Why QBism is not the Copenhagen interpretation and what John Bell might have thought of it (arXiv + video)that he gave in Vienna 3 months ago. The conference was dedicated to John Bell and the 50th anniversary of his theorem. I agree with many statements that Mermin is making (and was making, in recent years) about the foundations of quantum mechanics. But that's not the case of some of the focal points in this talk.
First, I don't think that John Bell was such an important physicist that we should spend too much time with speculations what he would think about some ideas proposed after his death. Bell didn't discover quantum mechanics, he wasn't even in the top 100 ring of its co-founders and their first-generation followers, and his expectations about the fate of quantum mechanics turned out to be wrong. He didn't coin QBism and related concepts, either. Would Bell like QBism? Yes, no, who cares?
It was a conference about Bell which may (at least partially) explain why Mermin cared.
Also, the talk was meant to be nice to Fuchs and Schack, two guys behind the QBism meme, which may explain why Mermin tried to present QBism as a new idea – even though it is not a new idea – and, in fact, as an idea that the notorious critics of proper quantum mechanics such as Bell may have liked - even though Bell and others would almost certainly hate QBism, too.
But let me discuss the 14 pages of the arXiv preprint a bit more systematically.
The first page contains the abstract. On the second page, you find a dedication to a recently deceased colleague of Mermin's, Mr Chester, who worked hard to keep Mermin attached to the ground. There are some quotes on the third page. For example:
Our students learn quantum mechanics the way they learn to ride bicycles (both very valuable accomplishments) without really knowing what they are doing.This is quite a provoking quote because it is a direct attack on the "shut up and calculate" pedagogic approach to the teaching of quantum mechanics. Recall that Mermin coined the "shut up and calculate" dictum as a slogan that expressed similar dissatisfaction as Bell presented above.
John Bell, a 20/8/1980 letter to Rudolf Peierls
Well, I think that Bell was wrong about this pedagogic issue, too. It is indeed wise to learn things like quantum mechanics by a similar general strategy as the usual strategy by which we learn to ride bikes. Some months ago, I was teaching my nephew and niece (5- years old at that time) to ride the bike, too, and I would claim much of the credit about the ultimate success. How should the kid learn to ride the bike without the auxiliary wheels on the sides?
Believe me or not but I actually remember the moment when I was 3 and I found it impossibly difficult to abandon the side wheels. The bike cannot stand vertically, can it? As you know, when the bike reaches a sufficient speed, it actually doesn't fall (although the precise reason, especially the question whether the angular momentum conservation law is enough, is always able to ignite a controversy). It is totally paramount for the kid to experience this miracle. You must push him or her quickly or instruct them to ride down the hill and they suddenly learn – much like I did – that the bike may stand vertically if the speed is sufficiently high. The kid instinctively and quickly learns that when she or he reaches that speed, things just start to work. The kids actually know how to pedal – this is almost never the stumbling block. They may continue to have some problems with starting but this may be learned as an improvement of the basic skill.
I would insist that the basic mystery is about the trust or confidence in the bike and its balance at a high enough speed. The case of quantum mechanics is analogous. The students should shut up and calculate certain things that give clear results, in some cases directly verifiable by experiments. These practical skills then define their basic constraints into which their philosophy has to fit. If the philosophy doesn't agree with the practical skills, more precisely with the verifiable predictions, then it's wrong.
If someone tries to teach quantum mechanics or how to ride a bicycle in the opposite way – encouraging the kid or the student to develop lots of philosophies how the bike or the microscopic world should behave, before they actually know how to control the bike or quantum mechanics even in the simplest cases – then you may be pretty sure that they will converge to a wrong philosophy about the bikes or the microscopic world and they will never be able to learn the things properly. In science, correct philosophy may only be obtained by extracting some general properties of the genuine, down-to-earth, practical insights that have been accumulated. Well, to be sure, big principles sometimes do "produce" the whole theoretical framework. Einstein did it in this way twice. But you should understand that if you're less smart and lucky than Einstein – and a vast majority of the people are less smart and perhaps also less lucky – then the "product" will only be a pile of rubbish. Philosophy cannot come first in science research or in the learning of science.
Be sure that certain famous men agreed with my comments about teaching – and not just teaching of quantum mechanics – entirely. In a story about the awful textbook forcing kids to repeat "Energy/wakalixes makes it go", Richard Feynman would say that abstract notions such as energy should only be taught when the kids have already learned how things work at some simpler, more instinctive, more heuristic, less abstract level – when they have already internalized a sufficient body of the data:
What they should have done is to look at the wind-up toy, see that there are springs inside, learn about springs, learn about wheels, and never mind "energy." Later on, when the children know something about how the toy actually works, they can discuss the more general principles of energy.There are two more quotes from Bell's letters to Peierls in Mermin's paper that I didn't find too interesting. Incidentally, at the end of the paper, Mermin mentions that Peierls himself disliked the term "Copenhagen interpretation" because this term incorrectly suggests that there also exist other "interpretations" of quantum mechanics. There aren't any. I completely agree with Peierls. Quantum mechanics is a well-defined theory, not an abstract painting inviting diverse interpretations of critics who are mostly high (or drunk). The whole idea that quantum mechanics should be supplemented with (new) "interpretations" is a delusion by drug addicts who think that the amount of drugs in physics is too low.
Back to the paper.
Mermin says that QBism is at least as deep a revolution in all of science as cubism was a revolution in arts. Whether people should celebrate cubism is debatable because arts is always debatable. But science is not arts and the idea that QBism is fundamentally new is ludicrous. The tools and rules of Bayesian inference have been around for centuries and the "fundamentally epistemic" and "intrinsically probabilistic" attitude to quantum mechanics has been a part of the theory from its beginnings in the 1920s, too. QBists may like some new words and they may like to rewrite some equations involving probabilities in a different way but the paradigm isn't really bringing any conceptually new ideas.
I agree with the fourth page of Mermin's paper. The page says that our knowledge about the world ultimately boils to the personal experience. So it is true that any legitimate scientific argument must fundamentally boil to the personal experience because that's how we get the information about anything. There is nothing wrong about this ultimate reduction to the personal experience because it's really unavoidable and the scientific theories have no duty to add something on top of the personal experience, such as the objective reality.
At the same moment, this observation doesn't mean that the external world only exists in one's head. Mermin says that the accusations of "solipsism" are silly because the focus on the "personal experience" in no way implies that there should only be one user of quantum mechanics. It doesn't imply that one shouldn't expect any agreement between the different users, either. In quantum mechanics, one may prove lots of this agreement. Quantum mechanics just disagrees with the claim that the maximum knowledge exists and is universal for everyone – it disagrees with the existence of the "objective reality".
Pages 5 and 6 offers some QBist-like quotes by Sigmund Freud, Niels Bohr, and others. However, what Mermin means by "QBist-like" is so vague that of course pretty much everyone has sometimes said things that would be "QBist-like" by this vague definition. But that doesn't mean that people actually agree about all key questions. At the bottom of Page 6, Mermin reveals that he isn't addressing the talk to those who refuse QBism but to those – like your humble correspondent – who maintain that QBism is nothing new (relatively to the Copenhagen "interpretation").
On Page 7, he admits that QBism and Copenhagen agree that the mathematical objects such as the wave function are not "objective properties of a physical system" but "tools to think about the physical system". But Mermin wants to claim that they're not the same thing. Why? A bizarre answer appears on the bottom of Page 7:
A fundamental difference between QBism and any flavor of Copenhagen is that QBism explicitly introduces each user into the story, together with the world external to that user.This is almost laughably silly, of course. Copenhagen said exactly the same thing. It just used different words (well, German words) for the same concepts. The "fundamental difference" between QBism and Copenhagen is that Copenhagen called him the "observer" while QBism (or at least Mermin) calls him the "user" (or, as on the next page, "agent"). His role is exactly the same, however. He always assumes that there exists an observable physical object – or "external world" containing such objects (it's again the same thing, the two formulations differ by using the singular and plural only) – whose properties may be obtained by observations and the equations of quantum mechanics may predict the probabilities of different outcomes of future observations.
Mermin writes several paragraphs about this alleged "difference" between QBism and Copenhagen. These paragraphs contain the (bizarre) word "user" many, many times, but he completely avoids the word "observer". But foundational discussions on Copenhagen have always contained many copies of the word "observer". Much of Copenhagen is about the "observer". Mermin has either forgotten about this fact or he is demagogically avoiding the word "observer" because he knows that most readers would then realize that Mermin's claims about the "users" and "observers" who have nothing to do with each other are completely silly. That's particularly the case of statements such as
Science is about the interface between the experience of any particular person and the subset of the world that is external to that particular user.Holy cow, Copenhagen school's theory of quantum mechanics is all about this interface, too. Mermin has used several new words such as "users" and "agents" in order to make old ideas look new. But even when it comes to this amusing trick, he failed in the case of the word "interface". If you look at Wikipedia's entry describing the Heisenberg cut, you will see that the Heisenberg cut is defined as a "hypothetical interface", by almost exactly identical words that Mermin ascribed to QBism above. There is an observer/user who cares, there has to be an external object/world that he observes, and the theory predicts properties he cares about.
The exact location of the Heisenberg cut doesn't really affect the predictions as long as you do everything right. The only condition is that you are not allowed to place the cut to "too long length scales" because small enough (or otherwise "too quantum") objects simply have to be described by quantum mechanics, with all of its superpositions etc., and the classical approximation is not good enough. But as long as you describe all the "unavoidably quantum" features of the world using the quantum variables (as the observed object), and only the "clearly classical-like" limit in terms of the observers, the Heisenberg cut will perform its job correctly. One may also say that the minimum "subset" of the world that has to be described quantum mechanically is determined by the decoherence scale. If some quantities are quickly decohering at the accuracy you need, you may treat these observables as if they were observables in classical physics – but you don't have to. The full quantum description always works for anything in the external world.
I find Mermin's claim that there is no "interface" like that in Copenhagen doubly ironic because the whole Copenhagen school has often been criticized for the very existence of this "interface". In fact, if you continue to read this essay, you will see that Mermin has criticized Copenhagen for this "interface" as well! ;-) At this point of the paper, however, he promotes the very same idea – in fact, using pretty much the same word – and he sells it as a "difference from" or an "advantage of QBism relatively to the Copenhagen interpretation". Bizarre. It is not a difference from the Copenhagen interpretation. It is one of the defining features of the Copenhagen interpretation.
Landau-Lifshitz and "interpretations" of quantum mechanics
You know, I have read most of the volumes of the Landau-Lifshitz textbook on theoretical physics - when I was a high school student and later a freshman (I still have the Slovak edition of several volumes). It is an excellent textbook by a Soviet genius and his student. The legend has it that Landau wrote the whole initial volumes "in his head" while he was spending a year (1938-1939) in an NKVD prison (be sure that NKVD was worse than KGB). However, not a single word was actually written by Landau which produced the popular witticism:
Landau-Lifshitz: not a word from Landau, not a thought from LifshitzSomeone may have been inspired by this slogan when she invented a similar witticism (surely an inaccurate one because GS wrote most of the Volume I and the chapter on the light-cone gauge superstring field theory in Volume II) about a 1987 textbook on superstring theory:
Green-Schwarz-Witten: the cover is green, the fonts are schwarz (black), and it's written by Witten.But back to Mermin's claims about the Landau-Lifshitz picture:
LL: "It must be most decidedly emphasized that we are not discussing a process of measurement in which the physicist-observer takes part."That's supposed to be Mermin's justification for his conclusion that Landau and Lifshitz – while they are using a version of the "Copenhagen interpretation", as Mermin admits – contradict the QBist picture of physics. But this justification is totally indefensible.
Mermin: They explicitly deny the user any role whatever in the story.
The problem is that the quote by Landau and Lishitz isn't saying that the word "observer" is prohibited (an interpretation by Mermin on Page 10) or that there are no observers or that there are no physicists (WTF?) or that the properties of physical objects (in the initial state) may be "learned" without an observation. They just say that they don't discuss the particular process of observation or some philosophical fairy-tales about this process anywhere in the book. They don't discuss it because it is not needed to understand what the physical laws can actually calculate!
Calculate the energy spectrum of the hydrogen atom, for example (I could really pick any example). It tells us what energies of the hydrogen atom are allowed. We may measure the spectrum in certain ways, using various apparatuses. But the precise procedure employed to measure the spectrum isn't needed in the calculation of the spectrum. The measurement is actually needed for any "user" to know what the state of the hydrogen atom (or any other physical system, object, or a feature of the external world) is. But the measured observables make sense and may be probabilistically predicted independently of (and before) the measurement.
It's an entirely anti-Copenhagen – and anti-QBist – misconception that much of the complicated stuff that needs to be explained by (new) equations is happening during the measurement and the characteristic quantum flavor of the microscopic world arises because of some special features of this measurement. The measurement is needed to find the value of an observable and it inevitably affects the measured object. The measurement apparatus works because of some physical laws that may be explained quantum mechanically, too. But the observable describing the measured object doesn't depend on any complicated properties of the measurement. When it comes to predictions, the measurement plays a completely trivial role and may be entirely eliminated from the difficult quantum mechanical calculations – and Landau and Lifshitz are indeed eliminating this philosophically flavored "observer side" of the problem very systematically!
And indeed, the fundamental thesis of QBism – i.e. Copenhagen – that the probabilistic nature of the wave function is intrinsic means that one shouldn't invent any "mechanisms" that would "transform" the wave function into "what we really observe". No extra "mechanism" is needed. No extra chapters or volumes of the Landau-Lifshitz textbook (about the measurement) are missing. The wave function of the object of interest already contains the information that the observer needs. By a measurement, he may actually find out but the measurement is only producing one of the results whose probabilistic distribution was completely determined before the world knew which apparatus we would use to measure the observable. When you calculate the probability amplitudes for the process that you care about, your physics derivation is complete, perhaps except for a trivial squaring according to Born's rule. That's how it works in Copenhagen=QBism. If you know the probability that an observable has one value or another, you know everything that you may in principle know about an observable before you actually measure it. What you know after the measurement is probabilistically determined by this distribution. There's nothing else that should be discussed concerning the value of the observable at a given moment. The (Bayesian) probabilities are not constructed out of little wheels and gears. They are fundamental quantities by which the theory (quantum mechanics) organizes the knowledge and laws.
That's also what Landau and Lifshitz mean in another quote that, as Mermin incorrectly believes, support his claim that the Landau-Lifshitz attitude was incompatible with QBism:
By measurement, in quantum mechanics, we understand any process of interaction between classical and quantum objects, occurring apart from and independently of any observer.They just define the measurement as the physical process that is allowed to be used by an observer to "learn" about something, but it doesn't have to be used by anyone. Instead, what matters in their definition is the external appearance of the process. One may define words, in this case "the measurement", in any way he wants as long as it is internally consistent (and this one is). This definition was good for their purposes because they wanted to avoid nonsensical distracting debates about things like consciousness. (Well, some people think that the more rubbish about consciousness they add, the deeper their discussion is. Such people may be very self-confident but they're dumber than Landau's thumb which mostly explains that their priority is exactly the opposite than Landau's.)
This doesn't contradict the claim that a "user" is the only one who will actually find the insights of quantum mechanics "useful". There are lots of processes but the only way to find out about the processes and objects is for a "user" to "use" the laws of quantum mechanics and make predictions about the external world that are, in practice, revealed by the observations. The revelation of a particular state of a particular object (feature of the external world) is subjective and user-dependent but the laws governing the relationships between the experiences of any observer are objective. The Landau-Lifshitz textbook is exactly a "user manual" for a user to use the world, a term that Mermin connects with Fuchs and Schack.
Again, I find Mermin's claims extraordinarily ironic because the very point of these Landau-Lifshitz propositions was to make the reader certain that all these derivations may be interpreted as derivations about the "external world". The claim that one may still assume that there is an "external world" aside from the "observers" (and if you check the Landau-Lifshitz quotes again, they surely do agree with the existence of both!) was mentioned by Mermin as a defining feature of QBism on the bottom of Page 7. Now, he uses the very same combination to argue that Landau and Lifshitz contradict QBism! It's very strange.
Maybe the issue is that Landau and Lifshitz didn't "introduce" the observer sufficiently warmly, with the Slavic bread and salt. But Landau and Lifshitz have surely agreed that they are observers who are observing things and using the laws of physics to predict and explain the phenomena. They just correctly emphasized that none of the special features of some observers are needed to calculate the spectra of atoms and molecules, the black body curve, or anything else they cover in the volumes about quantum mechanics. All this mathematical beef of quantum mechanics is entirely independent of the properties of the observers who may use it which is why it was possible for Landau and Lifshitz to "cut" this layer from the textbooks entirely. This "shut up and calculate" approached is fortunately used by most textbooks on quantum mechanics. This doesn't represent any "omission" because bullšitting about the spiritual or otherwise philosophical "interpretations" of a physical theory, while possible, doesn't belong to science. Science explains and predicts natural phenomena – that we inevitably learn about by our observations, measurements, perceptions, call it in any way you want, it's always the same thing.
I am sure that Fuchs and Schack also agree that it is possible to entirely eliminate any discussion about the properties of observers or details of the observational processes when we want to calculate the spectrum of an atom, a molecule, a piece of metal, transition amplitudes or probabilities in a scattering, or anything else in physics (I mean every individual section of the Landau-Lifshitz textbook, for example). The observer is needed for a particular observation but he doesn't affect the laws of physics – the properties of the external world may be calculated independently of any particular observer. For spectra, one only needs to know "which system" we talk about. For transition amplitudes etc., we need to specify the initial (and final) states. But every and any observer may find his object in this initial state and when he does, he may use the universal – objective – laws of physics to deduce what will happen with what probability. The laws what may exist and what will probably occur assuming some facts about the situation (and these are appropriate topics for a physics textbook) are objective; the individual particularly felt "situations" and "outcomes" are according to quantum mechanics subjective.
The idea that there is any contradiction between QBism on one side and Landau-Lifshitz (or any "user" of proper quantum mechanics) is fallacious. If there is a difference, it's at most in the formalities, like the fact that Landau and Lifshitz didn't offer the bread and salt to the visiting "users". But physics isn't about the formalities and bread and butter that someone may offer. Physics is about the claims about the phenomena and there is no difference between Copenhagen – as articulated by Bohr, Heisenberg, or Landau, among others – and QBism.
On Page 8, Mermin also says that a pair of humans cannot be a "user" according to QBism. It depends. A human experiences the external world because of his eyes, ears etc. There are some "rather straightforward biowires", the nerves, in between, and the eyes are clearly just another gadget. Of course that an observer such as the Director General of CERN may view the CERN employees as additional ears and the channels used to communicate the measurements to him are the "nerves". There is no fundamental difference. The true consciousness may reside in Heuer's head only but what one identifies as the "limbs" that collect the data or experiences is flexible. This is nothing else than a discussion about the Heisenberg cut again. Perhaps, it is combined with some puzzling neuroscience. After all, a human is just a collective of cooperating cells so if the cells may (aside from this dull individual "consciousness") share "collective consciousness" in this way, it must surely be at least qualitatively possible for groups of people or societies to share "consciousness", too. To judge these things in a qualified way, one has to apply the laws of physics and biology on the cells, mammals, and societies. There is no strict boundary anywhere and quantum mechanics isn't obliged to give you an algorithm that counts "conscious souls". A "user" may choose "himself" to include the eyes, limbs or employees – or not. What's important are not the other cells, eyes, limbs, or employees. What's important is the structure of truth values about observables (or their real c-values) that a particular user works with and cares about. As long as we obtain trustworthy information about the observables in the initial state, we may calculate the odds of future observations, too. That's true if "we" are a "single cell", a "single human", the "ATLAS collaboration", or the "mankind". These comments were meant to describe proper – Copenhagen-discovered – quantum mechanics but I am convinced that Schack and Fuchs would agree that QBism doesn't allow to count souls or to decide when the individual consciousness of cells is replaced by their shared one, or things like that. QBism is about the abstract structure to assign values or truth values (in general, probabilities) to observables and propositions.
Mermin's criticism of Landau and Lifshitz culminates as follows:
Their insistence on eliminating human users from the story, both individually and collectively, leads them to declare that “It is in principle impossible... to formulate the basic concepts of quantum mechanics without using classical mechanics.” Here they make two big mistakes: they replace the experiences of each user with “classical mechanics”, and they confound the diverse experiences of many different users into that single abstract entity.There is no mistake. Any observer of a type similar to the humans "runs" on classical physics. This is no speculation and we don't need to emit vague, inconclusive conspiracy theories such as a recent New York City talk by Scott Aaronson who has "conjectured" that decoherence was needed for "consciousness". (Just to be sure, I emphasize that I basically agree with Scott – I just don't like how he makes certain obvious things sound like speculations.) What I say may be proven absolutely easily and rigorously. An observer is someone or some agent that considers some answers about the external world – answers to questions he ultimately cares about – to have classical No/Yes (0/1) values. They commute with each other which makes them by definition classical. So what one sharply perceives is always classical, pretty much by definition, and it is Mermin, and not Landau and Lifshitz, who are fundamentally and importantly wrong.
Bohr is criticized "less" than Landau and Lifshitz but Mermin still doesn't like many ideas of Bohr's, e.g. Bohr's assumption that the classical measurement aparatus is single and large. However, it's simply true that if there are several disconnected (constructed of mutually commuting observables e.g. in distant, cluster-decomposed parts of the world in a quantum field theory) classical apparatuses, they may always be treated as components of a single disconnected apparatus. In particular, all practically observable entangled states of these objects may be described as classical correlations – to the same extent to which the objects themselves are classical. The information about the relative phases of the practically distinguishable microstates is lost, so all the practically measurable observables commute with the density matrix and the eigenvalues of the density matrix may be interpreted as the classical probabilities.
On Page 9, Mermin criticizes the "shitty split". Oh, I see, it is a "shifty split". Doesn't matter. What's important is that it is the Heisenberg cut which has already been discussed – and it's exactly the same thing as the "interface" that Mermin enthusiastically embraced as the basis of all science in the middle of Page 8. Now this "glorious interface" becomes a "shitty split" and is being criticized as the source of the world's evil. Can someone not see the self-evident inconsistency in Mermin's comments?
The anti-Copenhagen screaming acquires a profoundly populist tone on Page 10. Mermin approvingly quotes Bell who said that "physics was not limited to piddling laboratory tests". We're told to believe that the Copenhagen folks' picture of the world doesn't admit other measurements than boring and piddling measurements in a laboratory. Are you serious? Do you really believe that Bohr didn't know that you may observe things on a sunny beach, while drinking wine with a beautiful woman? Do you really think that Bohr's or Heisenberg's rules of the measurement break down outside a clean observatory or laboratory (or at least a clean lavatory)? I can't see anything else in Mermin's would-be argument than a cheap populist trick to dehumanize the founders of modern physics. Of course that they knew everything that they're claimed to have been ignorant about.
In the next paragraph, Mermin tries to picture the difference between Copenhagen and QBism as the contrast between "classical and personal":
Because outcomes of Copenhagen measurements are “classical”, they are ipso facto real and objective. Because in QBism an outcome is a personal experience of a user, it is real only for that user, since that user’s immediate experience is private, not directly accessible to any other user.The problem with this Mermin's would-be argument is that the two descriptions of the personal experience don't contradict one another at all. The perceptions are personal and subjective but they are classical, too. They are classical because the Yes/No truth values are numbers, either zero or one, and those commute with each other. Because everything that an observer is actually aware of commutes with each other, the perceptions are classical.
However, the identification of the subset of the commuting observables describing the external world isn't universally unique in general. Different observers may in principle talk about mutually commuting observables in their classical limit that, however, refuse to commute with the observables in another observer's classical picture.
By definition, a classical description of the external world is one in which the observables – the things that may actually be observed – commute with each other which implies that they may be thought of as objectively knowable. But this definition doesn't say that they have to commute with observables in completely different (different observers') descriptions! Also, the adjective "classical" doesn't mean that "just the probabilistic knowledge" of the observables must be impossible. Indeed, classical statistical physics is a classical theory where one deals with the probabilistic knowledge about the classical observables. So this attempt to find a disagreement between QBism and Copenhagen is completely wrong, too.
On Page 11, we can hear a mantra:
Asher Peres’ famous Copenhagen mantra, “Unperformed experiments have no results”, becomes the QBist user’s tautology: “Unexperienced experiences are not experienced.”The interpretation where the mantra is a tautology is undoubtedly clearer. But it is not true that this clearer picture was unknown to Heisenberg and others. It's really how they started. The positivist attitude – they would use the word "positivism" for pretty much the same ideas as QBism – was that one should only discuss the observations of observables that were actually observed because observations of observables that no one has observed haven't really taken place. ;-) And science isn't obliged to talk about things that may be viably assumed not to exist. You may see numerous equivalent comments in the quantum founding fathers' papers. I protest against the attempts to steal the credit from the heroes of science who have actually realized, discovered, and intensely utilized all such insights.
Mermin's superficial verbal tricks to steal the credit from Bohr, Heisenberg, and pals continue on Page 11 where he blames them for saying that the wave function encodes the observer's "knowledge". The right word should be "belief", as in QBism, Mermin argues. Holy crap. They meant exactly the same thing. Moreover, they haven't used the word "knowledge" but a German counterpart which isn't quite identical. If I measure something with a perfect apparatus, my "belief" that the result is a particular number or Yes/No is so reliable that I can call it "knowledge". On the other hand, if my knowledge has a probabilistic character, like the knowledge of an observable that I will measure in the near future, the probabilistic knowledge is "unsharp" enough so that we may call it a "belief" instead. Both these aspects of knowledge/belief appear and both words are equally legitimate. At any rate, these verbal exercises by Mermin show how important it is to shut up and calculate. If you start to criticize Bohr et al. because they didn't use the word "belief" for "probabilistic knowledge", then you shouldn't just shut up but you should – please allow me to say it a bit more politely - shut your dirty f*cking mouth that is full of šit. Physics isn't about which word among near synonyma we use to describe the same idea and even if it were, Mermin has presented no evidence, not even a sociological one, that his choice of words is better.
Mermin adds that no discoverer or advocate of proper quantum mechanics has ever considered the wave function to be tied to an observer. All of the sane and important ones have. It has always been their point and it's what the century of deluded criticisms of the Copenhagen picture has been all about. To suddenly say – after this unjustifiable criticism – that the quantum founders had never discovered the observer-dependence of the wave function is what I call chutzpah. Among dozens of billions of people who have ever lived on Earth, they're the first ones who must be praised for those discoveries. You may find tons of flavors of the propositions which ultimately differ by the personalities only. For more "new logic is needed", see von Neumann – Hugh Everett got most of his fame by adding a left-wing, multi-world bias to his plagiarism of von Neumann. Wigner was a bit more spiritual than others but the beef was the same. And so on.
On Page 12, Mermin claims to possess a new understanding of outcomes that are certain i.e. have probability one. Mermin accuses all physicists before QBism from thinking that the existence of probability-one outcomes requires the existence of a deterministic mechanism. Wow, this is so stupid. If an outcome has probability one, it is certain. If such an outcome may be derived to be certain (i.e. probability one) just from assuming certain things about the initial conditions, then the outcome is determined by the initial conditions. That's what the word "determine" means. If one could talk about these certain quantities only, one could say that the mechanism that produced the outcomes was "deterministic". However, there are other measurements and their outcomes whose probability is strictly between zero and one. They are not certain which is why there is no deterministic mechanism that would predict outcomes of all measurements. All founders of quantum mechanics and most of their physicists-contemporaries (including EPR) were aware of these trivialities of mathematical logic and anything that contradicts these trivialities is self-evidently wrong.
I didn't find the extra quotes by Bell at the end intelligent or witty or provoking or novel so let me omit that part.
At any rate, this sport of spitting on the founders of quantum mechanics who got everything that matters correctly is deeply pathetic. If we lived in a scientific world, they would be celebrated and their priority and their victory in all the intellectual confrontations would be universally known. Instead, we are drowning in the mud of ambiguity and downright hostility, even when we listen to people who should know better.
This is not a scientific world.