Quantum theory is one of the most profound discoveries of humanity. In my view, it’s on a par with Cuban cigars and single malt whiskey. The theory has been hugely successful in showing us the inner workings of the universe. However, it is plagued by serious problems and provokes fiery debates, such as the giant argument that erupted on the Cosmic Variance blog a couple of weeks ago. One of these problems is the unanswered question of how the everyday macroscopic world emerges from its strangely behaving microscopic constituents.

Let me put it this way. Nobody doubts that the moon is circling the Earth even when we’re not looking. In the microworld, however, quantum objects can behave as if they were created by the mere act of observation. It goes like this: you open a box and you see an atom there. Can you infer that it was there the moment before you looked inside? The mind-boggling answer is NO! This was proven theoretically in 1964 by John Bell and later confirmed experimentally, beginning with the work of John Clauser and others in the early 1970s (described in a classic *Scientific American* article by Bernard d’Espagnat). If the same were true for the moon, it would be like saying that the act of looking into the night sky made the moon appear.

So where does this leave us? Physicists believe that quantum physics should describe all objects big and small, yet the microworld is very different from the macroworld. We have to explain how the latter is the result of the former.

With some colleagues I recently proposed and then generalized a new solution. Interestingly, our explanation for how the classical world arises from the quantum world does not invoke “decoherence” or any of the other mechanisms proposed before.

Decoherence is probably the best-known model introduced so far to explain how the microscopic world’s quantum weirdness disappears. Decoherence is the loss of information about a quantum state to the surrounding environment to the point where the state has lost many of its quantum properties. It had been believed that decoherence is ubiquitous in large systems and happens extremely fast. However, recent research done by my colleague at the Centre for Quantum Technologies in Singapore, Vlatko Vedral, indicates that decoherence can be much slower in large systems than was expected (more about this in Scientific American’s June cover story “Living in a quantum world”).

In our recent paper, we take a different approach. We consider how measurements work in the macroworld, finding that some quantum features are simply unobservable. Most remarkably, this approach shows that something called quantum nonlocality disappears for objects big enough to contain roughly the Avogadro number of atoms—the number of atoms you’d expect in a few grams of matter.

The idea of quantum nonlocality goes back to a seminal paper by Albert Einstein, Boris Podolsky and Nathan Rosen (EPR) published in 1935. They described a simple thought experiment which showed that quantum theory cannot be reconciled with two common-sense and very reasonable assumptions about the fabric of reality, namely, locality (there is no “spooky action at a distance”) and realism (measurements reveal objective properties of physical systems).

Their ingenious test exploits the notion of “correlations,” something that we encounter in everyday life. For example, if a clock in New York strikes midnight, then we know that at the same time a clock in Singapore strikes noon. The EPR experiment tests what kind of correlations one can observe between measurements performed on quantum particles, highlighting that quantum correlations can be stronger than the logic of locality and realism dictates. Since then, experiments have confirmed quantum theory’s predictions, forcing physicists to accept that the fabric of reality is nonlocal, nonrealistic or both!

The problem then becomes how to reconcile the local realism of the macroscopic world—the moon and other big objects really do follow these two common-sense assumptions—with quantum theory. We solve this problem: we show that macroscopic observables are always local realistic even when the underlying states are quantum.

When one measures large systems, it is impossible to do it precisely. For instance, if you have a gas in a container, you can measure its pressure, temperature, etc., but you cannot measure the velocities of all the particles of the gas. This translates technically to a limited set of observables that are average properties of the system.

Our result derives from this concept of macroscopic observables being a kind of average. There is a limit to the number of quantum correlations each particle can have with another, which is referred to as the “monogamy” of quantum correlations. The concept is simple: if particles A and B exhibit correlations of the kind predicted in the EPR experiment then A and B can only have local and realistic correlations with other particles.

This monogamous behaviour extends to correlations between larger groups of quantum particles, which is the main idea behind our result. Imagine you are making a macroscopic measurement between two regions in space, A, containing quantum particles A1, A2, A3, etc., and B, containing B1, B2, B3, etc. The measurement samples all possible pairs. Due to monogamy, as you increase the number of particles, the overall strength of the correlations measured dilutes. For instance, A_{i}B_{j} may be strongly correlated but then A_{i} and any other B-particle exhibit only local realistic correlations (see figure). Analysing the statistics, we find that local realism emerges for macroscopic correlations without us needing to invoke any other mechanism.

This doesn’t mean there are no macroscopic quantum effects, just that any such effects will be too weak to demonstrate themselves in as bizarre a way as they do on the microscopic level. Whatever quantumness is present can never be strong enough to contradict local realism. It is, after all, reassuring to know that the moon is there when you don’t look at it. What the remaining quantumness in the macroscopic world means is unclear at this stage. The debates will surely rumble on.