Sabine Hossenfelder on free will and “superdeterminism” of quantum mechanics

December 23, 2021 • 9:30 am

I had a bit of a hard time fully understanding this absorbing 20-minute video by physicist Sabine Hossenfelder, but I think I get most of it. The main problem I had was understanding the notion of “superdeterminism” in quantum mechanics (QM) and what it really means for things like the famous double-slit experiment.  But, like reader Darrell, who sent it to me, I think you need to listen. She might convince you that quantum mechanics isn’t really indeterministic!

Hossenfelder is intrigued by the notion of libertarian free will (which she rejects) and maintains that a belief in this sort of dualism was held by many physicists working on QM. As you probably know, interpretations of quantum mechanics have differed historically, with some having maintained that QM is truly indeterministic. (Hossenfelder defines “determinism” as the system in which “everything that happens is a result of what happens before”.) Most advocates of QM think that it is not deterministic, but inherently indeterministic. Einstein never believed that, rejecting that idea with his famous assertion that God doesn’t play dice with the universe.

As far as I knew, “Bell’s theorem” and subsequent tests of it completely rejected any determinism of quantum mechanics and verified it as inherently indeterministic. But, as Hossenfelder argues in this video, this is not so.  She argues that a sort of “superdeterminism” holds in quantum mechanics, so that, in the end, everything in the universe is deterministic according to the known laws of physics.

I’m not quite sure what “superdeterminism” means is on the level of particles, but it appears to be something like this: “What a quantum particle does depends on what measurement will take place.” And once the measurement system is specified, somehow a quantum particle is determined to behave in a certain way. That’s what I don’t get.

But my inability to understand it may be because the idea of superdeterminism is inherently mathematical (she gives a simply equation for “superdeterminism of quantum physics”). Like in QM itself, everyday interpretations of superdeterminism might not make sense. Any reader who understands the concept is invited to explain it below. (Briefly, if possible!)

At any rate, Hossenfelder agrees with Einstein: there is no dice-playing, and quantum mechanics is deterministic. But she still rejects libertarian free will (see here, here, and here).

But the part that especially interested me beyond superdeterminism is that many physicists rejected such deterministic interpretations of QM simply from their own emotional commitment to dualistic free will. For if determinism be true everywhere, say some physicists, then free will cannot be true. Indeed, Bell himself believed in libertarian, you-could-have-chosen-otherwise free will, while Einstein, a hardnosed determinist, didn’t. As I’ve reported before, physicist, atheist, and Nobbel Laureate Steve Weinberg also believed in libertarian free will. He sat next to me at the Moving Naturalism Forward meeting in Stockbridge, MA several years ago, and after I gave my spiel on the nonexistence of libertarian free will, Weinberg told me that he didn’t accept that his behaviors were determined by the laws of physics.

What I find fascinating is that physicists were conditioning their ideas and research directions on a philosophical belief that humans must have libertarian free will. Perhaps that impeded the ideas of “superdeterminism”.

I have no dog in the indeterminism vs. superdeterminism interpretation of QM; I don’t know enough.  That’s my fault, and it’s probably my fault that I don’t fully understand Hossenfelder’s explanation of superdeterminism in the video. She is a great communicator of science, and except for that puzzling bit, I greatly enjoyed her clear explanation.  (A transcript of her video is here.)

So I’m with Hossenfelder in our rejection of libertarian free will, which is the most common view of free will. I don’t give a hoot about compatibilism, which I see as a matter of semantics that is far less relevant than accepting the implications that pure naturalism—including any quantum indeterminism—has for society and for human behavior.

Weigh in below, but watch the video first. It’s excellent, especially in how it interweaves science with an a priori personal commitment to libertarian free will.

And if “superdeterminism” of QM is now widely accepted, let me know.

h/t: Darrell

112 thoughts on “Sabine Hossenfelder on free will and “superdeterminism” of quantum mechanics

  1. I think I need to read the paper she mentions in the video. It will be hard to get through the mathematics, but her arguments in this and other videos are fairly (though not definitively) convincing to me. I’d like to see their formal presentation, even if I can’t grasp it fully.

    As for Weinberg and Bell thinking that free will is possible, well…as Carl Sagan said, “Intellectual brilliance is no guarantee against being dead wrong.”

    1. Like Jerry I don’t have the physics chops to be confident in understanding the physics itself, but the part about some physicists holding human free will as a premise to any explanation of QM, including Bell, is pretty convincing given the quotes she shows.

      I’ve long noticed how fervent many physicists and interested amateurs are in their rejection of any QM interpretation that results in QM being deterministic, especially Everett. I’d never heard about Bell’s (and other physicists’) view that free will is an assumed premise, but it could help explain what I’d call an institutional bias against Everett and similar interpretations.

      1. I agree. It seems entirely an emotional response, by people one would wish were able to rise above such things. Sean Carroll has a much more nuanced view that I rather like. He says that CLEARLY everything is deterministic, but that it can be USEFUL to invoke terms like “free will” when talking about macroscopic things like humans, as long as one is clear that it’s a short hand…like using the term chair to describe something rather than getting into all the quarks and leptons of which the chair is made. That was, if I recall, in his live discussion with Sam Harris, which was a good discussion. I think Sam’s view (and PCC(E)’s) are the most thoroughly convincing and useful approaches, though, and they avoid the emotional confusion that seems inherent in allowing some pseudo-free will like compatibilism.

        I have started reading the first paper Sabine links, “Rethinking Superdeterminism”. I’m only 4 pages in so far (I’m at work), but what little mathematics I’ve encountered so far, while slightly esoteric, is fairly well explained. But I have a long way to go…

    2. I don’t agree with Weinberg’s philosophical predilections – they surprised me in fact – but I note that shininess of new toys of the lure of new physics is no guarantee against being dead wrong either. Hossenfelder has long polished her ability to be convincing, but I don’t see it get much traction in her field.

  2. As I’ve reported before, physicist, atheist, and Nobbel Laureate Steve Weinberg also believed in libertarian free will. He sat next to me at the Moving Naturalism Forward meeting in Stockbridge, MA several years ago, and after I gave my spiel on the nonexistence of libertarian free will, Weinberg told me that he didn’t accept that his behaviors were determined by the laws of physics.

    Hmmm, might you have misinterpreted his stance Jerry? Elsewhere he has stated views that are straightforwardly deterministic and compatibilist.

    E.g.: “The only meaning I can give to free will is that we sometimes do things because we decide to do them. What difference does it make if those decisions can be traced to processes in the brain of which we are not conscious?”

    1. Well, all I can say is what I remember of our conversation, and that was a while back.

      W: Do you mean to tell me that I couldn’t have ordered X instead of Y? (I don’t remember the items.)
      C: Yes, that’s what I said
      W: I don’t believe it.

      I reported it at the time; I’ll try to find it. I don’t think I “misremembered”, as I was shocked at the time. Maybe he hanged his mind.

      1. Interpreting that depends on what one means by “could have” ordered X. Maybe Weinberg was thinking something like: “if my brain had been in the state such that I wanted X, then I could have (would have?) ordered X, and if my brain had been in the state such that I wanted Y, then I could have (would have?) ordered Y”.

        Your question could then have intended as something like: “if your brain had been in the state that you wanted Y, could you then have ordered X?”, but that’s not how a compatibilist would interpret the question (unless there was also some other factor, such as the item being too expensive, or having run out, or similar).

        Here’s another Weinberg quote that seems deterministic/compatibilistic (from a 2002 review of a Wolfram book, though the original is paywalled):

        “I have never been able to understand the inconsistency that other people find between free will and a completely deterministic view of nature. Free will to me means only that we sometimes decide what we do, and we know that this is true by the same sort of mental experience that convinced Descartes that he existed, but we have no mental experience that tells us that our decisions are not inevitable consequences of past conditions and the laws of nature.”

        1. That does seem to be a plausible interpretation of empiricism (there is no empirical difference between making decisions and other deterministic systems) vs philosophism (there are philosophical differences). As I note on the vacuum density prediction in another comment Weinberg refused the philosophical interpretations

      2. Hi. Superdeterminism is the premises that the intial conditions of the universe(big bang) was set up in a certain way so that all human choices/decisions and measurements millions of years in the future will go a certain way to give us the impression that quantum mechanics requires indeterminism or non-locality when really it is a local-realist theory hidden by the correlations of our choices and the Big Bang conditions being a certain way to prevent us finding the truth. As you can see, the super bit of superdeterminism makes it utter nonsense. Google “superdeterminism conspiracy” to get the low down on what the physics community think of it

  3. I’m going to start saying that I am a total amateur in the subject but the book Something deeply hidden by Sean Carroll provided me with a good basis for understanding better the QM phenomena.
    As far as I understood the Bell theorem disproves the notion that there can be hidden variables that would allow us to do deterministic predictions, tackling the notion that textbook QM is an incomplete theory.
    Everett’s flavour of QM states that the wave function that describes the system contains a superposition of all possible states and when a measurement is done branching of the universe occurs where both states exist, therefore returning determinism to a seemingly indeterministic probabilistic outcome

  4. [ commenting before reading/listening in order to subscribe ]

    “Everything is determined”

    true – for sufficiently determined, and sufficiently defined “things”… are there “not things” to which I can point? Dark matter can be pointed at, yet – is it a “thing” as we know it at this moment?

    “God does not play dice with the universe”

    true – but how do we know it is only one sufficiently determined universe, or if it is being “played” with, let alone who made god?

    I jest I jest! This is fun stuff!

    1. Dark matter is observed in many complementary but agreeing ways, so it exists. The observational constraints makes it appear as heavy, non- (or weakly-)EM interacting particles – so quantum field objects – but that could be wrong. That “we don’t know (yet)” on specifics of its nature doesn’t make it less of “a thing”.

  5. I’ve seen the claim that most members of the US National Academies of Sciences are hard determinists – and I agree with the view. However, does anyone have numbers to support the claim? It’s been a long while since I last saw it referenced.

  6. I am glad someone is transcribing the YouTube videos they make – not that it helps them, I guess… not that all YouTube videos should be transcribed, of course…

  7. Like in QM itself, everyday interpretations of superdeterminism might not make sense. Any reader who understands the concept is invited to explain it below.

    Here’s an attempt (though other commenters here will do better):

    QM in its “default” interpretation is indeterministic (dice-throwing at the “wavefunction collapse” stage).

    So maybe some “hidden variables” determine the outcome of the dice throw, making it deterministic again.

    Bell’s inequalities then show that any such hidden variables cannot be “local” (meaning, they must depend on instantaneous information from some other spatial location — which is a bit weird).

    So how can one get round that? Well, one possibility is retro-causation, that is, information from the future determining the outcome — that is even more weird.

    “Superdeterminism” preserves determinism and locality by, instead, simply declaring that none of the outcomes that would violate Bell’s inequalities, and thus be incompatible with determinism and localism, actually do occur, because the whole system is set up such that they don’t occur, and that’s just the way it is (being deterministic, one can achieve this by simply specifying the initial starting conditions of the entire system to make it so). Most people find this “explanation” unsatisfying.

    Personally I see nothing wrong with indeterminism and dice throwing.

    [Though this does not give anything like “free will”, since there is no “will” at the particle level. Personally I don’t think that the QM determinism vs indeterminism debate has anything to do with the “free will” debate. Neither of those gives you dualist/libertarian free will and both of them are compatible with compatibilist free will. ]

    1. I agree entirely, because this is what “superdeterminism” meant in other QM discussions I have read. And this also agrees with one of Hossenfelder’s definitions, superdeterminism = violation of statistical independence. As you point out, one way to achieve the inequality-violating correlations is by specifying the initial starting conditions in just the right way. Like your “most people”, I find this unsatisfying, outrageously so. It basically amounts to positing a giant coincidence. That’s no way to do science.

      I can see why John Bell called this a “free will assumption” but agree with Hossenfelder that it’s a misnomer. Sure, one of the variables in the correlation is the choice of experiment made by the scientist, but that shouldn’t be the focal point of the objection. Rather, the objection comes from positing a giant coincidence.

      HOWEVER! There may be another way to get the inequality-violating correlations, maintain a revised version of locality, and have a deterministic theory. The Two State Vector Formalism allows forward-going (in time) and backward-going quantum states to co-determine outcomes. Because both time-directions are restricted to the speed of light, a revised kind of locality still applies. As for free will, if you think carefully about what such physics would imply, you’ll see that the “problem” of free will would not exist, because such “causality” does not make any one time/event the master and all others its slaves.

    2. I agree with all of that. After perusing the Wikipedia article on “superdeterminism” (which mentions Hossenfelder’s work, explaining her interest in making the video) I think the common analysis may be trending towards a more damning “implausible”.

      It seems genuine randomness is at the core of relativistic quantum field theory, taking it from an application of classical equiprobabilty in quantum mechanics to a requirement of preserving laws for different observers in field theory. See my long article response for a reference on how to show that the current physics is equivalent to such a mechanism (as well as a quote on superdeterminism).

      As for physiology, we are biochemical machines. Which on cellular scale of neurons translate to stochastic chemistry of large, crowded molecules. But still obeys the same laws. A requisite that philosophy doesn’t have – it has too many masters and can say nothing.

  8. I don’t think the concept of superdeterminsim is helpful. I think “free will”, as physicists use the term, is just shorthand for spooky action at a distance. The “delayed choice” Bell experiments demonstrate this. She says the only thing that matters is which measurement was actually made, not the ones which were considered but not made. Right, but the other particle had to choose the right state before anything physical was arranged to make the measurement of the first particle. This could very well be true, but if so, it means we can’t explain the dynamics of particles by any local physical theory; So superdeterminism gets us no farther than David Bohm’s old nonlocal hidden variable theory, and has the same problems.

    1. I think “free will”, as physicists use the term, is just shorthand for spooky action at a distance.

      I don’t think there is any set way in which physicists (en masse) use the term “free will”. At the level of physics there is nothing that can be sensibly called “free will” — like “hunger” or “pleasure”, having a “volition” (to use a more neutral word) is something that applies at the level of animals.

    2. I read it as superdeterminism means that the experimenter is part of the system, and the physical arrangement of the experiment is thus determined by prior events. Everything about the experimenter and the experiment is determined by what came before. There have never been nor ever will be any experiments based on randomness in any way, shape, or form. It is all bound to happen.

      1. But this is an empty declaration, even if true. It has no effect on the problems of science. What we would like to do in science is explain or predict the dynamics of what will happen based on what has happened. And in QM we don’t lnow how to do this without invoking nonlocal hidden variables.

        1. This nails it on its head.

          Your comments in combination with Wikipedia’s article on “superdeterminism” makes me realize it is a modern variant of Einstein’s philosophical “block universe” which replaces time with a spatial dimension. As such it invalidates local dynamical theories by ignoring the pesky negative topological signature (for starters).

          It starts out as “implausible” and ends up as “rejectable” (even though I don’t know how to do it).

  9. Do these physicists still ‘measure’ reality or has the measurement itself become reality? After reading this kind of advanced particle-stuff, I often begin to wonder whether apples turn into bananas when I do not look and turn back into apples as soon as I check whether they did.

  10. I don’t think anyone can “reject” freewill other than in a sense of holding an argument in ones head and being able to verbalize it.

    Every waking moment of almost everyone is imbued with a sense of being able to act with free will.

    Every society works because we assume people are free willed agents with choice, not deterministic automaton who sometimes act randomly

    1. >Every society works because we assume people are free willed agents with choice, not deterministic automaton who sometimes act randomly

      Is it the telelogical fallacy when people use a desired outcome (‘society works because we assume…’) to justify the assumption of a truth (that free will exists)? I know many people who have preferential outcomes whose success was contributed to by a belief in untrue things. Does the positive outcome justify an untrue belief?

    2. Every society works because we assume people are free willed agents with choice, not deterministic automaton …

      No, not really, more like: “Every society works because we treat people as free willed agents with choice, and that fact that that “will” is then the product of determined lower-level processes is irrelevant to that purpose”.

      (Meaning that things would be just the same, and society would work the same, if it were actually true that the “will” is the product of determined lower-level processes — and indeed it is true!)

      1. Well said.

        It seems like the maximally human way to treat other humans is to pretend they have free will, even though it does not exist. Unfortunately, there is no convenient way to inject that line into the political dialogue, and we are stuck with people dehumanizing each other (What do you expect? Of course she would say that! She’s [demographic]). Once we acknowledge publicly that there is no free will, I expect human rights violations to skyrocket (‘I was programmed to do what I did, and I prevented him from doing what he was programmed to do.’). I could use a little more faith in humanity this month.

        1. You don’t even have to pretend that they have something that they actually don’t.

          You just use the words we have for things that do actually exist, thus “will” as in “volition”, as in people do actually want things, and “free” as in the distinction between being in jail versus not, or the distinction between a gun-to-the-head versus no coercion.

          And my prediction for the effect on society of “acknowledge publicly that there is no free will” (in the dualistic/libertarian sense) is absolutely nothing at all — everything would stay the same (bar some minor alterations to commentary).

          1. > And my prediction for the effect on society […] is absolutely nothing at all

            I’ve just seen too many people on the New Left and Populist Right both doing everything they can to protect their chosen group at the expense of others (the 1%, non-citizens, etc.). Once political spin doctors start chanting that we have to protect ourselves in this dog-eat-dog world, and no one has free will to make a decision, some kind of war is almost inevitable.

            Deterministic science fiction tends to be extremely depressing and occasionally dystopian. Deterministic fiction just isn’t uplifting.

            > You just use the words we have for things that do actually exist

            In a rigid philosophical context, I would continue to do so. In a political context, it just doesn’t seem appropriate. It feels artificial. Having to reframe political conversations every single time feels unnatural and constrained. I’ve seen a few other ideological experiments with the English language (look at E-prime!)

            1. But when it comes to everyday life (as opposed to abstract philosophical or theological contexts), the “compatibilist” meaning of those words — the mundane meanings in line with my previous comment — are exactly how people already do use those words.

              Large swathes of people in, for example, Scandinavia, already accept that how people are and how they act is a product of their past, and no it hasn’t caused any wars.

          2. Exactly. The meat machines that we are, all of our choices can be traced to the operation of the brain and its underlying physics. Whether the physics is deterministic or indeterministic is largely irrelevant, as in neither case does it give us ‘free will’ in any libertarian sense.

            However, we can still (and routinely do!) meaningfully talk about our choices and actions being ‘free’ or ‘coerced’ or ‘restricted’- and that is what is meant by ‘compatibilism’.

            I am free to type whatever I want into this comment box right now (thanks Jerry!), but if I were banned from the site (or if my fingers were taped together) I would lose that freedom.

    3. The trouble is that free will is an incoherent concept. Even if you posit the existence of some sort of soul that is not constrained by the laws of the Universe, it can’t work.

      When you make a decision, you look at the information you have and apply some though processes to it and then you make a choice. If you could rerun the process starting with the exact same information and the exact same you, you will make the same choice because you’ll go through the exact same thought process. Where’s the free will?

      Adding some randomness in doesn’t help. You might start making some different choices, but I don’t think anybody would argue that sometimes making random decisions counts as free will.

      1. > The trouble is that free will is an incoherent concept. Even if you posit the existence of some sort of soul that is not constrained by the laws of the Universe, it can’t work.

        Solid point. I think we gave up on that when the scientific establishment stopped taking psychic/telekinetic phenomena seriously. The only was a non-deterministic immaterial ‘soul’ could move a deterministic material object (including the very neurons in your brain!) is by using telekinesis.

  11. Superdeterminism just means that the initial state of the universe is set up such that, while the universe operates on a local hidden variable theory, its results appear as if they are not. To sum it up, it posits a cosmic conspiracy that fools us into thinking the universe is not working under a local hidden variable theory. To me, that means we’re doing away with science entirely. Or at the very least, it smells to me like an article of faith.

    Bell’s theorem, barring superdeterminism, only rules out local hidden variable theories, instead of deterministic theories. Hidden variable theories are a strict subset of deterministic theories, which local ones are another strict subset of.

    The Everettian approach, which just says that everything works according to quantum mechanics, still seems the most promising to me, because we continue to use quantum theory to explain large-scale phenomena, which collapse theories can’t explain, and nonlocal hidden variables had its hidden variables as epiphenomena, which are causally disconnected from the mechanics.

      1. No there’s really nothing else to it. It is that crazy (‘t Hooft wouldn’t be the first prominent physicist to espouse crazy idea at the end of his carreer…).

        Invoking this kind of conspiracy would indeed destroy science because it can be invoked in any context and therefore explains absolutely nothing.

        Interestingly, it is a similar epistemological dead end than multiverse ideas but in reverse. Instead of saying that anything and everything happens for sure an infinite number of times, it says that only what we observe could have happened and it couldn’t have failed to happen no matter what we could have tried to experimentally obtain (without explaining why it happens the way it did rather than some other way…). In both cases, it is totally unpredictive and unfalsifiable in principle and therefore explains absolutely nothing at all !

        1. Maybe multiverses is a dead end for philosophy, but it is popular in cosmology and – contrary to your claim here – has passed tests. That fact is somewhat buried in both popular and vocal physics discussions, so you have to go to the original papers. Steven Weinberg first proposed the theory and made the correct prediction of the vacuum energy density, then it became observed.

          How popular is it?

          The recent eBOSS 20 year galaxy survey summary cosmological paper shortlist it as hypothesis [ , reference 224 ].

          At the high precision found here, cosmic acceleration remains most consistent with predictions from a cosmological constant. A deviation from consistency with a pure cosmological constant perhaps would have pointed toward specific dark energy and modified gravity models. However, since many of these models have parameter choices that make them indistinguishable from ΛCDM, those models all can be made consistent with our observations. Nevertheless, the observed consistency with flat ΛCDM at the higher precision of this work points increasingly toward a pure cosmological constant solution, for example, as would be produced by a vacuum energy fine tuned to have a small value. This fine-tuning represents a theoretical difficulty without any agreed-upon resolution, and one that may not be resolvable through fundamental physics considerations alone [224,225]. This difficulty has been substantially sharpened by the observations presented here.

          The recent article on multiverses describes it as “tough to say” [ ].

          Physicists suspect that eternal inflation is generic, meaning a consequence of most, if not all, models of inflation. So, following this suspicion, if inflation is correct, then eternal inflation is also likely correct, and the multiverse might be real.

          Let me improve on the laudable uncertainty of that article by noting two things.

          First, the recent BICEP3/Keck observations preferred a simple Higgs like scalar field for inflation (and claims we will know either way within a decade). In that physics eternal inflation is generic and with the observed parameters applicable, consistent with Weinberg’s theory. [“Squeezing down the Theory Space for Cosmic Inflation”, APS Physics, 2021.]

          Second, in any case cosmologists accept that inflation preceded the Hot Big Bang of LCDM cosmology since the last two decades [“Pathways to Discovery in Astronomy and Astrophysics for the 2020s (2021)”, Appendix C: Report of the Panel on Cosmology, NAS US Decadal Survey].

          The question of what process set the Hot Big Bang in motion and created the seeds of structure has been with us for many decades. Early theoretical developments, together with observations over the past two decades, have established the inflationary paradigm as the dominant picture in the field.

          I’m a bit tired of this (and of the untestable “direct/indirect” label of personal incredulity levels), as it appears besides similar claims on dark matter and dark energy. Both of which has been similarly observed by many means and undergone tests: LCDM predicts 5 major parameters, inflation predicts 6! I don’t personally know the history of the “falsified” claim of philosophical unfalsifiability, but I recently read that two physicists made a personal and vocal early campaign against it – they personally didn’t like it. If so, it is a bit more understandable that the rest of us are confused.

          But we don’t need to be, as I have described amply above.

          And In any case, it seems that apart from the personal vitriol and irrelevant and in any case here erroneous philosophy, the area is well and truly alive. Inflation explains a lot that other theories can’t, at least as simple (see reference 245 to eBOSS paper). Including explaining the value of vacuum density then (high) and now (low and at a habitable level for our local universe). Will it stay that way, can we make more tests than what today appears like the early stoichometric test of atom theory, i.e. the cosmological constant prediction, being consistent and compelling but not persuasive? “Throughout the nineteenth century, some scientists had cautioned that the evidence for atoms was indirect, and therefore atoms might not actually be real, but only seem to be real.” Future history may tell.

          1. Inflation is a purely speculative and unnecessary idea whose sole motivation was pseudo problems of imaginary magnetic monopoles and fine tuned initial conditions. Standard hot big bang cosmology explains our observations equally well without inflation if you accept that the initial conditions were fine tuned. I don’t see any problem with fine tuned initial conditions, especially since the inflaton potential needs to be fine tuned itself !

            No one has ever measured any inflaton field, which remains a speculative invention. And if this idea implies that anything and everything happens for sure an infinite number of times, then it is a completely empty idea that explains nothing and it is therefore dead. That’s called the measure problem and it is deadly indeed. Whether you like it or not, whether you like philosophy or not…

            Unfortunately, that’s the sad results of theorists indulging themselves in all kinds of unwaranted speculative degrees of freedom in the late seventies and eighties and led physics in these sorts of dead ends (same thing with string theory by the way). It’s a real mess.

            1. I mostly agree.

              While there is some circumstantial support for inflation, it is not really firmly proved. Many people believe in it because it can solve the flatness problem. While it can indeed solve some incarnations of the flatness problem, it cannot solve them all. However, many experts agree that the flatness problem doesn’t actually exist. So many, in fact, that I wrote a review article discussing them:


        2. Multiple blockquote fails, but I can’t repost such a long comment, and I hope it can be parsed even if it looks awful.

        3. without explaining why it happens the way it did rather than some other way

          By the way, I don’t understand this, even though it seems irrelevant when mechanisms are proposed, predictions are made and tests passed. But if it is a lack of understanding of the area, selection bias is a known constraint on cosmology.

          Selection bias was explicitly nullified in early cosmology that relied on the “cosmological principle” long before it became an observation and so a theory requisite that passed further tests by way of the cosmic background radiation and it basis in homogeneity and isotropy. It was also used to explain the carbon cycle of fusion before the triple-alpha process resonance was observed (Hoyle’s hypothesis of stellar nucleosynthesis).

          However, it is in LCDM cosmology its greatest strengths lies. Besides predicting the vacuum energy density by having dark energy set up so that galaxies will form briefly before (mainly dark) matter dilutes towards zero density it seems it may do the same for dark matter.

          Quantum field theory proposes that the dark matter particle mass is set by galaxies forming briefly before dark matter evaporates towards zero density [ ].

          In a World First, Physicists Narrow Down The Possible Mass of Dark Matter
          We may not know what dark matter is, but scientists now have a better idea of what to look for.

          Based on quantum gravity, physicists have worked out new, much more stringent upper and lower mass limits of dark matter particles. And they have found that the mass range is way tighter than previously thought.

          This means that the dark matter candidates that are either extremely light or heavy are unlikely to be the answer, based on our current understanding of the Universe.

          “This is the first time that anyone has thought to use what we know about quantum gravity as a way to calculate the mass range for dark matter. We were surprised when we realised no-one had done it before – as were the fellow scientists reviewing our paper,” said physicist and astronomer Xavier Calmet of the University of Sussex in the UK.

          From the abstract:

          The lower bound comes from limits on fifth force type interactions and the upper bound from the lifetime of the dark matter candidate.

          Dark matter doesn’t seem to be stable when put in combination with gravity. And so our existence is yet again dependent on local conditions.

          1. “By the way, I don’t understand this”

            Why would the unvierse conspire to consistently produce violations of Bell’s inequalities and the rules of quantum mechanics in particular rather than something else ? Superdeterminism doesn’t tell. There’s a conspiration for producing something specifics for completely unknown reasons.

      2. I’ve tried reading ‘t Hooft’s writing on it and he is just confused. He doesn’t actually believe in a superdeterministic theory but a (generic) local hidden variable theory, which has been ruled out by Bell’s theorem.

    1. CS Peirce wrote a really great paper about this sort of thing in 1892, “The Doctrine of Necessity Examined”. Discussing with an imagined opponent who thinks, for example, that the result of every throw of every die ever is pre-determined from all eternity, he says this:

      “Very well, my obliging opponent, we have now reached an issue. You think all the arbitrary specifications of the universe were introduced in one dose, in the beginning, if there was a beginning, and that the variety and complication of nature has always been just as much as it is now. But I, for my part, think that the diversification, the specification, has been continually taking place. Should you condescend to ask me why I so think, I should give my reasons as follows:”

      I recommend reading the whole paper for his reasons.

    2. I don’t know what “explain large-scale phenomena, which collapse theories can’t explain”. But I have seen at least one theory* explaining collapse by relativity making it on par with time dilation and length contraction – odd, but not breaking laws. So preserving laws I assume it can explain phenomena on all scales, same as relativity in general does.

      Maybe you could expand on your claim, give references et cetera?

      *I have posted the reference earlier many times under this article already, so I’m loath to do it again. I’m not married to it…

      1. How do collapse theories explain macroscopic things doing quantum stuff? E.g. superfluids? Nano-scale superpositions? The CMB?

        As far as I can tell, no one actually uses these collapse theories to do anything. They’re all just cute models that attempt to explain few-particle phenomena with no application to anything physicists do. Physicists use quantum theory even on cosmological scales, which basically grants Everettianism.

        I only looked at the paper after writing the above, but I think a serious issue to the relativity-causes collapse conclusion is that the paper assumes collapse happens from the outset:

        the explicit conservation that obtains for Alice and Bob’s Stern-Gerlach spin measurement outcomes in the same reference frame holds only on average in different reference frames, not on a trial-by-trial basis.

        Something that does not assume collapse would not obtain an on-average conservation, but strict conservation, e.g. with pilot wave theory and the many-worlds interpretation.

  12. Fascinating video. As to your quandary over “what ‘superdeterminism’ means … on the level of particles,” I can bring to bear two trenchant observations — Sabine Hossenfelder has a cool name and an even cooler accent. 🙂

    1. Ken — If you like her name, and her accent, you should check out her music videos. Since it’s nearly Christmas, I can recommend Sabine’s “Ave Maria” for atheists!

  13. The easiest concrete way to see what superdeterminism means is to consider the experiments of Alain Aspect which confirm that Bell’s inequality is violated by quantum mechanics. In the first experiments the separate polarizers of Alice and Bob were rotated to different positions by oscillators running at different frequencies. But this was seen as a loop-hole. To apply Bell’s theorem the polarizer settings had to be statistically independent. This is how “free-will” became involved, because it was often discussed as Alice and Bob choosing the settings. But nobody every choose the settings in the actual experiments because they that to be selected outside each others light cone, spacelike, to rule out any interaction. The settings were done by electronics because they had to be done within nano-seconds of one another. So what was the loop-hole? The electronics used had been sitting in the lab, able to interact before the experiment so the polarization selectors and the photon source could have conspired earlier, and the assumption of statistical independence would be violated. So in the last form of the experiment, instead of using some local electronic random number generator, star light from stars on opposite sides of the celestial sphere, stars hundreds of light years from Earth and each other, was used to set the polarizations.

    Superdeterminism is the proposition that even the light from those stars is not statistically independent because tracing back to beginning of the universe there was a time when the matter/energy of those stars, and the particle source, were not space-like separate and hence could have conspired to violate Bell’s inequality. Put that way it’s like “God did it.” It can explain anything and everything as simply being the deterministic outcome of the initial conditions of the universe. To be more than a mockery of a theory, it needs to be filled out with some explanation of why this conspiracy and and it’s consistency. So far that is not forth coming.

    1. “Superdeterminism is the proposition that even the light from those stars is not statistically independent because tracing back to beginning of the universe there was a time when the matter/energy of those stars, and the particle source, were not space-like separate and hence could have conspired to violate Bell’s inequality. Put that way it’s like “God did it.” ”

      Well put. Superdeterminism is complete nonsense.

  14. “So I’m with Hossenfelder in our rejection of libertarian free will, which is the most common view of free will. I don’t give a hoot about compatibilism, which I see as a matter of semantics that is far less relevant than accepting the implications that pure naturalism—including any quantum indeterminism—has for society and for human behavior.”

    Right on. Indeterminism, wherever it might exist in nature, can’t add to the control and responsibility we already have under determinism. Seeing that we don’t have the ability to have done otherwise in actual situations (as opposed to counterfactual) in a way that would have given us more control than under determinism, can help to mitigate reactive attitudes premised on the idea we do have that ability (libertarian free will). And this in turn can condition the sorts of responsibility practices and social policies we judge to be fair and humane. Here’s a recent lecture on the implications of Jerry’s “pure naturalism,” what I call pragmatic determinism:

    1. Specifically, this may break Da Roolz: “Please do not use this site to promote your project, book, website, and so on, or to raise money for your cause.”

  15. I have to admit I found her exposition somewhat confusing.

    The best exposition of Bell’s work I know of is a paper by N.David Mermin, “Is the moon there when nobody looks?” – it’s so goo that Feynman wrote him a fan letter:

    As I understand it, Bell’s theorem is not about determinism at all but about whether a realistic interpretation of quantum mechanics is possible. According to the orthodox view, certain pairs of quantities, position and momentum being the best known example, not only cannot be known simultaneously but do not even have well-defined values simultaneously. It was this that Einstein objected to, and discussed in a famous paper with Podolsky and Rosen, “Can Quantum Mechanical Description of Physical Reality Be Considered Complete?”. In the variant invented by David Bohm, he discussed two electrons whose total spin is zero allowed to become widely separated in space. According to QM, spin in the x and y directions are similar to position and momentum, in that they do not simultaneously have well defined values. However, if we measure the x values of electron 1 we know for certain that electron 2 will give the opposite value, even if we do not measure it, because the total angular momentum is zero. If we do the same for the y values, the same is true. Since the electrons are assumed to be so far apart that no signal can travel between them even at the speed of light, we seem forced to conclude that, contrary to QM in the orthodox interpretation, we can simultaneously know the x and y spin components of electron 2 by measuring those of electron 1. As EPR said, “If, without in any way disturbing a system, we can predict with certainty (i.e., with probability equal to unity) the value of a physical quantity, then there exists an element of reality corresponding to that quantity”

    Bell’s brilliant idea was to devise an experiment in which the spins of the two electrons are measured at angles other than a right angle. In this case, a “realistic” interpretation in which the spin components in any direction have well-defined values gives one value for the correlation between the two elements, and quantum mechanics gives another. When the experiment is carried out, QM gives the correct answer. The only way in which the realistic interpretation can be saved was thought to involve some kind of non-locality to explain the exact correlation between spins measured in the same direction – local realistic theories have to be given up.

    Insofar as I understand superdeterminism, it is an attempt to reinstate local realism by denying Bell’s tacit assumption that measurements at each detector can be chosen independently. Bell explained this very clearly in a BBC interview:

    “There is a way to escape the inference of superluminal speeds and spooky action at a distance. But it involves absolute determinism in the universe, the complete absence of free will. Suppose the world is super-deterministic, with not just inanimate nature running on behind-the-scenes clockwork, but with our behavior, including our belief that we are free to choose to do one experiment rather than another, absolutely predetermined, including the “decision” by the experimenter to carry out one set of measurements rather than another, the difficulty disappears. There is no need for a faster than light signal to tell particle A what measurement has been carried out on particle B, because the universe, including particle A, already “knows” what that measurement, and its outcome, will be.”

    It’s an option, but personally I agree with Bell that it is implausible, and would go even further and say that there is in fact not the slightest reason to believe it.

    Incidentally, I don’t think it’s correct to describe Einstein as a hard-nosed determinist, despite the playing dice stuff. His problems with QM may be divined from a letter that his very close associate Pauli wrote to Max Born at Einstein’s request (Born Einstein letters, 31 March 1954)

    “In particular, Einstein does not consider the concept of ‘determinism’ to be as fundamental as it is frequently held to be (as he told me emphatically many times), and he denied energetically that he had ever put up a postulate such as (your letter, para. 3) : ‘the sequence of such conditions must also be objective and real, that is, automatic, machine-like, deterministic’. In the same way, he disputes that he uses as criterion for the admissibility of a theory the question: ‘Is it rigorously deterministic?’
    Einstein’s point of departure is ‘realistic’ rather than ‘deterministic’, which means that his philosophical prejudice is a different one. “

    Apologies for rambling on a bit, but it’s a fascinating topic.

      1. Is the observer dependent time dilation there when nobody looks? I guess it depends on your frame of mind. 🙂

        In any case, I wasn’t aware of the full context to Mermin’s work, but you may find it interesting that a recent work has showed that it – when seen from relativistic field theory of spin entanglement – is equivalent to other realistic physics. See my own response to the article which I posted before I saw this discussion, or go directly to “Answering Mermin’s challenge with conservation per no preferred reference frame”, Stuckey et al., Scientific Reports 2020. I don’t think you have to accept the theory, just observe that it shows a possible equivalence.

  16. The talk of the incoherence in Libertarian Free Will reminds me:

    In a recent debate with a Christian Apologist I pushed on the idea of God’s Free Will. Because God presents a particular problem for the Christian and their claims about Libertarian Free Will. When faced with the Problem Of Moral Evil, typically Christians using the Free Will Defense will say that the Libertarian Freedom to actually choose to do evil or good is NECESSARY for morally relevant free will, hence OF NECESSITY God had to grant His creatures this ability in order to produce morally relevant beings.

    BUT…we are also told that God Himself has Libertarian Free Will, in fact He’s the ultimate expression of Libertarian Free Will – nothing external to God causes His actions! – and yet God’s Perfectly Good Nature entails God ALWAYS CHOOSES THE GOOD.

    So…what gives?

    I pointed out how this at least implies SOME type of determinism to explain God’s never choosing to do evil. At the very least you have to say something like “God’s Own Nature Determines His Actions” because if you DON’T, then on what basis can you possibly tell anyone you know what God would or wouldn’t, could or couldn’t do? If we can be confident that God will never choose to start torturing babies for pleasure…on what basis? WHAT PREDICTS THIS if not, at a minimum, that His nature determines He can never do so?

    The apologist had no answer for how a being with Libertarian Free Will can be predicted to never choose evil…and simply said “that’s something we Christians have Faith in.”

    Basically, a “have your cake and eat it too” response. God gets Libertarian free will, but never actually chooses evil, explanation: “don’t have one: Faith.”

    (If the Christian admits that God never does evil because of His determinant nature, yet this is COMPATIBLE with being Free Willed, then the Christian must also admit God could have created beings whose nature determined their good actions, while being free willed, hence God creating beings – humans – who actually chose evil is incompatible with the claims to his Perfect Goodness).

    1. I’d been suspicious of his prefect goodness since he threw Adam and Eve out for seeking knowledge. Knowing a little girl who died of leukemia at age 12, sealed it for me.

      1. This is not quite right. Adam and Eve were thrown out for tasting the fruit from the tree of the knowledge of good and evil. Not knowledge per se. It could be argued they were thrown out for thinking in terms of good and evil. For example thinking, all of a sudden, nudity was somehow wrong.

    2. Yes exactly. For closely related reasons, having Libertarian Free Will would suck big time. I would have to move far away from my wife and child. Wouldn’t be able to predict how I would treat them.

      1. The thing is, even if the Christian refuses to admit of any determinism in God’s actions, and even if the mechanics of Libertarian Free Will remain unexplained, they are caught in a bind because however Libertarian Free Will works, God purportedly has it, yet never chooses evil acts, which breaks the logical necessity between “having Libertarian Free Will” and “being able to choose to do evil.” They can no longer make this “necessity” argument for the Free Will defense. IMO.

        1. But couldn’t the Christian reply that God cannot, or will not, choose to do evil because that would be going against his essential nature of being omnibenevolent, and hence be self-contradictory. God’s actions are ‘determined’ by his nature, just as an electron’s behavior is determined by its properties.

          1. YF,

            Yes, that was my point.

            Christians are always appealing to God’s Nature as a way of explaining what God is like, what God can or can not do etc. So I was saying “Ok, in light of your free will defense, let’s zero in on that and look at the implications. Basically, the only way the Christian can tell us that we can have ANY expectation of what God is like, is by appealing to God’s nature. That God’s “Benevolent” nature determines His actions. Fine. But then this entails God could have created beings with a Benevolent nature who always chose good.
            “But that would have made us robots, not morally relevant beings!” the Christians always tell us. And yet…this describes their own God!

            They can’t actually get out of this dilemma! The Christian apologist I mentioned simply did a Hail Mary – he wouldn’t admit that God’s nature determines God’s actions, because that of course smacks of determinism of a sort, which they CAN’T have as they hold to Libertarian Free Will*. So he just said “somehow God’s actions are reliably good, even though He has Libertarian Free will…but I can’t explain how that is.”

            *(There are versions of Libertarian Free Will that allow a sense of determinism, where the choices are determined by the agent, but the agent’s actions are not externally determined. But that still presents the dilemma I’m talking about).

  17. From what I’ve read, the “mainstream” view of superdeterminism is that it’s a crackpot theory. (I myself am not qualified to hold an opinion.) But you don’t need to believe in superdeterminism in order to be a determinist.

    Bell’s theorem proved that there are no non-local hidden variables. Advocates of David Bohm’s “pilot wave” theory argue that there are “non-local” hidden variables. This is a deterministic theory that is distinct from superdeterminism.

    If you want a cogent explanation of the points I’ve raised, I’m afraid you’ll have to look elsewhere. 🙂

  18. No, philosophical “superdeterminism” is not widely accepted , and cannot be so as long as the different theories of “interpretations” are untestable. I tend not to listen to Sabine Hossenfelder that – albeit doing papers – has always been fringe and tended towards the market of “communicator of science” by marketing her own ideas.

    So I’m not going to support her marketing by watching that video, which would be superfluous for the science at hand. It is a gods-of-the-gaps argument of a perceived loophole in Bell tests, arguing for implausible hidden variables of local theories.

    In quantum mechanics, superdeterminism is a loophole in Bell’s theorem. By postulating that all systems being measured are correlated with the choices of which measurements to make on them, the assumptions of the theorem are no longer fulfilled. A hidden variables theory which is superdeterministic, hence, can fulfill Bell’s notion of local causality and still violate the inequalities derived from Bell’s theorem. …

    Physicists Sabine Hossenfelder and Tim Palmer have argued that superdeterminism “is a promising approach not only to solve the measurement problem, but also to understand the apparent non-locality of quantum physics”.

    You can’t have relativity and superdeterminism at the same time. Unless you subscribe to the hidden variables that experiment rejects and add the implausible need for them to be sensitive to what they were exclusively introduced to be, potential residual influences from the last phase change of the universe (the Hot Big Bang). [If you don’t understand what I wrote, read the article: “Superdeterminism”, Wikipedia – I’m not making this up, the superdeterminists are.] But famously it is the basis of marrying non-relativistic classical quantum mechanics with relativity to get modern quantum physics of quantum field theory.

    I believe I have noted before that there is now a Einsteinian framework to replace “interpretations” with the analogous idea of making observed constancy of the light speed in vacuum universal applied to the observed constancy of Planck’s constant. [“Answering Mermin’s challenge with conservation per no preferred reference frame”, Stuckey et al., Scientific Reports 2020.] The result is that “non-locality of entanglement” and the measurement problem in combination is dissolved, the latter as a relativistic effect of “wavefunction collapse”, on par with time dilation and length contraction. It may not be correct, but it shows that there isn’t a new type of fundamental mystery here, it could be the familiar type of oddities resulting from having different observers but the same laws.

    Finally, since Steven Weinberg was among the parents to effective quantum field theory he fits oddly into this physics analysis. Since he rejected philosophical interpretations of his prediction of the value of the vacuum energy density (“the cosmological constant”) I was mildly surprised to note that he thought string theory would be able to replace field theory. Perhaps he, like Sabine Hossenfelder, felt that seemingly odd problems may have exciting new physics explanations. But it behooves us to look at the already known before we reject it because of the lure of a shiny new toy – we are not cats.

  19. I thought of a good analogy for superdeterminism (though posting it now is likely too late for anyone to read it!).

    Suppose we lived in a universe where, when we throw a dice, it always gives either 1, 3 or 5, and never 2, 4 or 6. And suppose that everything we knew about dice and physics and how the world works suggests that all 6 numbers should be equally likely. So the lack of 2, 4 and 6 would then be a big puzzle.

    The superdeterminist would then say: easy, it’s simply that the universe is absolutely deterministic, and it just happens to be the case that the initial conditions of the Big Bang were such that, as the determined outcome plays out, 2, 4 and 6 never occur. Essentially, all the starting points that would have led to 2, 4 and 6 simply didn’t exist, only those leading to 1, 3 and 5 exist.

    Would anyone find this convincing?

    1. I wouldn’t. It sounds like your argument is convincing. But Gerard ‘t Hooft is probably smarter than I am so if he believes in superdeterminism my guess is that there is something more to it.

  20. Two thoughts cross my mind with this video:
    1) Let’s say I use another allegedly random quantum phenomenon to choose the timing, location and type of measurement then I’m not sure how this impacts anything, irrespective of the truth of suerdeterminism.
    2) This also gives an insight into what eminent scientists (some) might think of as free will. People who object not being able to choose completely freely are somehow mistakenly confounding a random choice with free will. Would compatibilists still think their choices are somehow independent of cause even if that cause is a quantum phenomenon?

    1. No. As a compatibilist I think my acts of “free will” are determined by my genetics, my experience, my education, my circumstances, as encoded in my brain and body. So they are mine, even though they are determined in the physics sense. Of course there’s still a little randomness in the physics, like K40 decays in the blood, and an occasional cosmic ray.

      1. So you believe that having been determined to have taken on some flavour of responsibility or perhaps ownership is “free will”?

        Regarding potassium decay, the fact that we can apply the first order law to its decay is pointing to something not being completely random. But I agree the moment a K40 atom in your banana decays seems nigh on impossible to predict.

        1. Obedience to first-order kinetics implies exactly that the decay process is random. For any given radionuclide, the rate of decay (decay events per unit time) at a moment in time depends only on the number of radioactive nuclei present at that time. As far as is known, no intervention has the ability to alter the kinetics away from randomness.

          And the moment that any given K-40 nucleus in your banana will decay is completely impossible to predict, not just nigh on. All you can say is the longer you watch it, the more likely you are to observe it signaling a decay event.

          1. I don’t think first order rates necessarily imply randomness. Most physical processes follow first order rates. Like diffusion. Or imagine you have a well mixed tank with water flowing through it. Add a spike of salt to the tank and the you will find a sinusoidal decay curve that decays at a first order rate. Is this random too? Say are first order iodine clock reactions random?

            If something were to follow, say half or second order rates would they not be random? Could it be simply chaotic?

        2. Notice i put it in scare quotes. I don’t think the “free” means anything. It’s sensible to talk about my will versus someone else’s will. But free invites the question, “Free of what?”

            1. The only meanings I can infer for “free” are:

              (1) The legal meaning of “not coerced” by threat or force.

              (2) Unpredictability. No one, not even you, can reliably predict what you will do in given situation.

  21. 1) The point of superdeterminism is to assume that the other allegedly random quantum phenomenon will conspire to rig the experiment outcome.
    2) Your free choice appears random to me, if I cannot predict it.

  22. All persons occurring below are accomplished physicists, none are entirely science pundits!

    An abstract of Lev Vaidman: “It is argued that the lesson we should learn from Bell’s inequalities is not that quantum mechanics requires some kind of action at a distance, but that it leads us to believe in parallel worlds.”

    The final of many references to Bell, on p.330 (of 530) in David Wallace’s closely argued book “The Emergent Multiverse”, are in the following paragraph, whose follow-up needs to be read for further explanation:

    “But in Everettian quantum mechanics, violations of Bell’s inequality are relatively uninteresting. For Bell’s theorem, though its conclusion arguably entails …action at a distance …. , simply does not apply to the Everett interpretation. It assumes, tacitly, among its premisses that experiments have unique, definite outcomes.”

    Sabine H. is an excellent expositor, but she completely avoids the (quasi-classical, not fundamental) many-worlds in the interpretation of Everett. Elsewhere she disagrees with many-worlds, but (to descend momentarily to ‘argument by authority’) Weinberg does not, nor does David Deutsch, nor does Sean Carroll, the latter two having written extensively in its favour.

    Has anyone here thought long and hard about free will as it would or would not occur in the many-worlds interpretation?

    1. I’ll look up Wallace’s book, but it makes no sense to say experiments have unique outcomes AND all possible outcomes occur.

      1. Is it an idea from the early 20th c. – the photon goes through every path AND we observe the interference pattern – because not all paths are equally likely…. Hawking’s book uses a soccer ball metaphor.. … hmmm but “unique” perhaps is not the word to use… unless the distribution is the “unique, definite outcome”.

        1. In many-worlds it is “THE” photon only in that all the involved photons relevant here, one in every different world resulting from the split, are initially identifiable with each other. In this form the idea is from 1956, at least as far as anyone (i.e. Everett) having the idea and the guts to write it down for physicists to make judgement.

            1. Very late, and perhaps I misunderstand you—but except for the interference between the ‘mates’ (say) of a photon (say) in the Mach-Zehnder apparatus (again, say–so it is realized that this is just one such experiment of many indicating the likelihood of the Everett splitting)—but your ‘disappear’ surely depends on for ‘whom’ it is disappearing: I see only the effect of the photon remaining in my world, but the mate of me sees only the mate of my photon in the mate world– the situation is completely symmetric. There is nothing special about me and my world (except to me and others here). Of course often there is no such interference, which is normally detectable only in specially controlled experiments (e.g. also in quantum computers and in the Feynman lectures famous discussions of ‘slits’).

              And so the Everett interpretation seems to disobey common sense, as discussed elsewhere here. But the Copenhagen is simply logically incoherent, the fig leaf interpretation.

              To repeat myself, David Deutsch’s books are really the best initial sources.

              And afterwards David Wallace’s book, for much careful argumentation on the two difficult issues needing explanation, particularly how probability works when one adopts the many-worlds interpretation. That is Part II in a 500-page book, pp. 113-256, so rather substantial, to say the least. The other thorny issue, the Born Rule, comes earlier and is easier. Sean Carroll (and coauthor I think) have a different approach to deriving the latter, if I remember correctly.

              I don’t think single sentences do any justice at all to these matters.

              1. You misunderstand my point. If you think of a photon going thru every path in a different “world” in say a double slit experiment, then there will be no interference dark lines. They have to be different paths in the same world. Deutsch makes the same mistake when he claims the many worlds must exist because the universe is not big enough to realize a quantum computers computation. He ignores the fact that the quantum computation depends on destructive interference to suppress all the wrong answers; so the computation must take place in one world.

                Once you assume that your model of reality assigns probabilities to states that must vary continuously and reversibly then QM follows, including the Born rule (c.f. Lucien Hardy). So the multiple worlds interpretation has two problems. The first is explaining how having everything happen, but in different worlds, implies a probability measure on what happens. The second is whether this model is “the” world, or is it just what we can say about the world (to quote Bohr), c.f. Chris Fuchs.

                Bohr was very careful to say that collapse of the wave function was just an instrumentalist limitation on what we could model. And he was perfectly correct when he said that science can only be done in a classical world in which there are definite outcomes that are recorded and everyone can agree on. And that’s the measurement problem in QM. When, where, and how is that transition from quantum to classical. MWI tries to deny a problem exists by just saying everything happens. Decoherence theory has gone part way to explaining it, but a gap still remains, c.f. Zeh “Quantum Darwinism”.

        1. Agreed, except that maybe the word “fan” when applied to physicists like Deutsch, Vaidman and Carroll, and Weinstein who agreed more weakly, could be strengthened to something less similar to the term ‘fans’ of pop musicians and to Trumpist morons.

      2. Replying to Brent:
        But it does make sense to say that an experiment has a unique outcome in each one of the many worlds of the splitting it issues forth, and that, because there are enough such worlds, any possible outcome does occur in at least one such world.

        I take your “makes sense” to mean ‘is not a logical contradiction’; and your “possible” to mean ‘is among the predictions assigned a non-zero probability by quantum theory’.

        I think Deutsch’s two books, less technical, are better read before reading Wallace’s technical one, where dealing with the quantum meaning of probability needs considerable length.

        As for so-called ‘common sense’, quantum theory in any serious interpretation certainly violates that, as does Special Relativity, as did Copernicus in his time. Too bad Hossenfelder completely ignores the Everett so-called interpretation, and so cannot be taken seriously by me. Big deal!

        Sorry for the lateness here and just above.

  23. OOPS, I misinterpreted Hossenfelder badly! And so did many other commenters. H/t Mike Smith for discussing her argument and revealing the mistake.

    Hossenfelder writes in her paper:

    I believe that many people find violations of Statistical Independence confusing because such violations have frequently been portrayed in the literature in an extremely misleading way. Superdeterminism is supposedly a “conspiracy” in which the detector settings are influenced by hidden variables in just the right way to give the results of quantum mechanics. But that Statistical Independence is violated tells us nothing about the freedom to choose measurement settings, unless you want to say that the measurement setting tells you what the measurement setting is, which is correct, but not very deep. In a superdeterministic theory, the measurement setting, θ, can be chosen in the same way you can chose it in normal quantum mechanics. It’s just that what the prepared state will evolve into a depends on those measurement settings.

    A key part of Hossenfelder’s view is future-input dependence, meaning that the way a quantum state evolves depends (statistically) on the measurement settings it will encounter in its future. So when I wrote about the Two-State Vector Formalism in my comment above, I was accidentally on the right track! See section 7 of her paper. I can’t resist pointing out though, that her insights on causality in section 7 lead, not to a rejection of free will as such, but to a diagnosis of why the “problem” of determinism for free will and the “need” for a Libertarian “solution” is based on bad physics.

    1. I don’t think that is the standard superdeterminism. That’s a different theory that includes retro-causation like Klauber’s. Superdeterminism says that the experimenter’s choices are determined along with everything else, and they are statistically correlated with the measurement outcomes which were also determined in the past.

      1. Is there an official definition of “superdeterminism”? Wikipedia says “Some authors consider retrocausality in quantum mechanics to be an example of superdeterminism, whereas other authors treat the two cases as distinct. No agreed-upon definition for distinguishing them exists.”

  24. Yes, she says that the state depends on the future. She also argues for superdeterminism, where it is all determined by the distant past. The quantum state and the measurement settings are both determined in a way that fools us about what is going on.

    1. If you assume strict determinism (no free will) then there’s no distinction between determined by the past and determined by the future. Any global state determines all the future and all the past.

  25. Hossenfelder makes a powerful argument for superdeterminism. She is one of the most clear thinking physicists that I have come across. In my book Notfinity Process: Matter-In-Motion 2nd Edition (2021), I cite her paper “Free will is dead, let’s bury it.” I write: “Theoretical physicist Sabine Hossenfelder of the Frankfurt Institute for Advanced Studies shows the impossibility of free will. Hossenfelder denies the generally understood concept of free will, defined as the possibility of an alternative action having been available. Her position is based on the view that the decision-making process cannot alter the immutable laws of physics, in which all events are an outcome of electromagnetism, gravitation and the strong and weak nuclear force. As everything derives from the differential equations for the fundamental laws of nature, it is possible using these and a system’s configuration to calculate what happens at all other times using classical mechanics. She argues that including QMT in that calculation introduces an irreducible element of randomness into some events accounting for that aspect of one’s undetermined future. She rejects the argument that quantum indeterminism allows for free will on the grounds that this simply means some randomness is present, which could be fundamental or represent deliberately ignoring information on short distances. Regardless, there is no will or freedom.”

  26. The part of Hossenfelder’s argument I don’t get is how she claims that “Superdeterminism is local” under her understanding. She seems to underplay the oddities of her view. If I’m understanding this correctly, the particle in some sense ‘knows’ about aspects of the experimental setup ‘before’ encountering them directly (ie, by local causal encounter). It seems as if something about the particle is non-locally informed or made aware (maybe time-wise or in virtue of other ‘hidden’ variables present for the system) however lately or freely the setting are chosen. This could all be right, but those would be strange properties that the particle would have access to in some way. SH seems to be replacing ‘spooky action at a distance + free will(/statistical independence)’ with ‘no spooky action + no free will + spooky foreknowledge/awareness of the experimental setup by the particle/system before running it’. It seems a stretch of the concept of locality to claim something like that ‘it’s a local fact about the particle that it knows/is aware of facts about the setup that do not seem local (in either time or space) at all (under a reasonable understanding of the term “local”).’ So how is Sabine understanding the term in the context of SD?

    1. It’s only “foreknowledge” in the sense that all aspects and results, including the position the experimenter will select to measure, are “known” in advance because all the things involved have chains of interactions to common causes in the distant pass and so have been determined. All those interactions were local; no faster than light.

      1. Ah, right, thanks, that is surely what SH means by “local”: no superluminal causation. Still though, this level of preordination seems no less spooky or weird than faster than light communication, at least to me (and Bell, I’d guess). I’m not entirely sure Sabine is committing herself to this level of preordination though? The minimal form of denying statistical independence is just to claim that the ‘hidden variables’ are somehow correlated with the measurement settings. One of the possible ‘somehow’s is that everything that happens traces back to initial conditions, but it might be possible some other way, it’s just mysterious what that other way might be. I take her point that however the variables and settings are correlated, this doesn’t entail an ‘inability to do science’ but in worlds with more preordained conspiracies of results it would be possible that science could be wildly more wildly mislead than in worlds that are less preordained, it seems to me.

        The issues to do with ‘undermining science’ remind me of some criticisms of Everettian interpretations of QM that go something like this: if there are worlds for every possible result of every experiment, then it seems like there would be worlds where the most unlikely results happen *every single time* some experiment is performed and in such worlds science would be grossly mislead at least and perhaps impossible.

        1. The problem with superdeterminism is that every law-like pattern we observe is just explained by, “That’s the way the initial conditions determined it.”. Unless we have some additional theory that constrains the initial conditions and what they can lead to, the theory is vacuous. A theory that can explain anything fails to explain at all.

Leave a Reply