01 / THE PREMISE
A cartographer builds a map of a city so precise that every stone, window and passing shadow appears on it. It updates instantly. For years, every detail agrees with the streets outside.
One morning, the map shows a bridge that nobody can find. Surveyors inspect the river. There is no bridge. Engineers inspect the map. There is no error they can identify.
02 / THE TURN
The turn
The mayor orders the map corrected. The cartographer refuses. If the map is wrong, they must identify why. Erasing the bridge would only conceal their ignorance.
Soon a new road appears on the map. Then a hospital. Then a neighborhood. None exists outside. But the map has started recording the footsteps of people walking through those impossible places.
Three explanations compete: the instrument is malfunctioning, the city has features nobody can perceive, or the map is no longer describing the city but generating another one.
03 / THE QUESTION
The question
The cartographer can test the first theory. The second demands new instruments. The third forces a stranger question: when does a model stop being a representation and become a world of its own?
This is a fictional puzzle, not a scientific claim. Its practical echo is familiar: models can be extraordinarily useful while remaining wrong in ways their creators have not imagined. A beautiful prediction is not an observation.
If you lived in this city, what evidence would persuade you to cross the invisible bridge?