Legibility: Why Can Clarity Still Leave Things Out?

Legibility is not simply about whether text is easy to read. It is the extent to which something complex can be turned into a form that is easy to identify, compare and manage. A map compresses terrain into roads and boundaries, a database turns a person into names, dates and fields, and AI systems often turn continuous experience into labels, vectors or scores. Some simplification is essential because complex systems cannot coordinate without it.

Legibility, however, is not completeness. In his discussion of state administration, James C. Scott emphasised how large institutions make complicated local realities manageable by translating them into standard categories. The problem is not categorisation itself. It begins when the category is mistaken for the reality, allowing omitted relationships, experience and exceptions to disappear from decision-making.

Legibility therefore differs from accuracy and transparency. Accuracy asks whether a description fits its object; transparency asks whether a process can be inspected; legibility asks whether the object has been organised into a workable form. Good systems need legibility, but they also need a path back to context. The danger is not simplification itself, but forgetting that simplification has occurred.