non-growing ecosystem. For the economy to persist, it must respect the carrying capacity of that broader ecosystem. [This is the limits to growth argument.]
Two different types of knowledge
Scott contrasts two types of knowledge:
- techne — technical, universal knowledge that can be written down and taught formally; and
- mētis — practical, local knowledge that is learned through experience.
Techne and mētis are both Greek terms. States have tended to privilege techne while dismissing mētis as backward or unscientific. However, both forms of knowledge have their place.
Technical, universal knowledge (techne)
Techne is universal knowledge that tends to be characterised by impersonal, quantitative precision and verification. The scientific method is one example. By its nature, techne can be transmitted quite easily at scale, and taught as a formal discipline.
A recurrent theme of Western philosophy and science, including social science, has been the attempt to reformulate systems of knowledge in order to bracket uncertainty and thereby permit the kind of logical deductive rigor possessed by Euclidean geometry.
— James Scott in Seeing Like A State
At its most rigorous, techne is based on logical deduction from self-evident first principles. Philosophy, for example, is a discipline that exists as a realm of pure thought, beginning in the mind or on a blank sheet of paper. Scott describes utilitarianism as an example of philosophers trying to make ethics into a “natural science”.
Another example is economics, which developed techniques to isolate variables that can be expressed in numbers, such as GDP. Invention and certain types of risks (e.g. ecological dangers) are treated as exogenous, because they are too difficult to measure.
Limitations of techne
Although techne can be very powerful, it also has limitations:
- Limited variables. Experimental science isolates just a few key variables, because each new variable increases the number of possible interactions exponentially. Scientists must therefore radically simplify their experiments if they want to produce verifiable and universal results. This inherently favours monocultures. Polyculture and intercropping (i.e. growing different crops in the same field) introduces too many variables for researchers to easily test.
- Local variation gets averaged away. For example, standard practice in soil analysis is to gather soil from several parts of the field and combine them into an “average” sample. But this average is often a statistical fiction. Soil can be exceptionally variable even within the same field. The recommended fertiliser application may not be right for any part of the field.
- Short-term bias. While the scientific method itself doesn’t necessitate a short-term perspective, institutional and commercial pressures make long-term experiments incredibly expensive. Nearly all agricultural studies last one or at most a few seasons. However, those who argue for polyculture claim it is superior over decades, because of the long-term ecological costs of monocropping.
- Conditions may change. Some routines or methods may appear “universal” and look very efficient so long as the environment remains stable. But if the environment changes, such methods can fail badly.
Example: Potato growing in uncontrolled conditions
The Andes has incredible ecological variety. Traditional potato farmers in the Andes started by taking their field and its surroundings as a given, then selecting the varieties that are likely to do well in that setting.
By contrast, in scientific potato growing, researchers start with a new cultivar or genotype, and then try to alter the field conditions to meet that genotype’s requirements. This usually requires nitrogen fertiliser and pesticides, as well as large-scale irrigation. But the potato genotypes that thrived under controlled, scientific conditions usually failed within 3-4 years under real-life conditions.
This attempt at total control is an invitation to disorder. And the rule seems to be that the more rigid and exclusive is the specialist’s boundary, and the stricter the control within it, the more disorder rages around it.
— Wendell Berry, The Unsettling of America
Practical, local knowledge (mētis)
Mētis represents a wide array of practical skills and intelligence for dealing with a volatile uncertain environment. These skills are very difficult to write down, because much of the skill operates beyond our conscious awareness. You can’t write down the instructions for riding a bicycle and expect a beginner to learn just by reading them. Many crafts or trades involving mētis are therefore taught through apprenticeships.
Example: Implicit mētis in diagnosing syphilis
Mētis knowledge is often so implicit and automatic that even those who hold it are at a loss when asked to explain it.
Around 1900, one doctor is said to have had a spectacularly high success rate in diagnosing syphilis in its early stages. He himself could not explain why. So two other doctors were tasked with observing his patient exams over several weeks to see if they could spot what he was picking up.
Eventually, they realised the doctor was unconsciously registering a slight eye tremor in patients with syphilis. That eye tremor later became a universally recognised symptom of syphilis.
The litmus test for mētis is practical success. People who used the advice only care that it works, without stopping to figure out why. For example, South American Indians knew that the bark of the cinchona tree could cure malaria, without knowing what its active ingredient was (quinine) or the causal mechanism behind it. The knowledge reflected in mētis is almost always local—the experience that is most useful for one place does not transfer fully to another environment.
Example: Mētis for when to plant corn
According to one legend, a piece of Native American mētis was to plant corn when the oak leaves were the size of a squirrel’s ear. In some years, this might be earlier or later, but it turned out to be a reliable formula for avoiding a frost in the places where this mētis prevailed.
Here, the knowledge likely could be translated into more universalistic scientific terms (e.g. plant when the soil hits a particular temperature or moisture level). But that advice is less practical.
Jobs involving responses to emergencies and disasters also require mētis, because each accident is complex and unique. Examples include firefighters, paramedics, and rescue squads. There are rules of thumb, but knowing which rule to use requires mētis, as does knowing when to ignore the usual “rules”. The best responders are those who have experienced many different situations.
Mētis is *not* “traditional knowledge”
The term “traditional knowledge” may imply knowledge that is rigid, and practised only out of custom or tradition. It may even bring to mind “backward” superstitions or old wives’ tales.
But the key test for mētis is practical efficacy. When a new technique works, it spreads rapidly. Because mētis is so local and decentralised, it is very much open to new ideas. There is no centralised gatekeeper. Moreover, since mētis is usually not written down and passed down orally instead, it naturally evolves over time—just like language.
Limitations of mētis
- Hard to transfer. Since mētis is hard to write down, it is confusing for state officials. High modernists therefore ignored or sidelined this valuable knowledge when imposing their plans.
- Not very democratic. Artisans and guilds often held a monopoly on their knowledge. Standardised knowledge can make certain skills more broadly available.
- False inferences. Since many indigenous cultivators carried out their research without proper experimental controls, they could draw the wrong conclusions from their observations.
- Failure to see the bigger picture. Local knowledge can overlook the bigger picture. Some methods used by indigenous people only worked at low population densities, and caused damage to their ecosystems.
Combining both forms of knowledge
We shouldn’t see techne and mētis as being opposed to each other. Scott is not making a general argument against the experimental techniques of modern science, nor is he claiming that traditional medicine is superior.
Rather, it’s about knowing the limits of your tools, and understanding when to use each one. Many scientific breakthroughs (e.g. vaccines, and modern malaria treatments) were built on top of mētis. The problem with rationalism is not its use of techne, but its failure to recognise any other forms of knowledge. By contrast, mētis does not claim to be universal; it is inherently pluralistic.
The state’s view ends up shaping reality
The builders of the modern nation-state do not merely describe, observe, and map; they strive to shape a people and landscape that will fit their techniques of observation.
— James Scott in Seeing Like A State
Knowledge and power
When the state devalues local knowledge, it also erodes local power. Similarly, when it elevates centralised, technical knowledge, it increases the power of those who hold that knowledge. As such, high modernism is highly appealing to technicians, planners, engineers, and the bureaucratic intelligentsia. [I think causality could also go the other way around. While Scott seems to suggest high modernism appeals to these groups because they stand to gain power from it, I think it’s just as likely that people naturally drawn to high modernist ideas are more likely to become planners, engineers, bureaucrats, etc.]
Example: Scientific management disempowers artisan producers
Frederick Taylor understood that his mass production methods would empower managers and destroy artisans’ mētis.
Only factory managers had knowledge of the whole process, while workers executed just a tiny part of the process. Workers accordingly became a weakened and more pliable population. While this boosted efficiency, it particularly boosted the control and profits of those in power.
While many agricultural innovations of the 20th century might appear technical and neutral, they had rather profound effects on power. Most state projects to modernise agriculture ended up consolidating power in large, centralised institutions and diminishing the power of cultivators and their communities.
Example: Large-scale farming creates dependencies
Small-scale farming makes use of farmers’ place-specific knowledge about rainfall, soils, and the peculiarities of their particular plots. They may even choose different crop mixes to suit their household’s particular labour mix and circumstances—e.g. a farm with many able-bodied young workers may grow more labour-intensive crops; a farm whose workers migrate out for part of the year may choose crops that require little care throughout the year.
But small-scale farmer’s knowledge is almost entirely local. When he is moved to a different place, his local knowledge becomes useless.
Since large-scale farming makes farms interchangeable, local farmers instantly become interchangeable, unskilled labourers, dependent on those providing seeds, fertilisers, and pesticides:
- Hybrid seeds are either sterile or do not breed “true”, meaning farmers must buy new seeds every year from the company that controls the parent plants.
- High-yielding varieties of wheat, rice, and maize require abundant irrigation and commercial fertiliser to reach their potential.
- Since high-yielding varieties lack biological diversity, each generation is likely to succumb to pests. Farmers then need to buy pesticides to counter them.
Facts on paper reshape facts on the ground
Abstractions such as maps and official data don’t just summarise facts; they also have the power to transform them. This happens through tax incentives, grants, and subsidies for those who fit the state’s schema, and penalties for those who do not. And since government officials’ productivity is determined by legible facts, this can lead to problems like Goodhart’s law.
Example: Last names in the Philippines
While the development of permanent last names usually took several generations in the West, it was sped up when Westerners colonised other populations.
In 1849, the Spanish issued a decree requiring Filipinos to take on permanent Hispanic last names. Each local official was given a “supply” of surnames for their region (there were whole towns with surnames beginning with the same letter). To ensure that Filipinos wouldn’t just ignore their last names, any documents that didn’t use the official names were treated as null and void.
Courts of law and state bureaucrats rely mostly on written records, so an error in an official document can be more powerful than an unreported truth. When written records are the only way to assert reality, fictitious facts-on-paper prevail over facts on the ground.
High-modernists seek to transform the environments in which people live. But they end up transforming the people themselves. Complex and diverse environments produce a resilient and flexible population that takes initiative and adapts to new challenges. Narrow, planned environments, by contrast, foster populations that are less innovative and resourceful. Ironically, the state ends up producing exactly the populations that need close supervision.
If the facts—that is, the behavior of living human beings—are recalcitrant to such an experiment, the experimenter becomes annoyed and tries to alter the facts to fit the theory, which, in practice, means a kind of vivisection of societies until they become what the theory originally declared that the experiment should have caused them to be.
— Isaiah Berlin, “On Political Judgment”, as quoted in Seeing Like A State
My Review of Seeing Like A State
I really didn’t like Scott’s writing style, which involves lots of unnecessarily complex words, long, passive sentences and unnecessary repetition. But I pushed through because I’d heard many people recommend this book, and it does have some valuable insights.
I confess I have a tendency to favour technocratic expertise, abstract reasoning and centralised knowledge. But in recent years, I’d grown more aware of its limits, so I think I was fairly receptive to many of the ideas in this book. I found the stuff about monoculture vs polyculture fascinating. I just wish the book had been more even-handed.
Scott does make an effort to sound balanced, but it feels pretty token. He is clearly deeply sceptical of technocratic, top-down governance. However, I think he straw-mans technocracy. In the context of urban design, he writes:
Technocracy, in this instance, is the belief that the human problem of urban design has a unique solution, which an expert can discover and execute. Deciding such technical matters by politics and bargaining would lead to the wrong solution. As there is a single, true answer to the problem of planning the modern city, no compromises are possible.
— James Scott in Seeing Like A State
But I’ve never thought that the argument for technocracy came from some belief that there is a “unique solution” to anything. Personally, I think the strongest argument for technocracy is because most complex issues have plenty of objectively bad answers (as in, answers that have little overlap with reality), and it’s very easy to land on one of those bad answers if you haven’t considered an issue deeply. And most people, understandably, simply don’t have the time or the desire to consider most issues deeply. Technocrats, by contrast, are paid to consider issues deeply — and the good ones do.
Scott also fails to engage with problems that seem to require centralised coordination. He praises “shifting cultivation” (aka “slash-and-burn” agriculture) as being rational and efficient in the areas where it was practised, but it sounds like it only works at low population densities and has rather high carbon emissions. Similarly, he never addresses the challenge of interstate competition. Defending against external threats requires exactly the level of centralised mobilisation that Scott distrusts, and is a big reason why states became so powerful in the first place.
Where I do agree with Scott is on the need to combine centralised knowledge with local, practical knowledge through genuine consultation. Unfortunately, the book doesn’t really explore how to do this. While there are a few rules of thumb towards the end, this is <5% of the book. The first 95% just bangs on about how bad high modernism is. Moreover, the rules of thumb boil down to platitudes like “be humble”, “take small steps”, and “plan on surprises”. This isn’t super practical. As it stands, there’s a real risk that readers come away from Seeing Like A State without much more than a general vibe that “top-down planning is bad”.
Let me know what you think of my summary of Seeing Like A State in the comments below!
Buy Seeing Like A State at: Amazon <– This is an affiliate link, which means I may earn a small commission if you make a purchase through it. Thanks for supporting the site! 🙂
If you enjoyed this summary of Seeing Like A State, you may also like:
4 thoughts on “Book Summary: Seeing Like A State by James C Scott”
An excellent summary, well done, P! I really like how you provide the general principles and then you give an example in the coloured boxes.
This summary is especially helpful as I think I would not read this book (esp since you say it is a hard read!).
It sounds like to me as well that there needs to be a balance of techne and metis — and the more interesting inquiry may be when good technocrats think about these issues deeply and plan well.
Thanks Lizzie! I did find the book a challenging read, but I’ve also seen many people highly recommend it, so mileage may vary 🙂
Enjoyed this – thanks for writing!
I’ve had this book on my to-read for ages and now I’m not going to bother.
It’s interesting how popular this book has been in SF tech culture. Especially given that software and recommendation algorithms are arguably shifting how humans pay attention. They likely have more sophisticated models of the population of a country than the state does
Yeah, I think the book is popular among libertarians, and they are quite prevalent in SF tech circles. Good point about large tech companies probably having a more sophisticated view of the population than governments do. I guess their rebuttal would be that they don’t have the same degree of coercive power that states do.