First major attempt to chart researchers’ views finds interpretations in conflict

Physicists disagree wildly on what quantum mechanics says about reality, Nature survey shows

First major attempt to chart researchers’ views finds interpretations in conflict

Quantum mechanics is one of the most successful theories in science — and makes much of modern life possible. Technologies ranging from computer chips to medical-imaging machines rely on the application of equations, first sketched out a century ago, that describe the behaviour of objects at the microscopic scale.

But researchers still disagree widely on how best to describe the physical reality that lies behind the mathematics, as a Nature survey reveals.

At an event to mark the 100th anniversary of quantum mechanics last month, lauded specialists in quantum physics argued politely — but firmly — about the issue. “There is no quantum world,” said physicist Anton Zeilinger, at the University of Vienna, outlining his view that quantum states exist only in his head and that they describe information, rather than reality. “I disagree,” replied Alain Aspect, a physicist at the University of Paris-Saclay, who shared the 2022 Nobel prize with Zeilinger for work on quantum phenomena.

To gain a snapshot of how the wider community interprets quantum physics in its centenary year, Nature carried out the largest ever survey on the subject. We e-mailed more than 15,000 researchers whose recent papers involved quantum mechanics, and also invited attendees of the centenary meeting, held on the German island of Heligoland, to take the survey.

The responses — numbering more than 1,100, mainly from physicists — showed how widely researchers vary in their understanding of the most fundamental features of quantum experiments.

As did Aspect and Zeilinger, respondents differed radically on whether the wavefunction — the mathematical description of an object’s quantum state — represents something real (36%) or is simply a useful tool (47%) or something that describes subjective beliefs about experimental outcomes (8%). This suggests that there is a significant divide between researchers who hold ‘realist’ views, which project equations onto the real world, and those with ‘epistemic’ ones, which say that quantum physics is concerned only with information.

The community was also split on whether there is a boundary between the quantum and classical worlds (45% of respondents said yes, 45% no and 10% were not sure). Some baulked at the set-up of our questions, and more than 100 respondents gave their own interpretations (the survey, methodology and an anonymized version of the full data are available in supplementary information at the foot of this page).

“I find it remarkable that people who are very knowledgeable about quantum theory can be convinced of completely opposite views,” says Gemma De les Coves, a theoretical physicist at the Pompeu Fabra University in Barcelona, Spain.

Nature asked researchers what they thought was the best interpretation of quantum phenomena and interactions — that is, their favourite of the various attempts scientists have made to relate the mathematics of the theory to the real world. The largest chunk of responses, 36%, favoured the Copenhagen interpretation — a practical and often-taught approach. But the survey also showed that several, more radical, viewpoints have a healthy following.

Asked about their confidence in their answer, only 24% of respondents thought their favoured interpretation was correct; others considered it merely adequate or a useful tool in some circumstances. What’s more, some scientists who seemed to be in the same camp didn’t give the same answers to follow-up questions, suggesting inconsistent or disparate understandings of the interpretation they chose.

“That was a big surprise to me,” says Renato Renner, a theoretical physicist at the Swiss Federal Institute of Technology (ETH) in Zurich. The implication is that many quantum researchers simply use quantum theory without engaging deeply with what it means — the ‘shut up and calculate’ approach, he says, using a phrase coined by US physicist David Mermin. But Renner, who works on the foundations of quantum mechanics, is quick to stress that there is nothing wrong with just doing calculations. “We wouldn’t have a quantum computer if everyone was like me,” he says.

Copenhagen still reigns supreme

Over the past century, researchers have proposed many ways to interpret the reality behind the mathematics of quantum mechanics, which seems to throw up jarring paradoxes. In quantum theory, an object’s behaviour is characterized by its wavefunction: a mathematical expression calculated using an equation devised by German physicist Erwin Schrödinger in 1926. The wavefunction describes a quantum state and how it evolves as a cloud of probabilities. As long as it remains unobserved, a particle seems to spread out like a wave; interfering with itself and other particles to be in a ‘superposition’ of states, as though in many places or having multiple values of an attribute at once. But an observation of a particle’s properties — a measurement — shocks this hazy existence into a single state with definite values. This is sometimes referred to as the ‘collapse’ of the wavefunction.

It gets stranger: putting two particles into a state of joint superposition can lead to entanglement, which means that their quantum states remain intertwined even when the particles are far apart.

The German physicist Werner Heisenberg, who helped to craft the mathematics behind quantum mechanics in 1925, and his mentor, Danish physicist Niels Bohr, got around the alien wave–particle duality largely by accepting that classical ways of understanding the world were limited, and that people could only know what observation told them. For Bohr, it was OK that an object varied between acting like a particle and like a wave, because these were concepts borrowed from classical physics that could be revealed only one at a time, by experiment. The experimenter lived in the world of classical physics and was separate from the quantum system they were measuring.

Heisenberg and Bohr not only took the view that it was impossible to talk about an object’s location until it had been observed by experiment, but also argued that an unobserved particle’s properties really were fundamentally unfixed until measurement — rather than being defined, but not known to experimenters. This picture famously troubled Einstein, who persisted in the view that there was a pre-existing reality that it was science’s job to measure.

Decades later, an amalgamation of Heisenberg’s and Bohr’s not-always-unified views became known as the Copenhagen interpretation, after the university at which the duo did their seminal work. Those views remain the most popular vision of quantum mechanics today, according to Nature’s survey. For Časlav Brukner, a quantum physicist at the University of Vienna, this interpretation’s strong showing “reflects its continued utility in guiding everyday quantum practice”. Almost half of the experimental physicists who responded to the survey favoured this interpretation, compared with 33% of the theorists. “It is the simplest we have,” says Décio Krause, a philosopher at the Federal University of Rio de Janeiro, Brazil, who studies the foundations of physics, and who responded to the survey. Despite its issues, the alternatives “present other problems which, to me, are worse”, he says.

But others argue that Copenhagen’s emergence as the default comes from historical accident, rather than its strengths. Critics say it allows physicists to sidestep deeper questions.

One concerns the ‘measurement problem’, asking how a measurement can trigger objects to switch from existing in quantum states that describe probabilities, to having the defined properties of the classical world.

Another unclear feature is whether the wavefunction represents something real (an answer selected by 29% of those who favoured the Copenhagen interpretation) or just information about the probabilities of finding various values when measured (picked by 63% of this group). “I’m disappointed but not surprised at the popularity of Copenhagen,” says Elise Crull, a philosopher of physics at the City University of New York. “My feeling is that physicists haven’t reflected.”

The Copenhagen interpretation’s philosophical underpinnings have become so normalized as to seem like no interpretation at all, adds Robert Spekkens, who studies quantum foundations at the Perimeter Institute for Theoretical Physics in Waterloo, Canada. Many advocates are “just drinking the Kool-Aid of the Copenhagen philosophy without examining it”, he says.

Survey respondents who have carried out research in philosophy or quantum foundations, studying the assumptions and principles behind quantum physics, were the least likely to favour the Copenhagen interpretation, with just 20% selecting it. “If I use quantum mechanics in my lab every day, I don’t need to go past Copenhagen,” says Carlo Rovelli, a theoretical physicist at Aix-Marseille University in France. But as soon as researchers apply thought experiments that probe more deeply, “Copenhagen is not enough”, he says.

What else is on the menu?

In the years after the Second World War and the development of the atomic bomb, physicists began to exploit the uses of quantum mechanics, and the US government poured cash into the field. Philosophical investigation was put on the back burner. The Copenhagen interpretation came to dominate mainstream physics, but still, some physicists found it unsatisfying and came up with alternatives (see ‘Quantum mechanics: five interpretations’).

In 1952, US physicist David Bohm resurfaced an idea first touted in 1927 by French physicist Louis de Broglie, namely that the strange dual nature of quantum objects made sense if they were point-like particles with paths determined by ‘pilot’ waves. ‘Bohmian’ mechanics had the advantage of explaining interference effects while restoring determinism, the idea that the properties of particles do have set values before being measured. Nature’s survey found that 7% of respondents considered this interpretation the most convincing.

Then, in 1957, US physicist Hugh Everett came up with a wilder alternative, one that 15% of survey respondents favoured. Everett’s interpretation, later dubbed ‘many worlds’, says that the wavefunction corresponds to something real. That is, a particle really is, in a sense, in multiple places at once. From their vantage point in one world, an observer measuring the particle would see only one outcome, but the wavefunction never really collapses. Instead it branches into many universes, one for each different outcome. “It requires a dramatic readjustment of our intuitions about the world, but to me that’s just what we should expect from a fundamental theory of reality,” says Sean Carroll, a physicist and philosopher at Johns Hopkins University in Baltimore, Maryland, who responded to the survey.

In the late 1980s, ‘spontaneous collapse’ theories attempted to resolve issues such as the quantum measurement problem. Versions of these tweak the Schrödinger equation, so that, rather than requiring an observer or measurement to collapse, the wavefunction occasionally does so by itself. In some of these models, putting quantum objects together amplifies the likelihood of collapse, meaning that bringing a particle into a superposition with measuring equipment makes the loss of the combined quantum state inevitable. Around 4% of respondents chose these sorts of theories.

Nature’s survey suggests that ‘epistemic’ descriptions, which say that quantum mechanics reveals only knowledge about the world, rather than representing its physical reality, might have gained in popularity. A 2016 survey¹ of 149 physicists found that only around 7% picked epistemic-related interpretations, compared with 17% in our survey (although the precise categories and methodology of the surveys differed). Some of these theories, which build on the original Copenhagen interpretation, emerged in the early 2000s, when applications such as quantum computing and communication began to frame experiments in terms of information. Adherents, such as Zeilinger, view the wavefunction as merely a tool to predict measurement outcomes, with no correspondence to the real world.

The epistemic view is appealing because it is the most cautious, says Ladina Hausmann, a theoretical physicist at the ETH who responded to the survey. “It doesn’t require me to assume anything beyond how we use the quantum state in practice,” she says.

One epistemic interpretation, known as QBism (which a handful of respondents who selected ‘other’ wrote down as their preferred interpretation), takes this to the extreme, stating that observations made by a specific ‘agent’ are entirely personal and valid only for them. The similar ‘relational quantum mechanics’, first outlined by Rovelli in 1996 (and selected by 4% of respondents), says that quantum states always describe only relationships between systems, not the systems themselves.