We think we see the world objectively—yet every "map" has already chosen its center, its margins, and its silences for us. How does space enter the mind, and how does it shape us back?
2026 · Book Recommendations · No. 25
These four books each seize one mechanism in the human–space relationship—and they collide head-on. Brotton pries open the map itself: no world map is neutral; each paints one standpoint as "the objective world." Diamond swings to the opposite pole: he reduces the fate of civilizations to the geographical endowments of continents—environment, not race, decided who conquered whom. Yi-Fu Tuan turns inward: our love for a place (topophilia) is not about scenic quality but an emotional bond woven by the senses and cultural memory. Solnit walks to the map's edge: the things that matter most usually lie beyond every map, and the only way to find them is to get lost first. Finish this issue and you'll carry four different lenses for "space"—objective and constructed, determined and felt, known and unknown.
| Book | Author | Year | The one thing it nails |
|---|---|---|---|
| A History of the World in Twelve Maps | Jerry Brotton | 2012 | No world map is neutral—each paints the power and faith of the mapmaker's age as "the objective world" |
| Guns, Germs, and Steel | Jared Diamond | 1997 | The vast gaps between civilizations are rooted in continents' geographic endowments (axes, species), not racial intelligence |
| Topophilia | Yi-Fu Tuan | 1974 | Our love and attachment to a "place" is not objective environmental quality but emotion built by the senses and culture |
| A Field Guide to Getting Lost | Rebecca Solnit | 2005 | What matters most lies beyond the map—and the way to find it is to have the courage to get lost first |
Brotton is a historian of Renaissance cartography. He picks twelve world maps spanning two thousand years (from Ptolemy, the medieval Hereford Mappa Mundi, the Islamic geographer al-Idrisi, to Mercator's 1569 projection, to Google Earth) and dissects each, arriving at an unsettling conclusion: flattening a spherical Earth onto a plane makes it mathematically impossible to preserve area, angle, and distance all at once—so every world map "lies"; the only question is what it chooses to sacrifice.
What matters is not just technical distortion but that every trade-off carries a standpoint. What goes in the center, which projection, who gets enlarged and who shrunk—none of it is neutral. Mercator's projection keeps straight compass bearings for navigation, at the cost of grossly inflating the high latitudes—so Greenland looks as big as Africa, when Africa is in fact fourteen times larger. In the colonial era that inflation conveniently made Europe and North America loom huge. Brotton's refrain: a map disguises the power, faith, and commercial interest of its maker's age as "the objective world itself." Medieval maps put Jerusalem dead center with east on top, because they charted theology, not geography; today Google puts you at the center, because it sells advertising organized around you.
Hence his sharpest cut: a map's power comes precisely from its pretense that it is merely describing. The moment you accept a map, you accept its preset center and margins—without ever noticing that was a choice. We cannot do without maps; yet none can truly "represent" the world. That is not the flaw of one bad map—it is the fundamental paradox of maps as such.
Massive and dense; the choice of twelve maps tilts toward European and West Asian cartographic traditions, with comparatively little on Chinese or Indigenous-American spatial worldviews (despite the "world" in the title). Some chapters lean more on erudition than insight, reading like twelve standalone monographs whose through-line the reader must thread alone.
Brotton's mechanism applies directly to the AI era: every large model, every training set, every dashboard is a "map"—with its own projection, its own center, its own silences. An LLM "flattens" a corpus dominated by English and the Western web into one world map, where non-Western experience becomes a drastically shrunken "Greenland." One thing to try next week: take an AI output or team dashboard you rely on and run a "Mercator interrogation" on it—who is placed at the center? Who is enlarged, who shrunk? What does it leave unsaid (the silences)? Write it down, and you'll find that what you've been treating as "objective fact" is a map with a standpoint. For making technical decisions and judging an AI's trustworthiness, that step matters more than tuning one more parameter.
The book answers a question from Diamond's New Guinean friend: "Why is it you white people developed so much cargo and brought it to us, but we black people had little cargo of our own?" The old answer—racial differences in intelligence—Diamond calls both wrong and dangerous. His alternative is a thoroughgoing geographical argument: the tens-of-thousands-of-years gap between civilizations is rooted in the natural endowments of continents, not in the peoples themselves.
The mechanism has two links. First, the endowment of domesticable plants and animals. Worldwide, the large mammals humans could domesticate are few (cattle, horses, sheep, pigs, a mere handful), and high-yield wild grains cluster in a few regions. The Fertile Crescent happened to hold wheat, barley, sheep, and goats all at once—once farming produced a food surplus, it could feed non-farming artisans, bureaucrats, and soldiers, and writing, metallurgy, and standing armies followed. Sub-Saharan Africa and the Americas lacked that hand of cards—not because the people were inferior, but because the same domesticable species simply weren't there.
The second link is Diamond's most elegant insight—the orientation of continental axes. Eurasia stretches east–west: along one latitude, climate, daylight, and seasons are alike, so crops, livestock, and technology could spread along that axis with little obstruction. The Americas and Africa run north–south: getting from equator to temperate zone means crossing deserts, rainforests, and violent climate shifts, so maize domesticated in Mexico took thousands of years to inch north. The same invention spread into shared wealth across Eurasia, but in the Americas geography sliced it into mutually isolated islands. Guns, germs (from long cohabitation with livestock), and steel were all, in the end, snowball effects of that geographic starting point.
The price of geographic determinism is that it nearly erases the agency of culture, institutions, religion, and historical contingency—why the same geography yielded wildly different fates in different eras is hard for the book to answer. Anthropologists fault it for compressing tangled history into a straight "environment → outcome" causal line, and its explanatory power outruns its predictive power. Read it as a powerful lens, not the whole truth.
Diamond's mechanism is unusually practical for raising a school-age child: rather than pushing effort, design the "geography"—let good habits diffuse along a low-friction east–west axis like crops. How much a child does a thing often hinges less on willpower than on the environment's coefficient of friction. To try next week: pick a habit you want her to build (reading, piano, free drawing) and put the relevant thing within arm's reach, on the same "latitude"—the book on the sofa arm rather than the top shelf, the piano always open, paper kept on the desk; meanwhile move high-friction distractions (tablet, snacks) off to a faraway "different climate zone." You're redrawing the geographic axes of her living space so the diffusion you want is easy and the diffusion you don't is hard. That's closer to Diamond's real lesson than lecturing: structural endowment often decides the outcome earlier than individual effort does.
Tuan is a Chinese-American geographer and the founder of "humanistic geography." Where Diamond treats geography as a cold objective force, Tuan turns to the opposite pole: geography is first of all an experience that is felt and charged with emotion by people. He coined the word "topophilia," defined as the emotional bond between people and place.
His core argument: our love or aversion toward an environment is not a response to objective environmental quality, but something constructed by the senses, cultural memory, and personal history. The same mountain is four different mountains to a hunter, a painter, a geologist, and an exile. City-dwellers' romantic longing for "wilderness" exists precisely because they don't have to survive on it—those who actually wrest a living from the wild regard it with wariness. Landscape is never a neutral backdrop; it is layered over with our needs, fears, and imaginings.
Tuan traces, with great delicacy, how people use all the senses (not just sight, but smell, sound, touch, temperature) to translate space into "place"; how a childhood courtyard, a hometown dialect, a particular quality of light can elevate an ordinary location into an emotionally dense symbol. He also concedes honestly that topophilia is often not the strongest human emotion—many people are indifferent to the very environments that shaped their lives—yet once activated, it can make a single place carry the weight of an entire life.
A 1974 book, methodologically an erudite essayistic sweep rather than empirical research; it cites widely across times and cultures but offers few testable claims, and several of its generalizations about cultures and gender now read as broad and dated. The structure is loose—more a wandering than an argument. Yet as the founding work of "subjective geography," its perspective remains irreplaceable.
Tuan's mechanism hits a side of the "AI super-individual" path that's easily missed: as work moves to the cloud and attention is sliced up by screens, your emotional bond to concrete "places" is thinning—and that bond is precisely a root of depth and steadiness. A small experiment to try next week: for the work that most needs focus (writing, deep reading, thinking about your deep interests like consciousness or Buddhism), fix a single embodied "place"—the same desk, the same lamp, the same band of morning light—and do only this one thing there. Through repetition, let it grow into Tuan's "emotionally dense symbol": the moment you sit down, your body knows which state to enter. This is not bourgeois ritual; it's using topophilia to build a stable anchor for attention. A bonus gift for your child: take her to deeply experience one stretch of nature or one old place, over and over, so a real "homeland" can grow inside her.
Solnit is one of the most important essayists writing today. This is not a how-to guide but a set of meditations on being lost—extending geographic loss into a metaphor for epistemology and life. The book's guiding thread is a question she copied from a student, said to come from ancient Greece.
Her core claim runs against the grain: the thing whose nature is "totally unknown" to you is often exactly what you most need to find—and the way to find it is to get lost. This names a blind spot the "control space" logic of Brotton and Diamond can't reach—you cannot search a map for a destination you can't yet name. Every genuinely important discovery, transformation, and creation happens beyond the known boundary, in that territory without a map. Over-planning and demanding certainty everywhere is locking yourself inside the known, never to meet the unknown.
The book's loveliest image is "the blue of distance": why are far mountains blue? Because light at the blue end of the spectrum scatters and is lost in the air before it reaches you—the blue you see is the light that got lost. From this she makes blue the color of desire and distance: part of what we long for is beautiful precisely because it is far off, not yet in hand; the moment we draw near and grasp it, the blue recedes. Being lost, then, is not pure loss—it is also desire, possibility, and open space. Solnit asks us to revalue "losing" and "not knowing," to see them not as problems to be eliminated but as a necessary blank space in a life.
Lyrical and associative; its structure is a collage of private memory, art criticism, and nature observation, offering no actionable "method"—readers wanting clear argument or a to-do list will find it diffuse and over-scattered. The "guide" in the title is ironic: it can't teach you "how" to get lost, only infect you with a new way of seeing loss.
Solnit offers an antidote for the efficiency-and-certainty-minded technologist: deliberately reserve part of your time for "getting lost." The "Meno problem" is especially sharp for researchers—you cannot keyword-search a problem you can't yet name, so purely goal-driven, KPI-driven exploration is structurally bound to miss the truly new. To try next week: carve out a block of goal-free, output-free "lost time" each week—read a book entirely unrelated to your current project, walk a route you've never taken, let AI lead you off into an unfamiliar field, and forbid yourself the question "what's this good for?" This bears directly on your deep-interest terrain (consciousness, complexity, Eastern and Western philosophy): breakthroughs often come from unknown crossings, and the unknown cannot, by definition, be planned. Trade "I precisely completed my plan today" for the occasional question—"did I stumble into anything off-plan today?"
Test whether it's really a "map": can you state its "projection choice"—what it sacrificed to foreground something else? If you can't, you're still treating it as the world itself, not as a standpoint-laden representation. True media literacy is the ability to see, behind every act of "looking," the mapmaker who never appears.
Run a substitution test: change "he can't do it" to "if I stood inside his structural constraints, would I get the same result?" If the answer is "probably yes," then what needs changing is the geography, not the person—redesigning the friction and endowments of the environment often beats flogging willpower. But don't slide to the other extreme: geography sets the odds; it doesn't cancel agency.
Self-check with the "Meno problem": are the goals you're pursuing all drawn from a list you can already name? If so, you're probably spinning efficiently inside the known while shutting out the thing "whose nature is totally unknown, yet may matter most." Healthy exploration should include a patch deliberately free of KPIs, where getting lost is allowed—that is Solnit's "door into the dark."