The Reality of the Wavefunction

A couple of months ago we spoke with quantum physicists Martin Ringbauer and Alessandro Fedrizzi of the University of Queensland, in Australia, on the podcast, about their experiment looking into the nature of the wavefunction. Their results lend support (though not quite definitively) to "Psi-ontic models" that say that, if there's an objective reality, then the wavefunction is real. Psi-ontic models include Many Worlds, Bohmian mechanics and collapse models. That's as opposed to "Psi-epistemic models" that say that the wavefunction just reflects our ignorance about the state of reality. I recently wrote an article for Nature that rounds up various experiments--including that Australian one--that are trying to uncover quantum reality. The full article, "What is Really Real?" on Nature's site, so please take a look. (The image is taken from Nature's twitter feed: @NatureNews.)

When I attended a quantum foundations meeting in Erice, Italy, sponsored by COST, last March, I was surprised to find that the participants were mainly split between Bohmians and Collapse Model fans. Anecdotally, I'd say that Many Worlders dominate FQXi meetings. I've already highlighted collapse models on the blog and podcast, but I wanted to take this chance to flag up a podcast special recorded in Erice with Shelly Goldstein of Rutgers University and Jean Bricmont of the Catholic University of Louvain, who explain what Bohmian mechanics is and present their case for why they believe that it makes the most sense -- and talk about why they feel it hasn't had a fair hearing over the years.

In the Nature article, I mention the oil droplet experiments that have been getting a lot of attention over the past couple of years because they appear to show droplets "walking" along an oil surface, guided by their own ripples, analogous to the predictions made by Bohmian mechanics. (See the "Pilot Wave Hydrodynamics" forum thread, suggested by John Merryman.)

But there were a couple of other tests to note, which didn't make the final cut. Bricmont pointed me to the first, which was published in Science. Bohmian mechanics, unlike standard quantum theory, says that particles have definite locations, even before they are observed, and it makes predictions for the paths taken by these particles as they move through, for instance, the double slit experiment. In 2007, Howard Wiseman, a quantum physicist at Griffith University, came up with a way to sneakily track the paths taken by photons in a double-slit experiment using "weak measurements" that allow physicists to quickly peek at an experiment while it is in progress. This disturbs the photon's location slightly, so no single measurement can indicate where the photon would have been, had it not been observed. But by repeating the test many times, it is possible to build up a statistical picture of the "classical trajectories" taken by photons through the apparatus, which is just what a team led by Aephraim Steinberg at the University of Toronto, Ontario, did, in 2011.

The paths they found tended to correspond with those calculated using the Bohm model, a similarity that Steinberg told me he found "thought-provoking". The experiment, he said, "underscores the elegance of the Bohm model," though he added that it does not serve as proof of the interpretation. That's because the trajectories are not seen directly while the experiment is in progress, but inferred after the experiment has taken place and all the results are in. A Many-World's fan could thus argue that the trajectories only represent where the particles would have been *if* they took classical trajectories, rather than showing the definite trajectories they actually took. Bricmont agrees that experiment doesn't prove Bohmian mechanics, but feels that the similarities are striking enough that the results should inspire more physicists to take another look at Bohmian theory.

Owen Maroney, a physicist at Oxford University, also highlighted a paper by Samuel Colin and Antony Valentini at Clemson University in South Carolina, who last year analysed a modified version of the model and calculated that it would have led to a subtly different pattern of quantum fluctuations in the early universe than is predicted by conventional quantum theory and the Many-World's interpretation (arXiv:1407.8262v1). These signatures could show up in measurements of the cosmic microwave background currently being made by the Planck satellite.

As I mention in the Nature story, there are hopes that it could be just a matter of months before an experiment is carried out that completely rules out Psi-epistemic models and favours Psi-ontic models. But even if that is successful, there is another quantum interpretation that I didn't cover in the article that would not be affected one way or the other by such tests, even though it is not a Psi-ontic model: quantum Bayesianism (or QBism). This is a relatively recent model based on classical Bayesian probability theory聽that rejects the notion that the results of quantum experiments can directly access an external objective reality that is independent of the agent making the measurements. According to Q-Bists, the results of quantum measurements are intimately tied to the presence of an agent and serve to change the agent's degree of belief about what their personal future experience will be.

I asked Ruediger Schack, a QBist at Royal Holloway University in London, about why QBism escapes. He explained that is because the definitions of Psi-epistemic and Psi-ontic -- and the wavefunction reality tests -- are valid within a specific framework, called an ontological model (by Nicholas Harrigan and FQXi's Robert Spekkens). "In this framework, outcome probabilities of measurements are determined by some real property lambda. Such models used to be called hidden variable models," said Schack. QBists, however, do not subscribe to this framework. The Copenhagen interpretation also lies outside this framework.

Maybe it's time to do a survey. What's your favourite quantum interpretation? And you think there will ever be a test that could help people choose between various interpretations? Could there be a test that would make you change your mind?