David Mermin on Quantum Bayesianism

Many physicists, when they get older (and, in some unfortunate cases, long before that), have the tendency to reduce their powerful brains back to the era of Newton or ancient Greece and "undo" their knowledge of quantum mechanics. People like Gerard 't Hooft – and even, to a much lesser extent, Steven Weinberg and Leonard Susskind – start to pay lip service to fundamentally deluded ways to squeeze the laws of quantum mechanics into the straitjacket of classical physics, using one (or many!) of the several popular, comparably misguided strategies: hidden variables of one kind or another (including the Bohmian pseudoscience), collapse mechanisms (including GRW), the many-worlds interpretation (with some parallel universes that "really exist" just like other planets), and others.

David Mermin is an optimistic counterexample. His views have been evolving. Mermin is the actual originator of the "shut up and calculate" dictum often attributed to Feynman. But I think that when he said it for the first time, he coined it in order to humiliate Feynman's – and generally orthodox – positivist attitude towards these questions. Over the years, he got fully converted to the Copenhagen school's positivist, intrinsically subjective understanding of the quantum phenomena. He learned how to love Bohr. He realized that Einstein was just wrong in his debates with Bohr, and so on.




Now, he published a piece in Nature

Physics: QBism puts the scientist back into science

which argues that the scientist (observer) cannot quite be eliminated from the scientific process, as classical protophysics and physics has been – for a very long time, successfully – assuming. Instead, since the switch of physics to quantum mechanics, physical theories have been predicting probabilities of some particular experiences by the observers which means something about the relationship between the observer and the external world.




When experiments with the same initial conditions are repeated many times, we may interpret probabilities in the frequentist way – as a fraction of the repetitions that lead to a desired outcome. But before a single repetition, we always care about the probability because it quantifies our subjective belief that the outcome will obey a condition. Before a single experiment, the probabilities are Bayesian ones. The "collapse" is just a change of our expectations – subjective Bayesian probabilities – about our future experiences. It happens in our heads, and so on.

The same thing has been said by your humble correspondent many times and Mermin says it in new ways.

The particular buzzword "Quantum Bayesianism" (or "Qbism") is meant to describe this 2001 preprint (and 2002 PRA paper) by Carlton M. Caves, Christopher A. Fuchs, Ruediger Schack. It describes probabilities in quantum mechanics in the way they are. Your humble correspondent would probably agree with all papers in literature presenting Qbism. I just have a trouble with the "credits", with the claim that "Quantum Bayesianism" is some really new 21st contribution to physics (and also with the prominent role that is given to Thomas Bayes). It is really just a new brand that describes the very same thing that the Copenhagen school understood well. They just didn't expect that the meaning of probabilities in quantum mechanics – which is really simple and obvious for anyone who is not prejudiced – would remain a source of controversy among professional physicists for at least 90 years so they didn't write long essays and they weren't inventing new words to describe the same thing.

The 2001 paper has about 200 citations but relatively to the huge traffic in the industry of deluded, classical misinterpretations of quantum mechanics, it is not enough. So in a 2013 poll about the foundations of quantum mechanics, "Copenhagen" got over 40% while "Qbism" only received 6% of the votes.

Mermin nevertheless believes that not only all the mysteries have been solved but because of his crusade that began when he became a professor emeritus, the booming industry of misinterpretations of quantum mechanics will evaporate. It would be nice if he were right and successful but I doubt it because the human stupidity knows no borders, limits, genders, races, ages, calendars, or nationalities. It is everywhere. Only two observable things are infinite, the universe and the human stupidity, and I am not sure about the former.

Mermin also recorded an interview. Click at this podcast link, press the "»" button ("next segment") three times, and you will jump directly to the conversation about the "subjective science".

He also claims that Qbism solves "the Now" problem in philosophy, a discussion between Einstein and Carnap on whether or not the special awareness of the present is accessible by physics. One may say that this vague problem is the same thing as the presentism vs eternalism battle. And even though relativity wants us to look at things "eternalistically" because the whole spacetime is the only invariant thing, quantum mechanics really overrides it and does give the present a special status – because it assigns a special role to the observers' experiences and those are attached to some particular moments (or periods of time). I think that not-quite-well-defined debates of this sort become meaningless philosophical babbling after 5 sentences or so which is why I try not to go over this limit. The Now problem may either be viewed as an analogy (one that even exists classically) of the "existence of subjective experience" or as a special example of the latter. In both cases, the problem is "reification" of something (fallacy of incorrect attribution of "real existence" to some auxiliary theoretical constructs such as the spacetime or the quantum state, in these two cases).