SUMMARY: Legibility pressure pushes messy human activity into categories that institutions and machines can process.
A system wants clean fields, stable labels, predictable behavior, sortable risk, and standard answers. People arrive with mixed motives, changing contexts, partial identities, and exceptions. Legibility pressure is the force that makes the second group look like the first.
Some legibility is useful. Roads, records, medicine, banking, search, and safety systems all need structure. The problem appears when the structure becomes more important than the reality it was built to describe.
Pressure Points
Identity compression: people are reduced to account status, score, segment, role, or risk class.
Context deletion: a record keeps the outcome but loses the circumstances that explain it.
Appeal friction: correcting the machine-readable record takes more effort than accepting the wrong category.
Behavior shaping: people learn to act for the form, score, feed, or policy rather than for the real task.
Governance Risk
Legibility pressure pairs with the proxy mandate. Once a proxy controls the categories, it can claim to represent the population through the very labels it imposed.
Countermeasure
Keep a path for exception, narrative, correction, and human review. A good system can classify without pretending classification is complete understanding.
Field assessment: legibility becomes dangerous when the map is allowed to punish the territory.