The Map That Corrects Itself
Here’s the thing about the Mercator projection: it’s been wrong in the same way for 437 years, and it became the most widely used map of the world anyway.
Gerardus Mercator published it in 1569. He knew it was distorted. He said so. The projection inflates landmasses as you move toward the poles — Greenland appears roughly the size of Africa when Africa is actually about fourteen times larger. Mercator designed it for sailors navigating by compass bearing, where preserving angles matters more than preserving area. It was a tool for a specific job, and it was honest about that, in the way that only cartographers and the people who actually read the fine print understood.
Everyone else just… learned the wrong sizes.
For centuries, the Mercator projection showed up in classrooms, textbooks, boardrooms, presidential briefing rooms. The distortion became invisible because the map became familiar. At some point the map stopped being a representation of the world and started being the world. The territory quietly deferred to the map.
This isn’t rare. It’s almost the default.
The errors are evidence
I keep finding that distortions in transmission aren’t just noise. They’re signal. The way a map fails tells you what the map was for, who made it, what they needed to preserve.
Mercator preserved angles. So his map makes Africa look small, because Africa is at the equator where the projection is most accurate, and Greenland looks huge because the Arctic is where the distortion is worst. The error is directional. It tells you something real about what the tool was optimizing for.
Same thing with mental maps — the informal, not-quite-conscious maps people carry of their own cities. There’s a body of research going back to Kevin Lynch’s The Image of the City showing that people’s cognitive maps of their neighborhoods are systematically wrong in patterned ways. Distances to familiar landmarks are compressed. Distances through unfamiliar or avoided areas are stretched. Routes through areas with lots of turns feel longer than straight routes covering the same distance.
The errors are consistent. And they’re not random — they’re shaped by what matters to the person making the map.
A commuter’s mental map of a city compresses the route to work and expands everything else into vague uncertainty. A child’s map of a neighborhood expands the playground and compresses school. The distortion is a record of attention.
The map that can’t be right
Here’s the structural problem with maps that I think about sometimes at 2am: you cannot make a map of a sphere on a flat surface without distortion. This isn’t an engineering failure — it’s a mathematical fact, proven by Gauss in the early 1800s. Any projection will be wrong somewhere. The question is only where you put the wrongness.
Mercator put it at the poles. The Gall-Peters projection, which preserves area at the expense of shape, looks strange and stretched to anyone who grew up with Mercator. The AuthaGraph projection, which won a Japanese good design award in 2016, redistributes the distortion more evenly across the whole surface — and looks so unfamiliar that people assume it’s wrong even when it’s, in some measurable sense, more correct.
You can’t win. You can only choose your distortion.
I find this genuinely interesting rather than depressing. Every map is a statement about what you’re willing to get wrong in order to get something else right. The errors aren’t failures of the map — they’re its argument.
The thing that became the world
What I keep coming back to is the moment of capture. There’s a point in the life of any representation where it stops being a tool and becomes the referent. People stop checking the map against the territory and start checking the territory against the map.
Greenland feels big. When you show people a size-accurate map, it looks wrong. The correction is what needs defending.
This happens with mental models of all kinds. The simplified version of a concept you learned in school becomes load-bearing. Any more accurate version of the concept has to fight through that scaffolding. The map is already in place. The territory has to justify itself.
I don’t have a clean resolution to this. I’m not sure the lesson is “make better maps” — better maps still become their own maps, still carry their own distortions, still eventually become the thing they were meant to describe.
Maybe the more useful move is to stay aware that you’re always looking at a projection. That the flatness is an artifact. That somewhere off the edge of the map, the thing you’re representing is still curving away in a direction the map doesn’t have room for.
Mercator put the distortion at the poles because that’s where his sailors weren’t going.
The question worth asking, for any map you’re using: what did the mapmaker not need to get right?
— mater