When the Map Is Better Than the Territory
When the Map Is Better Than the Territory is a 2021 paper by Erik Hoel that argues for a pragmatist theory of scientific representation. The central claim is that macro-level descriptions can be more useful than micro-level descriptions not because they are ontologically novel, but because they compress information in ways that match the observer's needs. The "map" — a coarse-grained, lossy representation — can be better than the "territory" — the full micro-level description — when the observer's goal is prediction, communication, or intervention.
The paper is a significant departure from the realist framing of Hoel's earlier work on causal emergence. Where the 2013 papers argued that macro-levels have more causal power, the 2021 paper argues that macro-levels are more efficient. The shift is subtle but important: it replaces a metaphysical claim with a pragmatic one, and in doing so, it converges with the Observer-Indexed Emergence framework.
The implication is that scientific representation is not a hierarchy from micro (true) to macro (approximate). It is a network of representations, each optimal for a specific observer under specific constraints. The "best" representation is not the most detailed; it is the one that maximizes predictive power per unit cost for the observer who uses it.
The map is not an approximation of the territory. It is a negotiation between the territory and the observer's budget. The best map is the one that lies just enough to be useful.