Physicists start saying farewell to reality
Quantum mechanics just got even stranger
[This is my pre-edited story for Nature News on a paper published this week, which even this reserved Englishman must acknowledge to be deeply cool.]

There’s only one way to describe the experiment performed by physicist Anton Zeilinger and his colleagues: it’s unreal, dude.

Measuring the quantum properties of pairs of light particles (photons) pumped out by a laser has convinced Zeilinger that “we have to give up the idea of realism to a far greater extent than most physicists believe today.”

By realism, what he means is the idea that objects have specific features and properties: that a ball is red, that a book contains the works of Shakespeare, that custard tastes of vanilla.

For everyday objects like these, realism isn’t a problem. But for objects governed by the laws of quantum mechanics, such as photons or subatomic particles, it may make no sense to think of them as having well defined characteristics. Instead, what we see may depend on how we look.

Realism in this sense has been under threat ever since the advent of quantum mechanics in the early twentieth century. This seemed to show that, in the quantum world, objects are defined only fuzzily, so that all we can do is to adduce the probabilities of their possessing particular characteristics.

Albert Einstein, one of the chief architects of quantum theory, could not believe that the world was really so indeterminate. He supposed that there was a deeper level of reality yet to be uncovered: so-called ‘hidden variables’ that specified any object’s properties precisely.

Allied to this assault on reality was the apparent prediction of what Einstein called ‘spooky action at a distance’: disturbing one particle could instantaneously determine the properties of another particle, no matter how far away it is. Such interdependent particles are said to be entangled, and this action at a distance would violate the principle of locality: the idea that only local events govern local behaviour.

In the 1960s the Irish physicist John Bell showed how to put locality and realism to the test. He deduced that they required two experimentally measurable quantities of entangled quantum particles such as photons to be equal. The experiments were carried out in the ensuing two decades, and they showed that Bell’s equality is violated.

This means that either realism or locality, or both, fails to apply in the quantum world. But which of these cases is it? That’s what Zeilinger, based at the University of Vienna, and his colleagues have set out to test [1].

They have devised another ‘equality’, comparable to Bell’s, that should hold up if quantum mechanics is non-local but ‘realistic’. “It’s known that you can save realism if you kick out locality”, Zeilinger says.

The experiment involves making pairs of entangled photons and measuring a quantum property of each of them called the polarization. But whereas the tests of Bell’s equality measured the so-called ‘linear’ polarization – crudely, whether the photons’ electromagnetic fields oscillate in one direction or the opposite – Zeilinger’s experiment looks at a different sort of polarization, called elliptical polarization, for one of the photons.

If the quantum world can be described by non-local realism, quantities derived from these polarization measurements should be equal. But Zeilinger and colleagues found that they weren’t.

This doesn’t rule out all possible non-local realistic models, but it does exclude an important subset of them. Specifically, it shows that if you have a group of photons all with independent polarizations, then you can’t ascribe specific polarizations to each. It’s rather like saying that in a car park it is meaningless to imagine that particular cars are blue, white or silver.

If the quantum world is not realistic in this sense, then how does it behave? Zeilinger says that some of the alternative non-realist possibilities are truly weird. For example, it may make no sense to imagine ‘counterfactual determinism’: what would happen if we’d made a different measurement. “We do this all the time in daily life”, says Zeilinger – for example, imagining what would happen if we’d tried to cross the road when that truck was coming.

Or we might need to allow the possibility of present actions affecting the past, as though choosing to read a letter or not affects what it says.

Zeilinger hopes his work will stimulate others to test such possibilities. “I’m sure our paper is not the end of the road”, he says. “But we have a little more evidence that the world is really strange.”

Reference
1. Gröblacher, S. et al. Nature 446, 871 – 875 (2007).