[xaetium]

In these types of quantum reality-probing experiments, any problems of experimental design that affect the validity of the findings are referred to as loopholes, as though there's some awkward legal wrangling going on, because the experiments were conceived originally to determine whether the controversial Bell's inequalities hold. The inequalities were designed to test Bell's theorem which states that any hidden variables (things not yet observed that have a causal influence on experimental outcome) are required to be non-local if they are to hold with the predictions of quantum mechanics. Non-local here means 'spooky action at a distance'.

Showing the inequalities to be violated (incorrect by experiment) was originally controversial because Einstein and Bohr had differing notions of what the quantum mechnical theory implied about reality. They engaged in a lengthy, open discussion about it which was never resolved. Einstein believed in local realism, in which there is no spooky action at a distance and properties like position and momentum exist even when not being measured. Bohr, on the other hand, insisted that there simply wasn't an underlying reality and that only when measurements are made are properties like position and momentum condensed out of the quantum mechanical reality. So, you see, the significance of the experiment is in line with the underlying nature of reality; by closing another loophole, we get closer to what's what.

[The rest here is historical context.]

The familiar refrain, "God does not play dice," is almost always taken out of context - within its original statement, Einstein was also talking about a kind of telepathy required with it - the non-local aspect of quantum mechanics. Einstein said in 1954 'it is not possible to get rid of the statistical character of the present quantum theory by merely adding something to the latter, without changing the fundamental concepts about the whole structure'. He was saying he lost conviction in using a hidden variable theory to replace quantum mechanics.

Bohr's view, like Einstein's later view, is more in line with modern thinking. A team led by Aspect in 1981-82 ruled out either locality or objective reality, by testing the inequalities experimentally. This left possible a non-local reality. In 2006, a group tested Leggett's inequality, and showed it to be violated, which refined experimentally what the nature of reality is, though showed only that realism and a certain type of non-locality are incompatible, without ruling out all possible non-local models. (Nature, April 2007) Aspect remarked that philosophically, the 'conclusion one draws is more a question of taste than logic'.

[stangeek]

OK - but what's the difference with previous experiments? Is it that they did it with a single photon? Or is it because they managed to do it from two remote laboratories?

[xaetium]

It may be the combination is new; I don't know the exact state of the field, but: This experiment uses a single photon, so they don't have to sample multiple times and make a statistical analysis on that part. If they did, that might open the efficiency loophole. The communication loophole isn't opened, as they are in sufficiently distant labs, with short enough measurement frames, but that's been done before.

As far as I can tell, the disjoint measurement loophole doesn't apply here, either, as it opens when correlations are drawn from multiple samples; here there's one. I'm not sufficiently expert to tell whether the rotational invariane, or other loopholes are closed here. Can anyone shed some light on this?