Legibility and Mētis

The idea in plain words

In eighteenth-century Prussia and Saxony, foresters wanted to calculate how much timber a forest would yield. Natural forests, with their tangle of species, ages and undergrowth, were impossible to count. So they planted new ones in straight rows of a single species, all the same age, easy to measure and harvest. For a generation, yields were excellent. Then the second and third generations sickened. The simplified forest had lost the soil, insects, birds and variety that had kept the old one healthy. The foresters had made the forest legible, and in doing so had removed the things they did not know they depended on.

The political scientist James Scott opened Seeing Like a State (1998) with this story. Legibility is the simplification that lets a large organisation see what it governs. Mētis is Scott’s name for what legibility destroys: practical, local, adaptive know-how that cannot be reduced to rules.

Where it comes from

Scott studied a series of grand schemes, from scientific forestry to planned cities, collectivised farms and compulsory villagisation, that failed in similar ways. Each began with a state that needed to see its territory and people in standard terms: surveyed land, permanent surnames, standard measures, grid-plan cities. Legibility, he argued, is a precondition of modern administration, and much of it has been beneficial. The disasters came when legibility was combined with what he called high modernism, a confident belief that a rational plan could replace the existing order wholesale, and with the power to impose it on people who could not resist.

The word mētis comes from ancient Greek. In the Odyssey, Odysseus is called polymētis, a man of many wiles: he survives by cunning, improvisation and reading situations, rather than by strength or rules. The French classicists Marcel Detienne and Jean-Pierre Vernant studied the concept in 1974 and showed how it named a whole form of intelligence: the navigator’s feel for the sea, the hunter’s anticipation, the politician’s sense of timing. It is the knowledge of how to act in situations that never quite repeat.

Scott used the word for the practical skill of farmers, craftsmen and communities: knowing when to plant from the look of the soil, or how to run a neighbourhood. This kind of knowledge is local and adaptive, and it cannot be written into a manual without losing much of it. Jane Jacobs had made a similar argument about cities in 1961: the busy, mixed, apparently disorderly streets that planners wanted to clear were in fact working systems, with an order planners could not see.

What it does to the mind

Legibility is seductive because it gives the person at the top a feeling of understanding and control. A dashboard of metrics, a clean organisational chart and a standard process all make a complicated reality look manageable. The danger is that the simplified picture comes to be mistaken for the reality, and decisions are made to improve the picture.

Mētis, by contrast, is hard to defend in meetings. It is held by people lower in the hierarchy, it is expressed in stories and judgement rather than numbers, and it often looks like inefficiency or resistance to change. When a new system arrives, the informal workarounds that kept the old one functioning are often the first casualties, and nobody notices until things break.

Where it shows up

The databases essay shows a deliberate trade: the relational model made data legible, with clean tables and a standard language, at the cost of some flexibility, and later NoSQL systems restored some of the mess on purpose. Large reorganisations and efficiency drives, such as the one described in Meta’s Year of Efficiency memo, face the legibility problem directly: what is visible in headcount and spending may not be what makes teams work. The metric becomes target pattern is legibility’s most common failure in product teams, and Goodhart’s law explains why.

What it asks of someone deciding

Before replacing a working but untidy system, study it. Ask the people who run it what it does that the documentation does not mention.

Prefer changes that can be undone and adjusted. Scott’s recommendation for planners was to take small steps, favour reversibility, plan for surprises and rely on human inventiveness. That is good advice for product and organisational change as well.

Respect the objection. Legibility is also how organisations become fair and accountable. The aim is not to reject measurement and standards, but to remember what they leave out.