Joy Christian's colourful exploding balls experiment

[Heinera]

Just a quick question for Joy Christian: Does the conservation law for angular momentum still hold in your theory? (I'm afraid I wasn't able to deduce that by myself)

t
Yes.

Thanks! Does that mean that in your proposed experiment, we could just measure the angular momentum in one wing of the experiment, and use conservation of angular momentum to deduce the result of the other wing? (Assuming we prepare the experiment so that the angular momentum is zero at the outset of each run). I suspect this would considerably reduce the cost of the experiment.

[Joy Christian]

Just a quick question for Joy Christian: Does the conservation law for angular momentum still hold in your theory? (I'm afraid I wasn't able to deduce that by myself)

t
Yes.

Thanks! Does that mean that in your proposed experiment, we could just measure the angular momentum in one wing of the experiment, and use conservation of angular momentum to deduce the result of the other wing? (Assuming we prepare the experiment so that the angular momentum is zero at the outset of each run). I suspect this would considerably reduce the cost of the experiment.

Absolutely not. The whole point of the experiment is to detect the effects of the topology of the physical space S^3 on the angular momenta and the correlation between them. Any short cuts of the kind you are suggesting would contaminate the effects of topology and reduce the correlation to those calculated by Peres.

[Heinera]

Ok. So if the initial state was zero angular momentum, and the measured angular momentums of the two wings did not add up to zero, angular momentum would still somehow be conserved?

[Joy Christian]

Ok. So if the initial state was zero angular momentum, and the measured angular momentums of the two wings did not add up to zero, angular momentum would still somehow be conserved?

Yes. The only difference between my setup and that of Peres is the difference between the symmetry groups SO(3) and SU(2), both of which conserve spin momenta.

[Heinera]

Ok. So if the initial state was zero angular momentum, and the measured angular momentums of the two wings did not add up to zero, angular momentum would still somehow be conserved?

Yes. The only difference between my setup and that of Peres is the difference between the symmetry groups SO(3) and SU(2), both of which conserve spin momenta.

Thanks! I always suspected that the conservation laws didn't literally mean that something was actually conserved. The renowned journal "Social Text" has lots of interesting papers about this issue.

[gill1109]

Ok. So if the initial state was zero angular momentum, and the measured angular momentums of the two wings did not add up to zero, angular momentum would still somehow be conserved?

In Joy's experiment, we observe the two colourful spinning hemispheres with video cameras or other sensors, do a lot of image analysis, and come up, in the k'th run, k = 1, ..., N, with two directions of angular momentum denoted s_k and minus s_k. So in the processing of the data (the video films) we *assume* conservation of angular momentum.

We do this N times and store the N directions s_k in a computer file.

Then we pick directions a and b and compute A_k(a) = sign(a . s_k) and B_k(b) = sign(b . - s_k). The "dot" here standing for inner product, thinking of these directions as unit vectors in R^3. Presumably the directions a, b and s_k will be represented in our computer software by two angles (inclination and azimuth), see https://en.wikipedia.org/wiki/Spherical_coordinate_system, but any representation will do.

Then we calculate E(a, b) = 1/N sum_k A_k(a) B_k(b).

Now we pick some other pairs of directions, and repeat. Joy has repeatedly told us that we can use the same set of N directions of angular momentum +/- s_k, k=1, ..., N, for each pair of measurement directions a, b.

Do take a look at Joy's short experimental paper http://arxiv.org/abs/0806.3078, or (just Section 4 of) the longer paper http://arxiv.org/abs/1211.0784. Everything is written out very explicitly in both places, exactly the same instructions.

To be precise, the directions +/- s_k are the directions of spin at the moment the hemispheres reach a certain distance from their initial location. The computation involved here is the reconstruction of the actual (classical) motion of a spinning flying object from a number of video films of the object, taken from different angles. Joy thinks it would help to do the experiment in zero gravity and in a vacuum but probably it will work out OK in ordinary terrestrial conditions. Joy says there will be no "non-detections". But it doesn't matter since we either get a +/- s_k or we don't. The measurement directions a and b have zero influence on the detection.

Basically we are *measuring* the local hidden variable s_k and then only after that, *deducing* the measurement outcomes by plugging in our estimate of s_k.

Betting on the outcome of this experiment reminds me of the saying "taking candy from a baby".

[FrediFizzx]

Betting on the outcome of this experiment reminds me of the saying "taking candy from a baby".

Is that why you won't do the 100K pounds? LOL! So how much are you willing to lose?

[gill1109]

Betting on the outcome of this experiment reminds me of the saying "taking candy from a baby".

Is that why you won't do the 100K pounds? LOL! So how much are you willing to lose?

Fred

If Joy actually truly wants to do the experiment exactly according to the description in his experimental papers by which we shoot off and film N exploding balls, measure N pairs of directions of angular momentum s_k and minus s_k, and then compute E(a, b) according to the recipe in his papers using the same N directions s_1, ..., s_N for each pair of settings, then I will happily bet 100K Euro, because I can't lose. More if Joy wants to put more on the table.

If Joy decides after all to follow the usual CHSH protocol whereby each of the N runs is allocated by chance to one of the four usual measurement setting pairs, so that four correlations are each computed according to the recipe in his papers using a disjoint subset of about one quarter of all of the N runs, then depending on how large N is, I will happily bet say 5 000 Euro. I'll make sure that N is so large that I have less than 1 in a million chance of losing.

Either way: first the arrangements of the bet are formalized including a careful description of the protocol of the experiment and the protocol for determining who has won. It has to have the approval of our three man board of adjudicators as well as of both Joy and me, otherwise there is no bet. The bet is on the usual CHSH style combination of four correlations based on two settings for Alice and two for Bob. Half-way between 2 and 2 sqrt 2 will be the dividing line between me and him winning.

[gill1109]

We now have a three-man adjudication committee in place: Andrei Khrennikov, Hans de Raedt, Gregor Weihs. Next stage will be: finalise protocol, contact experimenter, start publicity offensive (crowd-funding) to pay for experiment.

This will be the experiment of the century and guarantee either a Nobel prize or an igNobel prize.

[FrediFizzx]

It will be the experiment of the century if it shows macroscopic violation of Bell's inequality. If it doesn't, Joy still has the true explanation of quantum correlations in the microscopic realm. The quantum experiments themselves support that. If the experiment is done with properly prepared macroscopic bunches of photons, my bet is that it will show violation of Bell. No doubt about it. I'm still a bit skeptical of a purely mechanical experiment but it must and should be done to find out for sure.

[gill1109]

It will be the experiment of the century if it shows macroscopic violation of Bell's inequality. If it doesn't, Joy still has the true explanation of quantum correlations in the microscopic realm. The quantum experiments themselves support that. If the experiment is done with properly prepared macroscopic bunches of photons, my bet is that it will show violation of Bell. No doubt about it. I'm still a bit skeptical of a purely mechanical experiment but it must and should be done to find out for sure.

It will be a purely classical-mechanical experiment. Joy wants to prove there's nothing "quantum" about "quantum correlations", he claims they result from the torsion of space. Because we live in S^3 not R^3.

I don't understand how this could make a difference in the laboratory. The lab scale is "local". Locally, S^3 is R^3. We are not sending our ping-pong balls to the end of the universe and back.

Anyway, the quantum folk haven't even done a good experiment yet. It's still 5 years off as far as single photons are concerned. It must be 50 years off for big bunches of photons.

[FrediFizzx]

Big bunches of photons = EM radiation. It's easy. Just a specialized EM pulse. Should be no big deal. And it will be fully macroscopic. The lasers feeding the parametric downconverter can do macroscopic EM pulses no problem. The problem will be in the filtering. But as I always say, experimenters are very clever and most likely could figure it out. I would say it should take them no longer than 6 months to figure it out. I had a paper somewhere on my computer on a partially macroscopic experiment. I will try to find it.

[gill1109]

Big bunches of photons = EM radiation. It's easy. Just a specialized EM pulse. Should be no big deal. And it will be fully macroscopic. The lasers feeding the parametric downconverter can do macroscopic EM pulses no problem. The problem will be in the filtering. But as I always say, experimenters are very clever and most likely could figure it out. I would say it should take them no longer than 6 months to figure it out. I had a paper somewhere on my computer on a partially macroscopic experiment. I will try to find it.

Well, the experiment I'm betting on with Joy, is the one described in his experimental paper(s). Classical mechanics. Spinning hemispheres.

Your experiment: if it would take them only 6 months to figure out and since a positive result (no loopholes, CHSH style repeated, random, binary setting choices, binary outcomes) would give whoever does it the Nobel prize ... why didn't they do it yet??? There is no rule saying these experiments need to be done with pairs of photons.

[Ben6993]

Hi Richard

If S3 is locally R3, then I think I can see what you mean, ie you might need to go a long way in space to get get away from the local R3 to get the torsion effect of S3. But that is an interpretation of Joy's torsion of space as a long scale curvature of space. However, Joy seems to have denied that previously. The torsion seems to require an extra spatial dimension and that dimension is available at every point in space. So there is no need to travel far away to see an effect. I think that Joy previously wrote that the local R3 is a hypothetical tangential space at a point in S3?

I interpret that extra dimension as a compactified dimension, as in string theory's hypothetical compactified dimensions. The extra dimension is available everywhere in space but because it is compactified it only gives rise to two results, say +1 and -1. Just as travelling very close to speed c in the x direction would compactify the x direction, so presumably all of our (x,y,z) space is
travelling close to speed c in the extra dimension and the extra dimension, ie the spin dimension, is compactified at every point in (x,y,z). The electron has access to the spin dimension and that gives rise to its 4π cycle.

Fred has mentioned partial macroscopic experiments. My opinion is that a BEC might show the microscopic effect. If all the bosons in a BEC have the same state then they would all have the same relationship to the extra dimension, eg all in the +1 spin state. A macroscopic body on the other hand would contain particles with both spin states and I doubt that it would show the microscopic effect. I am not certain if an EM pulse is equivalent to a BEC, but I guess it is and so would not be a full test of a macroscopic effect.

[gill1109]

Joy describes his experiment in a short experimental paper http://arxiv.org/abs/0806.3078, and again in Section 4 of the longer paper http://arxiv.org/abs/1211.0784. The whole point of his experiment and of our bet, is that prior to those publications, it would not have occurred to anyone to use anything except classical mechanics in order to predict likely statistics of the experiment.

Yet Joy predicts we'll see the singlet correlations.

I predict we won't.

That is what we are going to bet on.

Doing the experiment with a BEC (two of them?) or with two bunches of photons would be another story. Another topic. Nobody has done a rigorous *and* succesful Bell-CHSH experiment with *any* physical system yet, at all. The experimenters believe that they are closest to doing it with single photons. That is the only physical system for which succesful experiments have closed all the major loopholes, albeit not simultaneously.

It would be a bolt out of the blue to the physics community if the first succesful rigorous experiment used "subjects" for which everyone so far has found classical physics to be perfectly adequate. But that is what Joy is betting on.

[Joy Christian]

Hi Guys,
I am busy working on another paper, so I will be brief. In agreement with Richard, I very much like the experiment done with unambiguously classical system, like Ping-Pong balls---for example. The whole point of my argument is that there is nothing quantum about the quantum correlations, and this can be best proved by an unambiguously classical experiment. Well, actually there is more to my argument than just that, but I have explained all that in my papers and books, so I won't go into that again. The only other thing I would like to stress is that there is nothing special about the hemispheres. In fact, experimentally, they would be problematic, because they would wobble violently, inducing all sorts of inertial effects, which would wipe out the strong correlations. But experimenters are clever people and they may come up with something like a "chemical bond", say made out of two spherically symmetric "Ping-Pong" balls. The aim would be to reduce all undesirable effects as much as possible. They are the only reason why I would loose the bet. If the experiment is done competently, then Richard's 5,000 euros are already mine.

[FrediFizzx]

Well, I say if an experiment with a form of "entangled" EM radiation show violation of Bell, then Joy wins in my book also. Of course if a ping pong ball type of experiment shows violation, then it is for sure "game over" for Bell's theorem. No doubt about it. As far as the bunches of photons thing goes, these people are working on it big time.

"Investigating macroscopic quantum superpositions and the quantum-to-classical transition by optical parametric amplification"
http://arxiv.org/abs/1202.5518

But I think they should concentrate more on other kinds of pre-selection filtering rather than this single photon amplification scheme.

I'm still looking for that other paper that was a half macroscopic photon Bell experiment. But here is another paper on the subject,

"Bell Inequality Tests with Macroscopic Entangled States of Light"
http://arxiv.org/abs/1109.4832

[Mikko]

If the experiment is done competently, then Richard's 5,000 euros are already mine.

That's a big if. You need to carefully define when the experiment is acceptably done. For example, what to do with the runs that fail to determine one of the spins or the determined value is too impresise to be usesful, and what exactly is the required presision.

[gill1109]

If the experiment is done competently, then Richard's 5,000 euros are already mine.

That's a big if. You need to carefully define when the experiment is acceptably done. For example, what to do with the runs that fail to determine one of the spins or the determined value is too impresise to be usesful, and what exactly is the required presision.

We are going to carefully define all these things. Anyway, read Joy's paper. If we follow his paper closely there will not be too many problems. If Joy wants to depart from what he wrote in his papers I may or may not have issues with that.

The k'th pair of hemispheres fly apart, a battery of video cameras film both. The films are analysed using image analysis software. The result is a direction s_k (for the k'th pair of hemispheres) such that one is spinning in direction s_k, the other in -s_k ... according to the software. Joy and his experimenter can determine how they do that. I don't mind.

The direction s_k is stored in a computer file.

Repeat N times, get N directions s_k and -s_k, all in one computer file.

Then we go home with our computer file of these N pairs of directions.

We now pick two directions a and b and *compute* E(a, b) = 1/N sum_k sign(a . s_k) sign( - b . s_k)

Now do this again for some other pairs of measurement directions, same set of N already reconstructed spin directions s_k and -s_k.

I am insisting on the usual four pairs for a CHSH experiment (a, b), (a, b'), (a', b), (a', b') and I propose that the winner is determined by whether or not CHSH exceeds the half way mark between 2 and 2 sqrt 2.

The adjudicating committee will adjudicate if there are issues between us, which our prior written agreement does not cover. e.g. we don't specify what computer platform and what computer language to use for the calculations. Suppose it turns out that Joy wants the computations done in Mathematica on a PC, I want R on a Mac, and suppose that makes a difference who has won. Then the board of adjudicators will decide.

[gill1109]

I thought of a solution to the instability of spinning hemispheres. We take two rubber balls which are quite squeezy so that when you push one against a flat surface it takes on the form of a ... squashed ball, squashed against a flat plate, ie roughly ... a hemisphere. Two are glued together but little explosive charges let them shoot away from each other, regaining their spherical shape... no wobbles!!!!