Explaining John Bell & Anton Zeilinger's Attempt to Prove No Definite Reality

[thinkandmull]

I was wondering if someone could explain to me how John Bell and Anton Zeilinger have attempted to prove there is no definite reality in the sub-atomic world. How could it ever be proven there are no hidden variables that humans just don't or can't know about? If I am not mistaken, Einstein for example did not prove there is no ether, just that we can't detect it.

THANKS

 

[Heinera]

I suggest you start with this paper:

http://cern.ch/maltoni/PHY1222/mermin_moon.pdf

And then you can continue with this one, also by Mermin:

https://upload.wikimedia.org/wikipedia/commons/6/68/Variables_ocultas_y_los_teoremas_de_Bell.pdf

 

[Nugatory]

You will also find some good stuff, including a link to Bell's original paper and some simpler but analogous arguments at http://www.drchinese.com/Bells_Theorem.htm, maintained by our own @DrChinese . Read what you find there, follow Heinara's references, and if you still have questions, search this forum for some of the 93 megabazillion previous threads on the subject - chances are that any objections/questions you have have been discovered before. And this book by Louisa Gilder is a good layman-friendly historical summary.

And then if it's still not clear... Ask away.

Bell's theorem is surprisingly subtle for something so simple in hindsight. The problem was starkly posed by Einstein in 1935, and it was nearly three decades after that before Bell noticed the implications of the correlations between three different observables, as opposed to two.

 

[Nugatory]

And one more must-read: https://www.scientificamerican.com/media/pdf/197911_0158.pdf

 

[thinkandmull]

What general argument can I give to lay people who say "there is no way to know if there are hidden forces at work"?

 

[Weddgyr]

What general argument can I give to lay people who say "there is no way to know if there are hidden forces at work"?

Assumming they mean local hidden forces/variables, you can say that any such forces, when measured, produce data which must satisfy certain constraints. Quantum mechanics produces data which violates these constraints, and so cannot be the result of such hidden forces. You can add that the constraints themselves are given by Bell's Theorem and are accessible to even the modestly mathematically inclined.

 

[RUTA]

I was wondering if someone could explain to me how John Bell and Anton Zeilinger have attempted to prove there is no definite reality in the sub-atomic world. How could it ever be proven there are no hidden variables that humans just don't or can't know about? If I am not mistaken, Einstein for example did not prove there is no ether, just that we can't detect it.

I think the easiest way to see how QM can imply no counterfactual definiteness is Mermin's explanation of the Hardy experiment. I give an overview here: https://www.physicsforums.com/insig...elayed-choice-no-counterfactual-definiteness/

Mermin also argued against the existence of "instruction sets" aka "counterfactual definiteness" in his paper: Bringing home the atomic world: Quantum mysteries for anybody. American Journal of Physics 49, 940-943 (1981). I give an overview of that paper and its quantum liar implication here: https://www.physicsforums.com/insights/quantum-liar-experiment-instantiation-mermin-device/

Mermin also wrote a simple explanation of how the GHZ experiment implies no instruction sets in his paper: Quantum mysteries refined. American Journal of Physics 62 (10), 880-887 (1994). I give an overview of that paper and its Many Histories implication here:
https://www.physicsforums.com/insights/greenberger-horne-zeilinger-experiment/

 

[phyzguy]

What general argument can I give to lay people who say "there is no way to know if there are hidden forces at work"?

The link given by Nugatory in post #4 is quite accessible to the lay person, and doesn't require any advanced mathematics, just logical thinking. It is old, and since that time there have been many more experimental tests that show that the universe obeys quantum mechanics and violates the Bell inequalities, and hence cannot be locally realistic.

 

[thinkandmull]

Is this getting into philosophical questions though? A theist could still believe that God is the unknown factor, or an Indian mystic an impersonal force he calls Brahman. Is QM really able to refute these claims? I think that was what Einstein was trying to say.

 

[thinkandmull]

Or maybe Einstein however believed the unknown laws could be found apart from God. He did like to read Spinoza though, who thought the world was God in a sense

 

[RUTA]

Is this getting into philosophical questions though? A theist could still believe that God is the unknown factor, or an Indian mystic an impersonal force he calls Brahman. Is QM really able to refute these claims? I think that was what Einstein was trying to say.

The philosophical issue here is one of ontology, not epistemology, so no one is questioning the method of physics. Physicists have a working ontological model whenever we employ our equations, usually they're tacit and widely agreed upon. In the "weird" QM experiments of interest here, ontology is exactly what's being discussed. Such discussions are germane to progress in physics, e.g., phlogiston, ether, atoms, etc., so physicists of the highest order have engaged this topic. The goal of such discussions from a physicist's perspective is to generate new theories, not to create despair or advocate for religious belief.

 

[.Scott]

What general argument can I give to lay people who say "there is no way to know if there are hidden forces at work"?

Imagine writing three sets of 1's and zeros each 1000 bits long, sets A, B, and C. Then you compare the sets - matching up the first bit with the first bit, the second with the second, as so on. When you compare A with B, you get a 10% difference. When you compare B with C, you also get a 10% difference. But when you compare A with C, you get a 30% difference. The problem is how can A be 30% different from C when it's only 10% different from B and B is only 10% different than C. There should be a core 80% to 90% of the bits that are common with each other - so A should be no more than 20% different from C.

In very general terms, that is the problem present when the Bell Inequality is violated. You are measuring the differences among three groups of particles, and the result is "unreal".

 

[thinkandmull]

Oh now I get what the argument is. It's similar to Schrodinger's cat. In the world of classical physics though, does quantum mechanics apply? I don't see that atomic phenomena alone can prove that my spouse's body does not exist unless someone is seeing it.

 

[Nugatory]

Oh now I get what the argument is. It's similar to Schrodinger's cat. In the world of classical physics though, does quantum mechanics apply? I don't see that atomic phenomena alone can prove that my spouse's body does not exist unless someone is seeing it.

Quantum mechanics applies to macroscopic and classical systems just fine. If you calculate the quantum mechanical behavior of a large number of individual particles assembled into a classical object, you'll come up with the classical result. It has to turn out that way, because classical physics provides an accurate description of the behavior of objects composed of a large number of particles. Therefore any correct theory of the behavior of particles has to predict that they behave classically when assembled into the large collections we call classical objects - otherwise it wouldn't be a correct theory of the behavior of particles. (This was understood in a heuristic sort of way from the beginning, but it took a lot longer before the non-trivial math was worked out in the middle of last century - google for "quantum decoherence").

In particular, quantum mechanics does not say that your spouse doesn't exist, or that the moon isn't there when no one is looking, or especially that Schrodinger's cat is both alive and dead (or neither dead nor alive, or in any other weird state). Either there's a classical dead cat in the box or there's a classical live cat.

 

[RUTA]

Quantum mechanics applies to macroscopic and classical systems just fine. If you calculate the quantum mechanical behavior of a large number of individual particles assembled into a classical object, you'll come up with the classical result. It has to turn out that way, because classical physics provides an accurate description of the behavior of objects composed of a large number of particles. Therefore any correct theory of the behavior of particles has to predict that they behave classically when assembled into the large collections we call classical objects - otherwise it wouldn't be a correct theory of the behavior of particles. (This was understood in a heuristic sort of way from the beginning, but it took a lot longer before the non-trivial math was worked out in the middle of last century - google for "quantum decoherence").

In particular, quantum mechanics does not say that your spouse doesn't exist, or that the moon isn't there when no one is looking, or especially that Schrodinger's cat is both alive and dead (or neither dead nor alive, or in any other weird state). Either there's a classical dead cat in the box or there's a classical live cat.

I would add there is no cut between the classical and the quantum, you can get a cat or your spouse into a superposition state if its screened off. It's been done with molecules of over 100 atoms (http://arstechnica.com/science/2012...ith-big-molecules-approaches-the-macroscopic/). And if you do that, the cat or spouse can exhibit no counterfactual definiteness, which is what I assume you mean by "doesn't exist."

 

[.Scott]

I would add there is no cut between the classical and the quantum, you can get a cat or your spouse into a superposition state if its screened off. It's been done with molecules of over 100 atoms (http://arstechnica.com/science/2012...ith-big-molecules-approaches-the-macroscopic/). And if you do that, the cat or spouse can exhibit no counterfactual definiteness, which is what I assume you mean by "doesn't exist."

Not that I have a cat or spouse that I would like to "screen off", but how long would you need to screen them off before there would be noticeable affects of superpositioning?
----edit----
Actually, that article you cited is based on the double slit experiment. So... If I was tossing 5Kg cats at two slits, I would want the wavelength to be wider that the cats - say 50cm. With λ = h/p, p = h/λ, the momentum would be Planck's constant divided by that 50cm = 6.62607004 × 10^-34 m² kg / s / 0.5m = 1.325 × 10^-33m kg / s.
Dividing by 5Kg to get velocity: 1.325 × 10^-33m kg / s / 5Kg = 2.65 × 10^-34m/s.
Okay, to get a good interference pattern, the feline trajectory should be roughly 20λ or more. That's 10 meters. So the transit time for the cats would be 10m / (2.65 × 10^-34m/s) = 3.77 × 10³⁴ seconds = 12.0 × 10²⁶ years. Or about a hundred million billion times the age of the universe.

 

[Nugatory]

IAnd if you do that, the cat or spouse can exhibit no counterfactual definiteness, which is what I assume you mean by "doesn't exist."

That's not quite what I meant - I was responding to the original poster's use of the term "doesn't exist" and his misunderstanding of Schrodinger's cat.

However, you are making a point that I agree with - quantum mechanics does not say that these weird macroscopic states cannot exist, it says that they won't happen (except when we make extraordinary efforts to temporarily suppress decoherence in what are still very small, barely macroscopic systems). There's an analogy with classical statistical mechanics here: statistical mechanics doesn't say that a volume of gas at equilibrium cannot violate ##PV=nRT## (it can - at any moment every particle has a 50% probability of moving towards the edge of the container so we might get lucky and find a large majority of them are doing so), it says that the probability of this happening is so low that for all practical purposes it never happens.

 

[fresh_42]

If you calculate the quantum mechanical behavior of a large number of individual particles assembled into a classical object, you'll come up with the classical result.

Not that I dare to claim having understood both statements in their most inner essence, but it seems to me worth debating it:
http://phys.org/news/2015-12-quantum-physics-problem-unsolvable-godel.html

 

[RUTA]

Not that I have a cat or spouse that I would like to "screen off", but how long would you need to screen them off before there would be noticeable affects of superpositioning?
----edit----
Actually, that article you cited is based on the double slit experiment. So... If I was tossing 5Kg cats at two slits, I would want the wavelength to be wider that the cats - say 50cm. With λ = h/p, p = h/λ, the momentum would be Planck's constant divided by that 50cm = 6.62607004 × 10^-34 m² kg / s / 0.5m = 1.325 × 10^-33m kg / s.
Dividing by 5Kg to get velocity: 1.325 × 10^-33m kg / s / 5Kg = 2.65 × 10^-34m/s.
Okay, to get a good interference pattern, the feline trajectory should be roughly 20λ or more. That's 10 meters. So the transit time for the cats would be 10m / (2.65 × 10^-34m/s) = 3.77 × 10³⁴ seconds = 12.0 × 10²⁶ years. Or about a hundred million billion times the age of the universe.

The numbers are correct, but far more problematic for actually making quantum cats is keeping them from interacting with the environment during the incredibly long time it takes them to get from Source to detector. It's certainly not technologically feasible to screen off a cat. My point is that QM says there is no classical-quantum boundary in principle, even if such boundaries exist in practice.

 

[.Scott]

The numbers are correct, but far more problematic for actually making quantum cats is keeping them from interacting with the environment during the incredibly long time it takes them to get from Source to detector. It's certainly not technologically feasible to screen off a cat. My point is that QM says there is no classical-quantum boundary in principle, even if such boundaries exist in practice.

I think we both make the point that it's not practically feasible. Personally, I was more impressed with the problem of keeping the cats (and the investigators) alive for that long. But certainly keeping the cats isolated in a weightless environment is a stretch as well.

 

[votingmachine]

What general argument can I give to lay people who say "there is no way to know if there are hidden forces at work"?

One analogy is to take a typical coin toss. It has two possible ends: Head or Tail. And if you toss a coin 100 time, the number of heads plus the number of tails is 100. There is an underlying classical reality, if you catch the coin on your arm, and cover it, it has a value, even though that value is unknown.

You would be disturbed if you flipped the coin 100 times, and ended with 40 heads and 40 tails. Or if you ended with 60 heads and 60 tails. The numbers don't add up to 100. That would violate the essential assumption that the coin flip state has defined properties that become fully known. Apparently, some of the time, the properties were unknown.

Classical probabilities for things with defined properties will have that critical feature that all of the possible outcomes have a probability sum of 1.0000... Bell's inequality is at its most basic, a re-statement of that, with the INequality being that if you end up with a number that is not equal to 1.000... then you are dealing with a system that cannot have determinate classical properties.

The path in physics traces back to the EPR paper, which attacked the uncertainty principle, and quantum mechanics, calling it an "incomplete" description of the underlying reality. The uncertainty principle is not a complete description, in the classical sense because underlying property sets are uncertain. But what experimental evidence shows us, is that the description is as complete as it gets. Einstein's suggestion was to take two identical things (entangled particles) and measure the properties of one without disturbing the other, then measure the other. Then a complete description would be possible. If EPR was right, the outcomes of separate measurements would obey the math of simple classical probabilities. Instead the experiments have outcomes where the probabilities don't add up to 1.000...

The hidden forces that people want to assume, are ones that bring the description to completeness, and the probabilities to classical ones. The experiments are pretty conclusive that such a rendering is not possible.

 

[Ilja]

I was wondering if someone could explain to me how John Bell and Anton Zeilinger have attempted to prove there is no definite reality in the sub-atomic world. How could it ever be proven there are no hidden variables that humans just don't or can't know about?

I would not like to say something about Zeilinger, but Bell definitely has not attempted to prove that there is no reality. His aim was different. He has known a hidden variable theory - Bohmian mechanics, or de Broglie-Bohm theory.

This theory has a problem: It has also a hidden preferred frame, and the hidden variables have, similarly hidden, causal connections which happen faster than light. This raises the question: Are there better hidden variable theories, theories without hidden faster than light causal influences? The answer is NO.

But this is support for the hidden variable theory which has such FTL causal influences. Because having such influences is no longer an argument against the theory: Every hidden variable theory has such FTL causal influences. So, if somebody objects against dBB theory that it has such FTL effects, the appropriate answer is "so what?".

 

[RUTA]

Retrocausal interpretations need not employ FTL mechanisms. See my Insight "Understanding Retrocausality"