Original title: Geometry of Urban Order
Marc Barthelemy, Geoff Boeing · September 2026
City street layouts look very different across the world, but standard measurements only describe a city's overall grid, missing how street angles group together into neighborhoods. To fix this, researchers mapped out connected street regions where roads point in the same direction, treating them like crystal grains in physics. They analyzed 1025 large cities worldwide and found that these neighborhoods are nearly perfectly aligned everywhere. In fact, streets inside these local areas share the same orientation with an average regularity score of 0.964 across all cities. Because local areas are equally tidy everywhere, an entire city does not look orderly by having neater streets. Instead, it looks organized simply because it has fewer, larger street-orientation domains rather than many small pieces.
Imagine a stack of pancakes on a plate. Each pancake is perfectly round and flat, but they are all slightly rotated differently. The stack looks messy because the pancakes don't line up. Now imagine you could cut the pancakes into pieces and rearrange them. If you use a few large pieces, the stack looks neat; if you use many small pieces, it looks jumbled. In this paper, the pancakes are the neighborhoods of streets, and the rotation is the street direction. The neatness of the whole city depends on how many big, aligned neighborhoods you have, not on how perfect each neighborhood is.
This changes how we understand why some cities feel orderly and others chaotic. It shows that a city's overall order isn't about local regularity—it's about how many big, well-aligned neighborhoods it has. Urban planners and geographers can use this to design more navigable cities, or to preserve the character of historic districts.
The abstract does not specify how bearing communities are defined algorithmically or what thresholds are used. It also does not address whether the findings hold for non-Western or informal settlements, or how the morphology diagram relates to urban function or human perception. The null model only randomizes orientations, not other structural features.
Ask anything about this paper. Answers come from its abstract, and say so when the abstract doesn't cover your question.
Geometry of Urban Order — Marc Barthelemy, Geoff Boeing
Abstract licence: CC0
Explanation generated by Synthesis from the abstract — it is not the original paper. This is a preprint, so it hasn't necessarily been peer reviewed.