What Is a Map? The First Technical Question

What Is a Map? The First Technical Question

Every map begins with a lie.

That is not a complaint. It is a description. The moment you draw a line on a piece of paper to represent a road, a river, or the edge of a continent, you have already done three things: you have shrunk the world, you have flattened its curves, and you have replaced its infinite detail with a small collection of marks. A satellite photograph, at first glance, looks less like a lie than a hand-drawn sketch — but it is no less a compression. It still chooses a scale, a projection, a palette, a resolution. The choices are different; the act of choosing is the same.

This chapter asks the first technical question of the book: what exactly is a map, and what technical compromises must every map make? It is a question that took cartography thousands of years to learn how to ask precisely, and the answer — that a map is a controlled, scaled, symbolic, two-dimensional representation of spatial phenomena — sounds dull until you realize what it forbids. It forbids the perfect map. Not as a failure of effort, but as a consequence of geometry and information theory. The very act of mapping is the act of trading away something. Understanding which trade-off you are making, and why, is the technical foundation on which every later chapter of this book rests.

We will start with a definition that is sharper than the dictionary. Then we will meet the four variables every cartographer must set — scale, projection, symbolization, and generalization — and we will use a flowchart to see how they fit together. After that, we will confront a piece of nineteenth-century mathematics, Gauss's *Theorema Egregium*, which proves that no flat map of a curved surface can be both conformal and equal-area at once. We will work through a hands-on exercise in which you draw a 1:5,000 map of your own bedroom and keep a distortion log. Finally, we will look back across thirty thousand years of mapmaking — from a stick in the dirt to a Landsat-8 satellite — and notice that the trade-offs have not changed. The technology has. The compromises have not.

By the end of this chapter, you should be able to look at any map, on any medium, made in any century, and answer four questions about it: how big is it relative to reality, how was it flattened, what signs it uses for what it shows, and what was deliberately left out. Those four answers define the map. They also define you, the reader of maps, as a participant in the same act of compression that has defined every cartographer since the Paleolithic.

1.1 A Map Is Not a Copy — It Is a Compression

Let us begin by getting a precise definition on the table. A map is a *controlled, scaled, symbolic, two-dimensional (or projected three-dimensional) representation of spatial phenomena*. Six adjectives — controlled, scaled, symbolic, two-dimensional, projected, representation — and each one is doing real work. We will unpack all six over the course of the chapter, but for the moment I want to draw your attention to the most important of them: *controlled*. A map is not whatever happens when you put geography on paper. A map is the result of a series of deliberate choices, made before the first mark touches the surface, about how reality will be reduced.

The cleanest way to think about this is as a problem in information theory, the branch of mathematics that Claude Shannon founded in 1948 with the paper *A Mathematical Theory of Communication*. Information theory is usually taught in the context of telephone lines and computer files, but it is, in disguise, the deepest theory we have about mapping. Shannon's central insight was that any message can be measured in *bits* — the smallest possible yes/no questions needed to specify it — and that a good communication system is one that packs as much meaning as possible into as few bits as possible, while tolerating as much noise as possible.

A map is a message. The world is the source. The map is the encoded version. The reader, looking at the map, is the decoder. The cartographer is the engineer who designs the encoder. The moment we see mapping this way, two things become obvious.

First, the world contains vastly more information than any map could ever hold. Consider a single square kilometer of real landscape. It contains the position of every leaf, every pebble, every ant, every photon of light bouncing off every surface. If you tried to record all of it, even at low precision — say, a 1-meter grid, with each cell holding a depth, a color, a temperature, a material type, and a timestamp — that one square kilometer would already require several gigabytes of storage. A single city block would be measured in megabytes. The entire land surface of the Earth, sampled at 1-meter resolution with 100 measurements per square meter, would require roughly $5 \times 10^{22}$ bytes, which is more storage than has been manufactured in human history. The world is not just big; it is informationally *unbounded*.

Second, a map must reduce that information dramatically to fit on a piece of paper, a phone screen, or the inside of your head. This is compression, and compression is the central act of cartography. Compression is what distinguishes a map from a photograph (which is largely uncompressed), a map from a 3D model (which preserves depth), and a map from a written description (which preserves narrative rather than geometry). Every map discards information. The craft of cartography is the controlled, deliberate, useful discarding of information.

There are two kinds of compression in information theory: lossless and lossy. Lossless compression, like a ZIP file, lets you recover the original data exactly. Lossy compression, like a JPEG image, throws away information that humans are unlikely to miss — and does so in a way that is irreversible. Maps are lossy compression. When a cartographer generalizes a coastline into

When a cartographer generalizes a coastline into a smooth curve, when a census map drops individual houses and keeps only districts, when a weather map replaces ten thousand station readings with a few isobars — the cartographer is doing exactly what a JPEG encoder does when it throws away high-frequency detail from a photograph. The difference is that the JPEG's standards are mathematically specified and invisibly applied, while the cartographer's choices are visible, deliberate, and politically consequential. Every discarded detail was once someone's home, someone's fishing ground, someone's sacred site. The compression is not neutral. We will return to this point several times in the book, especially in Chapter 7 on ethics and Chapter 12 on counter-mapping. For now, hold this idea firmly: a map is a lossy compression of the world, written by an engineer who decided which frequencies matter and which can be thrown away.

1.2 The Four Variables Every Map Must Set

To lose information deliberately, the cartographer needs four knobs. Twist them one way, and the map preserves shapes; twist them another, and it preserves areas; twist them a third, and the map favors some use case over another. The knobs are scale, projection, symbolization, and generalization. They are not independent, but for the sake of explanation we will treat them one at a time.

Scale is the ratio between a distance on the map and the corresponding distance on the ground. The conventional way to write a scale is as a unitless fraction: 1:5,000, 1:250,000, 1:50,000,000. The fraction's numerator is always one unit on the map; the denominator tells you how many of *the same* units that one unit represents in reality. So 1:5,000 means one centimeter (or inch, or any unit you like) on the map equals 5,000 of those units on the ground. A larger denominator means a smaller map of a larger area — that is called a *small-scale* map, and the language is counterintuitive at first but consistent once you rememb