Models Are Maps

All models are wrong. Some are useful. Models are maps—lossy compressions of territory. Confusing map and territory is the core epistemological error. Understanding this changes how you hold beliefs.

"The map is not the territory." Alfred Korzybski said this almost a century ago. It remains the most important thing to understand about knowledge.

What Maps Do

A map represents territory. It is not the territory.

A map of London doesn't contain London. It contains symbols that correspond to features of London. The symbols are useful because they preserve certain relationships: if X is north of Y on the map, X is north of Y in London.

But the map omits most of London. It doesn't contain every brick, every person, every moment. It can't. A map that contained everything would be the territory itself—useless for navigation.

The value of a map is what it leaves out while preserving what matters for its purpose.

Models Are Maps

Every model, theory, framework, and concept is a map of some territory.

• Newton's laws map mechanical motion
• Supply-and-demand curves map market behavior
• The periodic table maps chemical elements
• Your mental model of a friend maps their personality

Each preserves certain relationships while omitting others. Each is useful for certain purposes and misleading for others.

All Maps Are Wrong

By definition, maps omit information. They're compressions. Compression means loss.

Therefore all models are wrong—they don't capture the full territory. The question isn't whether a model is wrong, but whether it's useful enough to justify its distortions.

George Box: "All models are wrong, but some are useful."

Newton's physics is wrong—it breaks down at high speeds and small scales. It's still useful for building bridges and launching satellites. The map is good enough for certain purposes.

The Map-Territory Confusion

Problems arise when we forget the map is a map.

Reification

Treating abstractions as real things. "The economy" is a map of countless individual transactions. Optimizing "the economy" as if it were a thing, rather than asking what individuals experience, is map-territory confusion.

Over-extension

Using a map beyond its valid domain. Supply-and-demand curves work for commodity markets. Applying them to love, friendship, or meaning is using a map outside its territory.

Model lock-in

Unable to see territory that doesn't fit the map. If your only model is "people are rational," you can't see the systematic irrationalities. The map prevents seeing what's actually there.

Mistaking updates for errors

When territory contradicts map, sometimes the map needs updating. Sometimes you misread the territory. Map-territory awareness helps distinguish.

Multiple Maps

The same territory can have multiple valid maps.

London has a street map, a subway map, a population density map, a historical map. Each is valid. Each is useful for different purposes. None is "the true map of London."

Similarly, human behavior can be mapped by economics (incentives), psychology (cognition), neuroscience (neural mechanisms), sociology (social forces), evolutionary biology (adaptive pressures). Each map highlights different features. None is complete.

Understanding this prevents false debates. "Is behavior driven by incentives or psychology?" isn't a real question—it's two maps of the same territory. Both capture something. Neither captures everything.

Good Maps

What makes a good map?

• Accuracy: Preserved relationships actually hold in the territory.
• Compression: Omits enough to be simpler than the territory.
• Usefulness: Preserves what matters for the intended purpose.
• Clear scope: Explicit about what it covers and what it doesn't.

A map can fail on any dimension. Inaccurate maps mislead. Maps that don't compress aren't useful. Maps that preserve the wrong relationships waste attention. Maps with unclear scope get misapplied.

Living With Maps

We can't navigate without maps. Raw experience is too overwhelming. Compression is necessary.

But we can hold maps lightly:

• Remember that every belief is a map, not reality
• Expect maps to fail outside their scope
• Maintain multiple maps for the same territory
• Update maps when territory contradicts them
• Ask what each map omits, not just what it shows

This is intellectual humility operationalized. Not "I might be wrong" as empty phrase, but "this is a map, and maps are wrong" as structural awareness.

— Decoded by DECODER