How to Lie With Maps

Jorge Luis Borges’ short story On Exactitude in Science points out the perfect map has a 1:1 relation to the territory mapped: 1 map inch to 1 territorial inch. It’s easy to imagine even more perfect maps at greater scale; someone interested in an aspen forest’s root system or the neocortex’s neural arrangement might want maps at 2, 5 or 10 map inches to the territorial inch. In our modern world of bits, such maps are easy to produce (when ignoring the cost of gathering, vetting, and transmitting enough detail to keep the maps accurate). In the old-fashioned world of atoms, however, perfection is impractical, and maps end up scaling 1 map inch to 10s or 100s of thousands of territorial inches. Such radical compression brings abstraction into cartography: selecting features of the territory — for example, those that from a distance look like flies — to be omitted from the map. And abstraction doesn’t stop with feature selection. On a perfect map each feature looks exactly like it does in the territory (when ignoring map medium), but under abstraction representational details have to be omitted.

These two forms of cartographic abstraction — feature selection and feature representation — are the subject of How to Lie With Maps, emphasizing the mischief that can arise when abstraction is incorrectly applied or incorrectly interpreted. The first third of the book deals with feature representations — the forms, shapes and colors used to represent features on a map — and the problem of projecting a large spherical terrain onto a small planer map. The middle third deals with feature selection, both in the general case and for specific map types, such as topographical maps or soil-type maps. The final third deals with the various ways abstraction can be tilted this way and that to emphasize features to be promoted or obscure features that are embarrassing. Although the book calls out lots of lying, it also covers the less malignant distortions resulting from the trade-offs arising from abstraction, as well as from the mistakes that must occur when humans engage in an activity as exacting and as detailed as cartography.

The most serious criticism to lay against this book is that it’s ugly. I think most people would agree that most maps are beautiful, but you’d be hard-pressed to be convinced of that by flipping through this book. The drawings, in addition to being poorly designed, are poorly executed on software that doesn’t seem much more capable than Mac or MS Paint. The map excerpts can be cut a little more slack because they’re illustrating cartographic malpractice, but even then you’d expect some beauty because that’s what greases the eyeballs, making it easier for them to slide off the under-emphasized and onto the over-emphasized. Maybe it’s unreasonable to expect quality at the level of Edward Tufte’s books, but it’s completely reasonable to expect something better than How to Lie with Maps delivers.

Monmonier is writing for map users, to alert and inform them about the distortions that can arise from cartographic abstraction. But it’s easy to see he also has his eye on what he calls lay cartographers, people who can generate maps using software running on their personal computers. The more he writes into the book, the more his attention turns to lay cartographers, warning them about questionable features of desktop cartographic software and urging them to iterate over map design and construction. All of this is interesting, but it tends to split the focus between map users and map makers, which is less helpful to map users, particularly towards the end of the book.