Wednesday, November 8, 2023

Quantum Mechanics said to be Hilariously Ill-defined

Physicist Sean M. Carroll on his monthly AMA podcast:
Q. Can you talk about the future of many-worlds? What would have to be true, or what milestones would have to be achieved for the majority of the physics community to adopt many-worlds as a proper model of foundational physics?

A. I think there's two things going on, one is already happening. Which is, you just need to appreciate the need for a proper model of foundational physics, a proper model of quantum physics in particular. You know we have been muddling along with the Copenhagen Interpretation, which is hilariously ill-defined for a very long time now, but as technology improves, and as physicists are paying more attention to truly quantum phenomena, especially entanglement and measurement, and so forth, they are becoming easier to convince that we need to get quantum mechanics right. So whether it is many-worlds or something else, first we need to convince physicists that it is worth spending the effort to think carefully about the foundations of quantum mechanics.

He goes on to say the second thing, which is that someone has to find some practical utility to many-worlds, and no one has found any yet.

I do not see how a respectable professor can say anything so ridiculous.

QM is not a new or immature theory. Its worldwide impact is about a trillion dollars a year, counting all the semiconductors, lasers, liquid crystals, etc. Those products are built with an understanding of the basic science from textbooks on the Copenhagen interpretation of QM.

Physicists need to be convinced to get it right? What about all those people invested in that trillion dollar economy? Don't they want to get it right?

Carroll is telling them that they need to give up their theory for some many-worlds nonsense that has never shown any practical utility whatsoever.

Got that? On the one hand, a theory worth a trillion dollars a year. On the other, one worth zero. And Carroll is trying to get people to jump to the latter theory.

This is like saying: What would it take for people to stop driving cars, and using wormholes for transportation instead?

And the eminent professor answering: We need two things, to convince everyone that cars are hilariously ill-defined, and to find a way to use wormholes for transportation.

In a sense, he's right, in that people would use wormholes if they became more useful than cars. But there is no chance of that ever happening.

Here is a new article on many-worlds (MWI) being unscientific:

We show, in fact, that a whole class of theories -- of which MWI is a prime example -- fails to satisfy some basic tenets of science which we call facts about natural science. The problem of approaches the likes of MWI is that, in order to reproduce the observed empirical evidence about any concrete quantum measurement outcome, they require as a tacit assumption that the theory does in fact apply to an arbitrarily large range of phenomena, and ultimately to all phenomena. ...

The unblemished success of the theory in such ample range of phenomena is really staggering. Precisely because of that, it is suicidal to leave our best comprehension of such rounding success in hands of any interpretation that, due to its soaring ambition, is incapable of building itself on any concrete empirical ground, and therefore cannot but fall apart sooner or later. ...

In the specific case of MWI, there seems to be an almost religious sentiment that animates its supporters by believing that everything that exists is a single, “simple”, immutable, elegant mathematical object, which supposedly lives in an abstract Hilbert space. In this view, everything we observe and experience, including the space in which we move and live, would just be emerging from the only “real” entity – the universal wave function [44]. With the arguments exposed in this article, we then join Heisenberg here who, to similar claims put forward by Felix Bloch, once simply replied: “Nonsense, space is blue and birds fly through it.”

I do believe that many-worlds is completely contrary to everything we know about science. Believing in it is more backwards than believing in astrology or witchcraft.


  1. Sean is right. QM doesn't even define measurement, and that's why all the crackpot theories proliferate. The standard model is a QFT and not QM. QM can't even explain the Pauli exclusion principle. QM even offers a bogus position pseudo-operator. Field ontology dispels all the conventional mysteries and is completely local. I don't know who thought you could error-correct infinite-dimensional fields to create a quantum computer. QM is too abstract and it's obsolete.

  2. So we have a century of textbooks and a trillion dollar QM economy, and you and Sean say it is all bogus? Let me know when you find something better.

    1. WRONG! The standard model is a QFT. You are a century behind.

    2. Sure, fields are more fundamental than particles. Nevertheless, quantum textbooks talk about particles all the time, and there is a huge industry based on it.

    3. And it's bogus nonsense. QM uses finite-dimensional Hilbert space without relativity. That's beside the point that Wilson RG lattice theory can incorporate gravity into QFTs without much fuss.

  3. Gentlemen,

    Two brief points:

    1. To Roger: The Copenhagen interpretation of QM isn't the text-book QM --- or vice versa. The latter is based on the QM postulates. I have a compilation document about it, here: . Conflating the two has been a long drawn business.

    2. To MD Cory:

    > ``QM even offers a bogus position pseudo-operator.''

    No harm, within the QM context, in regarding it as an operator. Recall: Dirac's delta has been condemned, right from its inception, to live its entire life behind the integral sign. From the limit of (parametric) integrals to the integral of a (parametric) limit---that's the switch involved. The function to the distribution switch is just a part of it, a minor one, if you ask me.

    The position operator, even in NRQM, has Dirac's deltas as eigenstates. Why must it be pseudo? Just because the result of the action of an operator has a *compact* support doesn't mean that the operator isn't *defined* globally.

    > `` Field ontology dispels all the conventional mysteries and is completely local.''

    As to the ``field ontology,'' there has been a lot of discussion on the Fields vs. Particles in the context of (relativistic) QFTs. The consensus is that no unique ontology associated with today's (and yesteryears') QFT has been settled on. People keep arguing, but in a mature, sober, way.

    For the ``local'' part: No. If you have QFT in mind, then: Field operators *are* local, but the states *are* global. (Once again, it's the issue of the compactness of the support vs. the domain of the very definition itself.) The premise of the global definition, together with the requirement of the unitarity, directly implies that even the *relativistic* QFT must remain nonlocal --- in principle. Feel free to correspond privately with me, if you wish to pursue this issue further...

    Before closing, best compliments of the Diwali season!


    1. It's a pseudo operator because the underlying reality is fields. They even recently confirmed Unruh radiation.

      Read my letter to a relative at IBM against QCs:

    2. Yes, QM can be presented as postulates. But eventually the theory connects with real world experiments, and then the book usually follow the Copenhagen interpretation.

    3. Roger,

      Have to be a bit elaborate!

      1. For connecting the theory with the experiments, text-books fall back on one of the *postulates* themselves, viz., the measurement postulate, i.e., Born's rule (with randomness). They do it without (much) explanation. (Any explanation wouldn't affect the calculations!)

      2. Parts of the Copenhagen Interpretation (CI) which textbooks do mention:

      2.1. They do mention the Correspondence Principle (typically, with the example of the principal quantum number getting arbitrarily large, which actually involves the mathematical pathology of a wavefunction that tends to get kinked everywhere and so loses being differentiable, even approaching multivalued-ness, but textbooks summarily ignore these pathologies --- or the faulty argument by Bohr).

      2.2. Yet, typically, they do *not* mention the Complimentary Principle at all, except perhaps in minor historical tid-bits or asides. In contrast, Bohr was big on it, very big. It formed one *integral* part of the CI. Detectors had to be classically describable, Bohr taught.

      2.3. Textbooks also don't emphasize another thing that Bohr was big on, namely, the in principle sudden-ness of the quantum jumps (which had irked Schrodinger a lot). Textbooks hardly even mention it, and if at all, they do it as a minor aside, of almost no consequence... (Not useful for calculations, you see!)

      2.4. Thus, note, they don't use the last two elements of CI, whether to connect the theory to the experiments or for any other purpose.

      3. A thing which textbooks highlight a lot, but is *not* necessarily a part of the CI is: the Uncertainty principle.

      But here's the good part. Textbooks don't highlight the HUP for its Heisenbergian spirit (as limitation on knowledge itself, etc.). They begin highlighting it early enough, just in order to make the student comfortable with the idea of the *commutator*. You see, eventually, in more advanced work (including in QFT), the commutator comes to play a big role. That's why, textbooks make a good faith attempt --- from the viewpoint of *maths* and *pedagogy*, but *not* because they want to promote HUP.

      But, though textbooks emphasize HUP a lot, it's doubtful if you can call it part of CI or not! Bohr in fact didn't think much of it! (Which had diappointed/dismayed Heisenberg!)


    4. Ajit, those are minor quibbles about the definition of the Copenhagen interpretation. Yes, there are some minor disagreements about the details. The other interpretations are pilot wave, many-worlds, etc. The textbooks do not do those. They do Copenhagen.

    5. Roger,

      Those may be minor issues for the discussion in a thread like this, but I am afraid, there are consequences for accepting the ``Textbook description = Copenhagen Interpretation'' equation.

      Let's face it: Physics textbooks, despite their criticism, still have credibility --- and a lot of it is well earned too.

      So, if one accepts that equation, then doing so ends up transferring this credibility to CI in particular. Seeking to do that is obviously to the subtle advantage of a certain *philosophic* position (as distinguished from a *physics* theory), viz., the Logical Positivism (LP).

      I was trying to prevent *that* too, apart from saying that textbooks, by and large, have tried to avoid a tilt to any one interpretation, specifically, CI.

      People are free (even welcome) to advocate LP if they wish to. In the process, they may even make references to validated physics theories, by way of *illustration*. But, a theory in philosophy cannot be validated by appeal to scientific/physics theories (even people do routinely attempt that). ``XYZ is *scientific*,'' they cry out! But it's wrong (and futile) to corral physics/science theories to advance philosophic positions.

      In this context, IMO, QM textbooks have done fairly well, though they do show a certain *implicit* slant towards CI. The equation doesn't quite hold, but there *is* that bias.

      Even then, a true CI-ist would be disappointed too: ``Don't just quote Born's postulate and use it in calculations,'' he would demand, ``talk also about the in-principle impossibility of reconciling the classical and the quantum realms, because that's the settled truth'' (as the Complimentary Principle demands). QM text-books have avoided adopting such positions explicitly/categorically. So, they don't do Copenhagen. That's the bottom-line. Good for physics, and good for philosophy, if you ask me!

      Let me sign off for now, though I will check out this thread tomorrow...


    6. Quantum jumps were dispelled by experiment. Also, collapse is not instantaneous.

    7. Furthermore, QFT is completely local in the Bell sense. Field commutators vanish. It's true to say that it's trivially nonlocal. That doesn't conflict with any classical analog or experience.

  4. Epi-cycles still work, and are still used to make orbital predictions within a certain margin or error, but that doesn't mean they are in any way scientific or remotely related to the actuality of orbits, planets, or stars.

    Models often have useful purpose without being remotely correct, as long as someone doesn't forget this, and start making up theories based off the model's functioning instead of the reality, it's ok. Quantum computing and climate change modeling are excellent examples of what happens when this simple truth is forgotten.

    1. Yes, it is scientific to use epicycles to make successful orbital predictions.

    2. Pragmatic or more easy? Sure. Scientific, no. Read up on the premises which are underlying the very concept of epicycles model.
      The universe does not revolve around the earth, and it sure ain't ok to say so out of convenience.

      Perfect circles within circles are not how the heavens go, unless you subscribe to the universe being geocentric construct. Sure you can hand tune your periods like a Swiss watch or any other hand fitted algorithm, but once again, a watch with wheels within wheels is not how orbits function. At all. All you have demonstrated is that it is possible to fit a tailored algorithm to any outcome, the algorithm is not what produces the outcome however, it's a hand tuned correlation, not a causation.

      "All models are wrong, but some are useful."
      --George E. P. Box

      " On two occasions I have been asked [by members of Parliament], ‘Pray, Mr. Babbage, if you put into the machine wrong figures, will the right answers come out?’ I am not able rightly to apprehend the kind of confusion of ideas that could provoke such a question."
      --Charles Babbage

      "A big computer, a complex algorithm and a long time does not equal science."
      -- Robert Gentleman

      "With four parameters I can fit an elephant, and with five I can make him wiggle his trunk."
      --John von Neumann

      I would require on pain of death all physicists or fools calling themselves scientists to take computer science and computer programming. It would literally force them into seeing the consequences of their ridiculous assumptions, and acknowledge the limits of actual calculation in reality, not the untethered hand waving they think is logic.

    3. Perfect circles within circles is how everything goes. That has been proved by Fourier analysis. And under general relativity as has been understood for a century, it is okay to take any frame of reference for convenience.

    4. Roger,

      Actually, the way I see it now, after a re-reading, two points:

      1. You actually have a very fine post here (I mean the main text).
      2. I unnecessarily took it to further a part of the subsequent to-and-fro.
      3. I also made a couple of typo's (regarding, who else, Bohr!)
      4. But of course, I had some good points, but, nevertheless, the point no. 1 is more valid. Just realized this night...
      5. I am making a copy of your main post (together with all the replies, including mine), and am going to respond, eventually, probably at my own blog.