June 15, 2025
I read this book because, as someone who works in tech, it’s a book that has become popular within tech circles. It’s still a little weird to me that an academic book by an anthropologist with an anarchist bent has found such a wide audience. Nevertheless, it’s a great book. “High modernism” as the ideology (and as Scott points out, aesthetic preference) that undergirds the efforts of states to “legibilize” the societies they govern clearly has a lot of purchase. Scott bolsters his ideas with an enormous number of examples: scientific forestry in Germany; Brasilia; collectivization of agriculture in the Soviet Union; villagization in Tanzania.
Given Scott’s anarchist politics, it is amusing to me that his description of authoritarian high modernist states and their manifest failures reminds me so much of Patrick Deneen’s book Why Liberalism Failed, as Deneen is a conservative Catholic. If we put Deneen’s critique in Scott’s terms, it would go something like the following: liberalism is the ultimate high modernist project, and its imposition on societies of ideas such as universal human rights have the same failure modes as the projects of authoritarian high modernist states. Both Scott and Deneen, for their own reasons, contrast high modernist projects, with their aspirations of universality, with knowledge embedded in local cultures.
Why is this book popular among techies? I’m not the first one to ask this question. There are several explanations floating around in that Twitter thread:
All fine explanations. I myself would chalk it up to influential people in tech discussing the book; Scott Alexander reviews it favorably. But I think there’s another, deeper reason for its popularity within the tech crowd. My take is that technologists’ fascination with Seeing Like a State betrays a (perhaps unconscious) intuition they have about the deep relationship between legibility and information technologies. Let me sketch an outline of what that relationship is.
What makes a society legible or illegible to a state? Scott describes the pre-modern state as being “blind,” tasked with governing an illegible society:
The premodern state was, in many crucial respects, partially blind; it knew precious little about its subjects, their wealth, their landholdings and yields, their location, their very identity. It lacked anything like a detailed “map” of its terrain and its people.
To legibilize the society which it governs, the state must simplify society, enough so that it can make a “map” and other representations that are useful for governance. Scott does not really contemplate the nature of these representations, opting instead to give many examples that hopefully convey the general idea. “Legibility” is in many cases literal: high modernists were obsessed with turning chaotic geographies into simple, highly regular grids. Oftentimes, the “maps” used by the state to represent society are actual maps.
Scott does not use this term, but I think there’s a simple answer to what these representations are: information. What the state sees when it observes the society it governs is data. Thus the measure of the legibility of a society is the state’s ability to represent it as information. Equipped with data about the society it governs, the state greatly increases its capacity to make interventions.
Once we’ve made the link between Scott’s notion of legibility and information, the relationship between information technology and the state becomes clear. Information technologies operate on information, and thus assume a particular kind of social organization for such information to exist. Computers can only store and manipulate data only when there is data to store and manipulate. Thus information technologies can only operate in a legible society. But societies are not born legible; they are made legible, often through state coercion.
Think of a database table containing people’s names. The database table has two columns, one for a person’s given name and another for their family name. This means that every entry in the database represents a person, and that for this person to be stored in the database they need to have a given name and a family name. But there is nothing in nature that confers people with a given name and a family name. Such uniformity assumed by the database must be made true; and as with the example Scott describes in the Spanish colony of the Philippines, this uniformity is often made true by the state. The database thus inscribes in its contents the state’s awesome coercive power.
The fact that information technologies can only operate within legible societies—societies made legible through state coercion, as Scott painstakingly outlines—is a profound irony that tech libertarians must wrestle with. It’s not just that the state has had a hand historically1 in the development of such technologies; it’s that such technologies would not be useful without the state. If software is truly “eating the world”, it is because we live in a world with powerful states that have made societies legible enough to be consumed by software.