What does quantum mechanics mean?

At a 2011 conference on the foundations of quantum mechanics, thirty-three participants were polled on what they believed. No interpretation held a majority. The most popular answer, the one associated with Copenhagen, took 42 per cent. Everett took 18 and information-based readings 24.

Asked about the measurement problem itself, 27 per cent called it a pseudoproblem and 24 per cent called it a severe difficulty threatening the theory. Asked how much the choice of interpretation is personal philosophical prejudice, 58 per cent said a lot.

Thirty-three people at one conference is a small sample, and the authors said plainly that it cannot claim to be representative of the communities at large. It was, to their knowledge, the most comprehensive poll of its kind at the time. In 2025 a survey by Nature drew 1,101 usable answers from researchers who work with quantum theory, and the Copenhagen reading again led without a majority, at 36 per cent.

Bell tests rule out the class of theories in which particles carry their answers locally in advance, provided the choice of settings was not itself somehow fixed beforehand. Interference experiments with larger and larger objects have pushed the boundary of where quantum behaviour stops, without finding it. The latest, in 2026, used sodium clusters of more than seven thousand atoms, and in 2023 the vibration of a sixteen-microgram crystal was put in superposition over a far shorter distance.

So this is not a domain where nothing is decidable. Candidate theories do get killed by measurement, which is why the survivors are the ones that are hard to tell apart. What the Bell tests killed is not any of the interpretations on offer, though. It is the wider class of local theories with pre-existing values, which none of them was.

Some proposals make different predictions. Objective collapse models add a physical process that destroys superpositions in large systems, and experiments are closing in on their parameters. One such model, the parameter-free version of the gravitational proposal of Diósi and Penrose, was ruled out in 2020 by a search under a mountain for the radiation it should have produced.

Others make identical predictions by construction. Between those, the choice is made on grounds of simplicity, of what one is willing to call real, and of taste, which is not a word physicists enjoy using about physics. In the poll, the largest single answer on that point was that the choice is a lot a matter of personal philosophical prejudice.

The equations describe smooth, reversible evolution. Observations produce one definite outcome. Nothing in the theory says where one ends and the other begins, or what qualifies as an observation.

Decoherence explains why superpositions become unobservable very quickly in any system touching its environment, which removes much of the mystery about why the world looks classical. It does not, on its own, explain why one outcome rather than another is the one that happened.