Projection and Perspective

Imagine cutting an orange peel and pressing it flat. Tears and overlaps make the problem visible. Mathematical projection replaces the torn peel with a systematic transformation. The result is more predictable, not lossless.

Ptolemy’s Geography described methods for representing the inhabited world with a graticule. Renaissance printed editions made those instructions newly repeatable. Later projection families formalised different constraints as navigation, surveying, thematic mapping and global comparison asked different questions.

The chronology is not a march from primitive to correct. Innovations remain useful because tasks differ.

The map reader should therefore look for the scope of the promise. “Preserves area” means a thematic comparison of regions can be proportional; it does not mean coastlines retain familiar shapes. “Preserves angles” supports local bearing relations; it does not make a world map a transparent picture of continental size.

Gerardus Mercator’s 1569 projection makes rhumb lines-courses of constant compass bearing-straight. That property gave the map a powerful navigational logic. A mariner could plot a bearing as a line on the chart, though real navigation still depended on instruments, observation, corrections and local knowledge.

At world scale, Mercator greatly enlarges high latitudes. The familiar comparison between Greenland and Africa exposes this area distortion: their apparent sizes approach one another on many Mercator maps even though Africa is vastly larger. The observation is important, but the moral is often simplified into “Mercator is bad.”

A projection suited to one task can be irresponsible for another. Straight rhumb lines do not justify using Mercator for a population cartogram or a classroom comparison of continental area. Conversely, area distortion does not erase its mathematical purpose. Critique becomes stronger when it connects a property to a use.

Web maps revived a spherical form of Mercator because it supports a convenient square, tiled, multi-zoom system. At neighbourhood scale its local angle behaviour is useful; at a zoomed-out world view its area politics return. Software can switch projections or disclose distortion. Familiar infrastructure is a choice, not an alibi.

Equal-area projections change what comparison feels natural. Regions occupy surface in correct proportion to one another, making maps of land cover, population density or ecological change less likely to grant visual weight through projection inflation.

But area fidelity does not settle political interpretation. A choropleth still uses chosen boundaries. Large, sparsely populated units can dominate the page. A national average can conceal local variation. Colour classes can exaggerate thresholds. Projection solves one geometric problem while leaving statistical and administrative choices intact.

The Peters controversy made these questions public in the late twentieth century. Arno Peters promoted an equal-area projection as a corrective to Eurocentric world maps, though the underlying cylindrical equal-area family predated his version. The debate mattered because it revealed that projection choice carries cultural consequence. It also showed how easily a technical property can be marketed as a complete ethical solution.

There is no socially neutral geometry, but geometry is not reducible to ideology. The ethical practice is to name both: the measurable property and the institutional use.

Projection is usually described as a transformation of the Earth. Perspective asks where the viewer stands. A bird’s-eye city view combines elevation, oblique sight and selective enlargement so streets and monuments can be seen together. It may resemble a view from the air while exceeding what any single eye could observe.

The viewpoint produces authority. From above, parcels align, routes connect and crowds become patterns. The city appears available to inspection. Yet oblique views can also foreground civic identity, giving towers, walls and sacred buildings more space than strict geometry would allow.

Satellite imagery seems to complete the view from above, but it too is assembled: sensor bands, orbital paths, rectification, mosaicking, dates and cloud removal shape the image. “Perspective from space” is still a production pipeline.

The Theatre makes this visible. A map stages a world for a viewer. North-up, centred longitude, crop and zoom decide where the audience is placed. Changing the centre does not merely move objects; it interrupts the habit by which one view becomes the world.

Good projection practice begins with a sentence: this map needs to preserve what, for whom, at which scale? Navigation, cadastral measurement, continental comparison and polar climate analysis will produce different answers.

The next sentence names sacrifice. A conformal web map should warn against world-area comparison. An equal-area map should prepare readers for unfamiliar shapes. A centred azimuthal view should identify its privileged point. A local projection should publish its valid extent.

Interfaces can turn disclosure into exploration. A projection switch should not offer a decorative carousel of globes. It should state what changes: area, angle, distance, direction, interruption and centre. Reduced-motion and static readers need the same explanation in prose.

The goal is not a projection without perspective. It is a map honest enough to show its position. Distortion becomes informative when a reader can connect the shape on the page to the contract that produced it.