Testing Realism?

Wo links to this article in the most recent issue of Nature.

An experimental test of non-local realism

Most working scientists hold fast to the concept of ‘realism‘—a viewpoint according to which an external reality exists independent of observation. But quantum physics has shattered some of our cornerstone beliefs. According to Bell’s theorem, any theory that is based on the joint assumption of realism and locality (meaning that local events cannot be affected by actions in space-like separated regions) is at variance with certain quantum predictions. Experiments with entangled pairs of particles have amply confirmed these quantum predictions, thus rendering local realistic theories untenable. Maintaining realism as a fundamental concept would therefore necessitate the introduction of ‘spooky’ actions that defy locality. Here we show by both theory and experiment that a broad and rather reasonable class of such non-local realistic theories is incompatible with experimentally observable quantum correlations. In the experiment, we measure previously untested correlations between two entangled photons, and show that these correlations violate an inequality proposed by Leggett for non-local realistic theories. Our result suggests that giving up the concept of locality is not sufficient to be consistent with quantum experiments, unless certain intuitive features of realism are abandoned.

I’ve only ever read philosophers (like Tim Maudlin) on the Bell inequalities, and I don’t think I’ve ever read about the Leggett inequalities. So I shouldn’t be too snarky. But really, these judgments about comparative spookiness can’t be left to stand. They find non-local relations, of the kind we need to posit to give a realistic explanation of Bell inequalities, “spooky”. (I wonder if they find spatiotemporal relations spooky too.) But the thought that there isn’t really a world out there to perceive, and that our impressions of the world are to be explained in some way other than as the perception of a mind independent reality. That’s apparently not spooky at all. This doesn’t strike me as particularly plausible.

Comments from anyone who knows more about Leggett’s inequalities, or who wants to mock my fear of anti-realism, more than welcome.

6 Replies to “Testing Realism?”

  1. I’ve only read Maudlin too. But isn’t ‘spooky’ here an Einstein reference? (Hence it’s appearance in quote-marks). I seem to remember the full phrase is ‘spooky action at a distance’. So it seems like a stock-phrase one would use in setting up the problem.

  2. Here’s a Scientific American writeup of the experiments. It’s not very detailed at all (and I probably wouldn’t understand it if it were), but it sounds to me as though they don’t mean to strain at spooky non-locality and swallowing anti-realism, but rather are saying that even if you introduce nonlocal interactions that you still can’t rescue realism.

    Or, to quote the article, “nonlocality is not enough to save realism from quantum theory.”

    The bit wo quoted about Aristotelian logic sounds goofy to me, but I can’t access the article (and again, might well not understand it if I could) so I can’t judge. As the SciAm article says, there appear to be some nonlocal realistic models that aren’t refuted by these experiments.

  3. From a quick scan of the paper, the real targets here are hidden variables theories. The thesis they characterise as “realism” is ‘(1) all measurement outcomes are determined by pre-existing properties of particles independent of the measurement’. They then develop an argument that a natural class of non-local hidden variables theories is ‘experimentally excluded’. (It’s worth noting that Bohmian mechanics isn’t one of these theories.)

    While the terminology is quite misleading, it is fairly common I take it for realism and the existence of hidden variables to be conflated. This probably starts from Einstein’s worries that quantum mechanics was ‘incomplete’ because it entails that certain quantities lack definite values at certain times. Think about that lack for long enough and it might come to seem as if quantum mechanics is anti-realist, because a truly realist theory would have values for such quantities being well-defined at every moment.

    But that just isn’t anything like what we are normally talking about when we discuss realism, and it is perfectly possible to have a realistic quantum theory (in the philosopher’s sense): take the quantum state (that thing which evolves according to the Schrödinger equation) to describe the real state of the world, and adopt a no-collapse interpretation so that the Schrödinger equation tells us everything about how the system evolves, and then believe the resulting theory to be true. Such a world is weird, but it is real in any standard philosophical sense: surely that’s part of what makes the weirdness interesting.

  4. I mostly wanted to second Antony’s point. In the many discussions of Bell’s inequalities, there is (from what I can tell) a long-standing use of the label ‘realism’ for the view that measurement results are determined by the properties particles carry prior to and independent of observations of them.

    It seems to me we shouldn’t begrudge them their usage of ‘realism,’ as long as they don’t begrudge us philosophers the meaning(s) that we give the term.

  5. Antony is right that the sense of realism at issue is not the same as in philosophical debates about realism and anti-realism. Unfortunately, this confusion is propagated by the first paragraph of the article when the authors write:

    Physical realism suggests that the results of observations are a consequence of properties carried by physical systems. It remains surprising that this tenet is very little challenged, as its significance goes far beyond science.

    This wildly overstates the reach of the actual results; it is absurd to think that realism characterised in this manner is refuted by the experimental results they report.

    Moreover, the narrow sense of realism they show to be incompatible with quantum mechanics does not seem very interesting even from a physical perspective, as far as I can tell. They acknowledge that Bohmian mechanics is compatible with the results, and then at the end they say that instead of realism in the narrow sense being given up it is possible to drop the assumptions of “absence of actions into the past or a world that is not completely deterministic”. Given all of these exceptions, the position they attack seems to constrain the class of possible interpretations in a way that doesn’t touch any interpretatations actually under development. If that is right, it is far less interesting than the exaggerated claims of the first paragraph.

Leave a Reply