Quantum Perspectivalism by D.Dieks

[lwjjack]

TL;DR Summary

Should we take arguments like this seriously?

Dennis Dieks is a respected Phycisist who works in Utrecht University.He has been developing his unique understanding of quantum mechanics namely Quantum Perspectivalism,close to Rovelli's RQM.But his works seem highly unpopular .
https://link.springer.com/article/10.1007/s10701-019-00264-0
This paper of his argues that his Perspectivalism must be adopted by unitary approaches in order to explain WF experiments.
https://link.springer.com/article/10.1007/s10701-022-00611-8
This is the most recent conclusion and arguments of his approach.

 

[Fra]

I need to spend more time to read to extract his unique core points... but the topic is interesting and some quick reflections

"Perspectives can also be defined with respect to physical systems instead of agents."
-- https://link.springer.com/article/10.1007/s10701-022-00611-8

For me I only entertain the concept of "agents" where the agent IS a physical system. So "physical system" and "agent" are not in any way in contrast to me at least.

If he means that "agents" that are "humans" or otherwise fictive or imaginary information processsing context that escape the constraints of real process are not useful, then I agree with this viewpoint.

"In fact, all single-world probabilistic interpretations of unitary quantum mechanics that pick out definite-valued quantities by an appeal to decoherence or a similar process need this kind of perspectivalism to make sense of situations of the Wigner’s friend type. These interpretations have the same resources as QBism, or Quantum Pragmatism, for handling conceptual puzzles like Wigner’s friend and nonlocality; but they handle such puzzles in a perspectival realist way, without recourse to subjectivism."
-- https://link.springer.com/article/10.1007/s10701-022-00611-8

The problem is however not solved by saying that the agent is just a physical system (it is), but the problem is that this physical system can only de described from the perspective of another agent. If if there may exists a kind of "realist" description of the set of all agents, that explains their different perspectives in a way consistent with quantum mechanics in it's corroborated domain of small subsystem, the problem is to find this description, it is easy to end up in a finetuning situation that is out of hand for any agent; in which case we are still in "sin" as the model we have created, can't be represented by a real agent. And approximations of this description for any given agent will be an effective theory only. So what have be we gained in understanding?? If this is so, then it's this more or less the standard view of decoherence (rephrased perhaps but still) etc???

Perhaps I missed some details, but this is my quick impression of the authors view from skimming.

/Fredrik

 

[vanhees71]

TL;DR Summary: Should we take arguments like this seriously?

Dennis Dieks is a respected Phycisist who works in Utrecht University.He has been developing his unique understanding of quantum mechanics namely Quantum Perspectivalism,close to Rovelli's RQM.But his works seem highly unpopular .
https://link.springer.com/article/10.1007/s10701-019-00264-0
This paper of his argues that his Perspectivalism must be adopted by unitary approaches in order to explain WF experiments.
https://link.springer.com/article/10.1007/s10701-022-00611-8
This is the most recent conclusion and arguments of his approach.

Well, one of the most superficial measure for a theoretical-physics paper is to glance over the pages to get an impression of the formula-to-text ratio. Although it's very superficial already this simple method gives an impression, what to expect from such a paper, particularly if it's about "interpretations of quantum mechanics" ;-)). SCNR.

 

[lwjjack]

If I understand it correctly,in the 2019 paper the author gave a thought experiment that supposedly rules out the role of non locality in explaining Bell inequalities,and claims that his view is the only way to explain it.(for unitary interpretations)

 

[lwjjack]

Well, one of the most superficial measure for a theoretical-physics paper is to glance over the pages to get an impression of the formula-to-text ratio. Although it's very superficial already this simple method gives an impression, what to expect from such a paper, particularly if it's about "interpretations of quantum mechanics" ;-)). SCNR.

So you think these arguments shouldn't be taken seriously?

 

[DrChinese]

If I understand it correctly,in the 2019 paper the author gave a thought experiment that supposedly rules out the role of non locality in explaining Bell inequalities,and claims that his view is the only way to explain it.(for unitary interpretations)

I am extremely disappointed in this "new" interpretation. (As if we need yet another name to add to the growing list of personal interpretations.) A few comments on the paper:1. It uses as examples EPR-type pairs (where there are perfect correlations) rather than the more stringent Bell state pairs (where inequalities are violated). The difference is that the EPR pairs operate exactly as presented in the 1935 EPR paper instead of the post-Bell pairs of 1964 which demonstrate that quantum mechanics must be contextual (i.e. the quantum expectation values depend only on Alice and Bob's settings, and are non-separable). No one disputes the examples from the EPR paper, which incorrectly lead one to conclude that quantum observables have well-defined values prior to observation. They don't.2. I hate it when writers throw in elements of Alice or Bob's reference frames into a discussion of quantum predictions. Why do I object? Because such elements do nothing except obscure the picture of what is going on by introducing red herrings when we are discussing the usual spin/polarization entanglement.

There are no relativistic considerations! Specifically, the only input variables required for the quantum expectation value are Alice and Bob's settings. Dieks says:

"...because Alice and Bob are space-like separated, there exists another inertial frame of reference with a simultaneity relation such that in this frame Alice measures her electron first, then Bob measures his electron, and finally Alice’s measurement is undone only after both measurements have taken place."

The issue being that the quantum mechanical prediction is the same regardless of whether Alice measures first, Bob measures first, or if they were magically performed simultaneously. There is no known causal direction in the action (or updating of knowledge if that is your thing) except by assumption. So why mention inertial reference frames?

The only paradox here is that IF one assumes there is a single global causal direction. So that assumption must be wrong. It is certainly nowhere to be found in standard QM.3. It is flat wrong! "Let Alice measure the spin of her particle and let her record (or memorize) the outcome—we may think of Alice as fulfilling the role of the measuring device in the discussion at the end of Sect. 2. After Alice’s measurement a unitary restoration process takes place that locally undoes the measurement and returns Alice and Bob plus the two particles to their original total state (so that Alice loses all her information about the outcome). "

Guess what? There is no "unitary restoration process" by which Alice's pre-measurement state can be recreated. To the extent Alice makes her (irreversible) measurement, her particle is separated from Bob's - and there is no going back. I'd like to see the experiment Dieks imagines actually performed... because it can't (and he seems to even acknowledge that). And that is not interpretation dependent, it is present one way or another in all interpretations.

I realize that Dieks was specifically attempting to demonstrate that unitarity* must be wrong as an element of an interpretation in his example above. But you can't cite examples that are known to be wrong to support your argument. No one believes Alice can "forget" her measurement in the first place. We already know that even the possibility of extracting quantum information from an environment will lead to what is commonly labeled "collapse", so why bother with a straw man?

He concludes: "This further supports the necessity of a privileged frame in Bohm’s theory." Obviously it proves no such thing. There is no dependence on reference frame. Note that Alice's observed particle need not even exist at the same time (in any and all reference frames) as Bob's to be entangled with Bob's observed particle. Further, the Alice/Bob entanglement can be created before or after Alice and/or Bob's measurements. That is an actual experimental outcome, not a hypothetical one, and it demonstrates clearly that frame dependence is NOT a variable in entanglement experiments.* "A point that will play an important role later on is that the assumption of universal validity of unitary Schrödinger evolution entails that all quantum processes can in principle be reversed."------------------------If you are going to give us a new interpretation, at this point it really needs to directly address the state of the art experiments (and theory) being done by entanglement teams such as Zeilinger, Hanson, Pan, etc. These teams and many others are probing the extremes of quantum nonlocality. Such experiments should be readily explainable by any new interpretation. As it is, I believe many of the popular interpretations struggle to provide a credible explanation of things such as quantum teleportation without jumping through ever more implausible hoops.

 

[kurt101]

Note that Alice's observed particle need not even exist at the same time (in any and all reference frames) as Bob's to be entangled with Bob's observed particle. Further, the Alice/Bob entanglement can be created before or after Alice and/or Bob's measurements. That is an actual experimental outcome, not a hypothetical one, and it demonstrates clearly that frame dependence is NOT a variable in entanglement experiments.

I don't understand why you keep using the retrocausal argument, when you say
Here that you are agnostic on it. Here is the quote from you:
"Now please keep in mind that I am agnostic on the issue of retrocausallity itself. All I assert is that quantum nonlocality violates strict Einsteinian locality, by creating a context that spans spacetime. I think a better term than "retrocausal" is "acausal"."

Specifically when you say "the Alice/Bob entanglement can be created" "after Alice and" " Bob's measurement". When you admit that you are agnostic on the issue of retrocaulity, aren't you admitting that it is possible that this part of the experiment has a causal explanation? Like maybe this part of the experiment can be explained by some physical causal process related to the post selection part of this experiment?

FWIW, in the other thread that got shutdown (unfortunately as I did not get any chance for clarification) that was my only point in bringing up the paper is that the authors were explaining why causality is not ruled out for the case where entanglement is established after the measurement of Alice and Bob in the entanglement swapping experiment. I don't think the authors were refuting any QM predictions so I still don't understand what Peter Donis was saying when he said "This isn't a refutation of anything. It's a proposed claim about a new "loophole" in such experiments. Such claims are ubiquitous in the literature, but they always boil down to one thing: claiming that if an experiment is done that closes the claimed loophole, the predictions of quantum mechanics will be violated. In other words, such claims are made by people who don't want to accept that the QM predictions are correct." What QM prediction are these authors disputing?
Here is the paper again: https://link.springer.com/article/10.1007/s10701-021-00511-3#Sec21
if you want to point out to me what QM prediction the authors are refuting.

 

[PeterDonis]

I still don't understand what Peter Donis was saying when he said "This isn't a refutation of anything. It's a proposed claim about a new "loophole" in such experiments. Such claims are ubiquitous in the literature, but they always boil down to one thing: claiming that if an experiment is done that closes the claimed loophole, the predictions of quantum mechanics will be violated. In other words, such claims are made by people who don't want to accept that the QM predictions are correct." What QM prediction are these authors disputing?

They're saying that the experiments in question are vulnerable to what they call the "Collider Loophole". As the term "loophole" is usually used in this context, it amounts to saying that if the claimed "loophole" is closed, the violations of the Bell inequalities will go away--which would imply that the predictions of QM would be violated.

The authors of this paper claim that they are not disputing violation of the Bell inequalities, but only "the inference from violation of the Bell inequalities to AAD [action at a distance]". But later on, they say "The alternative is that the Bell correlations are an artifact of collider bias." But they don't actually present this as an "alternative" to their argument, but as part of their argument. Which to me means that, whether the authors want to admit it or not, they are claiming that if the "collider loophole" is closed, the Bell inequality violations will have to go away (because the authors claim those violations are "an artifact of collider bias", which closing the "collider loophole" would remove). Which, again, would imply that the predictions of QM would be violated.

As I said in the previous thread that you quoted, "loophole" claims like this have been common in the literature ever since experimental tests of Bell inequality violations were begun in the 1970s. And every single "loophole" claim has turned out to be wrong. My money is on this one being wrong too.

 

[lwjjack]

I am extremely disappointed in this "new" interpretation. (As if we need yet another name to add to the growing list of personal interpretations.) A few comments on the paper:1. It uses as examples EPR-type pairs (where there are perfect correlations) rather than the more stringent Bell state pairs (where inequalities are violated). The difference is that the EPR pairs operate exactly as presented in the 1935 EPR paper instead of the post-Bell pairs of 1964 which demonstrate that quantum mechanics must be contextual (i.e. the quantum expectation values depend only on Alice and Bob's settings, and are non-separable). No one disputes the examples from the EPR paper, which incorrectly lead one to conclude that quantum observables have well-defined values prior to observation. They don't.2. I hate it when writers throw in elements of Alice or Bob's reference frames into a discussion of quantum predictions. Why do I object? Because such elements do nothing except obscure the picture of what is going on by introducing red herrings when we are discussing the usual spin/polarization entanglement.

There are no relativistic considerations! Specifically, the only input variables required for the quantum expectation value are Alice and Bob's settings. Dieks says:

"...because Alice and Bob are space-like separated, there exists another inertial frame of reference with a simultaneity relation such that in this frame Alice measures her electron first, then Bob measures his electron, and finally Alice’s measurement is undone only after both measurements have taken place."

The issue being that the quantum mechanical prediction is the same regardless of whether Alice measures first, Bob measures first, or if they were magically performed simultaneously. There is no known causal direction in the action (or updating of knowledge if that is your thing) except by assumption. So why mention inertial reference frames?

The only paradox here is that IF one assumes there is a single global causal direction. So that assumption must be wrong. It is certainly nowhere to be found in standard QM.3. It is flat wrong! "Let Alice measure the spin of her particle and let her record (or memorize) the outcome—we may think of Alice as fulfilling the role of the measuring device in the discussion at the end of Sect. 2. After Alice’s measurement a unitary restoration process takes place that locally undoes the measurement and returns Alice and Bob plus the two particles to their original total state (so that Alice loses all her information about the outcome). "

Guess what? There is no "unitary restoration process" by which Alice's pre-measurement state can be recreated. To the extent Alice makes her (irreversible) measurement, her particle is separated from Bob's - and there is no going back. I'd like to see the experiment Dieks imagines actually performed... because it can't (and he seems to even acknowledge that). And that is not interpretation dependent, it is present one way or another in all interpretations.

I realize that Dieks was specifically attempting to demonstrate that unitarity* must be wrong as an element of an interpretation in his example above. But you can't cite examples that are known to be wrong to support your argument. No one believes Alice can "forget" her measurement in the first place. We already know that even the possibility of extracting quantum information from an environment will lead to what is commonly labeled "collapse", so why bother with a straw man?

He concludes: "This further supports the necessity of a privileged frame in Bohm’s theory." Obviously it proves no such thing. There is no dependence on reference frame. Note that Alice's observed particle need not even exist at the same time (in any and all reference frames) as Bob's to be entangled with Bob's observed particle. Further, the Alice/Bob entanglement can be created before or after Alice and/or Bob's measurements. That is an actual experimental outcome, not a hypothetical one, and it demonstrates clearly that frame dependence is NOT a variable in entanglement experiments.* "A point that will play an important role later on is that the assumption of universal validity of unitary Schrödinger evolution entails that all quantum processes can in principle be reversed."------------------------If you are going to give us a new interpretation, at this point it really needs to directly address the state of the art experiments (and theory) being done by entanglement teams such as Zeilinger, Hanson, Pan, etc. These teams and many others are probing the extremes of quantum nonlocality. Such experiments should be readily explainable by any new interpretation. As it is, I believe many of the popular interpretations struggle to provide a credible explanation of things such as quantum teleportation without jumping through ever more implausible hoops.

Anything on the arguments like"theories of such kind must need this perspectivalism"or "there are reasons to think such perspectivalism exists"?
And what is the difference between this
new interpretation and RQM?What could it mean that reality is made of "disagreeing perspectives" where observers could disagree even when they perform local measurements.

 

[Fra]

What could it mean that reality is made of "disagreeing perspectives" where observers could disagree even when they perform local measurements.

IMHO, it could for example mean that:

Agents with not only different priors but having different implicit prior information (ie a wider concept) simply make different optimal inferences and they all optimal (at average, as there is stochastics too). Right or wrong is the wrong question IMO because there is no judge! In nature i guess the one who lives is "right". Ie. Natural Selection.

This creates a tension between agents. Ie they have different expectations of their common reality. Depending on your stance as theory builder this is either bad or a blessing.

My stance is that this "tension" in perspectives is exactly what encodes and explains agent-agent interactions, and thus ultimately the fundamental interactions of matter.

But noone has yet cast interactions in this form. Its on the todo list

/Fredrik

 

[kurt101]

They're saying that the experiments in question are vulnerable to what they call the "Collider Loophole". As the term "loophole" is usually used in this context, it amounts to saying that if the claimed "loophole" is closed, the violations of the Bell inequalities will go away--which would imply that the predictions of QM would be violated.

The authors of this paper claim that they are not disputing violation of the Bell inequalities, but only "the inference from violation of the Bell inequalities to AAD [action at a distance]". But later on, they say "The alternative is that the Bell correlations are an artifact of collider bias." But they don't actually present this as an "alternative" to their argument, but as part of their argument. Which to me means that, whether the authors want to admit it or not, they are claiming that if the "collider loophole" is closed, the Bell inequality violations will have to go away (because the authors claim those violations are "an artifact of collider bias", which closing the "collider loophole" would remove). Which, again, would imply that the predictions of QM would be violated.

As I said in the previous thread that you quoted, "loophole" claims like this have been common in the literature ever since experimental tests of Bell inequality violations were begun in the 1970s. And every single "loophole" claim has turned out to be wrong. My money is on this one being wrong too.

The authors are saying the Delft experiment or any experiment that has the BSM test in the absolute future of measuring Alice and Bob is vulnerable to thinking there is true entanglement when it can be post selection entanglement (correlation, but no action at a distance). For other experiments where the BSM test is in the past of Alice and Bob's measurement the authors say entanglement must be real.

And I would go further and say that the case where the BSM test is done in the absolute future of measuring Alice and Bob is most likely because of post selection and to deny this would contradict the Bell result itself. If you expect the non-local result of the EPR experiment, you would also expect that the inverse of the EPR experiment would have to be true. The entanglement swapping experiment with BSM test in the absolute future is just the inverse of the EPR experiment; instead of starting with two photon states the same, you are ending with two photon states the same. So to deny this case can not be a post selection trick is the same as being in conflict with the actual Bell result of the normal EPR case. And if you think I am wrong, do the experiment in your head or on paper in reverse; start with the photons at the BSM test and make them go in the reverse direction; is this not exactly equal to the normal EPR experiment?

So my objection to Dr. Chinese statements is implying (whether he intended it or not) that this case shows retrocausality when there is no indication that it does and the expectations are that it does not.

 

[PeterDonis]

The authors are saying the Delft experiment or any experiment that has the BSM test in the absolute future of measuring Alice and Bob is vulnerable to thinking there is true entanglement when it can be post selection entanglement (correlation, but no action at a distance). For other experiments where the BSM test is in the past of Alice and Bob's measurement the authors say entanglement must be real.

But this makes no sense if the authors agree that, in all of these cases, the Bell inequalities can be violated, even if the "Collider Loophole" is closed. Violating the Bell inequalities is "true entanglement". "Entanglement" is not some extra property that might or might not be added to systems that can violate the Bell inequalities. It is what violates the Bell inequalities.

So the only way I can make sense of this criticism at all is as a claim that, since there isn't actually "true entanglement" involved in at least some of these cases, if the "Collider Loophole" is closed, the Bell inequalities will no longer be violated in those cases. Which is not what standard QM predicts; standard QM predicts that the "Collider Loophole" doesn't matter at all.

 

[PeterDonis]

do the experiment in your head or on paper in reverse; start with the photons at the BSM test and make them go in the reverse direction; is this not exactly equal to the normal EPR experiment?

No. You can't just "make them go in the reverse direction". The BSM test measurement is irreversible.

If you took the photons at the BSM test and measured their polarizations after that test was done, yes, you would find they were correlated. But you would have to do the same post-selection on the BSM test results to see the correlations, that you do on the Alice and Bob measurement results to see the correlations. So by your argument, if the Alice and Bob correlations are artifacts of post-selection, normal EPR correlations that violate the Bell inequalities must be artifacts of post-selection too. Is that your argument?

 

[DrChinese]

1. The authors are saying the Delft experiment or any experiment that has the BSM test in the absolute future of measuring Alice and Bob is vulnerable to thinking there is true entanglement when it can be post selection entanglement (correlation, but no action at a distance). For other experiments where the BSM test is in the past of Alice and Bob's measurement the authors say entanglement must be real.

And I would go further and say that the case where the BSM test is done in the absolute future of measuring Alice and Bob is most likely because of post selection...

2. So my objection to Dr. Chinese statements is implying (whether he intended it or not) that this case shows retrocausality when there is no indication that it does and the expectations are that it does not.

1. The 1 & 4 photons end up entangled if and only if a successful BSM occurs. That is not post selection, and I can explain exactly why that is so:

a. There is no question that when the original biphotons (systems consisting of an entangled photon pair) are created (i.e. pairs 1 & 2 and 3 & 4), they are maximally entangled. They will of course exhibit perfect correlations and violate Bell inequalities. They cannot be entangled with any other object due to monogamy of entanglement.

b. There is no question that when the final biphoton is created (i.e. pair 1 & 4), it is also maximally entangled. The pair will of course exhibit perfect correlations and violate Bell inequalities. The photon pair cannot be entangled with any other object due to monogamy of entanglement.

c. The "cause" of this change is the BSM, and nothing else. There can be no post-selection issue because there is one and only one point of interaction between the initial biphotons, and that is at the BSM. There is no statistical connection from the past of the 1 & 2 and 3 & 4 pairs that can connect to the subsequent statistics of the final 1 & 4 pair - precisely because of the monogamy of entanglement. So there is nothing to see in the way of post selection. The BSM is the crossover point from 2 biphotons down to 1 biphoton. If and only if the 2 & 3 photons click within a small coincidence window - such that they cannot be distinguished as to source - will the BSM succeed. They must be indistinguishable.

d. And if there were only post-selection on the 2 & 3 photons - and not some kind of quantum interaction - you should be able to simply post-select on the identical quantum attributes - such as polarization, arrival time, transmit or reflect through the Beam Splitter, wavelength, etc. - without allowing any interaction. The 2 & 3 photons would pass the post-selection tests... but would be distinguishable. Obviously, the BSM fails and the 1 & 4 photons do not pass the Bell test.

What kind of post-selection requires interaction? Either you are post-selecting on some list of criteria, or not.2. I never say there is "retrocausality". I have no idea if there is or isn't. What I can say is that there is NO observable change in outcomes whether the BSM occurs before or after Alice (and/or Bob) makes a measurement on the 1 photon (or 4 photon for Bob). Ordering changes nothing in the measurement context. (That is therefore the argument for asserting that QM is acausal.) All I know is that Einsteinian causality fails. *You* might see this as a claim to retrocausality, because you assume there is causality in nature.

What is generally agreed upon is that there is NO possibility of explaining the random outcome of a quantum interaction. There is nothing in the past context to explain it, and nothing in the future that explains it either. Again, what I can say is that in a Bell test: only Alice and Bob's relative settings (the measurement context) are relevant to the statistical outcomes.Of course, we have again somehow deviated from the thread topic.

 

[kurt101]

Violating the Bell inequalities is "true entanglement". "Entanglement" is not some extra property that might or might not be added to systems that can violate the Bell inequalities. It is what violates the Bell inequalities.

There are violations of the Bell inequality that can be shown to violate locality and there are violations of the Bell inequality that can't be shown to violate locality. The author is calling attention to the later in the case where the outside photons (that I am calling Alice and Bob) are measured before the BSM test is done. Note, in no way is the author disputing the non-locality bell violation in two halves of the experiment in any of the cases.

1. The 1 & 4 photons end up entangled if and only if a successful BSM occurs. That is not post selection, and I can explain exactly why that is so:

I am good with this statement when entangled means a violation of the Bell entanglement and does not imply direct non-local influences between 1 & 4.

a. There is no question that when the original biphotons (systems consisting of an entangled photon pair) are created (i.e. pairs 1 & 2 and 3 & 4), they are maximally entangled. They will of course exhibit perfect correlations and violate Bell inequalities. They cannot be entangled with any other object due to monogamy of entanglement.

Agree.

b. There is no question that when the final biphoton is created (i.e. pair 1 & 4), it is also maximally entangled. The pair will of course exhibit perfect correlations and violate Bell inequalities. The photon pair cannot be entangled with any other object due to monogamy of entanglement.

Agree with the caveat that the entanglement between 1 & 4 don't demonstrate direct non-local effects between 1 & 4 in the case where the BSM test is done after the measurement of 1 & 4.

c. The "cause" of this change is the BSM, and nothing else. There can be no post-selection issue because there is one and only one point of interaction between the initial biphotons, and that is at the BSM.

There are lots of photons at 1 & 4 that arrive at the same time that are candidates for being entangled. The experiment ignores any that don't show entanglement in the corresponding 2 & 3 at the BSM test. This process is what I mean by post selection. Obviously post selection is necessary to see the bell violation in all cases. I assume we agree on this.

So I can agree that using the BSM along with the post selection process I just described is the cause / reason that photons 1 & 4 are entangled (have the Bell violation). I am sounding like a broken record at this point, but just to be clear what my point and what I think the author's point is, is that entanglement between 1 & 4 in the case where the BSM test is done after the measurement of 1 & 4 does not prove that there is a direct non-local influence between 1 & 4.

There is no statistical connection from the past of the 1 & 2 and 3 & 4 pairs that can connect to the subsequent statistics of the final 1 & 4 pair - precisely because of the monogamy of entanglement. So there is nothing to see in the way of post selection.

I am not clear what you are saying here.

The BSM is the crossover point from 2 biphotons down to 1 biphoton. If and only if the 2 & 3 photons click within a small coincidence window - such that they cannot be distinguished as to source - will the BSM succeed. They must be indistinguishable.

Agree

d. And if there were only post-selection on the 2 & 3 photons - and not some kind of quantum interaction - you should be able to simply post-select on the identical quantum attributes - such as polarization, arrival time, transmit or reflect through the Beam Splitter, wavelength, etc. - without allowing any interaction. The 2 & 3 photons would pass the post-selection tests... but would be distinguishable. Obviously, the BSM fails and the 1 & 4 photons do not pass the Bell test.

I think the quantum interaction (BSM test) is necessary because it is impossible to determine if the photons are otherwise truly identical by attempting to measure its properties. So I don't agree with your characterization "you should be able to simply post selection on identical quantum attributes" is possible.

What kind of post-selection requires interaction? Either you are post-selecting on some list of criteria, or not.

I don't understand your question here, but I explained what I mean by post-selection above.

2. I never say there is "retrocausality". I have no idea if there is or isn't.

Ok

What I can say is that there is NO observable change in outcomes whether the BSM occurs before or after Alice (and/or Bob) makes a measurement on the 1 photon (or 4 photon for Bob). Ordering changes nothing in the measurement context. (That is therefore the argument for asserting that QM is acausal.)

I am afraid to use the term "acausal" because I was threatened with being kicked off the thread last time I used it even though I was just replying to your use of it. If you mean "not causal" then I think my objection remains, as I don't think the entanglement swapping experiment shows this for the reason the author of the paper we have been discussing brought up.

All I know is that Einsteinian causality fails. *You* might see this as a claim to retrocausality, because you assume there is causality in nature.

What is your definition of Einstein causality. Is this a good definition (posted from google search):
In both Einstein's theory of special and general relativity, causality means that an effect cannot occur from a cause that is not in the back (past) light cone of that event.

If this is what your definition is then I continue to disagree with you that the entanglement swapping experiment shows disagreement with this.

What is generally agreed upon is that there is NO possibility of explaining the random outcome of a quantum interaction. There is nothing in the past context to explain it, and nothing in the future that explains it either.

I thought it was a hot topic of debate on whether there is an underlying model for Quantum mechanics. Though maybe I misunderstand you. I am always looking to understand why someone comes to this conclusion beyond just because nothing has been found yet.

Again, what I can say is that in a Bell test: only Alice and Bob's relative settings (the measurement context) are relevant to the statistical outcomes.

Except in the entanglement swapping experiment, it is also the BSM test that is relevant.

Of course, we have again somehow deviated from the thread topic.

I don't like doing that either and would have preferred to finish this in the thread that you created. Right now I just want to get a common understanding and ultimately the truth on this.

No. You can't just "make them go in the reverse direction". The BSM test measurement is irreversible.

If you took the photons at the BSM test and measured their polarizations after that test was done, yes, you would find they were correlated. But you would have to do the same post-selection on the BSM test results to see the correlations, that you do on the Alice and Bob measurement results to see the correlations. So by your argument, if the Alice and Bob correlations are artifacts of post-selection, normal EPR correlations that violate the Bell inequalities must be artifacts of post-selection too. Is that your argument?

I would like to respond to this as you are not understanding the argument I am making, but I have run out of time right now. I obviously realize you can't actually go in the reverse direction, but thinking of the experiment in the reverse direction makes it very apparent why you get the bell violation between 1 & 4 when the BSM measurement is made in the absolute future.

 

[PeterDonis]

There are violations of the Bell inequality that can be shown to violate locality and there are violations of the Bell inequality that can't be shown to violate locality.

What definition of "locality" are you using?

You are aware, aren't you, that a very common definition of "locality" in the literature in this area is "does not violate the Bell inequalities"?

Physics is not word games. You can of course define "locality" however you want for your own usage, but if you want me to adopt your definition, particularly if it's a highly non-standard one, you're going to have to first meet the very heavy burden of convincing me that I should care.

 

[PeterDonis]

thinking of the experiment in the reverse direction makes it very apparent why you get the bell violation between 1 & 4 when the BSM measurement is made in the absolute future.

I don't understand how. But, as has already been commented, that discussion probably belongs in a separate thread.

 

[Fra]

I thought it was a hot topic of debate on whether there is an underlying model for Quantum mechanics. Though maybe I misunderstand you. I am always looking to understand why someone comes to this conclusion beyond just because nothing has been found yet.

I think there are many broad views represented here so maybe there are people looking for loopholes to Bells inequality in the original sense, but IMO maybe this was a broader hot topic 50+ years ago but hardly today.

A mystery still remains to understand the logic of the causality in the interactions that QM well describes. What Bell made clear is that the simplest possible explanation (igonorance type HV) that can be used to "average" otherwise deterministic mechanisms doesn't work.

But because that first explanatory idea of Bell failed, does not mean there isn't yet a deeper lession to learn, that does comply to the bell mechanisms. This is where I see todays research should focus, and I see that conicides with more general foundational questions on QM. It's bigger than just a resistance to give up on deterministic world view, and thinking god does not play dice. I see it as the quest to understand how parts of nature "knows things" about each other, and how that may be part of understanding their objective interactions also from the macroscopic perspective.

I think I see three types of stances here

I) People that is seeking solution to the original wishes of Bell or Einstens (by looking for loopholes)

II) People that think that Bell proves that category I ideas wrong once and for all, we are done with this questioning of QM weirdness and see no problems, and just work on other things thiking QM has proven it's excellennce over and over again and we should not bother mess with it unless we want to go into domains of philosophy

III) People that agree with category II in that category I ideas are proven wrong, but still see a remaing unsolved and important puzzle with QM we must solve that belongs to science and not pure philosophy

(I belong to categoty III if it wasn't obvious)
/Fredrik

 

[kurt101]

What definition of "locality" are you using?

You are aware, aren't you, that a very common definition of "locality" in the literature in this area is "does not violate the Bell inequalities"?

Yes I understand this. I am using John Bell's definition of locality. It can be shown that photons 1 & 2 violate locality and photons 3 & 4 violate locality and that photons 1 & 4 violate locality, but the violation of locality of 1 & 4 is different in that the violation does not imply a direct non-local influence between photons 1 & 4 in all cases. It is this difference that is important for interpretations and is what the paper is pointing out. Without recognizing this distinction, cause and effect interpretations could be ruled out by mistake. I am still not certain that is what Dr. Chinese is doing, but I want to make sure that this forum is being careful on this. We don't want to mislead anyone, correct? That is why I am so persistent on trying to understand the entanglement swapping experiment and is why I went through the process of writing a program to prove to myself that cause and effect is not ruled out by this experiment.

Physics is not word games. You can of course define "locality" however you want for your own usage, but if you want me to adopt your definition, particularly if it's a highly non-standard one, you're going to have to first meet the very heavy burden of convincing me that I should care.

I am using John Bell's definition of locality. I don't want to play word games either.

If you took the photons at the BSM test and measured their polarizations after that test was done, yes, you would find they were correlated.

This statement does not make sense to me. You can measure the polarization of photons 1 & 4 after the BSM test, but not 2 & 3, because photons 2 & 3 are destroyed in the BSM test measurement.

But you would have to do the same post-selection on the BSM test results to see the correlations, that you do on the Alice and Bob measurement results to see the correlations.

Agree

So by your argument, if the Alice and Bob correlations are artifacts of post-selection, normal EPR correlations that violate the Bell inequalities must be artifacts of post-selection too. Is that your argument?

No this is not an argument I am making. The argument I am making is that in the entanglement swapping experiment, the only results of measuring photons 1 & 4 that we look at are when the BSM test shows that photons 2 & 3 are entangled and that this is post selecting the answer to the question of "When does it look like photons 1 and 4 are entangled?". And in contrast to the EPR experiment, with the entanglement swapping experiment, specifically for the case where the BSM test is done in the absolute future of measuring photons 1 & 4, you can not say that being entangled (having a correlation that violates the Bell inequality) implies a direct non-local influence between photons 1 & 4.

A cause and effect process must be shown to be in principle reversible. If I can show that the entanglement swapping experiment is in principle reversible then it is consistent with a cause and effect explanation.

Lets do the entanglement swapping experiment in reverse as a thought experiment for the case where the BSM test is done in the absolute future of measuring photons 1 & 4.

1. Reading results of experiment show a bell violation of photons 1 & 4 when measurement angles differ.
2. BSM test on photons 2 & 3 show that photons 2 and 3 are identical in state.
3. Measurement of photon 1 changes the state of photon 2.
4. Measurement of photon 4 changes the state of photon 3.
5. Photons 1 and 2 meet and are identical in state.
6. Photons 3 and 4 meet and are identical in state.

The point of the paper and the point I am trying to show is that at no time did photon 1 & 4 influence each other in this sequence of events. Cause and effect is preserved.

Contrast this with the case where the BSM test is done in the absolute past of measuring photons 1 & 4 (again doing the events in reverse order):

First try to do this without 1 and 4 being allowed to influence each other directly:
1. Reading results of experiment show a bell violation of photons 1 & 4 when measurement angles differ.
2. Measurement at photon 1 changes the state of photon 2.
3. Measurement at photon 4 changes the state of photon 3.
4. BSM test done between photons 2 & 3 shows that photons 2 and 3 are identical in state.
5. At this point you are stuck, if photons 2 & 3 are the same, that means photons 1 & 4 must have started out the same in this reverse experiment which can't be true unless their measurement angles are the same.

Now do this case but allow an influence between 1 and 4 and you are ok:
1. Reading results of experiment show a bell violation of photons 1 & 4 when measurement angles differ.
2. Measuring photon 1 changes the state of photon 4.
3. Photon 4 is measured.
4. BSM test done between photons 2 & 3 shows that 2 and 3 are identical in state (no conflict with anything prior) and establishes a causal influence between photons 1 and 2 and establishes a causal influence between photons 3 & 4 and drops the causal influence between 1 & 4. In establishing the causal influence it implies 1 & 2 are now in the same state and 3 & 4 are now in the same state.
5. Photons 1 and 2 meet and are identical in state.
6. Photons 3 and 4 meet and are identical in state.

This thought exercise demonstrates that the entanglement swapping experiment is consistent with a cause and effect interpretation.

 

[PeterDonis]

A cause and effect process must be shown to be in principle reversible.

Where are you getting this claim from? Did you just make it up? Is it your personal theory?