by On my grandfather’s porch in North Loup, Nebraska, the ground kept its promises. Summer evenings the floorboards held the day’s heat, the corn stood in rows that ran clear to the horizon, and nothing in all that flat immensity so much as trembled. When a teacher first told me the Earth was spinning at close to a thousand miles an hour while racing around the Sun at sixty-seven thousand more, I pressed my bare feet against those boards and waited to feel the ride. The stillness came back like a verdict. Whatever the books claimed, my soles reported a stationary world, and a boy trusts his soles. Every child who has run that experiment has rebuilt, in miniature, the strongest objection the seventeenth century could throw at Copernicus. If the Earth turns, why does a dropped stone land at the foot of the tower and never a few feet to the west? Why does the gale of our orbital speed fail to strip the leaves from the trees? The objection deserved respect because it rested on evidence, on the honest testimony of every human body that had ever stood still. It took Galileo’s ship, and then Newton’s laws, to explain why that testimony, though honest, was worthless.
I have been listening lately to David Albert, the Frederick E. Woodbridge Professor of Philosophy at Columbia University, on the campus where I took my own graduate training in another art. Albert earned his doctorate in theoretical physics at Rockefeller before crossing over into philosophy, and his 1992 book Quantum Mechanics and Experience remains the clearest door into quantum strangeness I know. He refuses the two standard vices of quantum talk: the mysticism that inflates the weirdness into wellness-seminar fog, and the shut-up-and-calculate gruffness that pretends the weirdness away. In a conversation I have been replaying, he walks from the bare experimental facts to Hugh Everett’s many-worlds picture, and along the way he retells an analogy I want to hold up for you, because the analogy carries the whole argument, and because I believe it works right up to the moment it stops.
Albert teaches superposition with a contraption he calls a two-path apparatus, and you can build it in your head. Electrons carry a measurable property he nicknames hardness; test any electron and you find it hard or you find it soft, one or the other, every time. Route incoming electrons through a splitter that sends hard ones down an upper rail and soft ones down a lower rail, let the rails bend back together, and reunite them in a device that merges the paths while measuring nothing. Feed the machine electrons prepared in a second property Albert nicknames color, every one of them white, where white happens to guarantee fifty-fifty odds on any hardness test. Then ask the only question a sane person can ask: while the electron crossed the apparatus, which rail was it on?
Every candidate answer dies on the experimental table. Say it took the upper rail: an electron known to be on the upper rail comes out fifty-fifty white and black when you test its color afterward, yet electrons leaving the intact apparatus come out white a hundred times in a hundred. Try the lower rail, and the arithmetic fails again in mirror image. Claim both rails, and detectors stationed on the paths refute you, finding one whole electron on one rail or the other, with half-electrons appearing nowhere in the record. Claim neither rail, and the detectors refute you again, since they always find it in there somewhere. Albert’s conclusion, delivered with the calm of a man reading a verdict, holds that the electron enters a condition all its own, one the equations describe, in which, as he puts it, “there fails to be a fact of the matter” about its location in space. Superposition names that condition. Physics had discovered a new way for a thing to be.
The trouble refuses to stay small. The Schrödinger equation, the law governing all of this, is linear, which means the strangeness spreads by contact. Let a measuring device interact with an electron in superposition, and the mathematics hands you a device in superposition, a pointer with no fact of the matter about where it points. Let a physicist read the pointer, and the mathematics hands you a physicist in superposition. Erwin Schrödinger dramatized the contagion in 1935 with his cat, hung between living and dead by a poison vial slaved to a quantum trigger, and the joke has outlived every reader who first laughed at it. Physics has offered three sober escapes. Amend the equation so superpositions collapse on their own, as the Ghirardi-Rimini-Weber theory does. Supplement the equation with hidden machinery, particles holding definite positions all along, as Louis de Broglie and David Bohm proposed. Or believe the equation as written, follow the mathematics wherever it walks, and swallow what it says about pointers and physicists. The third road belongs to Hugh Everett III.
Everett laid that road out in the mid-1950s as a Princeton doctoral student under John Archibald Wheeler, and the manuscript’s fate should interest every author who reads me. The full dissertation ran long and pulled no punches. Wheeler, who admired the work and feared Niels Bohr’s disapproval in equal measure, made Everett cut it to a fraction of its length and sand off its combative edges before publication in Reviews of Modern Physics in 1957. Copenhagen shrugged anyway. Bohr’s circle treated the theory as a confusion and its author as an annoyance, and Léon Rosenfeld, Bohr’s lieutenant, would later pour contempt on Everett in private correspondence. The uncut manuscript sat in a drawer for a decade and a half. I have spent a publishing life watching gatekeepers trim authors down to sizes the establishment can ignore, and I recognize that story’s shape from a long way off.
The objection at the center of today’s essay arrived before the paper was even printed. Bryce DeWitt, the physicist who had organized the North Carolina conference whose proceedings carried Everett’s work, wrote to protest that the theory had to be wrong on the plainest ground available: he could not feel himself split. Everett’s reply, folded into a note added to the paper in proof, reached back three centuries for its ammunition. Critics of Copernicus had insisted the Earth could not move because nobody felt it move, and that criticism collapsed once Newtonian mechanics demonstrated that the theory asserting the motion also entailed that the Earth’s passengers would feel nothing. Such arguments fail, Everett wrote, once it is shown that “the theory itself predicts that our experience will be what it in fact is.” DeWitt conceded in a single word, touché, and then spent the rest of his career as the theory’s chief evangelist. He gave it the name it wears today, many worlds, in a 1970 article for Physics Today, and in 1973 he and Neill Graham finally published the uncut dissertation.
Albert retells the exchange with relish, and he pulls the parallel tight. Newton’s laws, F equals ma joined to universal gravitation, entail that the Earth hurtles around the Sun; the same laws entail that creatures riding the Earth in free fall would register no hint of the hurtling. The Schrödinger equation entails that a measurement splits observer and observed into branches; the same equation entails that an observer confined to one branch can never detect a sibling. Albert grants the idiom of splitting worlds only with hedges attached, calling the talk approximate, vague, permissible by degrees, and I will come back to those hedges, because they matter. First the logical shape of the defense deserves a name. A theory stands accused of predicting something enormous that nobody has ever witnessed, and the theory answers the charge out of its own resources, deriving our blindness from axioms already on the table. Call it the alibi inside the equation. No auxiliary assumption gets hired. No epicycle bolts on. The shield comes free with the sword.
The move carries a pedigree older than Newton. When Aristotelians argued that a stone dropped from a tower on a turning Earth should strike the ground well west of its release point, Galileo answered with the ship. Drop a ball from the mast of a vessel gliding across a calm harbor and it lands at the mast’s foot, since the ball shares the ship’s motion before it ever leaves your hand. Motion held in common turns invisible from inside. The 1632 Dialogue built that observation into a principle, and Newton’s laws gave the principle its mathematics. By then the pattern had its full structure: the accusing observation gets predicted by the accused theory, and the prosecution’s star witness turns out to testify for the defense.
An alibi this convenient should make any reader suspicious, because the woods are full of theories that explain away missing evidence. In 1857, two years before Darwin published, the naturalist Philip Henry Gosse argued in a book called Omphalos that God had created the world recently with fossils already planted in the rock, growth rings already inside the trees, and every appearance of deep age installed at the factory. Gosse’s hypothesis also predicts that our experience will come out as it does, and the book earned the contempt of believers and geologists alike. So what separates Everett from Gosse? Economy, and the location of the shield. Gosse purchased his immunity as a separate acquisition, an extra assumption whose only job was to eat the evidence against him. Everett’s immunity was already sitting in the till. The linearity that generates the branching is the linearity that seals the branches, and the sealing carries a technical name, decoherence, worked out in earnest by H. Dieter Zeh around 1970 and developed by Wojciech Zurek across the decades after. Once a measurement’s outcome leaks into the environment, into trillions of air molecules and a torrent of stray photons, the delicate phase relationships that would let two branches interfere are scattered past all practical recall, like a drop of ink stirred into a river. Interference between macroscopic branches stays possible in the mathematics and hopeless in the world, for the same reason an egg never unscrambles.
Those hedges of Albert’s live here. The worlds of the many-worlds interpretation earn their plural gradually. Before decoherence completes, talk of separate branches misdescribes the physics; after it completes, the branches behave as two whole histories, each deaf to the other, and the plural becomes, in Albert’s phrasing, more and more permissible. David Wallace, whose 2012 book The Emergent Multiverse gives the modern view its fullest defense, treats worlds the way biologists treat species: real patterns without sharp borders, as real as a hurricane and as impossible to timestamp. Ask for the instant when the world split and the Everettian will answer that the demand belongs to the wrong genre, like asking for a storm’s first gust. I report that answer as the honest state of the art, while noting that specialists still argue over how much decoherence settles, and I flag its cost: an ontology of approximations underneath a slogan about whole, self-sufficient universes. The poetry runs ahead of the bookkeeping, and Albert’s hedges tell me he knows it.
Everett himself never stayed to argue. Rebuffed during a single disastrous visit to Copenhagen in 1959, he left academic physics for the Pentagon’s Weapons Systems Evaluation Group, where he applied his mathematics to the logistics of thermonuclear war, and prospered at it. He chain-smoked, he drank, and he died of a heart attack in 1982 at fifty-one, still convinced he had been right. At his own instruction, his ashes went out with the household trash. The man who multiplied the universe asked to be discarded in this one, and I have read few colder sentences about the price of going unheard.
Literature, as usual, arrived early. In 1941 Jorge Luis Borges published The Garden of Forking Paths, whose fictional novelist constructs a book in which every possible outcome of every event occurs, each spawning further forkings without end. Everett, so far as anyone can tell, never read it. The physicists who came to the theory through DeWitt’s article walked into a garden a municipal librarian in Buenos Aires had mapped three decades before.
Here the analogy begins to strain, and the strain teaches more than the analogy does. Newton’s alibi was porous, and porous on purpose. Uniform shared motion hides; rotation leaks. The theory that excused our senses also issued a list of places where the Earth’s motion should peek through for anyone with instruments fine enough: a planet slightly flattened at its poles, trade winds bent sideways by the spin, starlight displaced by our orbital rush, nearby stars shifting against the deep background as we swing around the Sun. One by one the leaks were caught. James Bradley detected the aberration of starlight in the 1720s. Friedrich Bessel clocked the parallax of the star 61 Cygni in 1838. In 1851 Léon Foucault hung a heavy bob from the dome of the Panthéon on two hundred and twenty feet of wire and let Paris watch the plane of its swing rotate while the planet turned beneath it. The moving Earth graduated from a defended inference to an instrument reading, and the alibi expired because the case had closed.
Ask that graduation of the many worlds, and the theory refuses. By its own accounting, a branch that has fully decohered stays sealed for the remainder of time; thermodynamics turns the lock and swallows the key. The analogy is effective because the shield comes free with the sword, and no ad hoc hypothesis ever gets hired. As a promise of progress, though, the analogy is not effective, because Newton’s shield came stamped with an expiration date while Everett’s carries none, and an alibi that no sighting can ever test wins acquittals forever without once proving innocence. Everettians answer that the leaks exist upstream of decoherence, and that every interferometer on Earth works as a Foucault pendulum in miniature, catching superposition in the act before the environment slams the door. Their catches keep growing: whole carbon-60 molecules interfered with themselves in a Vienna laboratory in 1999, and by 2019 the record had reached molecules of roughly two thousand atoms. Collapse theories of the Ghirardi-Rimini-Weber family predict that superposition must break down somewhere on the road to macroscopic size, so every heavier molecule tightens the noose around collapse. David Deutsch adds a taunt from computing: run Shor’s algorithm on a mature quantum computer to factor a number hundreds of digits long, and the machine will exploit more computational pathways than the visible universe holds atoms. Where, Deutsch asks, was that number factored?
I accept every one of those results, and the gap remains. Interference experiments confirm superposition, and superposition belongs to every serious interpretation on the menu. A Bohmian reads the two-thousand-atom experiment without blinking; the pilot wave threads both paths while the molecule rides one. Foucault’s pendulum discriminated Copernicus from his rivals. The interferometer, so far, discriminates quantum mechanics from its proposed amendments and stays mute in the quarrel between Everett and Bohm. Our evidence keeps confirming the equation; it has yet to count the worlds.
A harder objection waits behind the popular one, and Albert has spent much of his career sharpening it. Probability. Quantum mechanics earns its keep through the Born rule, which converts the mathematics of superposition into betting odds: this detector fires with probability seven in ten, that one with three. If every outcome occurs on some branch, the sentence loses its meaning, since a thing that happens on every run of the experiment has no business carrying odds of thirty percent. Newton owed nobody a theory of gambling; his alibi never touched the subject. Everett claimed the statistics emerge on their own, and three generations of Everettians have labored to make the claim good: Deutsch through decision theory in 1999, Wallace through a book-length derivation, Charles Sebens and Sean Carroll through the self-locating uncertainty of a freshly split observer who cannot yet tell which branch she inhabits. Albert has answered the whole program with patient demolition. In 1988 he and Barry Loewer proposed a many-minds variant expressly to give the probabilities somewhere to live, and he has argued in the decades since that the mainstream derivations smuggle in the conclusions they claim to derive. On my reading those derivations remain not effective as answers, because each assumes an agent who already cares about branch weights, which is the Born rule wearing a bow tie. And here sits the irony: the man giving Everett’s analogy its most generous public reading has doubted for most of his career that Everett’s theory can pay its probability bill. That is teaching of a high order, the steelman built before the strike, and it shames the cable-news style of argument that flatters its own side first.
So what do we make of the argument? The live options in the foundations of quantum mechanics have held steady for decades: amend the equation and buy real collapse at the price of new physics nobody has detected; supplement it with Bohmian particles and buy definiteness at the price of machinery hiding beneath relativity; take it neat with Everett and buy elegance at the price of uncounted worlds and unexplained odds; or shrug with Copenhagen and buy nothing at all. My recommendation, for whatever an old dramatist’s judgment is worth in a physics quarrel, is to accept the Newton analogy at its honest size. It succeeds at the job Everett built it for, which was demolition. It takes the question my ten-year-old self asked on that porch, why don’t I feel it, and converts the question from a refutation into a prediction, using no parts the theory had not already paid for. Few maneuvers in the history of physical argument work that cleanly. Demolition, though, clears ground without raising a single wall. The worlds still owe us a parallax, or a frank admission that none is coming, and they owe us a theory of odds besides. Stop asking why you cannot see the branches; the equation answers that one. Ask why your Geiger counter clicks to the beat of the Born rule when the theory says every rhythm plays at once. That question stays open, and open questions are where physics keeps its future.
I think again of those Paris crowds, watching a brass bob swing while the marble floor crept beneath it, the planet confessing at last after three centuries of alibi. Some evening a demonstration that plain may arrive for the worlds next door, though the equation itself keeps whispering that the wire snapped the moment anyone looked. Back on my grandfather’s porch the floorboards have gone cool, the corn stands unmoved against the horizon, and my feet still report one stationary world. I believed my soles once and was wrong about the Earth. Whether I am wrong about everything else they report, no pendulum now swinging can say.
Leave a Reply