FOR a theory that has the world鈥檚 finest physicists baffled, quantum mechanics is fantastically successful. It has made possible computers, lasers and nuclear reactors and explained how the Sun shines and why the ground beneath our feet is solid. But it is also strange, frustrating and incomprehensible. It insists that the microscopic world is a shadowy realm where nothing is certain 鈥 where an electron can be in two places at once and photons at opposite extremes of the Universe can communicate by some kind of weird telepathy.
But some physicists are beginning to suspect that there鈥檚 another level of reality beneath the quantum world. Nobel prizewinner Gerard鈥檛 Hooft believes that underpinning quantum weirdness is an old-fashioned deterministic theory 鈥 one in which there鈥檚 a simple relationship between cause and effect. Antony Valentini of Imperial College in London has now gone even further. He thinks that quantum mechanics may not always have applied, and that in the early Universe matter danced to a different tune. What鈥檚 more, some non-quantum stuff may even have survived to this day, tantalising us with the possibility of eavesdropping on secure cryptographic channels, constructing computers which outperform even the fastest quantum computers and, most remarkable of all, sending signals faster than the speed of light.
The reason for believing in a deeper level is that quantum theory merely predicts the probable outcomes of measurements, not certainties. To Valentini, it鈥檚 a bit like an actuary predicting the probability that a man will die at a particular age. 鈥淭his does not preclude a deeper level of cause and effect, which could be used to predict precisely when a given man dies,鈥 says Valentini. 鈥淚t might depend on the detailed condition of his heart and arteries.鈥
Advertisement
Indeed, everywhere in physics where a theory predicts probabilities, physicists believe there is a deeper level of certainty. Everywhere, that is, except quantum physics. Why not there too? Most physicists would say that this deeper level of explanation 鈥 a lower stratum or 鈥渉idden variable theory鈥 鈥 is unnecessary because quantum mechanics already fits all known experimental results. 鈥淭hey鈥檙e saying quantum theory works now 鈥 why look farther?鈥 says Valentini.
Nevertheless, a few people have tried. One attempt is the 鈥減ilot-wave鈥 theory, proposed by French physicist Louis de Broglie in the 1920s and developed by American physicist David Bohm during the early 1950s. Whereas in quantum mechanics the wave function is nothing more than a mathematical convenience for calculating the probability that a particle will be found at a particular point in space, in pilot-wave theory the wave is real. It鈥檚 an invisible but physical wave that guides particles along, and has a current that drives the precise motion of the particle, just as an ocean current drives a piece of flotsam. This theory reproduces all the statistical predictions of quantum mechanics. 鈥淢ost physicists are quite sceptical about this interpretation 鈥 including myself,鈥 says Lucien Hardy of the University of Oxford. 鈥淏ut it is important because it establishes the possibility of giving quantum theory a so-called hidden-variable interpretation.鈥
However, most physicists are put off this interpretation by a property called non-locality 鈥 physical influences that travel faster than light. Of course, even conventional quantum mechanics assumes non-local effects. Between measurements, the spin of an electron can be loosely thought of as in a state of high anxiety, flitting randomly from spinning in one direction, dubbed 鈥渦p鈥, to spinning the opposite way, dubbed 鈥渄own鈥. This has a remarkable consequence if two 鈥渆ntangled鈥 electrons have a total spin of zero between them 鈥 that is, the spin of one is up and the other down. Nature forbids the total spin from ever changing. So if the electrons are separated and a measurement on one finds it spinning 鈥渦p鈥, the far-away electron must at the very same instant plump for spinning 鈥渄own鈥. And vice versa.
鈥淚t doesn鈥檛 matter if one electron is in a steel box buried under the sea floor and the other is on the other side of the Galaxy,鈥 says Valentini. 鈥淓ach will respond instantaneously to the other鈥檚 state, in total violation of Einstein鈥檚 cosmic speed limit, the velocity of light.鈥
Yet while it鈥檚 possible to think of non-locality as a quirk of quantum mechanics 鈥 something that鈥檚 peripheral to the meat of the theory 鈥 the same can鈥檛 be said for pilot-wave theory. Non-locality lies at its very core. Take those two electrons again. Pilot-wave theory says that the pair of particles we see moving about in three-dimensional space is actually the projection of a single system that exists in six-dimensional 鈥渃onfiguration space鈥. 鈥淭he two particles are connected because they are really a single, higher-dimensional system,鈥 says Valentini.
Most physicists remain uneasy about non-locality because in our everyday experience things do not seem to be inextricably linked. Any theory that places this at its centre seems suspect. 鈥榯 Hooft, of the University of Utrecht in the Netherlands, is dead against the idea of non-locality. Yet he thinks that a novel kind of hidden-variable theory might offer a way around it. His idea, formulated in the late 1990s, is that some kind of deterministic theory can be applied at the very smallest scales of space and time. If you could zoom in and observe events that last just 10-43 seconds, in an area no more than 10-35 metres across, you would find a classically predictable theory with no need for probabilities and uncertainty. 鈥榯 Hooft describes it as being like a game of chess played on a board with microscopic squares. Quantum mechanics is then a kind of statistical theory that tallies all the smallest-scale events to give a fuzzy average description of what鈥檚 going on.
He has several reasons for believing quantum theory is built on deeper foundations. One is our inability, despite 80 years of effort, to reconcile gravity with the quantum world. Superstring theory makes many claims, he says, but it鈥檚 far too vague to be even remotely acceptable. Another reason is more deep-seated. 鈥淛ust like Albert Einstein, I am unhappy about the fundamental statistical nature of the predictions of quantum mechanics,鈥 he says.
鈥榯 Hooft is still developing his ideas, but even if he鈥檚 right, there鈥檇 be no way of telling. By his reckoning we may never see the deterministic layer underneath quantum mechanics, or even be able to prove that it exists.
Which is why Valentini鈥檚 latest ideas are so appealing. He thinks we should find hard evidence that these solid foundations really exist.
Valentini believes that instead of rejecting non-locality, we should embrace it. He points out that in conventional quantum mechanics, a 鈥渟uspicious coincidence鈥 obscures non-locality. For example, you might think that by using pairs of linked electrons like the pair described above, you could create an instantaneous communication system that defied the rule against anything travelling faster than light. But, frustratingly, that鈥檚 impossible, because you can never know before a measurement which way an electron is spinning. So if one direction of spin codes for a 鈥1鈥 and the other a 鈥0鈥 and you want to send a 鈥1鈥, you can only be 50 per cent sure of sending a 鈥1鈥 鈥 a level of uncertainty, or 鈥渘oise鈥, that scrambles any message. 鈥淎lthough non-locality is a fundamental feature of quantum theory, nature provides precisely the amount of quantum noise necessary to make it unusable,鈥 says Valentini. 鈥淚s that simply a coincidence? I don鈥檛 think so.鈥
He uses a thermal analogy. If the whole Universe was in a state of thermal equilibrium 鈥 that is, characterised by a single temperature 鈥 heat could not do any work. It couldn鈥檛 move a piston, for example. 鈥淚t isn鈥檛 that heat intrinsically can鈥檛 do work,鈥 he says. 鈥淚t鈥檚 just that temperature differences are needed to do work.鈥 In this imaginary state of universal thermal equilibrium, random temperature fluctuations in any machinery would be of precisely the right size to make any small random temperature differences unusable.
Valentini suspects that quantum theory may merely describe a particular state of the Universe in which quantum noise acts like these random temperature fluctuations, making non-locality unusable and effectively preventing messages being sent faster than light. According to Valentini, in this special state we are unable to observe non-local signals because they 鈥渃ancel out鈥 at the statistical level. This could apply to any hidden-variable theory, but Valentini has done most of his work on a type of pilot-wave theory.
His ideas are certainly controversial. 鈥淭hese conclusions depend on a particular interpretation of pilot-wave theory which, whilst being perfectly respectable, has the support of only a small number of physicists,鈥 says Hardy.
But on the whole, physicists 鈥 including Hardy 鈥 do not dismiss it. 鈥淰alentini is a serious physicist and a very deep thinker,鈥 says Hardy. 鈥淚 am a big fan of Antony Valentini,鈥 says Lee Smolin of the Perimeter Institute for Theoretical Physics in Waterloo, Canada. 鈥淚 think his ideas are the most interesting and potentially true ideas concerning the foundations of quantum theory that I have heard for some time.鈥
If Valentini is right, the implications are profound. Just after the big bang, the Universe may have existed in a state in which non-locality was not cloaked by random noise, he says. Interactions between particles in this early Universe then rapidly caused it to relax into the special 鈥渆quilibrium state鈥 we find today. These interactions, Valentini suggests, imply that the pilot-wave currents driving particles along were so convoluted that they scrambled the particles鈥 probability distributions. This can be likened to interactions between hot gas particles 鈥 which on average transfer energy from fast-moving to slow-moving particles 鈥 causing the gas to relax into a state of thermal equilibrium.
In our world, the probable location of a particle is related to the square of the amplitude of its wave function. But in this early Universe, before quantum noise set in, probability distributions might have been more sharply defined than the square of the wave function. With less quantum noise to blur things, it would have been possible to locate particles with greater certainty. And since non-locality wasn鈥檛 blurred out, this means that at this time, signals could travel faster than light. For example, there would be less uncertainty about the spin state of an entangled pair of electrons, so a message could be encoded in electrons on one side of the Universe and sent to the other instantaneously.
Valentini has reason to believe this was the case. According to him, a split second after the Universe鈥檚 birth there were two competing processes going on. One was the interaction between particles 鈥 analogous to the interaction between molecules in a gas 鈥 which drove the Universe towards a noisy equilibrium. But this approach to equilibrium was countered by the tremendous expansion of the Universe which was pulling matter apart. Only when the expansion had slowed could particle interactions dominate, says Valentini, allowing matter to slip into the blurry, uncertain form we see today. This point was probably reached when the Universe was about 10-43 seconds old, he suggests.
With the transition occurring so quickly, you might think there could be no significant consequences. Not so, says Valentini. This transition could solve the puzzle of why far-flung parts of the Universe are at the same temperature and have the same matter density. How could they have influenced each other if there wasn鈥檛 even time for light to have travelled from one to the other? The standard solution to this conundrum is inflation, a hypothetical super-fast expansion of the Universe in which it arose from a volume so small that very early on all parts knew about each other. But if there was no speed limit, there is no puzzle.
There would be consequences for inflation too, if it really occurred. Quantum fluctuations in the fields that physicists believe drove inflation should be imprinted on the cosmic microwave background as small variations in temperature. 鈥淭hose variations may therefore reflect quantum fluctuations in the early Universe,鈥 says Valentini. 鈥淚f the actual fluctuations don鈥檛 obey the rules of quantum mechanics, we ought to be able to see the fossil imprint in the microwave background today.鈥 Data from NASA鈥檚 satellite observatory MAP could provide the answer next year, he says.
What makes Valentini鈥檚 theory even more surprising is that some non-quantum matter might have survived to the present day. Since the key to the transition to the equilibrium state is the interaction between particles, any particles that ceased to interact around the cut-off point about 10-43 seconds after the big bang could get left behind. In particular, Valentini suggests that some gravitons 鈥 the hypothetical carriers of the gravitational force 鈥 could have become isolated at about the time of the transition. In other words, gravitons left over from this time might still be in a non-quantum state today.
According to Valentini, there may be hitherto unknown non-quantum particles too. 鈥淚t鈥檚 conceivable they may even make up the invisible dark matter which dominates the Universe,鈥 he says. 鈥淢atter following familiar quantum theory could be a minor component of the Universe.鈥 Particles of non-quantum matter would look like normal particles, they鈥檇 simply not obey the statistics of familiar particles. The location of a particle trapped in a box, for example, would not be dependent on the square of its wave function: its position could be pinned down more precisely.
How could we test such an outlandish idea? Identifying gravitons that survive from the instant after the big bang seems unlikely, and even getting hold of dark matter might be difficult, to say the least. But it is conceivable that dark matter particles could decay into photons that preserve the non-quantum behaviour of their parents. If you could detect such photons 鈥 by pointing a telescope at a small region of dark matter 鈥 they would behave differently from quantum photons. Pass ordinary photons through a pair of slits, for example, and they produce distinct dark and light bands of interference. The bands produced by non-quantum photons, on the other hand, would be blurred. There鈥檚 even some possibility that non-quantum matter is being created in today鈥檚 Universe. Valentini鈥檚 guess is that gravity could shift matter that obeys quantum theory back to its primordial non-equilibrium state. This would probably take the ferocious gravity of a singularity in a black hole, though.
If we could somehow get hold of non-quantum matter, it would be magical stuff. For one thing we could violate Heisenberg鈥檚 uncertainty principle, which puts a limit on how accurately we can measure things such as the location of a particle. To locate a particle, it has to interact with something else, for example when a photon bounces off it in a detector. The problem is that there is an uncertainty even in the position of the photon. 鈥淗owever, if we had photons obeying a probability distribution sharper than that of standard photons, we could locate things with greater certainty,鈥 says Valentini.
This also means we could use the stuff to eavesdrop on secure cryptographic channels, says Valentini. Quantum cryptography is 100 per cent secure because any attempt at eavesdropping would be noticed. The simple act of reading the secret key transmitted as a string of quantum 1s and 0s introduces disturbances (New 杏吧原创, 2 October 1999, p 28). But if eavesdroppers possess non-quantum matter, they could beat the uncertainty principle and distinguish the state of the bits without disturbing them. This is because non-quantum particles contain less noise. Just a very weak interaction between them and the quantum bits 鈥 an interaction too weak to disturb the bits 鈥 is enough to leave a discernable signature in the non-quantum particles that could be used to decrypt the message.
And there鈥檚 more. Non-quantum matter would enable us to build a computer which massively outperforms 鈥渃onventional鈥 quantum computers. These hypothetical machines would exploit the fact that a particle such as an atom can be in many states at once 鈥 a so-called superposition 鈥 to do large numbers of calculations simultaneously (New 杏吧原创, 8 June, p 24). The problem is that you need a carefully crafted quantum program that concentrates the answer in a single branch of the superposition, from where it can be read.
So far such algorithms have been found for only a few specialised problems. But using non-quantum matter you could in theory access all the myriad parallel calculations of a quantum computer. It could be used to observe the computer鈥檚 quantum state without collapsing the wave function, enabling us to read the results of all the parallel computations.
But far more remarkable than all this would be faster-than-light communication. You could exploit non-locality without quantum noise getting in the way, using it to control robotic probes on planets at the other end of the Solar System in real time, for example. Troublesome time lags while instructions 鈥渃rawl鈥 at the speed of light across space would become a thing of the past. Why send humans on long, dangerous missions to Mars when robots, controlled from a comfortable lab on Earth, could do the job perfectly well?
And this sort of communication would force us to revise relativity theory, says Valentini. Contrary to what is suggested by Einstein鈥檚 theory, there would have to be an underlying preferred time 鈥 a sort of Universe-wide GMT.
Valentini will have a hard time convincing sceptics. But it could be worth it. 鈥淚t would mean that physics was finally making progress on a problem on which we have been stuck for many decades,鈥 says Smolin. 鈥淩ight now we鈥檙e staring into a sort of quantum fog,鈥 says Valentini. 鈥淚f we admit that an unexplored level might lie behind it, a whole new world comes into focus.鈥