EINSTEIN versus Bohr: one of the most famous bouts in science. For the years spanning the late 1920s and early 1930s, these two fought over the future of physics. Albert Einstein could not accept the outrageous randomness and unknowability of quantum mechanics, so he attacked the theory by devising a series of ingenious thought experiments. But whenever he seemed to have nailed an inconsistency at the core of quantum theory, Niels Bohr proved him wrong. Despite all its unpalatable ingredients, quantum mechanics won the day.
Bohr invariably demolished Einstein by using Werner Heisenberg鈥檚 uncertainty principle. Measure the position of an electron or any other quantum particle, and, according to Heisenberg, you will disturb its momentum. Measure its momentum, and you will disturb its position. So you can never know both the momentum and the position of a particle at once. Ever since Bohr used this idea to win his legendary victory, the uncertainty principle has stood as the conceptual heart of quantum theory.
Bohr and Einstein had to devise imaginary experiments to prove their theories, because the technology to do the real experiments just didn鈥檛 exist. That鈥檚 changed. On a table strewn with delicate lasers, Gerhard Rempe and his colleagues at the University of Konstanz in Germany have brought to life one of the most famous experiments that the giants of quantum theory argued over. And they are using it to turn history on its head.
Advertisement
Quantum mechanics is still standing tall, but it now appears that Niels Bohr won his famous victory with faulty arguments. He inadvertently misled Einstein, and for 70 years most physicists have misunderstood the most important physical theory there is. They have been labouring under the delusion that what makes quantum theory so weird is its inherent uncertainty, or fuzziness, but in fact another feature of the quantum world, a phenomenon called entanglement, is at the root of it all. So what has ended these decades of delusion?
The experiment is, in its logic, astonishingly simple. It is the well-known two-slit experiment, which shows up one of the quantum world鈥檚 deepest mysteries: how something can be both a wave and a particle.
The idea is to send a beam of particles towards a barrier with two slits in it and see where they hit a detecting screen beyond (see Diagram). According to quantum mechanics, the result at the screen is an interference pattern, a set of parallel dark and bright bands. This shows that the beams going through the slits act like waves, which either reinforce or cancel each other out depending on where they meet. The same pattern is built up, particle by particle, even if the beam is so weak that only one particle goes through per hour, say.
This is what so unnerved Einstein: how can a single particle interfere with itself? How does it know that both slits are open, and cooperate in forming the interference pattern? Quantum mechanics says that it must somehow split into two ghosts of a particle, one going through each slit, which interfere with each other on the other side.
Why not test this strange idea by simply looking to see which way the particle goes? Shine some light near the slits, and you will see a few photons bounce off the particle as it goes through one hole or the other, proving that the particle doesn鈥檛 go through both holes-but you鈥檒l still see an interference pattern. Surely this should show that the idea of a particle interfering with itself is nonsense?
But it doesn鈥檛. Using the uncertainty principle, Bohr and Heisenberg destroyed any hope that this ploy could work. To be able to tell which slit the particle goes though, the argument goes, you must fix its position to a precision better than the distance between the slits.
Split personality
Heisenberg鈥檚 uncertainty principle demands that if you pin down the particle鈥檚 position so precisely, you must increase the uncertainty in its momentum. Bohr said that this happens because the photons deliver random, uncontrollable momentum kicks as they bounce off the particle. This disturbance changes the position where the particle hits the screen by a distance that is about as large as the spacing between the interference bands, so the pattern inevitably gets smeared away. In other words, if you look to see which way the particle goes, there鈥檚 no interference, so Einstein鈥檚 hoped-for contradiction evaporates.
Physicists managed to do this thought experiment for real only in the early 1990s, but the results were exactly as Bohr and Heisenberg said they would be. If you look to see which way the particles go they stop acting like waves, and the pattern on the screen is a big blob, not an interference pattern.
For half a century, physicists have memorised, repeated and regurgitated this story of how the uncertainty principle acts as the invincible defender of quantum theory. Learning it is virtually a rite of initiation for aspiring physicists. Not surprising, then, that when Rempe and his colleagues reported the results of their experiment last September, there was consternation in the ranks. Bohr鈥檚 reasoning, their results prove, is based on a fallacy.
The essence of the new experiment was proposed in 1991 by Marlan Scully, Berthold-Georg Englert and Herbert Walther of the Max Planck Institute for Quantum Optics in Garching, Germany. A two-slit experiment works with any kind of quantum particle. But they suggested that atoms might offer an advantage. An atom has a variety of different internal states: a lowest energy ground state and a series of higher energy or 鈥渆xcited鈥� states. And these different states, they reckoned, could be used to record the atom鈥檚 path.
鈥淢uch of our experiment is based on that proposal,鈥� says Rempe. In the 1980s, physicists devised ways to cool atoms to within a hair鈥檚 breadth of absolute zero using laser light. 鈥淪cully and his colleagues came up with the idea because they could use cold atoms,鈥� says Rempe. The point about cold atoms is that they have long wavelengths, which makes their interference patterns relatively easy to observe.
Still, no one could make the experiment work until last year, when Rempe and his colleagues managed it with a few clever tricks. They didn鈥檛 actually send atoms through slits in a solid barrier, but instead split a beam of cold rubidium atoms using thin barriers of pure laser light (see Diagram). The beams overlap, but travel along slightly different paths, A and B. As in the classic two-slit experiment, the two beams then combine to create an interference pattern.
But then Rempe and his colleagues looked to see which path the atoms followed. The atoms going down path A weren鈥檛 interfered with, but those on path B were tweaked into a higher energy state by a pulse of microwaves. So the atoms, in their internal states, kept a record of which way they had gone.
The payoff is impressive. The microwaves have hardly any momentum of their own, so they can cause little change to the atom鈥檚 momentum-certainly not enough to smear away the interference pattern.
Yet the quantum world鈥檚 wave-particle balancing act still works. With the microwaves turned off, the interference fringes appear. Turn them on, so that you can tell which way the atoms went, and the fringes suddenly vanish. 鈥淓veryone believes that when an interference pattern is lost, it happens because a measuring device delivers random kicks to the particles. But there are no random kicks in our experiment,鈥� says Rempe. At least, none worth mentioning. Rempe estimates that, at worst, the microwaves deliver momentum kicks ten thousand times too small to destroy the interference fringes. The uncertainty principle isn鈥檛 proved wrong, because in this setup the measurement of position is very imprecise, but it can鈥檛 explain the results.
So what鈥檚 going on? Is the central story of quantum theory just that-a story? Or is this one experiment merely an unimportant curiosity? At the University of Cambridge, physicist Yu Shi is trying to find out. Motivated by Rempe鈥檚 experiment, he has taken another look at the early thought experiments in which Bohr 鈥渄efeated鈥� Einstein. And he has come to be less than impressed by Bohr鈥檚 analyses.
Each of these thought experiments was designed to portray a particular case in which the quantum world refuses to reveal both its wave-like and its particle-like faces at the same time. And in each case, Shi points out, Bohr discussed the physics using only the simple Planck and de Broglie relations. These are the rudimentary equations that connect a particle鈥檚 momentum and energy to its wavelength and frequency.
So Shi has reanalysed the thought experiments using the rigorous equations of quantum theory, which give the fullest description possible of a quantum particle. And he has found that despite everything Bohr said, the uncertainty principle never has anything to do with destroying the interference. 鈥淧eople think that Bohr was right, and Einstein was wrong,鈥� he says, 鈥渂ut this is far from the truth. Bohr鈥檚 idea that a momentum kick destroys the interference is wrong.鈥�
Shi鈥檚 point is that although momentum kicks seem to explain the classic two-slit experiment, it is just a happy coincidence of numbers. There is a far deeper mechanism at work: it is the getting of path information itself that spoils the interference, says Shi. Forget all vague ideas of uncertainty, and look instead to the far more precise notion of 鈥渜uantum entanglement鈥�.
Inextricably linked
Ordinarily, we regard separate objects as independent of one another. They live on their own terms, and anything tying them together has to be forged by some tangible physical mechanism. Not so in the quantum world. If a particle interacts with some object-another particle, perhaps-then the two can become inextricably linked, or entangled (see 鈥淏eyond reality鈥�, 快猫短视频, 14 March, 1998, p 26). In a sense, they simply cease to be independent things, and one can only describe them in relation to each other.
What does this do to a particle鈥檚 ability to show wave-like behaviour? By itself, an atom can act as a wave. In a two-slit device, however, it effectively splits its own existence, and goes through both slits. If these two ghosts of the atom move along their paths without running into anything, then they recombine and interfere at the wall.
But suppose you send a photon towards one of the slits. If an atom were there, the photon would simply bounce off and record the atom鈥檚 position. But because the atom鈥檚 identity is already split between the two paths, it makes the photon split too. A ghost of the photon bounces off the ghost of the atom at that slit, and a second photon ghost carries straight on. This is the essence of entanglement-the interaction pairs up each atom ghost with a corresponding photon ghost. Linked with their photon parasites, the two atom ghosts are mismatched, so the interference vanishes.
鈥淟oss of interference is always due to entanglement,鈥� says Shi, who sees in it the true origin of quantum weirdness. Quantum particles can split into ghosts that can move on many paths at once, and when they come back together we see wave-like behaviour and interference patterns. But reach into the quantum world, and you will inevitably attach disruptive partners to the quantum ghosts, partners that will spoil the reunion and make the ghosts act as if they were true particles.
Awe of Bohr
Why has it taken physicists so long to appreciate this? Shi suspects that is was simple confusion. In 1935 Einstein recognised that if two particles were entangled, doing something to one could immediately affect the other, even at a great distance. As a result, Einstein doubted that entanglement could be real. But since then, experiments have provided strong evidence that this 鈥渘onlocal鈥� linking of distinct parts of the world really happens (鈥淲hy God plays dice鈥�, 快猫短视频, 22 August 1998, p 26). Because this effect is so shocking, physicists have dismissed entanglement as a nonlocal effect, and missed its role in the simple experiments.
Overlooking entanglement, physicists have instead taken Bohr鈥檚 word as gospel. And most of them still share Bohr鈥檚 exaggerated opinion of Heisenberg鈥檚 uncertainty principle. 鈥淚 think people give too high a position to it,鈥� says Shi. He now sees the idea as little more than a conceptual half-way house- something that allows physicists to talk and think about quantum particles as if they were vaguely classical, and to 鈥渆xplain鈥� their strange ways with a fuzzy form of ordinary mechanics.
So for 70 years physicists have been explaining some of the most elementary of quantum happenings in terms of a principle that, although true, turns out to be irrelevant. But what does this mean for physics? For the day-to-day business of making quantum calculations, probably nothing. But for physicists鈥� understanding of why they believe what they believe, the implications are unsettling.
Physicists like to pride themselves on their questioning minds, and their fierce intellectual independence. But it seems that the uncertainty principle, and the nebulous picture of the quantum world it supports, owe much of their status to little more than Bohr鈥檚 exalted reputation as a physicist. 鈥淏ohr was notorious for being obscure in his writings,鈥� says Maria Beller, a historian and philosopher of science at the Hebrew University of Jerusalem. And yet many physicists, she points out, have referred to that obscurity as reflecting the 鈥渄epth and subtlety鈥� of his thought, even if they couldn鈥檛 really work out what he was saying. She cites the German physicist Carl von Weizs盲cker as a typical victim of what she calls 鈥渢he overpowering, almost disabling, impact of Bohr鈥檚 authority鈥�.
She tells the story of how after once visiting Bohr to discuss physics, von Weizs盲cker found himself wondering what Bohr had meant. 鈥淚 tortured myself,鈥� he recalled, 鈥渙n endless solitary walks.鈥� And yet von Weizs盲cker never considered the possibility that Bohr might be wrong. 鈥淨uite incredibly,鈥� Beller says, 鈥渉e wondered what must one assume and in what way must one argue in order to render Bohr right鈥�.
But maybe Bohr was so obscure that he couldn鈥檛 help but be right. He liked to talk about 鈥渢he profound truths鈥� for which 鈥渢he opposite is also the truth鈥�. And he also admitted that 鈥渆very sentence I utter must be understood not as an affirmation, but as a question鈥�. Perhaps physicists are beginning to catch on.