I’ve written a whole book on this subject in great detail, going into all the mathematics and all the major papers on the subject.
The issue actually comes down to an underdetermination problem in the mathematics. If you claim that the systems have definite states in the real world even when you’re not looking, you have to answer, what those states are. The intuitive solution is to say that the state of the system is in some configuration with non-zero probability in the Born distribution, but then this would imply that every non-communicating observable has a configuration in that probability distribution, which the Kochen-Specker theorem shows leads to an algebraic contradiction, so it cannot possibly be true.
The way out of the Kochen-Specker theorem is to privilege a certain measurement basis. The system has a pre-existing configuration on a particular basis, not all bases at once. When you take this approach, then the mathematics already automatically fixes functions of how the other measurement outcomes arise when you go make a measurement; they end up being emergent properties that only have reality in the physical process of measurement, but that physical process is well-defined and is a function of the system’s pre-existing configuration on the privileged basis.
The issue, however, is that all bases are symmetrical in quantum mechanics, so nothing actually tells you which basis to privilege, and since your choice of privileged basis defines what the physical properties of your system even are, then the physical properties of the system become ambiguous.
A materialist thus has to take one of two approaches.
- An conventionalist approach. Admit we can’t know what the actual ontology of the world is, but that doesn’t mean we cannot talk about the world’s ontology. We can speak of it under certain materialist frameworks conventionally, based upon what makes intuitive sense to think about and describe a certain experimental context, but with no claim that this is definitively the true ontology. This does not mean the ontology is subjective: you would still be acknowledging a true one exists, but you just can’t know it.
- A non-empirical approach. If you are comfortable with making arguments that go beyond what can directly be observed or detected, then you can justify the choice of a particular basis on grounds other than pure empiricism.
To expand on #2, I said all bases are mathematically symmetrical. This is true, although only Heisenberg’s original formalism of quantum mechanics actually treats them all the same. Schrodinger’s wave mechanics specifically gives privilege in representation to the position basis, because people find this way more intuitive, and so there is a strong non-empirical argument to justify the position basis, which then is how you end up with models like Bohmian mechanics. Relativistic quantum field theory also is intentionally formulated to privilege the configuration of fields, so you could argue for that as the basis as well. (Whether or not particles are “fuzzy” or not depends upon the basis representation chosen as privileged.)
Technically, yes, all bases are mathematically symmetrical, so your choice isn’t empirically detectable, but nobody treats them that way because there are obviously ways to think about the system that make sense and ways that do not. If you represent the system on the energy basis for example, the system could clearly be moving in observable ways yet would be static in the description if the total energy content of the system remains the same.
The positivist hates both #1 and #2, so they argue that we should just give up on the idea that things have definite configurations at all when nobody is looking. But no one has shown how you can make this consistent without descending to idealism. Most physicists either embrace the idealism or resort to a kind of instrumentalism, some claim that it can be made to work, mainly the Many Worlds’ proponents, but none have demonstrated any methodology of making it work; it is always self-contradictory or outright circular.
The conventionalist approach is also great for quantum computing, because it is incredibly rare for anyone to talk about changing basis representations in quantum computing. You pretty much always represent the system on the computational basis throughout; so the conventionalist approach, for a person working with quantum computers, can just take the computational basis of the computer as the privileged basis, and then you can represent it quite trivially easily as just a stochastic process whereby the qubits are just bits that have bit values at all times, and the logic gates are just stochastic perturbations upon those bit values.
The issue was pointed out decades ago by the Soviet physicist Dmitry Blokhintsev, that ψ is quite obviously a description of ensembles of systems. If you have large numbers of systems, the configurations, sampled over many runs of the same experiment, form clear physical patterns. But, given a single run of the experiment, you cannot predict how they will behave, and so then ψ becomes subjective, just referring to your “best guess” as to how they will behave based on previous knowledge of how they behave over an ensemble.
If you promote ψ, in an individual run of an experiment, to a physical object, then you are elevating something subjective to something physical. The consequence is that you end up with a reality that depends upon people’s subjective knowledge of it, and, as a consequence, it leads to a breakdown of objective reality, and there just isn’t any way to avoid that, other than recognizing that it was flawed to begin with to treat ψ as a complete description of an individual system. It is only a complete description of an ensemble of systems.
I used to see physicists as a beacon of truth until I started to learn physics and realize how much crackpottery is actually taken seriously in academia. I feel like we Marxists need to be more outspoken about how ridiculous contemporary 21st physics actually is.
In 1905, Einstein put forward his theory of special relativity. Despite people claiming it is “the most tested theory of all time,” it was actually mathematically and thus empirically equivalent to a theory Lorentz proposed in 1904, so it made no new predictions at all. However, it was simpler, so physicists liked it for its simplicity.
It’s simplicity, however, came at a cost. Einstein effectively promoted the speed of light limit to a universal principle for all things, a “speed of causality,” something Lorentz’s theory did not do. This means, technically, there is 1 way to distinguish between Lorentz’s theory and Einstein’s: if anything can violate the speed of light limit, then Einstein’s theory is ruled out.
In a paper published in 1964, the physicist John Bell pointed out that technically quantum mechanics can, and then in 2022 a Nobel prize was awarded for empirically verifying that these effects are real.
So, is special relativity ruled out? No, because physicists found a loophole. You see, Bell’s theorem works by measuring things in two different locations, bringing the measurements together, comparing them, then asking, “what was the system doing, when I wasn’t looking, to explain these results?” Bell’s theorem therefore requires you to make an inference about the state of the system when you were not measuring it.
Indeed, this leads into our loophole: the no-communication theorem. It proves that you cannot use these faster-than-light effects to send messages. Why? Precisely because they never show up on what is immediately measured in the instantaneous moment of the measurement. They only show up if you ask what the system was doing when you were not measuring it.
Hence, that is the loophole: say that nothing exists when you aren’t looking, or at the very least, claim that asking what the universe was doing is a meaningfully philosophical question that doesn’t belong in physics. This is known as “anti-realism” and has become the dominant paradigm in modern physics, and is why there is a “measurement problem” to begin with. Relativistic quantum field theory simply does not model reality independently of what shows up immediately on the measuring device in the instantaneous moment of measurement, and thus is not meaningfully a theory of reality and as a consequence has to take “measurement” to be a fundamental component of the theory, yet cannot even explain what these measurements are event measurements of.
This is why you have sometimes heard Bell’s theorem reframed as having ruled out not locality, but “local realism.” Physicists started to say it rules out “local realism” because if you say it rules out locality, then special relativity is debunked. But by saying it rules out “local realism,” then it implies you can salvage special relativity by adopting local anti-realism instead, which is the path that 21st contemporary physics has taken.
This, in my opinion, is a complete absurdity. If the empirical evidence rules out your model, claiming reality doesn’t even exist as a way to preserve the model is just religion, not science.
If you abandon this idea that we need to hold special relativity sacred, then it is quite trivial to fit quantum mechanics to a realist model. Indeed, Hrvoje Nikolic has even shown it is quite trivial to do it with relativistic quantum field theory. All you have to do is abandon the assumption that Lorentz covariance must be preserved, and instead adopt a weaker assumption that the empirical predictions must be preserved; i.e. it shouldn’t matter that it violates some sacred tenet if it makes the same empirical predictions.
Realist models do not need to hold the position that systems have no real properties until you measure them. They can define reality independently of the measuring device, and what you measure then just becomes a function of what was already there. Nothing comes into being at the moment of measurement.
There are things in this article I don’t agree with because I think they play into the anti-realist paradigm dominant in 21st century physics when in my opinion it is woo woo crackpot hogwash and needs to be criticized, rather than trying to “make sense of” it. It cannot be made sense of, it can be “brought down to earth” so to speak. It is inherently incoherent and has to be abandoned.
For example, you say:
I agree it is non-local and holistic, but not in the way you speak of, of measuring one “instantly determining” the other. Indeed, if you believe this occurs as a physical process, then it runs into mathematical issues.
Consider that your measurement device is so simple that it has 1 qubit of memory and you also measure something so simple it has 1 qubit of memory. Consider a circuit of the following form.
This would represent placing q[0] into a superposition of states and then measuring it onto the measuring device’s memory of q[1]. This circuit gives us the density matrix [0.5 0 0 0.5; 0 0 0 0; 0 0 0 0; 0.5 0 0 0.5]. If you, however, believe that “at the moment of measurement” the system “collapses” down to a definite state at random given by the Born rule, then it would instead be expressed by the density matrix [0.5 0 0 0; 0 0 0 0; 0 0 0 0; 0 0 0 0.5]. If the two density matrices are different one from one another, then a future evolution of the system can lead to different empirical predictions.
It then becomes necessary to answer the question as to whether or not a device consisting of 1 qubit recording 1 qubit from another system really qualifies as a measurement, because you get different empirical predictions based upon your answer. If yes or no, why or why not? Thus, we then run into a problem that Bell pointed out in his paper “Against ‘Measurement’” which is that any measurement-centric approach is actually physically ambiguous unless it rigorously defines what set of physical processes actually qualifies as a “measurement,” and any such theory will inevitably deviate from the predictions of orthodox quantum mechanics at the boundary as to what it defines a “measurement” to be.
Dmitry Blokhintsev also briefly discusses this in his book “The Philosophy of Quantum Mechanics,” that these “collapse” approaches break from the usual linearity of the theory, and so they cannot be made consistent with it in all regimes, and so he was convinced that the “correct” approach must be one that maintains the theory’s linear structure, so there can be no “collapse.”
However, Blokhintsev was not a “Many Worldist,” because he denied the “spreading out” to begin with. He argued that a superposition of states is an ensemble description of the system, not a description of the individual system, as if the state is literally in multiple places at once. If it is not in multiple places at once, if it never “spreads out,” then there is no reason for it to “collapse” when you make a measurement. It always contains some definite value in the real world, we just don’t know what it is due to the system’s random evolution.
Blokhintsev believed that the answer to this was indeed in holism, specifically, by interpreting quantum mechanics as a kind of holistic stochastic process. However, he never formalized that. Personally, I believe the formalization of this is in Jacob Barandes’ paper “The Stochastic-Quantum Correspondence.” He introduces a concept called unistochastic dynamics which allows you to derive the Born rule probabilities without invoking the wavefunction at all, by instead composing all unitary operators (describing the physical arrangement of the physical setup) that the particles pass through up to a given time interval.
What you get is an approach where you do not need to pretend things have no properties until you measure them, but that their properties simply change statistically in a way that depends upon the whole experimental arrangement. In the Mach-Zehnder interferometer for example, a single beam splitter has a 50% chance of reflecting the photon and a 50% chance of transmitting it, but if you have two beam splitters in a row, then the second beam splitter has a 100% chance of sending it down a single path.
This is usually explained by most contemporary physicists by saying that the particle just have no real definite location until you look. But in unistochastic dynamics, the explanation is instead that the second beam splitter’s behavior just changes because it is situated after another one, even though the piece of glass is identical between both of them. The dynamics of the particle, in this case how it reacts to the glass, depends upon the whole experimental arrangement, and so the same piece of glass can have entirely different behaviors in a different context.
The first piece of glass just has a 50% chance of reflecting or transmitting the particle, and the second just has a 100% chance of sending it down one particular path, because the two behave differently. Nothing requires you to deny that the particle had a position to begin with independently of you looking at it, and somehow acquires one only when you look.