A representation is never the thing represented, and confusing the two is the standard failure of models and metrics.
Korzybski's 1931 formulation. Borges's one-to-one map is the reductio: a map as detailed as the territory is useless, so every map is an omission, and the question is always which omission.
In practice
- A risk model that reports 3.2% is not reporting a property of the world; it is reporting a property of the model.
- A performance rating of 4.1 is a number about a form, and treating it as a number about a person is where appraisal systems fail.
- Further reading
- Wikipedia
- Subject
- Science & Epistemology
- Essay
- Rules for weighing claims