Does Every Choice Create a Parallel Universe? Not Quite
Does Every Choice Create a Parallel Universe? Not Quite
Talking to my teenage son these days is a little like driving someone else's Porsche — one wrong move and the ride stalls out. He's in high school now, deep in that stage where almost any topic can go sideways. Marvel movies are one of the few roads that stay open between us.
We watched Doctor Strange in the Multiverse of Madness together not long ago. On screen, characters step between universes, running into other versions of themselves who made different choices at the same forks in the road. Somewhere in the middle of it, I found myself asking something I hadn't really thought through before: do parallel universes actually exist, the way physics understands the term, or is that entirely Hollywood's invention?
The conversation behind this article started on an ordinary movie night.
What "Parallel Universe" Actually Meant in 1957
Every version of the multiverse Hollywood sells opens the same way: press a button, tear a hole in space, step through. Real physics never made that move. The idea closest to a genuine "parallel universe" started as a narrow question about what happens when someone measures a quantum system, not as a plot device.
The question was older than the answer. In 1926, Erwin Schrödinger published the equation describing how a quantum wave function evolves over time, and it worked beautifully — except on one point. It said nothing about why a measurement appears to produce only one outcome. That gap sat inside the theory for three decades before anyone tried to close it by deleting the collapse rather than explaining it.
The best-known attempt — the one that later picked up the name many-worlds interpretation — starts with Hugh Everett III and a 1957 Princeton doctoral thesis. Rather than add a rule that collapses the wave function into a single result, Everett proposed doing without such a rule altogether. On his account the other outcomes are never eliminated; they simply stop being able to reach each other. Whether that amounts to separate worlds is a reading of Everett's math rather than something he wrote down — a distinction that carries more weight than it first appears to.
The paper itself reads nothing like a multiverse story. Everett's step was to treat the observer and the measuring device as part of the same quantum system as the thing being measured, rather than standing outside it. A detailed account at the Stanford Encyclopedia of Philosophy traces how that single change reframed the measurement problem. The argument reached print that July as "Relative State" Formulation of Quantum Mechanics in Reviews of Modern Physics.
What strikes me, reading about it now, is how little ambition the paper has. It's a mathematical formulation aimed at removing one awkward postulate, nothing more theatrical than that. Even the phrase most people reach for today came later, and from someone else entirely.
Illustration of a young researcher at Princeton in the 1950s, the setting where Everett wrote his thesis. Not a photograph of Everett.
A Theory Everett Never Named
Everett never called his theory "many worlds." Somebody else attached that label to it well after the paper was in print — and the label is what survived.
What happened in between looks less like a discovery and more like a manuscript passing through several sets of hands. Research compiled in the Stanford Encyclopedia's entry on the many-worlds interpretation traces how Everett's Princeton work was reshaped along the way. The long draft circulated in 1956 under the title Wave Mechanics Without Probability, reaching several prominent physicists, Niels Bohr among them. It did not go over well. Criticism from Bohr's Copenhagen colleagues fed back through Everett's advisor, John Wheeler, and the thesis Everett ultimately defended was a far shorter document called On the Foundations of Quantum Mechanics. The journal version that July carried the shorter heading it's cited by today. The word "worlds" appears in none of them.
Bryce DeWitt supplied it. In a 1970 Physics Today article he was already describing Everett's math as producing a reality made of many worlds, and three years later he and Neill Graham made the phrase permanent with an anthology titled The Many-Worlds Interpretation of Quantum Mechanics — sixteen years after Everett's own paper.
Sixteen years is a long time for a theory to go without the name everyone now knows it by, and I keep wondering what the idea would look like today if DeWitt had reached for a duller phrase. A label can travel to places the argument behind it never reaches, and that gap is roughly how a technical paper on quantum measurement ended up as a movie premise. It's part of why science fiction keeps returning to the multiverse trope decade after decade: the phrase is portable, the math isn't.
None of which makes the theory less real. It makes the word less reliable — because "many worlds" turns out to name just one of several ideas that all get filed under "multiverse," as though they were the same thing.
Three ideas that share the word "multiverse" and very little else.
Three Different Ideas Wearing One Word
Say "multiverse" to most people and they picture a single thing: a mirror version of this world, slightly altered, somewhere out there. Physics uses that word for at least three separate ideas, and laid side by side they don't line up at all. That matters for the branching question, because two of the three have nothing to do with choices — or with quantum mechanics — at all.
The cleanest way to keep them apart is to treat them as levels rather than rivals, which is how the physicist Max Tegmark sorted them in a 2003 Scientific American essay and in a later survey of the same hierarchy. Three of his levels are the ones that keep surfacing in popular coverage; a fourth, about mathematical structures, rarely leaves the seminar room.
| Idea | What Question It Answers | What "Another Universe" Would Mean |
|---|---|---|
| Many-Worlds Interpretation | Why a quantum measurement appears to produce one outcome | Every outcome persists, in its own branch of one universal wave function |
| Eternal Inflation | Whether the early universe's inflationary phase ended everywhere at once, or only in patches | Separate "bubble" regions of space, usually expected to be beyond our observable reach |
| Beyond the Observable Universe | What lies past the edge of what light has had time to reach us | Under standard cosmology, expected to be more of the same spacetime — not a separate reality, just out of view |
These aren't three competing theories of the same kind, which is part of why the single word causes so much trouble. One is an interpretation of quantum mechanics, one is a class of cosmological scenarios, and one is a statement about how far light has had time to travel.
The machinery underneath differs just as much. In most modern accounts, many-worlds branching is grounded in entanglement and decoherence. Eternal inflation's bubbles are supposed to come from the dynamics of an inflating field. The third row needs no new mechanism at all — only more distance than light has had time to cross.
That third one is the easiest to mistake for science fiction, when it's really just geometry and light-travel time, not unlike how Polaris still guides navigators today even though the star we see isn't quite where it appears to be. Distance and delay change what's visible. They don't require inventing a second reality.
One more thing is worth clearing up, because it's the part the movies lean on hardest: none of the three hands you another version of yourself as a straightforward result. Eternal inflation doesn't require that any given bubble universe contain a copy of a specific person — that depends on assumptions about physical laws and starting conditions which current models simply don't settle. The beyond-the-horizon picture can produce duplicates, but only if space is genuinely infinite and matter is scattered across it in a sufficiently random way, and neither of those is established. "Another me" follows from assumptions, not from evidence.
Which leaves the branching version, and the question I actually had while the movie was still running: does a new universe split off every time somebody makes a decision?
Branching, as the many-worlds interpretation describes it: records that stop being able to reach one another.
How a Universe Is Supposed to Split
The short answer is no — or at least not in the way the word "decision" makes it sound. Branching, in this theory, comes from a quantum system becoming entangled with whatever measures it. Choosing a college or a career isn't a measurement in that sense, and the theory says nothing about producing one clean new universe per choice.
There's a further problem with the picture, which is that even inside the many-worlds interpretation, "splitting" isn't quite the right word. Nothing tears, and nothing moves. What happens is closer to two conversations in the same room becoming too loud to overhear each other.
When a quantum system is measured, the system, the measuring device, and eventually the surrounding environment become entangled — correlated in a way that ties each possible outcome to its own record. As that entanglement spreads outward into the environment, the different outcomes lose any practical ability to interfere with one another. The many-worlds interpretation treats each of those diverging, non-interfering records as an equally real branch.
That is a narrower claim than the branching imagery suggests. Decoherence explains why the branches stop interfering for any experiment we could realistically run; it doesn't establish that they could never, even in principle, come back into contact.
And it leaves the harder problem untouched. Once large-scale alternatives have decohered, recovering any interference between them is expected to be effectively impossible, which means an experiment run inside one branch has no realistic way of registering the others. That's why many-worlds is generally understood to reproduce the same experimental predictions as textbook quantum mechanics rather than to offer a direct test for other branches.
So the quantum version of the question runs into a wall — not a wall of principle, as far as anyone has been able to show, but a practical one solid enough that it may as well be. That strikes me as a stranger place to land than a flat impossibility would have been. The cosmological version, meanwhile, looked for a while like the more cooperative one. It had left something behind that an instrument could actually be pointed at.
Artist's impression of ESA's Planck spacecraft, built to map the oldest light in the cosmos.
What the Oldest Light in the Sky Did and Didn't Show
This is where the story is supposed to turn dramatic, and instead it turns almost anticlimactic.
The instrument was the European Space Agency's Planck. It launched on May 14, 2009, and mapped the entire sky five times over the next four years, across nine microwave frequency bands. It was never designed to hunt for other universes. It was built to read the oldest light in the cosmos more precisely than anything before it.
Planck's final data release came in July 2018. The standard six-parameter cosmological model, known as ΛCDM, continued to fit the measurements extremely well.
What the data did not turn up was any confirmed signature of a bubble-universe collision, or any other observation that would establish a cosmological multiverse. That light is, in a sense, an old photograph, much like almost every star you've ever looked at — information about a far earlier moment, not a live view of anything happening now.
It's tempting to read a well-fitting cosmological model as evidence against anything exotic. That reading doesn't survive contact with what the mission actually did. Planck wasn't looking for other universes, but the question wasn't beyond its reach either: some models predict that an ancient collision between bubble universes would leave a faint, roughly circular pattern in exactly this kind of light.
Searches of that kind have been run. Feeney and colleagues published what they described as the first observational tests of eternal inflation using WMAP data, found no collision signature, and pointed to Planck as the dataset that could test the idea more decisively. No confirmed detection has followed. A null result is a null result, not a verdict in either direction.
The Planck Collaboration's 2020 analysis of inflation used that same final dataset to tighten the constraints on many inflationary models without touching the deeper question. On the quantum side, the argument over probability, branch independence, and what an observer should expect to see inside the many-worlds interpretation is still active — unresolved not because nobody is looking, but because looking hasn't settled it.
What has been tested, repeatedly and carefully, is the structure of the early universe, not the existence of anything past it. That's the honest state of the evidence in 2026: real mathematical reasons to take the idea seriously, and nothing yet that confirms it. I'd rather have that answer than a tidier one.
The Question Underneath the Question
This next part is just my own read on it, not a physics claim. If every person on Earth needed a parallel universe for every choice they didn't make, the picture would spiral out into something far stranger than the tidy branching universes the movies show. But the question that's stayed with me isn't really whether other universes exist. It's why we keep reaching for the idea in the first place.
Somewhere out there, a universe like ours, holding another version of me who made the choice I didn't — who's living out whatever came after. Religion used to answer that kind of question: this life isn't the whole story, there's a world we can't see, and the choices we make carry more weight than they appear to. Lately, it seems like we're asking the same question in a different language: the vocabulary of physics, the imagery of movies, instead of the language of faith.
Maybe a parallel universe isn't first a theory about the universe at all. Maybe it starts as something closer to home: the need, built into most of us, to imagine one more version of the self.
The question that stays with me isn't really about other universes.
Sources & References
- Stanford Encyclopedia of Philosophy: "Everettian Quantum Mechanics"
- Stanford Encyclopedia of Philosophy: "The Many-Worlds Interpretation of Quantum Mechanics"
- Hugh Everett III, "'Relative State' Formulation of Quantum Mechanics," Reviews of Modern Physics, 29, 454–462, 1957
- Bryce S. DeWitt, "Quantum Mechanics and Reality," Physics Today, 23(9), 30–35, 1970
- Bryce S. DeWitt and Neill Graham, eds., The Many-Worlds Interpretation of Quantum Mechanics, Princeton University Press, 1973
- Max Tegmark, "Parallel Universes," Scientific American, 288(5), 2003
- Max Tegmark, "The Multiverse Hierarchy," arXiv:0905.1283
- S. M. Feeney, M. C. Johnson, D. J. Mortlock, and H. V. Peiris, "First Observational Tests of Eternal Inflation," Physical Review Letters, 107, 071301, 2011
- European Space Agency, Planck mission overview
- Planck Collaboration, "Planck 2018 results. I. Overview, and the cosmological legacy of Planck"
- Planck Collaboration, "Planck 2018 results. X. Constraints on Inflation," Astronomy & Astrophysics, 641, A10, 2020
Comments
Post a Comment