Six Impossible Things: The Mystery of the Quantum World

by John Gribbin

In 2008, Pozzi and another group of colleagues took a step further. They developed an experiment in which electrons could be fired one at a time through two genuine, nano-sized physical slits in a thin screen, to be detected on the other side in the usual way. As expected, the electrons arriving at the detector screen built up an interference pattern. But when the Italian team blocked off one of the slits and carried out another run of the experiment, there was no interference. The pattern on the detector screen was a simple blob directly behind the slit, just as you would expect to be produced by a stream of particles. How does an individual electron traveling alone through the experiment, through a hole in a wall, “know” whether there is another hole nearby that it might have gone through, and whether that hole is open or closed, and adjust its subsequent flight path accordingly?

We can only make analogies with things we have direct experience of, such as waves and particles. The physicist Arthur Eddington pointed this out in memorable fashion back in 1929. In his book The Nature of the Physical World, he said: No familiar conceptions can be woven around the electron ... something unknown is doing we don’t know what. [This] does not sound a particularly illuminating theory.

For example, probability. It was the German physicist Max Born who put the concept of probability, in the context of quantum mechanics, on a secure mathematical footing. But without going in to all the mathematics, we can get a feel for its importance using the example of electron spin (or tove gyre, as Eddington might have preferred). It is possible to describe, using the equations of quantum mechanics, an experiment in which an atom emits an electron that travels off through space (this is a real process called beta decay). In an idealized version of the experiment, the electron has a definite spin. It is either up or down. But there is no way to say in advance what it will be. There is a 50:50 chance of either possibility. If you do the experiment a thousand times, or simultaneously with a thousand atoms, you will find 500 electrons (plus or minus a few, maybe) with spin up, and 500 electrons with spin down. But if you catch a single electron and measure its spin you cannot tell which it will have until you look. Nothing surprising yet. But Einstein realized that something very surprising is predicted by the equations of the quantum theory for two electrons flying off in opposite directions.1 In certain circumstances, a conservation law applies, which says that the electrons must have opposite spin, one up and one down, so that, in effect, they cancel out. But the equations say that at the time the electrons are emitted from their source, they do not have a definite spi...

“Reality” is the idea that there is a real world that exists whether or not anyone is looking at it, or measuring it. Because of the probabilistic nature of the quantum world, Bell’s proposed experiment would need to involve measurements of large numbers of pairs of particles (such as electrons or photons) passing through the apparatus. The hypothetical experiment was designed in such a way that after a large number of runs, two sets of measurements would be produced. If one set of numbers was greater than the other, it would prove that the assumption of local reality is valid. This ratio became known as Bell’s Inequality, and the package of ideas as Bell’s Theorem. But if the other set of numbers was greater, Bell’s Inequality would be violated, which would mean that the assumption of local reality was incorrect. If quantum mechanics is correct, Bell’s Inequality must be violated. You can have a real world with spooky action at a distance. Or you can have locality, at the cost of saying that nothing is real unless it is observed.

But by the early 1980s, experiments had been carried out (using photons, rather than electrons) that proved that Bell’s Inequality is violated. Many more such experiments, with increasing technical sophistication, have confirmed this since. Local reality is not a valid description of the world; in John Bell’s own words, spoken at a meeting in Geneva in 1990, “I don’t know of any conception of locality which works with quantum mechanics. So I think we’re stuck with non-locality.” Einstein may have felt that “no reasonable definition of reality” could allow this, but the conclusion must be that reality is, in his terms, unreasonable. But the most impressive feature of all this is often overlooked. Although the jumping-off point for Bell’s Theorem was an attempt to understand quantum physics, and those words were spoken at a quantum physics meeting, these results do not apply only to quantum physics. They apply to the world—the Universe. Whether or not you think that quantum physics might one day be replaced as a description of how the world works, this will not change things. The experiments show that local reality does not apply to the Universe.

But to bring us back down to Earth, think again about the experiment with two holes. In the experiment, each electron seems to “know” how many holes are open, and where it is going. Does entanglement—spooky action at a distance—come into the story here as well? If a pair of photons flying in opposite directions are, in effect, part of a single quantum system, might we regard the whole double-slit experiment and the electron—all of the electrons?—as part of a single quantum system? Maybe the electron knows which holes are open because the state of the holes is also part of the state of the electron.

Bohr’s pragmatic approach extended to his interpretation. He said that we do not know anything except for the outcomes of experiments. These outcomes depend on what the experiments are designed to measure—on the questions we choose to ask of the quantum world (of nature). These questions are colored by our everyday experiences of the world, on a scale much larger than atoms and other quantum entities. So we may guess that electrons are particles and build an experiment designed to test this in an obvious way by measuring the momentum of an electron, thinking of the electron as a tiny pool ball. When we do so, lo and behold, the experiment measures the momentum of the electron, confirming our notion that electrons are particles. But a friend of ours has a different idea. She thinks that electrons are waves and designs an experiment to measure the wavelength of an electron. Lo and behold, her experiment gives a measurement of the wavelength, confirming her notion that electrons are waves. So what, says Bohr. Just because the electron behaves as if it were a particle when you are looking for particles, or as if it were a wave when you are looking for waves, doesn’t mean that it is either, let alone both. What you see is what you get, and what you see depends on what you chose to look for. It is meaningless, according to the Copenhagen Interpretation, to ask what quantum entities (such as electrons and atoms) are or what they are doing, when nobody is measuring them—looking at...

According to the CI that I was taught as a student, and that too many students are still taught today, as “the” way to “understand” quantum mechanics, an electron is emitted from a source—an electron gun—on one side of the experiment as a particle. It immediately dissolves into a “probability wave” that spreads through the experiment and heads toward the detector screen on the other side. This wave passes through however many holes are open, interfering with itself, or not, as appropriate, and arrives at the detector as a pattern of probabilities, higher in some places and lower in others, spread across the screen. At that instant, the wave “collapses” and turns back into a particle, whose position on the screen is chosen at random, but in accordance with the probabilities. This is called “the collapse of the wave function.” The electron travels as a wave but arrives as a particle. The wave, however, carries more than just probabilities. If the quantum entity has a choice of states it can be in, such as an electron that may be spin up or spin down, both states are somehow included in the wave function, the situation called a “superposition of states,” and the state the entity settles into at the point of detection, or interaction with another entity, is also determined at the moment the wave function collapses. In a lecture at the University of St Andrews in 1955, Werner Heisenberg said “the transition from the ‘possible’ to the ‘actual’ takes place during the act of obser...

This works as a method of calculating quantum behavior, as if things like electrons really did behave like this. But it also poses many puzzles. One of the most puzzling is a so-called “delayed choice” experiment, dreamed up by the physicist John Wheeler. He started from the fact that when photons are fired one at a time through the experiment with two holes, they still build up an interference pattern on the detector screen. But according to the CI, if a device is placed between the two holes and the detector screen to monitor which hole the photon goes through, the interference pattern will vanish, showing that each photon really did go through just one of the holes. The “delayed choice” comes in because we can decide whether or not to monitor the photons after they have passed the screen with two holes. Of course, human reactions are not fast enough to do this. But experiments have been carried out with automatic monitoring devices to do exactly this, switching the monitors on or off after the photons have passed the holes. They show that the interference pattern does indeed disappear when the photons are monitored, meaning that each photon (or the probability wave) only goes through one hole—even though the decision to monitor the photon was made only after it had passed the holes.

The quasar might be 10 billion light years away; the galaxy acting as a gravitational lens might be 5 billion light years away. But according to everything we know from experiment, what the photons were doing billions of years ago and billions of light years away is affected by what we choose to measure here and now. What is going on?

In essence, the Copenhagen Interpretation says that a quantum entity does not have a certain property—any property—until it is measured. Which raises all kinds of questions about what constitutes a measurement. Does human intelligence have to be involved? Is the Moon there if nobody is looking at it? Does the Universe only exist because human beings are intelligent enough to notice it? Or does the interaction of a quantum entity with a detector count as a measurement? Or where, in between those extremes, do you find the boundary between the quantum world and the “classical” world of good old Newtonian physics?

In many ways, de Broglie’s “pilot wave” interpretation is the most natural and obvious way to explain wave-particle duality. He proposed that the wave and the particle are both real, and that the wave (which became known as a pilot wave) guides the particle to its destination, like a surfer riding waves in the sea. In the experiment with two holes, the pilot wave spreads out through both holes and interferes with itself to make a pattern of interfering waves. Particles that are fired through the experiment start out with slight differences in speed or direction, so they end up surfing in slightly different directions, following the waves to build up an interference pattern on the detector screen. We measure the properties of the particles, but we can never measure the properties of the wave, only infer its existence from the behavior of the particles, which is hidden from us until they are detected. This kind of approach became known as “hidden variables” theory.

So the average distribution of everything in the Universe provides a frame of reference against which such changes are measured. Somehow, the “local” object is influenced by everything “out there.” Mach’s Principle tells us that a particle’s inertia is due to some interaction of that particle with all the other objects in the Universe. Just what that interaction is has long been a mystery. The pilot wave interpretation, and non-locality, may be the resolution of that puzzle. This leads to an interesting conclusion, which also features in another interpretation (Solace 3). The de Broglie–Bohm pilot wave interpretation applies to the whole Universe. The behavior of a single particle here and now depends on the positions of every other particle in the Universe at this instant.

As Schrödinger used to point out to anyone who would listen, there is nothing in the equations (including his famous wave equation) about collapse. That was something that Bohr bolted on to the theory to “explain” why we only see one outcome of an experiment—a dead cat or a live cat—not a mixture, a superposition of states. But because we only detect one outcome—one solution to the wave function—that need not mean that the alternative solutions do not exist. In a paper he published in 1952, Schrödinger pointed out the ridiculousness of expecting a quantum superposition to collapse just because we look at it. It was, he wrote, “patently absurd” that the wave function should “be controlled in two entirely different ways, at times by the wave equation, but occasionally by direct interference of the observer, not controlled by the wave equation.” Although Schrödinger himself did not apply his idea to the famous cat, it neatly resolves that puzzle. Updating his terminology, there are two parallel universes, or worlds, in one of which the cat lives, and in one of which it dies. When the box is opened in one universe, a dead cat is revealed. In the other universe, there is a live cat. But there always were two worlds that had been identical to one another until the moment when the diabolical device determined the fate of the cat(s). There is no collapse of the wave function.

Everett came up with the idea in 1955, when he was a PhD student at Princeton. In the original version of his idea, developed in a draft of his thesis, which was not published at the time, he compared the situation with an amoeba that splits into two daughter cells. If amoebas had brains, each daughter would remember an identical history up until the point of splitting, then have its own personal memories. In the familiar cat analogy, we have one universe, and one cat, before the diabolical device is triggered, then two universes, each with its own cat, and so on. Everett’s PhD supervisor, John Wheeler, encouraged him to develop a mathematical description of his idea for his thesis, and for a paper published in the Reviews of Modern Physics in 1957, but along the way, the amoeba analogy was dropped and did not appear in print until later. But Everett did point out that since no observer would ever be aware of the existence of the other worlds, to claim that they cannot be there because we cannot see them is no more valid than claiming that the Earth cannot be orbiting around the Sun because we cannot feel the movement. Everett himself never promoted the idea of the MWI. Even before he completed his PhD, he had accepted the offer of a job at the Pentagon working in the Weapons Systems Evaluation Group on the application of mathematical techniques (the innocently titled game theory) to secret Cold War problems (some of his work was so secret that it is still classified) and ...

The precise version of the MWI came from David Deutsch, in Oxford, and in effect put Schrödinger’s version of the idea on a secure footing, although when he formulated his interpretation, Deutsch was unaware of Schrödinger’s version. Deutsch worked with DeWitt in the 1970s, and in 1977, he met Everett at a conference organized by DeWitt—the only time Everett ever presented his ideas to a large audience. Convinced that the MWI was the right way to understand the quantum world, Deutsch became a pioneer in the field of quantum computing, not through any interest in computers as such, but because of his belief that the existence of a working quantum computer would prove the reality of the MWI. This is where we get back to a version of Schrödinger’s idea. In the Everett version of the cat puzzle, there is a single cat up to the point where the device is triggered. Then the entire Universe splits in two. Similarly, as DeWitt pointed out, an electron in a distant galaxy confronted with a choice of two (or more) quantum paths causes the entire Universe, including ourselves, to split. In the Deutsch–Schrödinger version, there is an infinite variety of universes (a Multiverse) corresponding to all possible solutions to the quantum wave function. As far as the cat experiment is concerned, there are many identical universes in which identical experimenters construct identical diabolical devices. These universes are identical up to the point where the device is triggered. Then, in some...