How Maps Distort Reality

Every flat map lies. The question is which lie is least harmful for the task at hand.

A globe is an accurate representation of Earth’s surface. The surface is curved. A map is flat. Flattening a curved surface requires stretching, tearing, or compressing it. None of those actions preserves every measurement. Every map projection chooses which measurements to keep and which to sacrifice.

This is not a design problem that cartographers have failed to solve. It is a mathematical theorem that proves no solution exists.

The mathematical constraint

Carl Friedrich Gauss proved the relevant constraint in 1827. His Theorema Egregium — Latin for “extraordinary theorem” — showed that the Gaussian curvature of a surface is an intrinsic property that cannot change if the surface is bent without stretching.

Earth’s surface has positive Gaussian curvature. A flat plane has zero Gaussian curvature. The two are mathematically incompatible. You can bend a sheet of paper (bending changes nothing about the zero curvature) but you cannot flatten a sphere without stretching or compressing it. Stretching changes the curvature. The theorem proves that the change is unavoidable.

This means that every flat map of Earth must distort at least one of the following properties: area, shape, distance, or direction. No single projection can preserve all of them. The choice of what to preserve is not a mathematical question. It is a practical one.

The oldest lie: the gnomonic projection

The gnomonic projection is the oldest documented map projection. Ancient Greek scholars developed it by projecting points on the surface of a sphere from the sphere’s center onto a tangent plane. The result has one remarkable property: every straight line on a gnomonic map corresponds to a great circle on the sphere, which is the shortest path between two points.

That property makes the gnomonic projection extremely useful for planning routes. A navigator can draw a straight line between two cities and read off the great circle route. The projection is otherwise almost unusable. It distorts area and shape so severely toward the edges that it can show only one hemisphere at a time. The poles are infinitely far away.

The gnomonic projection does not claim to be a general-purpose map. It is a tool for a specific task: finding the shortest path. It sacrifices everything except one property — that great circles are straight lines.

Preserving shape: the Mercator projection

Gerardus Mercator introduced his cylindrical projection in 1569. His goal was navigation. Sailing ships follow constant-compass courses, called rhumb lines or loxodromes. On a Mercator map, every rhumb line is a straight line. A navigator can set a compass bearing, draw a straight line, and follow it to the destination.

Mathematically, Mercator achieved this by progressively stretching the latitudes as they move away from the equator. The stretching factor is exactly the reciprocal of the cosine of the latitude. At the equator, the factor is one — there is no stretching. At sixty degrees north or south, the factor is two — everything is twice as wide as it should be. As the latitude approaches ninety degrees, the factor approaches infinity. The poles cannot be shown.

The Mercator projection is conformal. It preserves local angles and shapes. A small region on the map has the correct shape, even if its area is wrong. The property is invaluable for navigation because it means that the compass bearing you read from the map is the bearing you should steer.

The cost is area distortion that grows without bound toward the poles. Greenland appears larger than Africa on a Mercator map, even though Africa’s actual area is approximately fourteen times larger than Greenland’s. Europe appears much larger than South America, even though South America is about twice the size of Europe.

Contemporary cartographers criticize the Mercator projection not because it is mathematically wrong — it is not. It performs its intended function precisely. They criticize it because its area distortion conveys a misleading impression of relative size that has been interpreted as a statement about importance. Countries near the equator appear smaller than countries at higher latitudes. The distortion has been cited as reinforcing geopolitical biases that favor the Global North.

The Mercator projection was designed for a specific task: navigation. It does that task perfectly. The criticism arises when the same map is used for a different task: comparing the sizes of countries.

Preserving area: the Gall-Peters projection

An equal-area projection preserves the relative sizes of regions. Every square centimeter on the map represents the same amount of surface area on the globe, regardless of where it is located. The trade-off is shape distortion.

The cylindrical equal-area projection was first described by James Gall in 1855 and later independently reinvented and widely promoted by Arno Peters in 1967. It became known as the Gall-Peters projection.

In a cylindrical equal-area projection, the latitudes are compressed toward the poles to compensate for the natural east-west stretching that a cylindrical projection introduces. The result is that all Tissot’s indicatrices — small circles drawn on the globe and projected onto the map — have the same area. Area is preserved. Shapes are not. Regions near the equator appear stretched vertically. Regions near the poles appear compressed horizontally.

The Gall-Peters projection sparked a major controversy in the 1980s. Peters claimed that his projection was the only equal-area alternative and denounced the Mercator as a tool of Eurocentric imperialism. Cartographers pushed back. The projection had not been invented by Peters. It had been described by Gall nearly a century earlier. Peters’s historical claims were inaccurate. The academic community published corrections.

The controversy revealed something deeper than a dispute over credit. It demonstrated that map design is not a neutral technical exercise. The choice of projection carries implicit messages about which regions matter. Peters understood this and used it as a weapon. His historical inaccuracies do not erase the underlying point: every projection encodes values.

The compromise: Robinson and Winkel tripel

Not all projections are committed to preserving a single property. Compromise projections intentionally trade strict mathematical preservation for a balanced visual appearance. They do not preserve area perfectly. They do not preserve shape perfectly. They distort both, but minimize the total amount of distortion across the entire map.

Arthur H. Robinson developed his pseudocylindrical projection in 1963. He approached the problem empirically rather than analytically. Instead of deriving the projection from a mathematical formula, he constructed it by minimizing visual distortion based on his own assessment of what a world map should look like. The result is neither conformal nor equal-area. Distortion is mild near the equator and increases toward the poles, where the poles themselves are rendered as elongated lines rather than points.

National Geographic adopted the Robinson projection for general-purpose world maps in 1988. It remained their standard for a decade. In 1998, National Geographic switched to the Winkel tripel projection, developed by Oswald Winkel in 1921. The Winkel tripel averages the coordinates of two other projections — the equidistant azimuthal projection and the natural log of the homolosine projection — producing a map that fares better than Robinson in several distortion metrics, particularly in the polar regions.

Compromise projections are useful when the viewer’s task does not require precise measurements. A wall map in a classroom, a general-reference atlas, or a news article illustration benefits from a map that looks “right” even if it is not mathematically exact. The viewer sees a world that approximates the globe without obvious warping. The cost is that no measurement on the map can be trusted without knowing the specific distortion properties of the projection.

The geometry behind the categories

Cartographic projections are typically grouped by the geometric surface used to generate them. A cylindrical projection wraps a cylinder around the globe and projects the surface onto it. A conic projection places a cone over the globe and projects onto the cone. A planar (azimuthal) projection casts light from the center of the globe onto a flat plane tangent to the surface.

These geometric metaphors are not the actual mathematics used by modern cartographers. Contemporary projections are defined by systems of equations that do not correspond to any physical light source or wrapping surface. The geometric classification remains useful as a teaching tool and as a way to understand the general distortion patterns of different projections.

Cylindrical projections tend to preserve shape near the equator and distort area near the poles. Conic projections are most accurate along one or two standard parallels and become increasingly distorted away from them. This makes them well-suited for mapping mid-latitude regions that extend primarily east-west, such as the continental United States. Planar projections are most accurate near the point of tangency and distort everything else, making them suitable for mapping polar regions or hemispheres.

What all projections share

Despite their differences, all map projections share a common structure. Each one is a function that maps points on a curved surface to points on a flat plane. Each one introduces distortion. Each one preserves at least one mathematical property.

The distortion can be quantified using Tissot’s indicatrix, a tool developed by French mathematician Auguste Tissot in the 1800s. An indicatrix is a small circle drawn on the globe. When projected onto the map, the circle becomes an ellipse. The area of the ellipse tells you how area has been distorted at that point. The shape of the ellipse tells you how shape has been distorted. The orientation of the ellipse tells you how direction has been distorted.

By drawing indicatrices across a map, a cartographer can visualize the pattern and magnitude of distortion. A conformal projection produces indicatrices that are all circles — shape is preserved locally, though area may vary. An equal-area projection produces indicatrices that are all the same size — area is preserved, though shape may vary. A compromise projection produces indicatrices that vary in both size and shape, but within bounded limits.

The indicatrix makes the abstract constraint concrete. Every ellipse is a lie. The question is which lie serves the viewer’s purpose.

Why this matters beyond maps

Map projections are not a unique example of unavoidable distortion. They are a mathematical demonstration of a general principle: any representation of a complex system in a simpler medium must sacrifice something.

A two-dimensional diagram of a three-dimensional object distorts depth. A written description of a conversation omits tone, gesture, and timing. A statistical model of human behavior compresses individual variation into aggregate distributions. A database schema of a real-world process simplifies continuous, overlapping reality into discrete fields and tables.

The pattern is the same as the map: choose what to preserve, accept what must be distorted. The choice is not a bug. It is a structural property of representation.

The map projections that have survived — Mercator, Gall-Peters, Robinson, Winkel tripel — endure not because they are true. They endure because they are useful for specific tasks. Mercator is useful for navigation. Gall-Peters is useful for comparing national sizes. Robinson and Winkel tripel are useful for general reference. Each one is a lie that tells the right kind of truth.

Recognizing that a representation is distorted does not make it useless. It makes it inspectable. A map whose distortions are documented can be used with confidence. A map whose distortions are hidden cannot.

What remains unresolved

The mathematical theorem that governs map projections is settled. Gauss proved that distortion is inevitable. No amount of computational power, design skill, or mathematical ingenuity can flatten a sphere without altering measurements.

The practical question remains open: which projection should be used for which purpose? There is no universal answer. A navigation chart requires different properties than a classroom wall map. A thematic map showing population density benefits from equal-area preservation. A map showing flight paths benefits from straight-line great circles. A map in a news article may prioritize visual familiarity over mathematical precision.

The choice is never neutral. Every projection favors certain measurements and sacrifices others. The favoring is often invisible to the viewer. A reader who has grown up looking at a Mercator map internalizes its area distortions as fact. A reader who has grown up looking at a Gall-Peters map internalizes its shape distortions as fact. Both are looking at the same Earth. Both are seeing a version of it that has been filtered through mathematical choices they did not make.

The unresolved question is not mathematical. It is epistemic: how does a reader learn to read a map with awareness of its distortions, and how does a mapmaker communicate those distortions to the viewer?

Most world maps do not include a projection statement. A map that says “this map uses the Mercator projection” would help a viewer interpret its distortions correctly. Most maps do not. The convention treats the map as a window onto the world rather than as a constructed representation. The convention is convenient. It is also misleading.

Primary sources

• Gauss, C. F. (1827). Disquisitiones generales circa superficies curvas. Commentationes societatis scientiarum regiae gestae, 9, 93-141. The Theorema Egregium proving that Gaussian curvature is invariant under isometric bending and cannot change when a surface is flattened.
• Mercator, G. (1569). Nova et Aucta Orbis Terrae Descriptio ad Usum Navigantium Emendate Accommodata. The original publication of the Mercator projection, designed for maritime navigation with straight-line rhumb courses.
• Peters, A. (1970). Atlas equatorial. Kassel, Germany: Wochenschau Verlag. Arno Peters’s promotion of the cylindrical equal-area projection and its role in the 1980s cartographic controversy.
• Robinson, A. H. (1974). A Map Maker’s Handbook: Projections of the Earth and Planets. Land University College, London. empirical construction of the Robinson projection by minimizing visual distortion.
• Winkel, O. (1921). Neue Erdkugelkarten. Astro-dynamisches Institut, Potsdam. Development of the Winkel tripel projection as an average of two existing projections.
• National Geographic Society. (1998). Map Collection: Winkel Tripel Projection. Adoption of the Winkel tripel as the standard world map, replacing the Robinson projection used since 1988.