David Bohm, Quantum Mechanics and Enlightenment
blogs.scientificamerican.com
blogs.scientificamerican.com
I really do hope that we can find a way to make sense of everything in Physics, but Bohm's description of the infinite in the article is rather fascinating. Maybe there's no way out after all
https://www.youtube.com/watch?v=QcBpDVzBPMk
(In the end the most successfully interpretation has been the "shut up and do the math" one, but the interpretations help us find unexplored and contradictory sides of the theory and hopefully help to find new experimental possibilities)
Actually there is a reason: we see classic physics with our own two eyes every day. We know classical physics exists at the macroscopic level, and so a theory that simply explains the transition from microscopic to macroscopic physics and how this works.
This transition is unbelievably simple in Bohmian mechanics, since in principle, it's just classical physics with an extra term to account for quantum influences. The picture isn't nearly so simple for many other interpretations.
I like Bohm's book on Quantum Physics, but the pilot wave theory... well it doesn't explain anything new and the only problems it might solve are related to prejudices. Besides, the nature of that duplicity is never explained (and very ugly), so you're trading the ignorance you don't like for the one you like. In this kind of situations just go for the more simple and consistent theory as the only thing that matters is measurements.
(There's also the problem of getting to some form of QFT with it. Bell has written about this elsewhere, it's not pretty but I'm sure somebody has arrived to a working model by now.)
> well it doesn't explain anything new and the only problems it might solve are related to prejudices.
Correct, it actually explains everything old which prior interpretations like Copenhagen could not. I assume by "prejudices", you're referring to classical prejudices, but I think those prejudices are perfectly warranted.
> In this kind of situations just go for the more simple and consistent theory as the only thing that matters is measurements.
I disagree! John Bell formed the first no-go theorem because he was a fan of Bohmian mechanics. Without this focus on interpretations and quantum fundamentals, we may never have seen that landmark result or the series of subsequent no-go theorems which have all proven critical to our scientific understanding.
Science and physics does not consist solely of measurements; considerations of ontology are at least equally important because they inform the thought experiments which really drive scientific revolutions.
> Bell has written about this elsewhere, it's not pretty but I'm sure somebody has arrived to a working model by now
There are a few workable models, but Bell was correct to say that non-locality was the unsolved problem of quantum mechanics. Other interpretations let you "solve" it by sweeping it under the rug, but it always rears its ugly head elsewhere.
I'm not opposed to alternative formulations of QM, but I'd rather prefer people would try to come up with more fundamental theories than QM (if possible) instead of trying to get the same results with something else because for some ideological reason they don't like standard QM.
And it really is all about the measurements: Physics is an experimental science. You can think all you want and do crazy math and lots of gedanken, but we need at least a number and see what Nature thinks about it.
> You need spin to describe ferromagnetism, there's no room for that in a classical model of charges.
I suggest you actually read the link I provided. A classical theory has certain characteristics that distinguish it from something like a quantum theory. As I suspected in my initial reply to you, you have a bizarre notion of what constitutes a classical theory. Rest assured, a classical theory of spin and ferromagnetism is entirely possible.
As for measurements, physics is not solely an experimental science. Yes, theory must match experiment because empirical knowledge is what we're after, but measurements alone do not advance this agenda.
There are literally dozens of papers you can readily find that reproduce spin predictions using classical statistical mechanics. You denounce this as "model building", but every physical theory was model building.
> Try to build a classical theory of the photoelectric effect, it's impossible yet it opens the door of your elevator.
This is exactly what I was talking about, you consider that your atom has energy levels and then you plug in classical EM fields, sometimes that's good enough to get somewhere. But it can never be classical because there's no way to have energy levels in a classical atom.
Just take one positive charge and one negative charge (let's call electron to this one), then use Maxwell equations (you need the Liénard–Wiechert potentials) and try to calculate the energy of the electron, now it has a trajectory and everything. It's not for the faint-hearted (you have to learn classical electrodynamics first -whose consistency problem BTW has only been solved recently, or at least that's what Rohrlich thinks about it-). You won't get discrete energy levels, you won't even get a stable atom. Precisely this was the starting point of QM: Bohr coming up with a quantization rule for angular momentum and magical orbits where the accelerated classical electron does not radiate. Check "old quantum theory".
These problems were solved almost a century ago, the thing is people don't study enough or apparently have forgotten about them. They simply consider the QM postulates and then wonder why all that happened. Well, don't start the house by the roof, study what you should know already, the answers are there.
And this is what I'm talking about: this conversation is becoming an ever-shrinking pocket of you claiming, "well that's not classical", or "that can't be done in a classical theory", and me demonstrating that it has been done, toy theory or not.
I have no problem with discrete theories, I have a problem with people claiming something can't be done when it most assuredly can.
As for your derivation of old QM, I don't see how this is relevant. You seem stuck in the mindset that "classical theory" means building off of some known classical theory. As I've been saying, this isn't necessarily true. My discussion of classical theories are about the class of theories with various properties that differentiate it from something like QM.
People working on Bohmian mechanics are perfectly familiar with the history of quantum mechanics, see [1]. This doesn't somehow negate the fact that Bohmian mechanics can be viewed as a classical theory that nevertheless reproduces the predictions of QM. And it solves the radiating electron orbit problem just fine. Again, you're stuck thinking about this like classical mechanics rather than as a different classical theory that reduces to classical mechanics in the limit.
If there is a formulation of quantum mechanics which retains classical determinism, then just because it coheres so well with the behavior of everything we have directly observed (insofar as that's possible through the intermediaries of our senses) so far, it should be given a good amount of attention.
> but the pilot wave theory... well it doesn't explain anything new and the only problems it might solve are related to prejudices.
It can only be considered a prejudice if you're already pre-decided that non-classical physical descriptions are necessary before attempting to evaluate pilot wave theory. You're saying the prejudice is against non-classical behavior, right? Just put yourself in the shoes for a moment of a person judging (the modern formulations of) these theories, but at a point in time before any of them had a reputation (in an alternate history I mean). Would it be a prejudice to select from among essentially equivalent (predictive power) theories the one which doesn't force us to leave the then reigning classical paradigm?
If Pilot Wave theory and the more established quantum theories have comparable predictive power, but pilot wave theory can do the same thing without introducing new classes of behavior never observed at the macro level, then that's a point in its favor via Occam's Razor if nothing else.
About the pilot wave theory, well it came after the standard formulation of QM, so people didn't choose between them, and it's quite a patchwork that brings its own complications. Apparently as it includes classical trajectories some people love it, ignoring that you put by hand the actual QM wavefunction to make it work, then you fix things to get the old stuff back. I find all that weirder than standard QM.
I assumed that's what you're saying. I'm familiar with the viewpoint that classical physics is a pragmatic approximation to QM for macro phenomena.
> About the pilot wave theory, well it came after the standard formulation of QM, so people didn't choose between them
Of course. I was posing a hypothetical to demonstrate an error the argument you were making (assuming the conclusion as part of your argument).
> ignoring that you put by hand the actual QM wavefunction to make it work, then you fix things to get the old stuff back. I find all that weirder than standard QM.
I may be missing something here, and I haven't looked into the calculational details of pilot wave theory, but my understanding is that the same probabilities you ordinarily deal with in QM exist, but they are used to signify lack of knowledge of the observer of the state of the system (rather than saying that the system is literally in an indefinite state), so it makes sense that the wavefunction machinery would still be present, but used in a different way.
Bohmian Mechanics: Given Schrödinger's equation, Bohm's equation, the initial positions of the experiment and the measuring apparatus (assuming it is set on autopilot) and the initial wavefunction of the joint experiment and apparatus setup, then we evolve the system and the positions of the stuff in the apparatus will tell us the result of the experiment. Done.
Copenhagen Interpretation: We need Schrödinger's equation for the experimental setup, the initial wavefunction of the experiment, and some explicit timing for collapse as well as some stand-in for the eigenstates (we never actually collapse to an eigenstate or anything spiky, but rather to a nearby smooth state, such as a finite sum of Gaussians). We can optionally add-in the apparatus or observer wavefunction and collapse to something with an outcome involving those wavefunctions. It is kind of a mess to tell a computer how to do this, particularly if you have a number of experiments, possibly conditioned on the results of other experiments. It is not an independent time evolution. The Maths fail us in setting up a universe.
One could let this multi-experiment setup evolve until everything is done and then collapse at the very end. That leads to less of a mess of specification, but now one has the fact that we are saying reality does not have, in any, even approximate, sense, a definitive experience such as the one we are used to, at any point except for the last one. In some sense, this is the Many Worlds theory.
GRW is based on the idea of adding automatic collapse into the equations and that's fine and cool; it hits the wavefunction with a gaussian every so often. One also has to add in something else of an ontological nature, but that can be considered a philosophical issue (though it can matter what one chooses, say from a relativistic point of view). We at least have in GRW a Maths based evolution though it is different than the standard one.
The Copenhagen Interpretation is not an actual theory; there is no possible universe where it could be true or implemented. It can be made more specific, but it does require something more and the more one sticks to the Copenhagen spirit, the more of a "on high" specification mess, it becomes.
Both GRW and Bohmian Mechanics at least could be true.
No, that is simply throwing in the towel. Giving up should never be confused with success.
(Regardless, this is a great interview to watch if you are interested in physics)
In similar vein, here is Eric Weinstein [2] talking about why we can’t find the theory of everything — and has similarly...kooky...ideas as Bohm. He is short-String theory (https://www.edge.org/response-detail/25547). His main critique comes from the demands that the theory starts imposing about wanting 11 or 26 dimensions.
All this in some level is super exciting and suspenseful, as I do believe Physics is on the cusp of a “new big idea from a lone genius” type revolution, that might push things forward!
[1] https://www.youtube.com/playlist?list=PLFJr3pJl27pItxUFeZlx9...
I.e. why couldn't it be that a pair of photons are generated with their properties already in such state that, when measured, a correlation is present? the photons don't communicate, they are already in the appropriate state when measured.
http://www.scholarpedia.org/article/Bell%27s_theorem#Bell.27...
In a letter to Max Born he wrote about Bohm’s theory: “That way seems too cheap to me.”
This is a very, very big question. Its a little pity he didn't meet Sadhguru.
I should also say that it is not taught in schools, papers about it are not published in serious journals and the field is just not an active field of research. It's quite likely that all the people trying to develop this idea are crackpots.