Support Provided ByLearn More
Physics + MathPhysics & Math

The Many Worlds Theory Today

Half a century has passed since 27-year-old Hugh Everett III published a version of his Princeton Ph.D. dissertation in a leading physics journal, introducing the scientific world to his radical theory of parallel universes. In what ways did the theory break from existing theories of the day? How has it fared in those five decades, and where does it stand today in the physics community? Journalist Peter Byrne, author of the biography The Many Worlds of Hugh Everett III, answers these and other questions in this interview.


Hugh Everett

Receive emails about upcoming NOVA programs and related content, as well as featured reporting about current events through a science lens.

Hugh Everett in 1964, when he was 34 years old and was working at the Pentagon. He had abandoned quantum physics, to which he would never return.
Courtesy Mark Everett

Granting permission

Support Provided ByLearn More

NOVA: What is the Many Worlds theory?

Byrne: The Many Worlds theory basically gives physicists permission to think of the entire universe as quantum mechanical. That's what Wojciech Zurek, a renowned quantum physicist at Los Alamos lab, told me. And that is a real break with how physicists had been thinking of the universe ever since quantum theory got started in 1900. For about 50 years, until Hugh Everett came along, physicists divided the universe into two different worlds. One was the indeterministic microscopic world, where elementary particles fly around. Two was the deterministic macroscopic world, which is the world of our experience, where objects are large, where cause and effect are linked; in physics, this is called the "classical" world.

The quantum world builds the classical world. Everything in the classical, macroscopic world is composed of microscopic particles acting in unison. But in quantum physics, for 50 years, the only way that physicists could interpret or think about the work they were doing was to say that anything that happens in the microscopic world only has meaning in terms of how it is looked at as a large object. We cannot even talk about what happens in the microscopic world, because it is so indeterministic that we can never lay our fingers on what is actually going on.

Everett broke with that. He said—and he wrote the mathematics to back it up—that we can look at the entire universe as quantum-mechanical. We do not have to have an arbitrary division between the classical and the quantum, an arbitrary division that exists because people had no other way to explain the results they were getting. For many years before Everett introduced his theory, people had thought about this problem, but Everett was the first to propose a logically consistent way of removing the barrier.

cover of Nature
A half century after he shared his radical idea with the world in the July 1957 issue of a leading physics journal, Everett and his Many Worlds theory finally get their day in the sun. Here, the cover of Nature, July 5, 2007.
Reprinted by permission from Macmillan Publishers Ltd: [Nature] 5 July 2007

Everett's argument that the universe is quantum-mechanical is logically indisputable in and of itself. It's an "interpretation," which is a fundamentally different animal than what physicists call formalism. Formalisms are mathematical devices that show you how to operate experiments but that do not need to pop up any kinds of meanings beyond simply If you do A, you get B. Interpretation tries to tell you why If you do A, you get B. Everett was telling you why macroscopic large objects emerged from this microscopic quantum world.

How did he do that?

Well, the device he used to do that—and I don't want to get too technical here, but it's hard not to use this term—is a universal wave function. A wave function is basically just a mathematical list of every possible configuration of a quantum object, like a hydrogen atom. A universal wave function lists every possible configuration of every single elementary particle in the universe. And there are a lot, so you can't actually write it down! The way you symbolize a wave function is with the Greek letter psi. It's kind of like a U with a stake going down through the middle. The first time he ever saw this symbol, Everett's son Mark said, "What's that little devil's pitchfork?" It does look like a little devil's pitchfork.

So Everett came up with this universal wave function, which is just [Austrian physicist Erwin] Schrídinger's equation for describing how elementary particles move around writ large—that is, applied to the whole universe! And it makes beautiful mathematical and logical sense. Actually, it's very much in use in physics today. However, it has consequences to it that people were and remain uneasy with, which basically is that everything that is possible happens. This assertion, which Everett backed up mathematically, solves, according to him and his supporters, the so-called measurement problem, which is kind of the "dirty little secret" that has been afflicting quantum mechanics since it was invented and then formalized in 1920s.

Peter Byrne
"Before I started looking into this, I would have thought it was crazy," Peter Byrne says of Everett's theory. "Now I wouldn't be surprised if it's true."
Courtesy Peter Byrne

Dirty little secret

What is the measurement problem?

It has to do with the only other technical term I'll use if I can get away with it—the concept of superposition. Think of a superimposed photograph, a piece of photographic film that has been exposed several times so you have overlapping images. That's an analogy for superposition in quantum mechanics. If you think about an elementary particle like an electron, before you look at it, it could be at any number of positions in the device that you've got it trapped in.

Now, the wave function that describes all of the positions that that electron can be at does not say that any one of those positions is more or less real than any other position. They have an equal reality. That is, before you actually measure the electron, it could show up in any of the places the wave function allows it to. When you measure it, though—when you interact with it, when you observe it—it takes one position. We see only one position. In quantum mechanics, however, there's only a probability that it will have a certain position, so if you measure it again in exactly the same way, you might get another position.

"Wild as it sounds—a person splitting into numerous copies of herself—Everett's theory has not been shown to be mathematically incorrect."

Say the first time you measure it, the quantum-mechanical formulas pop up a 30 percent chance that the particle is at position X. The next time you do the measurement, exactly the same way, you might get a 70 percent chance that the particle is at position Y. That means that there's a 100 percent chance that it is at either X or Y. Not only is it a chance, but according to quantum mechanics, it actually is at both X and Y before you measure it. But when you measure it, it's at either X or Y, and if you measure it a million times, 30 percent of the time it's going to be at X and 70 percent of the time it's going to be at Y. This is the measurement problem nobody has ever been able to explain.

The dirty little secret.

Right. Many, including Everett, have claimed, however, that they could explain why you only get one result in our classical world when you interact with a superposed quantum object. But quantum systems composing even a gram's worth of carbon contain trillions of atoms that are quantum-mechanical objects all interacting with one another, so it's just mind-boggling to think of how many possible configurations a gram of carbon could be in.

"Mind-boggling" sounds like an understatement.

It is. If you consider the universe as a whole, and you use Everett's universal wave function to list every single possible position that every single particle in the universe could possibly have in all of time, you have this huge superposition that describes the entire history of the universe.

Now, according to Everett, you have superpositions in macroscopic objects as well as microscopic, like a gram of carbon or the tip of your pencil. Those classical objects are composed of microscopic systems that are all in superpositions, but we don't see 100 million positions for a gram of carbon or the tip of your pencil, we only see one.

The founders of quantum theory, people like Niels Bohr and Werner Heisenberg and Paul Dirac and John von Neumann and others, were faced with this problem back in the 1920s. It was inexplicable using any kind of formula that they could come up with up, so they postulated—that is to say, they decided arbitrarily—that what happens is that the wave function "collapses."

Becoming jellyfish

This is the part of the Copenhagen Interpretation that Everett had a problem with, this so-called "collapse" of the wave function from the many to the one?

Right. The Copenhagen Interpretation basically says "Don't ask, don't tell." It says, "We don't know why, so we're going to do this: We're going to say that what happens is that the wave function loses all of those other possibilities and collapses into only one possibility." They couldn't prove it; it remains a postulate. It works to actually describe the universe that we see, and it works to consider that the wave function is collapsing when you're building quantum-mechanical devices. But it doesn't make any philosophical or logical or interpretive sense.

Everett, who was around Bohr in 1954 and talking to other people around Bohr—as well as to his own mentor, the famous physicist John Wheeler, who died this summer after a long, glorious career in physics—Everett looked at this and likely thought back to something Schrídinger had said a few years before in Dublin. Schrídinger had said physicists fear that if we don't have the collapse, "We should find our surroundings rapidly turning into a quagmire, or sort of featureless jelly or plasma, all contours becoming blurred, we ourselves probably becoming jellyfish."

What he meant by that was that, if we don't collapse it, then all these possibilities are going to start propagating, and there won't be any cause or effect anymore that we can follow from one event to another, and our selves, our physical beings, will start to become duplicated, and every possible position that a human body can be in will suddenly exist in classical reality.


Schrídinger looked at that and said "Ouch," precisely, and dropped that idea and went off into mystical explanations that didn't make any sense whatsoever.

Is this where philosophers, who you've said have given a lot of thought to Everett's theory in recent years, come in?

I'll get to them in a moment, but it's important to realize that Everett, like all good physicists, did not give theories of consciousness any magical powers in quantum mechanics. Because of the intractability of the measurement problem and several other similar paradoxes in quantum mechanics, some people, especially philosophers, have been attracted to the idea that human consciousness collapses the wave function. That human consciousness is the major actor in the universe, and that without human consciousness, the universe would not exist. Physicists like Everett who are materialist and realist thought that was bunk. They think human consciousness is a quantum-mechanical system like any other quantum-mechanical system. Personally, I agree with that.


Okay. Go on.

So Everett, coming from this point of view in which consciousness was not "privileged," in which he had a paradox that had a mathematical formula backing it up that was not solved by the prevailing notion of a collapse of the wave function, started to calculate using information theory, which had just been invented in the post-war years by Norbert Wiener, Claude Shannon, and some others. We say we live in the Information Age. Well, it was pretty much born in about 1948, when Shannon and Wiener and others put forth some remarkable theories that said that information has a physical reality independent of any kind of meaning that you might want to give it. And on the basis of that analysis—that information is physical—all modern technology has come into being.

And Everett knew about this work?

Everett was very up on cutting-edge ideas like that when he was writing his thesis in 1954, 1955. He took the basic analysis, that information is physical, and developed a mathematical argument showing how data correlates within itself. That is, what happens in a superposition is that the person looking at a gram of carbon that exists in a superposition of a billion different places at once does not collapse the wave function. The Schrídinger equation never ends, including in the classical world.

In order to demonstrate the consequence of this mathematically, Everett came up with a solution showing that the observer, the human being, correlates with every possible state that that gram of carbon, that pencil tip, could be in. So before the human being looks at the gram of carbon, the carbon is in all the millions or billions or trillions of possible states, and after the human looks at the gram of carbon, he or she is in one state. In Everett's theory, what happens in between, as it were, when the human actually looks at the carbon—or a clock or any other object—is that he or she splits like an amoeba. (The act of looking, that interaction, is just exchanging energy. A person looking at a clock, for example, is an energetic interaction, with photons of light bouncing off the clock and going into the person's eye.)

So, in Everett's view, when the human correlates herself—that is, interacts, exchanging energy with the gram of carbon or a clock or whatever—she splits like an amoeba. She splits into copies of herself, one for each element in the superposition. [See Everett's draft of his never-published amoeba analogy.]

And this split is what creates the "many worlds" of his theory?

Yes. And wild as it sounds—a person splitting into numerous copies of herself—Everett's theory has not been shown to be mathematically incorrect. God knows, people have tried. They have found some mathematical gaps, but you can't fault his basic mathematical logic, which made a powerful case that every time there is an interaction anywhere in the universe above a certain size, one of the systems splits in order to accommodate all of the elements and the superpositions that are contained in the wave function that describes the observed system. In other words, the basis for having multiple universes emerges from his solution of the measurement problem.

"If there's a wave function, there's a collapse. They don't want to think about it much further than that."

What about Schrídinger's "jellyfish" problem?

Good question! How did Everett get around Schrídinger's fear that we would all become jellyfish, like some kind of blobs of ourselves, walking down the street as if in a photographic superimposition? How did he get away from that? Well, Everett showed mathematically that there would be no contact between the copies, that each copy when it correlated to an element of a superposition in the object it was observing would in effect then go off on a track that was a completely separate universe from the other copies that were doing the same thing, correlating with other elements of the superposition. They would all be going off within their separate universes.

So this, in a rather big nutshell, is the Many Worlds Interpretation.

Yes, that's the Many Worlds theory. (Everett, by the way, didn't call it the Many Worlds theory. He called it the "'Relative State' Formulation of Quantum Mechanics.")

But the upshot of the Many Worlds theory is that this universal wave function describes a series of branching universes that make up what [physicist] David Deutsch calls the "multiverse," and that in these branching universes, there are beyond trillions of copies of you, of me, of Everett. There are branches in which Everett is still alive. There are branches in which we did things that we don't want to talk about, that we may have thought about but we never did. Well, guess what? I'm sorry to tell you, in Everett's theory, you actually did it, because everything that is physically possible happens in some branch of the multiverse.

Multiple theories

I'll have to be more careful. Not everyone buys his theory, though, right? The Many Worlds Interpretation has detractors.

Well, people do not necessarily agree with the mathematical conclusions that he came to. Some people I've been talking to recently, very high-powered quantum physicists and mathematicians in Europe and the United States, think that he made some elementary mistakes concerning probability in his theory. But they don't think that those mistakes are fatal to it, and so you have a situation today in which there are several competing interpretations of quantum mechanics. Everett's is one of them; it's on the front burner. There are some others, the interpretation of David Bohm, for instance, which basically is a much more classically oriented interpretation but which also makes a lot of sense when you think about it.

Do any scholars in the field today just dismiss Everett's theory outright?

Yeah, there are lots of people who just dismiss it outright. I saw a book a couple years ago by the famous science writer Martin Gardner. The title of it was Are Universes Thicker Than Blackberries?, and it was like a diatribe against any type of multiple-universe theory, of which there are many. Everett's is not the only one. There are at least four different types of multiple universes that are relevant and considered seriously in physics.

People spotted weaknesses in Everett's argument early on, however. Bryce DeWitt certainly spotted it when he was looking at it carefully in the early 1970s. Here was the problem: If all possible physical events occur, then what happens to probability? We're making a measurement that says that there's a 30 percent chance that this electron is at position X, but if we believe in a many-worlds theory, we believe that actually there's a universe in which it's at every single possible position it could ever be at. So how do we assign a probability value to it? It's a very deep question that philosophers in particular have been struggling with a lot in the last 20 years especially.

Did Everett acknowledge this weakness?

Everett claimed to have made an argument that proved that the standard probability measure came naturally out of his theory. Most people who seriously parsed his work think he made a mistake and that it doesn't emerge that naturally. People attempting to improve his theory have been trying to develop methods of showing that you can have an explanation for why we think there's probability in a universe in which everything that is possible happens. In the little branch that you think you're in, whoever "you" are, if you do a quantum measurement and you get 30 percent probability but, in fact, you live in an Everett universe, then why did you get a 30 percent probability?

Well, it's tied to the question of, "If everything exists in superpositions, why do we not see all objects existing in superpositions?" We live in a classical world. It has an arrow of time that goes in a certain direction. It has entropy, which is related to probability and information. And in order to make sense of the Everett theory, you really do need to explain why we think probability exists, at least in "our" branch.

Is this what has drawn philosophers in?

Yes. There are a lot of arguments that have been made along those lines, especially by philosophers at Oxford University. Simon Saunders, David Wallace (working with David Deutsch), and many others have kind of an Everett school over there. Last year they sponsored a conference called "Everett at 50." It was celebrating the 50th anniversary of his thesis publication, and physicists and philosophers of great renown from all over the world went there. I was fortunate enough to go, and they even asked me to talk about Everett there, which was fun.

For days they debated this question of probability in Everett, and a related weakness, which is that if you have these branches splitting off constantly every nanosecond all over the universe, going in all different directions, how does one universe, one branch, link itself to all these different states so that a coherent, single branch in which my history, say, Peter Byrne's history and life, which I remember as being singular, I remember not being born as a thousand people but as one person, how does that emerge? These are difficult questions that pertain to philosophical topics that have been discussed for centuries.

"People didn't attack his theory publicly, because it's very hard to attack Everett's logic."

And what about the Copenhagen Interpretation. Is that still viable?

A lot of people still hold with the Copenhagen Interpretation. If there's a wave function, there's a collapse. They don't want to think about it much further than that. Some people would say that there's no measurement problem whatsoever; they just legislate it out of existence. And there are various classical theories based on random processes that claim to explain how our universe emerges from this multiplicity of possibilities. But Everett's theory is the simplest of the lot, and it does not modify the Schrídinger Equation, which is the basic law governing quantum mechanics.

So while Everett's is not the only interpretation out there, it was the one featured on the cover of Nature in July 2007, celebrating the 50th year since the publishing of Everett's dissertation, because it has had this huge impact on modern physics. [See Everett's full dissertation, published online for the first time.] Even though nobody has ever been able to prove or disprove the existence of multiple universes, it has been useful in a number of areas of physics.

Putting to use

What areas?

Well, it's been very useful in cosmology, for instance, because the universal wave function gives you a method of calculating quantum-mechanical structures in, say, the beginning of time, at the moment of the Big Bang, without having to stand outside of the universe to do these calculations.

This was another huge flaw in the Copenhagen Interpretation and its collapse-of-the-wave-function postulate, because by saying that while we can't explain what happens to the superposition, and we know that in our classical world we have only one measurement result, we have to postulate that in order to get that one measurement result, we stand outside of the quantum object. So what Bohr and the rest of them were saying is that, when you're making a measurement, you the physicist are a classical object, the measurement is a quantum system, and when you make a measurement, the classical world trumps the quantum world. What it predicates is that whenever you make a measurement of the quantum system, you have to be external to it. You have to be outside of it.

What Everett did was he showed how you could make measurements and be inside of the quantum system; that the observer, the scientist, could consider herself to be a quantum-mechanical object, correlating with another quantum-mechanical object, that could calculate probabilities—which is what quantum mechanics does; the only thing it does is calculate probabilities—without having to stand outside of the system created by observer and object observed.

Now, in cosmology this is really important, because if you want to understand the early state of the universe, the inflation state where quantum mechanics is very orderly, you can't get external to it. You've go to be able to calculate from inside the wave function that describes the entire universe. So one of the greatest uses of Everett's theory today is in cosmology, not just in a technical sense, but in an interpretive sense, because if you're a cosmologist who wants to understand the universe from an understanding also that, as a person, you're inside the universe—because how could you be outside of it?—then this gives you a perspective from which to view that whole universe that you're trying to comprehend.

Whew, my head is spinning. What other applications?

Another application of it is in quantum computation. Quantum computation relies on what are called qubits, quantum bits, a few very elementary examples of which have been created in the laboratory. Although quantum computation itself is probably a ways off, it has been experimentally demonstrated that you can keep an information-processing device, in the form of an electron for example, in superposition, and in those superpositions you can process information and, through a complicated process, pop it out at the other end.

People who are building quantum computers don't necessarily have to believe that there are multiple universes. But they are faced with working with these quantum qubits that exist in what can easily be described as multiple universes. If they're not that, then nobody has any other way of describing how they're situated. And if you're David Deutsch, who's been one of the founders of the science of quantum computation, you will look at this situation and you'll say that it's proof that Everett's theory is correct. In fact, David Deutsch has said that quantum mechanics itself is proof that there are multiple universes, although he thinks of them in a more sophisticated way than Everett did.

Anything else? How else has his theory proved useful?

Well, I have to use a third technical term—don't be scared—called decoherence. It's called decoherence because that's the opposite of coherence, and coherence is a useful term in quantum mechanics because it can be applied to that electron that's existing in a quadzillion different positions, in a superposition of multiple states. We call that a coherent state. It evolves through time. It keeps all these positions that it could be in evolving as if they were separately evolving without having to collapse or anything like that, and it is not until it interacts with objects in the quantum environment, a human being for instance, that it collapses or decoheres.

"Quantum mechanics is the most successful physical theory in the history of humankind."

What that means is that, using Everett's analysis as a starting point, physicists like Zurek and Dieter Zeh at the University of Heidelberg, and James Hartle and Murray Gell-Mann and others, have developed a theory of quantum mechanics basically called "decoherence theory." It's not interpretation, but rather a technique that, while not exactly solving the measurement problem, does explain how the classical world can emerge from the quantum universe. Some decoherence theorists think there are multiple universes; some think that there's only one. All of them will tell you that they were inspired to go along the path of developing this theory because of having Everett's universal wave function as a useful tool.

A difficult birth

To step back a moment, initially Everett's theory wasn't well accepted when it was first published in 1957, is that right?

No, it wasn't. First of all, when his dissertation was printed in 1957, it was highly edited from his original version. All the colorful language was taken out. But physicists looked at it and a lot of them thought, "This is crazy." [Physicist Richard] Feynman went on record as saying, in essence, "Well, this is not possible because there can't be multiple universes."

However, people didn't attack his theory publicly, because it's very hard to attack Everett's logic. They did attack it privately. For instance, in 1956, before it was published, Wheeler and Everett sent a copy of the dissertation to Bohr in Copenhagen to see if he would agree that it was true. It wasn't likely that he would, because if he did agree he'd have to admit that he'd been wrong for decades about everything else.

As it happened, Bohr was pretty polite. He didn't attack it himself, but he assigned his acolytes to attack Everett, and actually for decades they took every opportunity they could to say that Everett was stupid, that his theory didn't work, that Everett didn't understand quantum mechanics, stuff like that.

Without backing it up with any solid mathematics.

No. All that they would do to back it up is to say that Everett couldn't be right because the wave function obviously collapses—but then nobody could ever prove that the wave function collapses. Nobody has ever seen a wave function collapse. On the contrary, people actually have seen almost-macroscopic systems called mesoscopic systems exist in quantum superposition. Buckyballs, for example, have been observed in superpositions. So what we have is experimental proof that large systems can exist in superpositions, just like microscopic quantum systems, and no proof whatsoever that the wave function collapses.

It seems remarkable that such brilliant people would create the wave-function collapse out of whole cloth just because they couldn't figure out anything else.

Actually, it makes perfect sense, and whole books and doctoral dissertations have been written on this, because, yeah, to the ordinary person, that would seem like a cop-out, but put yourself in their position. First of all, these European physicists had just broken with three centuries of classical physics that had been developed since the days of Isaac Newton, who said that the world is deterministic, that if you know the initial positions or conditions of any system at one point you can calculate where it's going to be after a certain passage of time.

When quantum mechanics came along, you had this problem where you could only calculate with probabilities, where you could not say that a quantum system existed in a certain position before you look at it, because you can't. All you can say is that it exists in a distribution of possible positions, and so you have the problem that the quantum-mechanical world is indeterministic. Our classical world is largely deterministic, and collapsing quantum indeterminism into classical images is the only way we can describe the quantum world, Bohr said, because we must use what he called "ordinary language."

Nonetheless, Bohr said, we have to be honest and admit that indeterminism is a basic force in the universe. We just cannot talk about it, you see, because it's inexplicable. So we have to postulate that the world we see is the only real world. It was John von Neumann who invented the mathematics of wave function collapse in the early 1930s, and Bohr went along with it.

But not everybody agreed with that.

Einstein, Schrídinger, and others wouldn't agree with that. They were attracted to a deterministic worldview that included probability. In Everett's theory, of course, everything that is physically possible happens. You have no room for probability, because everything happens. And yet the science of quantum mechanics is based on calculating probabilities.

"I think that's one of the reasons he went to the Pentagon—they had the best computers."

But when the founders of quantum mechanics, including Niels Bohr, were looking at their new, beautiful theory back in the '20s, they realized that getting one world out of many was a problem, but they couldn't explain it any other way except to say what was in front of their eyes, which is we see one world. We know that the quantum-mechanical world seems to exist in many, many possible universes, although they didn't put it exactly that way. But we have to be true to our own experience and say that somehow it gets from many events to the single event. And when they invented their postulate—which was not so much a mathematical construct as it was an interpretive, philosophical construct—it allowed them to use quantum mechanics.

In fact, there's nothing wrong with the postulate in terms of hurting the ability of the physicists to do their work. It enables them to do their work. But, as Everett said when he was being attacked privately before his thesis was printed, he said the Copenhagen Interpretation is a "monstrosity," with one reality for the quantum world and another for the classical. Many, many people agreed with him over time, and many people actually agreed with him before he said it but were afraid to say it because Bohr was a very powerful figure in the history of science.

A quantum-mechanical world

When you say it doesn't hurt their ability to do their work, you mean that you don't have to understand exactly how quantum mechanics works for it to be useful?

That's right. Quantum mechanics is the most successful physical theory in the history of humankind. Something like 70 percent of all the industrial capacity in the world is based in some way on quantum mechanics. It's used in your cell phone, your computer, GPS devices, lasers, anything that's digital and electronic, even in biology. DNA, for instance, can be dealt with on a quantum-mechanical level. And the science, the mathematics of quantum mechanics, is phenomenally successful in predicting what will happen if you shoot electrons at certain targets or try to make cell phone signals error-free.

To give you an example, in your television you've got a cathode-ray tube that shoots electrons at a screen. According to quantum mechanics, there is a chance that one out of every 137 electrons that you shoot out of that tube will go where you want it to go—which means that 136 of them are just going to be lost. Quantum mechanics shows you how to set up a device so that you can get a coherent picture using only one out of every 137 electrons streaming out of a cathode-ray tube. Billions and billions of them are just streaming out every second, so that's plenty.

It's all based on probability.

Right. Basically, if you know the probability that a particle is going to behave in a certain way, even if it's only one out of every 137 times or one out of every 10 million times, you can set up devices that will capitalize on that. And we do.

However, we cannot tell you why that happens. Not only do we not really know what probability is, we do not understand the fundamental motions in the microscopic quantum-mechanical universe. We do not understand how our world emerges from this universe.

So this, again, is the measurement problem that was bugging Hugh Everett.

This was bugging Everett, and one of the reasons that he wrote his thesis was to solve it. There were other reasons, but that was one of the primary reasons, because it intrigued him; it was a paradox. This guy loved solving paradoxes and puzzles, and this was like the ruling paradox of science, and yet nobody wanted to talk about it. In fact, Bohr's Copenhagen Interpretation basically told people Thou shall not talk about it, because you're not going to be able to solve it, and if you start thinking about it, you won't get any work done. Murray Gell-Mann said something like, "Generations of physicists were raised not to ask questions about the foundations of quantum mechanics," and it's true.

I've talked to any number of experimental physicists, and these guys are not philosophically oriented. They're very interested in getting results by manipulating elementary particles in certain ways—say, to make a quantum computer—but to do this, they almost have to take a dualistic attitude towards what's going on in the elementary particle world. Don Eigler at IBM told me recently, "When I look at an electron from a distance, it's a particle. When I look it up really close, it's a wave." Electrons and all elementary particles behave dualistically—as waves sometimes and as particles sometimes—depending on the environment that they're in, and your point of view.

"Everett also had to live with the fact that there were millions and billions of universes in which the nuclear wars that he was designing took place. No wonder he drank."

These are questions that experimental theorists have been essentially taught in school do not concern them. They're issues of philosophy, they're told, and there's a certain wisdom in that, because if people were puzzling over unsolvable problems, nobody would want to go into that line of work. Look at [Nobel Prize-winning mathematician] John Forbes Nash. In her biography of him, A Beautiful Mind, Sylvia Nasar says that what tipped him into schizophrenia was trying to solve the measurement problem.

Goodbye to all that

Why did Everett utterly abandon quantum mechanics and go into weapons development? My understanding from the NOVA film is that he was rebuffed by Niels Bohr and others right from the start, and it was so depressing to him that he gave up.

Well, that's true, but there were other factors involved. The long and short of it is that Everett came from a military family. His father was a colonel with tremendous logistical talents, and Hugh went to military school in high school. And his whole generation, many of them just back from World War II, were sent to college by the newly formed National Science Foundation, which was dedicating itself to educating thousands of people to be able to work in the military-industrial complex, especially in research and development of weaponry. So Hugh was sent to Princeton via the National Science Foundation.

His mentor there was John Wheeler, who was one of the inventors of the hydrogen bomb—he'd invented it the year before he met Everett—and he was a huge shaper and player in the military-industrial complex. Princeton was a center of military research, and Everett, it turns out, had a bent for doing military work. He was never that excited about working in academia, because military work, especially if you started doing it in the private sector, which he did after working at the Pentagon for a few years, paid a lot better than academic work.

Everett also didn't trust academia. Here he had come up with this remarkable idea that, 50 years later, is one of the most powerful ideas in physics, acknowledged by physicists everywhere in publications. And in his day either people attacked it viciously because they had their own kind of dogmatic pursuits to protect, or they didn't want to talk about it. I think the idea of working in academia kind of repulsed him after that.

So when did physicists start paying attention to the theory, and why? How old was Everett at that point, and was he able, before his death, to bask in the light from these guys?

Yeah, he was, in a way. He was 27 in 1957 when his thesis was published in Reviews of Modern Physics, and the editor of that issue was a cosmologist named Bryce DeWitt. Initially, DeWitt was put off by Everett's theory. He wrote to Wheeler and Everett that, if the universe is splitting, then why don't I feel myself split? Everett wrote back to him, Well, Copernicus made the analysis that the Earth was moving around the sun, undoing thousands of years of belief that the sun was going around the Earth, and people asked him, If the Earth is moving around the sun, then why don't I feel the Earth move? And DeWitt, who was well aware of the Newtonian reasons why they wouldn't, said "Touché" and then forgot about the theory for a while. [See Everett's original letter in response to DeWitt's concerns.]

In the late '60s, when he was working seriously in quantum cosmology, DeWitt was attracted to the universal wave function as an interpretive method of dealing with what was going on, and he started writing about it. In 1970, he published an article in Physics Today that set off a fairly intense series of letters back and forth in that publication, debating the theory. Then, in 1973, he went to Everett and said, "I know that there's a longer version of your thesis than the one that I printed in 1957. I'd like to publish it." It was 137 pages, whereas the thesis that was published in 1957 was about nine pages. [See the full dissertation.]

John Wheeler had made Everett cut three quarters of his thesis. Wheeler had had this dream that Bohr was somehow going to approve it, so he made Everett remove his direct attacks on the Copenhagen Interpretation as well as his provocative metaphors about splitting observers and bifurcating cannonballs (and, for some reason, a whole chapter on information and probability theory).

So a lot of the explanation of things that people considered to be weaknesses in Everett's theory were cut out of the version that people read. In 1973, DeWitt published the long version, along with the short version and some other papers, including one by himself, in a book called The Many Worlds Interpretation of Quantum Mechanics. He used the phrase "many worlds," because he thought it would be provocative and catchy, and it was, and Everett was pleased.

So did Everett get involved again in quantum mechanics at that point?

Everett didn't get involved in any debates about it, but he followed what was going on from afar. He corresponded with a few people, saying he still believed his theory was true. As the 1970s rolled on and more and more interest was taken in it, DeWitt and Wheeler, who were both at the University of Texas in Austin by then, invited Everett to come down and give a seminar on his theory. (David Deutsch was a young grad student there at the time.) So Everett packed his entire family into his car and drove from Virginia down to Austin and smoked like three packs of cigarettes during the seminar and was just really, really pleased to be there.


Even after he gave up on quantum mechanics, Everett did a lot of useful work, right?

Oh yeah. Mostly he spent his time writing groundbreaking computer programs. He invented databases that gave the theoretical foundation for relational software like Oracle and PeopleSoft, stuff like that. Other people later took those ideas and found a way to make millions—but that's another story. He invented an algorithm called the Everett Algorithm that is still in use today. It's a method of maximizing use of resources. You can also use it to design anti-ballistic missile systems.

"I wouldn't be surprised to find that Everett's theory is true, and I'm not going to say that it's not."

Everett also wrote one of the classic military game theory papers of all time in his first year as a grad student at Princeton. In fact, it's such a remarkable paper that when one of the founders of game theory, Harold Kuhn, put out a book 10 years ago on the greatest of all game theory papers, he included Everett's paper. Everett's game theory work, his work in logic with the algorithm that he invented, his work in quantum mechanics, his work in developing software—all these things are still impacting science and computation today.

Computers became his obsession.

Everett loved computers. I think that's one of the reasons he went to the Pentagon—they had the best computers. And he was put in charge of computers at the Pentagon when he was 27 years old. He was doing research and making policy recommendations that seriously shaped United States nuclear war strategy. For instance, he worked for the Weapons System Evaluation Group for eight years, from 1956 to 1964, and one of the things he was in charge of was writing a famous memo called WSEG #50. This memo advised President Kennedy as he was coming in on issues of strategic nuclear warfare, and basically it was the foundation of the next 15 or 20 years of nuclear weapons development and strategy. Everett was involved in designing software that would play war games, that would simulate nuclear wars and political crises, and he was deeply involved when the Cuban Missile Crisis, for instance, came up.

People turned to technocrats like Hugh Everett to design programs that would give them options. So interestingly enough, in his work as a military operations researcher, Everett's specialty was looking at alternatives in different situations. Given that his quantum-mechanical theory said that everything that is physically possible happens, and he believed in it, he also had to live with the fact that there were millions and billions of universes in which the nuclear wars that he was designing took place. No wonder he drank.

Byrne's take

You seem to have a pretty good grasp of what many of us would consider rather difficult material. Do you have a physics background?

No. In fact, I flunked out of algebra in high school. I heard about Hugh Everett from a friend of mine who is a physicist. I ended up writing a profile for Scientific American and getting a book contract with Oxford University Press to do the biography. [Editor's note: The book, The Many Worlds of Hugh Everett III: Multiple Universes, Mutually Assured Destruction, and the Meltdown of a Nuclear Family, appeared in 2010.] One of the great things that happened was that I hooked up with Everett's son Mark, and we uncovered Hugh's papers in Mark's basement, which we're still going through. It's a treasure trove.

I would imagine. What have you found down there? It must be a biographer's holy grail.

Oh, it is. It couldn't be better. I would have liked to have found evidence that Everett was doing more quantum mechanics after he left Princeton, but I think the evidence is to the contrary. He doesn't seem to have ever done any more quantum mechanics, which is a loss to physics.

Clearly you've learned a lot about quantum mechanics yourself during your research.

Yes. As far as I'm concerned, quantum mechanics should be for the masses. Basic elements should be taught in elementary school. I find it appalling that basically it's been co-opted by the military-industrial complex and the university complex, as it were, for 50 years or more. It has its own inaccessible jargon, and the experts, wonderful as they are, meet and discuss things that only they can understand.

Why do children need to know about quantum mechanics?

Well, they might want to participate in building the next round of computers, for one thing.

[laughs] My question wasn't entirely flip. I meant: Why are you for people learning about quantum mechanics so early in life?

Well, whenever it's age appropriate. I'm not an educator, and it's true my six-year-old son is in kindergarten, and I'm not teaching him any quantum mechanics or telling him about multiple universes, because I don't want him to be confused.

[laughs] Right.

What I'm saying is that there comes a time in education when quantum mechanics should be taught to kids before they graduate from high school, in my opinion, and it generally isn't. It's not made accessible in the curriculum to students in a broad way. I'm not mathematical, but I read these equations and I can follow them to a certain degree. I can follow the arguments. And I'm engaged in e-mail conferences with physicists and philosophers all over the world about this issue. They've been very helpful to me. They know I don't understand the math, but they're determined that I will understand the concepts. Between that and reading 150 books and scores of papers, I've maybe learned too much jargon, though.

You've generously kept it out of our discussion. One final question: What's your personal take on Everett's theory? Do you believe it?

Before I started looking into this, I would have thought it was crazy. Now I wouldn't be surprised if it's true, and if I had to bet on it, I'd probably do a 50-50. I mean, I'm not in a position to read quantum mechanics in its formalistic presentation in such a way that it could convince me. Some physicists and mathematicians who do quantum mechanics are convinced, and others aren't.

I think the arguments against Everett hold some water, but they're inconclusive as well, and I see that Everett's theory has had a material and positive affect on the development of science. It would be kind of crazy to say that the universal wave function is true but the rest of it isn't, because you really can't have one without the other. So I just have to say I wouldn't be surprised to find that Everett's theory is true, and I'm not going to say that it's not.

Editor's Notes

This feature originally appeared on the site for the NOVA program Parallel Worlds, Parallel Lives.

Major funding for NOVA is provided by the David H. Koch Fund for Science, the NOVA Science Trust, the Corporation for Public Broadcasting, and PBS viewers.