The Last Drafting Tables
In a narrow building on the Rue des Écoles in Paris, three floors above a boulangerie whose croissants perfume the stairwell each morning, Théodore Marchais bends over a drafting table that has not moved since 1961. The table is scarred with compass pricks and ink stains that form their own accidental cartography — a palimpsest of sixty years of work. Marchais is eighty-one. He is, by most credible estimates, one of fewer than forty people alive who can still produce a hand-drawn topographic map to professional survey standards.
He does not own a smartphone. He has never used GPS. When I ask him whether he worries about the future of his craft, he sets down his ruling pen with the deliberate care of someone who has learned that haste is the enemy of precision, and he looks at me for a long moment before answering. "Maps are not about where you are," he says finally. "They are about where you might go. The machine tells you where you are. That is a very different thing."
The distinction matters more than it might first appear. In the half-century since satellite navigation began its quiet colonization of human wayfinding, something has been lost that we are only now beginning to name. Cognitive scientists call it spatial agency — the felt sense of inhabiting a landscape rather than being located within it. Marchais calls it, more simply, knowing where you are.
"The machine tells you where you are.
That is a very different thing."
Encoding the Unmeasurable
Traditional cartography was never purely a technical exercise. The great map schools of the seventeenth century — the Dutch Golden Age workshops of Blaeu and Hondius, the French royal cartographers of the Académie des Sciences — understood that a map was simultaneously a scientific instrument and a cultural argument. The choice of what to include, what to omit, how to render the gradations of terrain, where to place the legend: each decision encoded a worldview.
Digital mapping has not escaped this subjectivity — it has merely obscured it. When Google Maps routes you around a neighborhood, or Apple Maps renders a park as featureless green, these are editorial choices made by algorithms trained on data that reflects the priorities of their creators. The difference is that Marchais's choices are visible, legible, arguable. The algorithm's choices are invisible, naturalized, unchallengeable.
"A hand-drawn map has a signature," says Dr. Yuki Tanaka, a cartographic historian at the University of Edinburgh whose recent monograph, The Authored Landscape, traces the philosophical implications of the shift from analog to digital wayfinding. "You can see the hand. You can see the decisions. With algorithmic mapping, the decisions are hidden inside the data, and the data is presented as neutral. But data is never neutral. It is always someone's data."
The Apprenticeship Problem
The crisis facing analog cartography is not merely philosophical. It is acutely practical. The skills required to produce a hand-drawn map — the ability to read a theodolite, to calculate magnetic declination, to render contour lines that communicate both elevation and the feel of terrain — take years to acquire and cannot be learned from a manual. They are transmitted body to body, eye to eye, across a drafting table.
Marchais has had three apprentices in his career. The first left after two years to work in digital cartography for a logistics company. The second emigrated to Canada. The third — a young woman named Camille Durand, now thirty-four — stayed. She is the only person he has fully trained. When I ask whether she will carry the work forward, he pauses again. "She is better than I was at her age," he says. "But she is one person."
From the Archive Selected works, 1923–2019
What We Lose When We Stop Getting Lost
There is a growing body of research suggesting that our increasing reliance on turn-by-turn navigation is measurably degrading our spatial cognition. A 2020 study published in Nature Communications found that people who regularly use GPS navigation show reduced activity in the hippocampus — the brain region responsible for spatial memory — compared to those who navigate by landmark and intuition. The effect is most pronounced in younger users who have never developed the habit of reading a physical map.
Marchais is not surprised by this. "When you draw a map," he tells me, "you must first walk the land. You must feel the slope under your feet. You must notice where the light falls in the afternoon, where the wind comes from, where the water goes when it rains. All of this goes into the map. When you read the map later, you are reading the land again. You are remembering with your body." He taps his chest. "GPS remembers for you. That is not the same as knowing."
As I leave his studio, Marchais is already back at his table, ruling pen in hand, drawing a contour line with the unhurried confidence of someone who has made this exact motion ten thousand times. The line curves around an imaginary hillside, encoding elevation, drainage, the probable location of a path. Somewhere, a landscape is being made legible. Somewhere, someone will hold this map and feel, for a moment, that they know exactly where they are.