Is action at a distance possible as envisaged by the EPR Paradox.

[my_wan]

But why?

Certainly not because the inequality is wrong, but because it's physical meaning may be taken with a far higher level of generality than what the empirical fact of it justifies.

 

[unusualname]

Certainly not because the inequality is wrong, but because it's physical meaning may be taken with a far higher level of generality than what the empirical fact of it justifies.

Maybe, but don't you agree that the subtlety of the arguments are more in the philosophically delicate vein than scientifically interesting? The resolution to your logical conundrums can't be scientifically useful since as you probably know, EPR type arguments have been refuted now using trivial 100% correlation results (see links above) rather than statistical ones.

If you find some doubt in the reasoning attached to the implications of results of Bell experiments, it's a bit like discovering there are logical issues in Newton's Principia.

Your ideas do seem deeply thought out and might make an interesting paper from a historical/philosopical perspective.

And after all the thread title does ask a specific question

 

[DrChinese]

a. Neither can you claim that HV models are lacking in the published literature.

b. Here are the questions again for which even a yes/no answer would be a breakthrough:
1) Do coordinate independent LHV models of BI violations exist that lack rotational invariance?
2) Does a rotational invariant probability function exist for any mechanism with a randomized rotation?

a. The only suitably realistic one is the De Raedt et al model. And by suitably I mean able to generate simultaneous values for a, b, c. (Keep in mind that it exploits the Fair Sampling assumption, which is generally not used by other groups.) ALL of the others are not realistic, i.e. they cannot generate a local realistic dataset. Examples of published failures would be Hess, Phillipp, Santos, Christian, Laudisa, Broda, Matzkin, Thompson etc. If they could provide a dataset, my assumption is that they would - which makes them failures. Many have also be attacked on other grounds (Santos, Hess, Phillipp possibly gaining some kind of records in that regard). De Raedt's, on the other hand, provides an event by event dataset, which I have personally verified. I would be glad to look at anything any of these other authors presents in the way of a realistic dataset. In fact, you can consider it an open challenge.

So there are none such remaining in the literature as far as I am aware. That is my claim.


b. As to your "questions": Sure, let's see if I can be clearer (although these questions are really weird to me):
1. No, because there are NO LHVs that violate Bell inequalities. See Bell.
2. I believe a linear one might work. But I am not an expert in this.

P.S. Your comment about my having tried to find loopholes in Bell ("suspect") is off the mark. It is no different that ways I may have tried to find loopholes in the HUP. Or in other aspects of QM or GR. Just because I follow them does not mean I don't try to push the envelope. No telling what I might learn from trying.

 

[DrChinese]

Certainly not because the inequality is wrong, but because it's physical meaning may be taken with a far higher level of generality than what the empirical fact of it justifies.

"No physical theory of local Hidden Variables can ever reproduce all of the predictions of Quantum Mechanics."

You think that is shows a physical meaning which is more general than justified empirically?

I don't see this as something which is empirical. Nor do I see it as more general than warranted. Nor am I aware of any physical theory which threatens this conclusion.

 

[my_wan]

"No physical theory of local Hidden Variables can ever reproduce all of the predictions of Quantum Mechanics."

You think that is shows a physical meaning which is more general than justified empirically?

I don't see this as something which is empirical. Nor do I see it as more general than warranted. Nor am I aware of any physical theory which threatens this conclusion.

Again you have placed a backward interpretation of my words.
1) Yes. BI violations are an empirical fact. That is ALL that empirical means in this case.
2) No, I didn't imply "empirical" meant physical in any way. It's merely a fact that BI violations are an empirical fact.

I don't see this as something which is empirical.

Do you even know what "empirical" means? It means it is an observable fact, independent of any interpretation placed on it.

Nor do I see it as more general than warranted.

Presumably your associating "more general" with the empirical fact itself. An empirical fact is only a repeatable observation. The generality of it applies only to the interpretation, not the empirical fact.

Nor am I aware of any physical theory which threatens this conclusion.

I'm not aware of any physical theory which threatens the color orange either. I didn't know it was possible to butcher a parsing of my sentences so badly. Is my English really that bad?

I give up...

 

[DrChinese]

Do you even know what "empirical" means? It means it is an observable fact...

Yes. For example: It is an empirical fact that the Bell proof is considered a theoretical/mathematical one, and is not subject to empirical confirmation as an observable fact. Which is why I used it in my statement the way I did. Which is in direct contradiction to your statement.

Listen, you obviously are going to hold your (non-standard) opinion regardless. So I see no benefit to this conversation to either of us. You are able to find your own references supporting realism and casting dispersion on Bell, so I can't help you there either. You are capable of reading the literature and making your own decisions on what you will accept or reject. Further, my references and reasoning are of limited value at this point. Since you clearly think your words make sense - and I do not - I don't see any point of intersection.

All I ask is that you label your opinions going forward as non-standard wherever they are. Otherwise, you will suffer the same fate as ThomasT: I will pick apart your statements because it is wrong for you to use PhysicsForums as a soapbox for personal pet theories. Please see Forum guidelines if you have any questions.

 

[my_wan]

Yes. For example: It is an empirical fact that the Bell proof is considered a theoretical/mathematical one, and is not subject to empirical confirmation as an observable fact. Which is why I used it in my statement the way I did. Which is in direct contradiction to your statement.

So here you claim an empirical fact is not subject to empirical confirmation! Or even an observation. Freaking ridiculous beyond belief!

http://www.merriam-webster.com/dictionary/empirical]1[/PLAIN] : originating in or based on observation or experience <empirical data>
2 : relying on experience or observation alone often without due regard for system and theory <an empirical basis for the theory>
3 : capable of being verified or disproved by observation or experiment <empirical laws>

Bell's theorem is built on an an ansatz, and the proof that follows is predicated on that ansatz. There is NOTHING empirical about any logic "considered" true by people, no matter how immaculate the logic. Even a well proved theorem by definition can't be an empirical fact. The empirical is what is observed, irrespective of any disagreements about what that observation actually is, or any axioms used to define it, or how many people agree with it. The empirical fact is X devices counted Y coincidences in excess of inequality Z, period. Your opinion, predudice, etc., no matter how shared with others does not make anything you "consider" an empirical fact!

 

[DrChinese]

Even a well proved theorem by definition can't be an empirical fact.

Try reading before writing; that is precisely what I said. After you re-read, I will gladly accept your apology.

Before you say stuff like "freaking ridiculous beyond belief" you might want to carefully consider whether the context deserves that. I have considerately tried to follow your reasoning for weeks. And believe me, your muddled thinking has not made that easy. (Muddled, by the way, is me being kind.) I realize that in YOUR mind, I have not answered your questions. And the fault lies with me to the extent I have not. Whenever there is a lack of communication, there are at least 2 frustrated people. But you are far off base with your recent comments, including any hint that I would evade an answer. I have been around here for a while, and I doubt there are many who have seen me shy away from anything. Probably quite the opposite, a few who wish I might step back more frequently.

Ah, perhaps we can both enjoy a glass of wine later and toast Bell in some fashion.

 

[DevilsAvocado]

DrChinese & my_wan

I’ve been following your last debate, and to me it all seems like a misunderstanding, maybe on both sides.

I think that you both have very interesting points and great knowledge, and you both clearly makes this thread much more interesting. DrC advocates the standard position, and my_wan tries to find new solutions to explain what happens in Bell test experiments.

You are both confirming the mathematical validity of Bell’s theorem.

I don’t see anything wrong in either position, and 'tension' can create good (or bad) things...

In an attempt to hopefully make DrC understand what my_wan mean when talking about "Rotational Invariance in EPR-Bell Experiments", maybe this video could be a clue...?

"[URL - Fair Sampling and Rotational Invariance in EPR-Bell Experiments

[/URL]


Guillaume Adenier, from Vaxjo University Sweden, also has a paper related to the video: http://arxiv.org/abs/quant-ph/0606122"

(Personally I don’t think that "Fair Sampling" is an issue, but that’s another question...)

 

[DevilsAvocado]

... In my mind, rotational invariable is getting the same answer in any rotated reference frame. There can be no preferred frame.

DrC, I have some questions and if you have the time to answer, it would be great (as a 'reward' maybe I’ll deliver a BIG surprise in coming days ). Many questions are more of 'verification' yes/no:

Q1) The photons coming out of the BBO crystal are entangled + entangled in polarization, so called superposition of (polarization) state (in the intersecting cone), right?

Q2) Does this entangled superposition mean that the two photons are QM coherent (pure state), and described by one single wavefunction?

Q3) Does the wavefunction collapse (or decohere) when the photon is measured in the polarizer?

Q4) When measuring the entangled photons with polarizers aligned, let’s say 0º on both, what coordinate system is used, to get the exact same angle, if the polarizers are separated by 10 km or more?

Q5) Is it practical to talk about exact polarizer angles for incoming photons, in respect of HUP?

Q6) In this thread we usually talk about up/down spin of entangled photons, but the correct term is right-handed and left-handed (clockwise/counter-clockwise), that correspond to the two possible circular polarization states of the photon (along its direction of motion), right?

Q7) The circular polarization in Q6 explains why the two entangled photons can be measured aligned on any angle 0 – 360º, right?

Q8) If the photon polarization is a result of spin, and thus direction of motion, would a change in the direction of motion also result in a change of the polarization?


Thanks in advance!

 

[my_wan]

I don't accept the fair sampling argument either, but that doesn't mean an argument here that was characterized as a fair sampling argument here actually was. Some inefficiencies may lead to some slightly distorted data, but experimentally today it's well withing QM constraints. I see no reason to question perfect QM predicted correlations in the ideal case.

I did notice, in my own computer models, that the standard deviations increased greatly at offset maxima. I used a sample group of 10k photons. The standard deviation increase occurred due to low sample size as the expected hits/misses approached 0 or 100%. I was modeling the ideal QM case with exactly known particle numbers and detections, with perfect correlations and ideal detectors. This is likely to vary at least some in less than ideal conditions, but I don't see that refuting, or even seriously questioning, the QM predictions. I'll stick with modeling the ideal QM predictions.

Some interesting points was made in the video wrt rotational invariance (@12:30). I'm not sure how to interpret the anomalies. It looks too consistent in the pdf graph, even across experiments, to simply be variations in standard deviation. Yet the correlation coefficient remain consistent with QM. Weird.

 

[my_wan]

DrC,
Perhaps my language was too unforgiving. It wasn't your denials, your opinions, your rejections, etc., that put me over the edge. That I welcomed, and even hoped for. I even hoped I could be backed in a corner with no other escape than a handshake. Regardless of how it might be interpreted, I have not demonstrated the side of the debate I took is factual. But neither does your interpretation come with overwhelming certainty.

The thing that triggered me was a complete lacking of answers to questions. When I expending great effort for an argument, and got a single sentence denial, which didn't even specify what was being denied, it was aggravating. Surely my articulation shared heavy burden in fault, but lacking any clues as to what needed improved, how could I even think of improving it?

I took great care early on, and several times since, to insure not only you but other realist here that I was a thin ice with my position in the debate. yet bias was assumed in every inane statement I made. Naturally, taking a realist position in the debate, I was by definition going to bias toward realism. But only to the degree needed for a proper debate.

So my apologies, for such strong innuendo. Not even the innuendo about what my bias indicated fully justified that. However, at the core, when you look past the emotional implications I imposed, there remains what I feel is a valid issue to express. I was simply at a total loss without being given a clue, even when I asked, how to clean up my articulation of the issues. So I must take at least as much responsibility, for lacking the tact to get the point across any other way.

 

[ThomasT]

Don't want to appear insensitive, but shouldn't someone point out that arguing about the validity of Bell's Theorem experiments seems pretty redundant now ...

Your comments don't appear to be insensitive. This is an argument based on arguable positions, not a quarrel based on emotion. I've benefited (as I assume others have) from your comments. So please stay tuned in and keep commenting.

... since multi-particle entanglement experiments have trivially demonstrated that local realism fails (and in particular that the EPR argument for it fails).

Multi-particle entanglement experiments have demonstrated that Bell inequalities are violated. There's absolutely no dispute regarding that. The discussion here is about what can be inferred from those violations. The contention of some is that Bell's LR ansatz doesn't actually model the experimental situation that it purports to model, so that inferences regarding reality from those violations are obviated. I, and others, think they've got a good argument.

Certainly not because the inequality is wrong, but because it's physical meaning may be taken with a far higher level of generality than what the empirical fact of it justifies.

Maybe, but don't you agree that the subtlety of the arguments are more in the philosophically delicate vein than scientifically interesting?

I think what you're saying is true. But the foundations of physical science are philosophical issues, not scientific issues. So, if Bell's argument against LR models of entanglement is logically flawed, then this is important.

 

[unusualname]

Y
Multi-particle entanglement experiments have demonstrated that Bell inequalities are violated. There's absolutely no dispute regarding that. The discussion here is about what can be inferred from those violations. The contention of some is that Bell's LR ansatz doesn't actually model the experimental situation that it purports to model, so that inferences regarding reality from those violations are obviated. I, and others, think they've got a good argument.

But multi-particle entanglement experiments rule out local realism without the use of inequalities, which is why I suggested that the debate on Bell experiments is redundant (scientifically speaking)

Multi-Photon Entanglement and Quantum Non-Locality

(direct download link here)Not saying it's not a worthwhile debate, but now that there are simpler experimental refutations of local realism than Aspect et al I don't think you can appeal to flaws in Bell to lend any support to counter arguments against this.

(Unless of course we subsequently find logical/philosophical issues with interpreting GHZ relations etc )

 

[zonde]

But multi-particle entanglement experiments rule out local realism without the use of inequalities, which is why I suggested that the debate on Bell experiments is redundant (scientifically speaking)

GHZ inequality experiments rule out only non-contextual local realism just like Bell experiments only more directly.

But GHZ experiments just the same use fair sampling assumption so they can't rule out contextual local realism.

But I would agree that debate about non-contextual local realism is redundant.

 

[my_wan]

GHZ inequality experiments rule out only non-contextual local realism just like Bell experiments only more directly.

But GHZ experiments just the same use fair sampling assumption so they can't rule out contextual local realism.

But I would agree that debate about non-contextual local realism is redundant.

I certainly do accept that BI violations experimentally rule out all non-contextual local realistic models. However, perhaps my definition of "fair sampling" is overly restrictive. Perhaps others can judge that. I take "fair sampling" to mean only that a small sampling of detections are representative. Yet many of the EPR mechanisms posited made no claims dependent on sample size and/or efficiency with which detections occurred, yet were labeled a fair sampling argument.

For instance, when a complete sampling is assumed, you can still index the properties to events in physically inappropriate ways, such that physically absurd situations are implied. How, when no problems with the validity of the sampling is assumed, only in the way they are indexed, can it be called a "fair sampling" argument?

Let's get inequality violations without correlation in a single PBS:
Let's assume a perfect detection efficiency in a single channel, 100% of all particles sent in this channel get detected either go left or right using a PBS. Consider a set of detections at this PBS at angle 0. 50% go left and 50% right. Now if you ask which left would have went right, and visa versa, these photons would have went at an angle setting of 22.5, it's reasonable to say ~15% that would have went left go right, and visa versa. Yet this same assumption indicates that at an angle of 45, ~50% that would have gone left go right (relative to the 0 angle), and visa versa. Yet, relative to angle 22.5, the 45 angle can only have switched ~15% of the photon detection rates. 15% + 15% = 30%, not 50%.

If you look at the above situation closely, with a single PBS and channel lacking any correlated pairs, any possible counterfactual assumptions about what the photons would have done if the setting had been different leads to exactly the same inequality violation as Bell posited. Thus the coincidences only show a repeatability of a local correlation free phenomena.

Fair sampling cannot be responsible when the same inequality issues can be reproduced under any possible, empirically realistic, counterfactual assumptions about what photons would have done in a single channel, correlation free, response to a single PBS.

It can be argued that this proves a lack of realism without even referring to EPR correlations, but any role of an FTL mechanism becomes moot, except that it's perfectly repeatable with a perfectly correlated particle. Yet this repeatability apparently indicates there's a deterministic mechanism for the local inequality present in the single uncorrelated particle beam case.

Thus any realistic model must model this local inequality deterministically, but attempts to do so forces you to choose a reference coordinate system, in which you can arbitrarily relabel the coordinate labels such that one or the other detector, when arbitrarily rotated, allows you to relabel that coordinate point as 0, to get the EPR case to work mathematically. Though I don't consider such arbitrary coordinate relabeling as having a real physical significance, since coordinate systems aren't physical constructs, I'm attempting to replace the deterministic 0/1 bits in my model with vectors, in the hopes of fixing the coordinate transform requirement when arbitrary detector settings are chosen.

Here's a simple outline of vector properties under rotation, perfectly matching what I presented, I'm attempting to take advantage of.
http://www.vias.org/physics/bk1_09_05.html

 

[my_wan]

A second issue I had involves the meaning of what a contextual variable is. It's been posited that contextual variables entails a lack of realism. I have a hard time seeing how this can be justified. Though attempts at inquiring came to naught. It seems to me to be conflating the notion of a non-contextual variable lacking an existential container with contextual variables. A contextual variable is not simply a variable lacking an existential container, it's a variable that is contingent upon the overall structure of other undefined variables.

A simple example would be the property rabbit, rock, carrot, etc. Our entire physical experience is built on contextual variables defined by the periodic table of elements, plus radiation for our observation of it. If we take a pool ball collision to be a real event, and the pockets on the pool table to be detectors, then the variables our pockets detect is highly dependent on how you rotate the pool table under the pool ball collision. Thus pool table pockets only detect contextual variables. Is the polarization settings of a polarizer/detector like the pool table rotation? Almost certainly, whether the variables are realistic or not.

So I reject the notion that contextual variables entails a lack of realism.

 

[DrChinese]

But GHZ experiments just the same use fair sampling assumption so they can't rule out contextual local realism.

But I would agree that debate about non-contextual local realism is redundant.

GHZ tests are not considered to rely on the Fair Sampling assumption. Now there is a kicker on this that may confuse folks. It is true that only a sample is used, so you might think the Fair Sampling issue is present. But it is not. The sample looks like this:

-1
-1
-1
-1
-1
-1
-1
-1
-1
-1
-1...

Where local realism predicts

1
1
1
1
1
1
1
1
1
1
1...

A little consideration will tell you that local realism is falsified in this case. Every case, individually, is a falsification.

 

[DrChinese]

Let's get inequality violations without correlation in a single PBS:
Let's assume a perfect detection efficiency in a single channel, 100% of all particles sent in this channel get detected either go left or right using a PBS. Consider a set of detections at this PBS at angle 0. 50% go left and 50% right. Now if you ask which left would have went right, and visa versa, these photons would have went at an angle setting of 22.5, it's reasonable to say ~15% that would have went left go right, and visa versa. Yet this same assumption indicates that at an angle of 45, ~50% that would have gone left go right (relative to the 0 angle), and visa versa. Yet, relative to angle 22.5, the 45 angle can only have switched ~15% of the photon detection rates. 15% + 15% = 30%, not 50%.

If you look at the above situation closely, with a single PBS and channel lacking any correlated pairs, any possible counterfactual assumptions about what the photons would have done if the setting had been different leads to exactly the same inequality violation as Bell posited. Thus the coincidences only show a repeatability of a local correlation free phenomena.

Fair sampling cannot be responsible when the same inequality issues can be reproduced under any possible, empirically realistic, counterfactual assumptions about what photons would have done in a single channel, correlation free, response to a single PBS.

I would agree that Fair Sampling is not a factor here.

And I actually agree that your example can be instructive about counterfactual reasoning. In fact, Eberly performed a neat analysis of this a few years back extending your idea. It did lead to the violation of a Bell type equality precisely because QM does not acknowledge the counterfactual case. But that would not lend support to your basic premise, because QM rules work when classical rules do not.

http://www.optics.rochester.edu/~stroud/cqi/rochester/UR19.pdf

I just recently completed a detail analysis of his derivation for a paper I wrote. I can reproduce this in a shortened manner if it is of interest. He concludes that it is meaningless to contemplate the results of experiments in which "data is not taken" (i.e. it is erased). This includes counterfactual cases.

I agree with you that there is a mystery in the 22.5 + 22.5 degree case you describe above (even though I may have given you a different impression in earlier posts). However, I am not clear how (or if) you can use that to demonstrate that local realism cannot hold.

 

[my_wan]

I would agree that Fair Sampling is not a factor here.

And I actually agree that your example can be instructive about counterfactual reasoning. In fact, Eberly performed a neat analysis of this a few years back extending your idea. It did lead to the violation of a Bell type equality precisely because QM does not acknowledge the counterfactual case. But that would not lend support to your basic premise, because QM rules work when classical rules do not.

http://www.optics.rochester.edu/~stroud/cqi/rochester/UR19.pdf

I just recently completed a detail analysis of his derivation for a paper I wrote. I can reproduce this in a shortened manner if it is of interest. He concludes that it is meaningless to contemplate the results of experiments in which "data is not taken" (i.e. it is erased). This includes counterfactual cases.

I agree with you that there is a mystery in the 22.5 + 22.5 degree case you describe above (even though I may have given you a different impression in earlier posts). However, I am not clear how (or if) you can use that to demonstrate that local realism cannot hold.

I'd like to make a distinction here that may be leading to a little confusion, involving the two separate notions in "local realism" What this thought experiment was meant to exclude was not local realism or lack of realism, but the legitimacy of a non-local signaling mechanism to define the inequality violation. To validate the no-signaling assumption. I can't see how such a signaling mechanism is legitimate, when the same inequality violation exist without any reference events in existence to communicate with.

This argument was meant only to remove the non-local issue while leaving the realism issue open. Would any object to this removing the non-local issue, without actually resolving the realism issue? Thus, if acceptable, for any realist to maintain realism, a mechanism needs defined for how HUP pulls off this magic both locally and deterministically. Which requires other arguments this thought experiment doesn't address.

To extend beyond the limited argument above involves rotational invariance, due to the arbitrary detector setting requirement imposed by Bell proponents, this effect must be appreciated if such a mechanism is at all possible:
http://www.vias.org/physics/bk1_09_05.html

 

[my_wan]

It should be noted that, in my modeling, I merely presume Malus' Law a priori. How Malus' Law pulls off this magic, even in an uncorrelated beam and single PBS, technically requires it's own explanation to claim anything more than a 'real' variable working toy model. Thus any success still does not say what is 'really' going on in such an experiment, it merely provides a toy model mimic that shows it's possible only in principle.

That's why I started thinking in vectors, because I wanted the toy model to resemble QM as close as possible while still using real variables.

Edit: The Born rule is a beast

 

[DrChinese]

So my apologies, for such strong innuendo. Not even the innuendo about what my bias indicated fully justified that.

Why apology accepted!

I promise I will do a better job of listening for your questions. I am not trying to avoid anything, and certainly will not shy away because the going gets tough. I may be wrong, and if I am then that is my opportunity to learn something new. Which is why I am here.

 

[DrChinese]

DrC, I have some questions and if you have the time to answer, it would be great (as a 'reward' maybe I’ll deliver a BIG surprise in coming days ). Many questions are more of 'verification' yes/no:

Q1) The photons coming out of the BBO crystal are entangled + entangled in polarization, so called superposition of (polarization) state (in the intersecting cone), right?

Q2) Does this entangled superposition mean that the two photons are QM coherent (pure state), and described by one single wavefunction?

Q3) Does the wavefunction collapse (or decohere) when the photon is measured in the polarizer?

Q4) When measuring the entangled photons with polarizers aligned, let’s say 0º on both, what coordinate system is used, to get the exact same angle, if the polarizers are separated by 10 km or more?

Q5) Is it practical to talk about exact polarizer angles for incoming photons, in respect of HUP?

Q6) In this thread we usually talk about up/down spin of entangled photons, but the correct term is right-handed and left-handed (clockwise/counter-clockwise), that correspond to the two possible circular polarization states of the photon (along its direction of motion), right?

Q7) The circular polarization in Q6 explains why the two entangled photons can be measured aligned on any angle 0 – 360º, right?

Q8) If the photon polarization is a result of spin, and thus direction of motion, would a change in the direction of motion also result in a change of the polarization?


Thanks in advance!

Don't know if I can answer all of these, but let me take a stab at a few off the top of my pointy head:

1) Yes, the intersection is where you find the entangled ones.

3) This is not clear. You can subsequently "erase" the measurement so I would say no.

4) Actually, I would say that a somewhat relative system is used. Consider that the environment does slightly affect the polarization. To be honest, I don't know all of the adjustments they make in a lab so maybe someone else can add here. The usual texts are not completely clear and leave out a lot of details.

6) There is circular AND linear polarization. One can be changed to the other by use of wave plates.

8) No, you can change directions without changing the polarization. This is frequently done with fiber.

 

[DevilsAvocado]

... let me take a stab at a few ...

Thanks for the answers DrC!

I know these questions are kinda weird and not easy to answer. However Q2 + Q3 is maybe the most crucial if I am about to deliver "something interesting"...

If I put it this way: If we compare with the Double-slit experiment and the superposition of one photon (or electron), and that this superposition "is lost" if we try to measure which slit it passes, and thus the Double-slit interference pattern is also lost.

Would you say that the photon superposition of polarization in EPR-Bell "is lost" in the same way, when we measure a 'fixed' photon polarization in the polarizer?

(not counting delayed quantum erasers etc)


EDIT: In Q4, could we assume that the experimentalist does a "later calibration/fine-tune" on angles, and when they have 100% correlation in data, they can be sure that this was a perfect parallel alignment...?

EDIT2: "6) There is circular AND linear polarization" - What is the (normal) case in EPR-Bell entangled photons?


P.S. I have found some real interesting "new" material from John Bell himself. I can’t see it on your site, so it’s probably news even to you. This is going to be a jaw-dropper to many of us in this thread...

 

[DevilsAvocado]

... Some interesting points was made in the video wrt rotational invariance (@12:30). I'm not sure how to interpret the anomalies. It looks too consistent in the pdf graph, even across experiments, to simply be variations in standard deviation. Yet the correlation coefficient remain consistent with QM. Weird.

Yes, I don’t know if this is of any value, but would you say that this is somehow 'related' to 'your' rotational invariance?

 

[my_wan]

I found this, which I completely overlooked till I seen it, even though it should have been obvious, which provides more experimental justification for assuming light has a distinct polarization even when its particle properties are being measured.
http://farside.ph.utexas.edu/teaching/qm/lectures/node5.html

It[/PLAIN] is known experimentally that when plane polarized light is used to eject photo-electrons there is a preferred direction of emission of the electrons. Clearly, the polarization properties of light, which are more usually associated with its wave-like behaviour, also extend to its particle-like behaviour. In particular, a polarization can be ascribed to each individual photon in a beam of light.

It goes on with a really easy read of some of the arguments I made, but placed within the framework of superposition of states. I was thinking I needed something extra beyond state superpositions to model EPR, but maybe not. Maybe all I need is to define a degree of superposition, defined by a pair of vectors, which transforms with different polarizer settings. Then simply predefine a detection limit defined by the product of those vectors at that detector setting. Perhaps I can even dispense with predefined detection limits and simply let the vector product range from -1 to 1, and select by which side of 0 the vector product produces. As noted, vector products are not rotational invariant for the reason pointed out here:
http://www.vias.org/physics/bk1_09_05.html
Even though the underlying real state is.

This would actually build the Born rule into the model. This requires defining the wavefunction as physically real, while all observables are not themselves representative of a real state, but a product projection of a real state.

Ah, well.. I'm just running my mouth again, going beyond what I can demonstrate atm. The electron ejection data remains interesting and relevant to potential realism models.

 

[my_wan]

Yes, I don’t know if this is of any value, but would you say that this is somehow 'related' to 'your' rotational invariance?

It's hard to say. It's not a direct analog, and I'm not sure how to interpret the anomalies. I can't say it's not related either. The most significant effect wrt rotational invariance I can point at the illustrate what I described is here:
http://www.vias.org/physics/bk1_09_05.html

The[/PLAIN] operation's result depends on what coordinate system we use, and since the two versions of R have different lengths (one being zero and the other nonzero), they don't just represent the same answer expressed in two different coordinate systems. Such an operation will never be useful in physics, because experiments show physics works the same regardless of which way we orient the laboratory building! The useful vector operations, such as addition and scalar multiplication, are rotationally invariant, i.e., come out the same regardless of the orientation of the coordinate system.

And this was done using only two very basic unit vectors. It says it "will never be useful in physics", but the Born rule imposes this very condition to define observables from the QM framework!

 

[DevilsAvocado]

It's hard to say. It's not a direct analog, and I'm not sure how to interpret the anomalies. ...

Okay, I think that in the video the focus was on "rotational invariance in the source", and you are focusing on the angles of polarizers, right?

The http://en.wikipedia.org/wiki/Born_rule" seems tuff...

There have been many attempts to derive the Born rule from the other assumptions of quantum mechanics, with inconclusive results.

 

[DevilsAvocado]

... This requires defining the wavefunction as physically real ...

Where can I buy a member card??

Seriously, what’s your opinion on a possible "wavefunction collapse" when the entangled photon (superposition polarized) is measured by the polarizer?

 

[my_wan]

Okay, I think that in the video the focus was on "rotational invariance in the source", and you are focusing on the angles of polarizers, right?

More specifically the observable values, but since those would depend on a vector product with its (complex) conjugate, which in turn is dependent of polarizer angles, it does come down to the angles of polarizers.

The http://en.wikipedia.org/wiki/Born_rule" seems tuff...

Yes, more than a little weird from the perspective of classical physics. Yet taking the wavefunction serious as a real physical state provides an EPR mechanism. Unfortunately the published thermodynamics models don't really deal with it explicitly. They tend to simply impose such weirdness on top a priori. So even if we simply assume the wavefunction is real, and claim some modeling success, it still doesn't fully answer the realism question.

Edit: Oops forgot to answer the question. I don't really take wavefunction collapses seriously, for reasons well outside this debate. I think it more likely the apparent collapse is the removal of the possible states evolved from a prior state, to leave on the actual state. It still requires the rules what a wavefunction can really do be analogous to what it could have done, but didn't, in some very fundamental ways.