Chapter 5

Updated 2026-07-02

"The classification of the constituents of a chaos, nothing less is here essayed." — Herman Melville, Moby-Dick, ch. 32, "Cetology" (1851)

---

There used to be a dream, and it was not a small one. The dream was that you could build a system of categories so complete and so exact that the name of a thing would tell you everything true about it — that classification and language could be the same act, so that to know a word would be to already know where it belonged in the order of everything that exists.

The dream was already ancient by the time anyone tried to engineer it. Aristotle had supplied the method two thousand years earlier: divide existence by kind, define each cut exactly, then divide again, until everything that is has one address in the order of things. A third-century philosopher named Porphyry drew the procedure as a picture — a tree, substance branching into body, body into living body, on down, limb by limb, to you — and for the better part of a thousand years, students learned what knowledge was by climbing it.¹ By the seventeenth century the tree had settled into the furniture of every educated mind: the obvious shape of truth, standing there with most of its branches still undrawn, waiting for someone with the patience to finish it.

In 1668, a London clergyman named John Wilkins tried to build it. Wilkins was the first secretary of the Royal Society, a man who spent his career around people inventing the modern idea of empirical proof, and he set himself a project stranger than any of theirs: a universal language in which the spelling of a word would encode its place in a complete classification of the universe.² He divided all of existence into forty top-level categories. Every word began with a syllable marking its category, then added a letter for the subclass, then a vowel for the species below that — a kind of taxonomy you could pronounce. De meant element. Deb meant the first of the elements, fire. Deba meant a portion of fire — a flame. Say the word and you'd said the shelf it sat on.

It is, on its own terms, a staggering piece of engineering. It is also, if you try to use it, the kind of system the world keeps refusing to fit. Wilkins needed a word for a particular kind of fish — small, scaly, reddish, river-dwelling — and his system obliged him with zana, a word that means exactly that and nothing else, built entirely out of its own classification. He needed a word for whale, and the system filed it, confidently, under category sixteen: "a viviparous, oblong fish."³ A scaly river-fish gets a beautifully exact name. A warm-blooded, air-breathing, live-bearing mammal gets filed as a fish, because Wilkins built his forty boxes around how a seventeenth-century gentleman expected the world to be shaped, and a whale looks, from a boat, exactly like a very large fish. The system isn't broken. It's working exactly as designed. The design itself was the mistake.

Three centuries later, the writer Jorge Luis Borges read about Wilkins's project and, instead of admiring it, used it to play a trick on every reader who came after him. In an essay about Wilkins, Borges produces — as if quoting a real source — a passage from "a certain Chinese encyclopedia" that divides all animals into categories including those belonging to the emperor, those that are embalmed, those that are trained, suckling pigs, mermaids, fabulous ones, stray dogs, those included in the present classification, those that tremble as if they were mad, innumerable ones, those drawn with a very fine camel's-hair brush, and those that, at a distance, resemble flies. It is one of the most quoted passages in the entire literature of classification — Michel Foucault opens an entire book with it — and it is, as far as anyone has ever been able to determine, something Borges made up.⁴ He credits it to "the unknown (or apocryphal) Chinese encyclopedist" himself, hedging in the very sentence that introduces it. The most famous proof that classification is arbitrary is a fabrication, performing its own point: invent a system absurd enough and people will still teach it in seminars, because absurdity is so much easier to recognize in somebody else's categories than in your own.

Borges wasn't only joking, though. In the same essay he describes a real institution doing, in earnest, almost exactly what his fictional encyclopedia does for laughs. In 1895, two Belgian lawyers named Paul Otlet and Henri La Fontaine founded an organization in Brussels with a genuinely staggering ambition: to catalog the whole of recorded human knowledge in a single numbered system, one slot for every fact in the world.⁵ They got partway there. Their scheme assigns the number 262 to the Pope. 263 is the Lord's Day. 282 is the Roman Catholic Church. 294 is a single shared shelf for Brahmanism, Buddhism, Shinto, and Taoism combined — four religions, one drawer, room enough for whichever happened to come up. And tucked into the system at 179 is a slot defined, in full, as: cruelty to animals; protection of animals; moral implications of duelling and suicide; various vices and defects; various virtues and qualities. That is not a typo or an aside. That is a real numbered category in a real classification, built by serious people, for the express purpose of containing whatever didn't fit anywhere else. Borges's verdict on the whole project, after laying out its numbers, is one word: it "resorts to chaos." He meant it as a criticism. It reads more like a diagnosis. Every sufficiently complete system, pushed far enough, grows its own catch-all drawer, and the drawer is where you can watch the system quietly admit what it can't do.

---

Long before Brussels, long before Wilkins, the world had already been demonstrating, methodically and without malice, how this goes.

In 1799, a sea captain shipped a strange specimen from the colony of New South Wales back to London: a furred, web-footed, egg-laying animal with what looked unmistakably like a duck's bill stitched onto its face. The naturalist who received it, George Shaw, did the obvious thing first. He took scissors to the seam where the bill met the fur, looking for thread.⁶ He assumed, reasonably, that he was holding a hoax — that some enterprising taxidermist had sewn a duck's bill onto the body of a mole or an otter to fool a credulous Englishman, the way mermaid skeletons were occasionally stitched together for sideshows. There was no thread. The platypus was real, and it spent the better part of the next century refusing to sit still inside the categories European science had ready for it: a mammal that lays eggs, with a bill, with venom, with a body that seemed assembled out of leftover parts from other animals' assembly lines. Nature hadn't read the taxonomy. It almost never has.

Nineteen years after the platypus reached London, a New York courtroom took up a stranger question, and the stakes were not zoological at all — they were financial. A whale-oil merchant named Samuel Judd had failed to pay the city inspection fee required on "fish oil." His lawyers argued, correctly, that whales are not fish. Whales are warm-blooded, they nurse their young, they breathe air through lungs — every fact a naturalist would put on the board pointed the same direction, and had for decades. The court in Maurice v. Judd heard testimony to that effect, and the jury, weighing the testimony against a century of everyday usage that had always called a whale a fish, sided with everyday usage.⁷ Whale oil was fish oil. The category was decided not by what a whale is made of but by what the law needed it to be — a fish, for purposes of an inspection statute, regardless of what it was for purposes of biology. It is the same move Wilkins made with his thirty-nine working categories and his one stubborn animal, run through a jury instead of a dictionary, with money instead of vowels at stake.

A few decades later, a sailor turned novelist staged the whole argument again, on purpose, as comedy. In the thirty-second chapter of Moby-Dick, Herman Melville has his narrator Ishmael pause the chase of the white whale to attempt something he calls, without irony, "the classification of the constituents of a chaos, nothing less is here essayed." Ishmael knows perfectly well that science had settled the whale's status — he even gestures at Linnaeus separating whales from the fish in print, decades earlier — and rejects the finding anyway. "Be it known that, waiving all argument, I take the good old fashioned ground that the whale is a fish, and call upon holy Jonah to back me."⁸ Having declared his allegiance to the folk category over the scientific one, with his eyes wide open, he then classifies the whale species he knows by an entirely different axis: book size. Folio whales. Octavo whales. Duodecimo whales — the largest sperm and right whales sorted next to the smallest porpoises, organized the way a printer would, because Ishmael is, helplessly, a man who thinks in books trying to describe an ocean. He never finishes the scheme. He leaves it standing unfinished on the page, the way, he says, the great cathedral at Cologne stood for centuries with the builders' cranes still parked on top of an unfinished tower — and calls the incompletion itself a kind of honesty.⁹ A system that admits it isn't done is telling you something a system that claims to be finished is not.

Two centuries after the platypus and nearly two after Ishmael's joke, the solar system did it again, this time by vote. In 2006, astronomers meeting in Prague faced a problem of their own making: better telescopes kept finding objects out past Neptune large enough to plausibly be called planets, and if all of them counted, the solar system was about to grow from nine planets to dozens, with no end in sight. The International Astronomical Union solved the problem the way the Brussels institute solved its 179th category — by inventing a new shelf. Pluto, and everything like it, would now be a "dwarf planet," a precise-sounding label invented specifically to hold the things the old definition no longer wanted to hold.¹⁰ The vote passed. Bumper stickers reading "Pluto is still a planet" appeared within the month. Nothing about Pluto had changed — not its mass, not its orbit, not a single fact a telescope could measure. Only the box had moved, and people grieved the box-move as if something real had been taken from the sky — because planet had never felt like a box to anyone. It felt like a fact, found rather than made, and the vote let the whole world watch a found thing get un-found by a show of hands.

---

None of this is a story about people being careless. Wilkins was not careless; he was one of the most rigorous minds of his century, and his system is in some ways more disciplined than anything that replaced it. Otlet was not careless; he and La Fontaine built an actual working card-index of the world's recorded knowledge, by hand, with millions of entries, decades before anyone had a machine to help. The platypus, the whale, Pluto — none of them were trying to embarrass anybody. The dream itself was the only flaw: that a system could be built once, complete, with a single right place for everything in it, and that the world would then hold still long enough to be filed.

It never does. A thing arrives that is partly a mammal and partly a fish by reputation; partly a planet and partly something the astronomers haven't named yet; partly her father's country and partly her mother's, on the same dinner table, in the same body, in a system built with exactly one slot per child. The categories aren't wrong to exist — they're wrong to assume they're finished. A man was about to build his whole career inside that same dream of total order — and another man, a few decades later, was about to take it apart in a library instead of a toy department.

---