What Makes a System Legible?
A seed essay about turning messy operational reality into a model people can use.
Legibility is not the same as observability. Observability helps answer questions from the signals a system emits. Legibility helps people know which questions are worth asking.
The difference matters in operational systems because people rarely lack information. They lack a useful shape for that information.
A Working Definition
A system becomes legible when a skilled operator can see:
- the actors and handoffs,
- the constraints and queues,
- the signals that compress judgment,
- the failure modes that repeat,
- and the actions that can change the next outcome.
The beautiful point is the relationship between those pieces. A metric alone is not enough. A diagram alone is not enough. The work is to connect evidence to a model and the model to an action.
This seed will become a longer essay with examples from reliability engineering, workflow design, and AI-assisted operations.