I A new realistic stochastic interpretation of Quantum Mechanics

[JC_Silver]

Whatever Barandes is doing would lead to the particle effectively tunneling and thus not following classical physics either.

I see, thanks!

 

[Fra]

OK, that's what I thought you were aiming at. But... how could that work? Actually, I agree that photons 2 & 3 (when indistinguishable at the BSM) do connect initially separate systems 1&2 and 3&4.

But there is no spacetime region in which (in some versions of the experiment) that connection could possibly influence the previous 1 & 4 measurements - everything is distant in terms of c.

This is no answer, but what is wrong with this conceptual outline that ias how I understand it?

1&2 share a "hidden variable"
3&4 share another "hidden variable"

noone else can get they without breaking the entanglement, ie only 1 and 2 themselves, "know" the keys. That is the premise. same with 3 & 4.

If we can escape bells inequality, it could explain bother experiments and correlations. By "matching" the keys of 2&3; we know 1&4 keys correlate as well but without know what they are. - no spacetime region needed.

But the tentative escape of bells ansatz is that it is not possible to make a markov division of the process indexed by these hidden keys; because the interaction itself is influenced by the fact that the keys are hidden. So bell theorem does not apply to such KIND of hidden variable.

Baranade has no such theory but argues for its possibility? That one has trouble to even imagine what such as theory would be like, is a separate problem.

/Fredrik

 

[iste]

I made an account just to answer here because I feel I have one point to add.

Most of Barandes' newer interviews he has done he doesn't touch on the "interpretation" aspect of his paper instead choosing to focus on the "wavefunction is not a real thing" side and I believe this is because he has possibly realized what some here have too, this is not an interpretation of QM, this is a reformulation of the math instead.

And I think that if what Barandes claims is true (I don't have enough knowledge in non-markovian processes to say if his math is correct), I believe it's worth looking into.

Assuming ALL of Quantum Mechanics is a special case of non-Markovian processes, it could be worthy to look into it to see if we can find fenomena where QM can't seem to find an answer (eg. Gravity) to see if the answer lies outside the Markovian assumptions.

I wouldn't call this an interpretation though, much less say it confirms that Pilot Waves are real.

I say the value of this paper is to argue from a new perspective that wavefunction is a math tool to predict the world and that they are not the real object.

There are definitely parts of his latest interview where he implies this has fully interpretational consequences, parts where he talks about no need for observers I remember distinctly, as well as other parts. Its very general but it is unambiguous that the theory, if true, implies particles are always in one configuration at a time. That 100% implies some kind of interpretation that is different to what the great majority of people currently think, and would completely deflate things like the measurement problem. There is no way he could talk about this deflating quantum mechanics or making it more boring if there wasn't some tangible interpretational element here, even if not everything is explained about why quantum mechanics would be that way. The wavefunction not being real is in and of itself part of interpretation. True, what he would be pushing here is a distinct formulation of quantum mechanics, but it directly implies an interpretation too - the kind it shares with stochastic mechanics.

That seems to be the growing consensus here, if the math is right. In that case, Barandes just found a useful duality between quantum and non-Markovian akin to the Osborn's rule or Wick rotations. If it works as an interpretation some of us are still "agnostic".

Of course some have considered that the mathematics might not be right or well funded either. For example Barandes would have to explain how to avoid the result of paper in post #327

If it was just like Wick rotation, Barandes wouldn't be going on about how the wave function isn't real or quantum mechanics becomes more boring or observers are deflated.

And if you look at the bottom of that paper at the conditions required to satisfy classical stochastic processes in terms of a two-time correlation function - I am pretty sure that at the core of Barandes' idea is that this kind of thing is violated. The classical two-time correlation function implies divisibility. So Barandes doesn't have to avoid the result because the indivisible stochastic process is not of the kind of stochastic process the paper is using to compare to quantum mechanics and saying is incompatible. The Barandes process is compatible.

This might be obvious but I honestly don't know the answer. How does Barandes' (and any other pilot wave interpretation) explain quantum tunneling? There shouldn't be any classical fenomena that allows for it, right?

The particle does not need to have enough energy to jump over the barrier because it is not following classical physics, it is just guided by the guiding wave function which can tunnel.

I don't know anything about this topic but I have seen it mentioned in the context of stochastic mechanics, e.g. this paper:

https://scholar.google.co.uk/schola...2266132&hl=en&as_sdt=0,5&as_yhi=1990&as_vis=1
https://www.mdpi.com/2218-1997/7/6/166

I quote from first paper but says similar in second recent paper:

"The energy in stochastic mechanics is a randomly fluctuating
quantity which is always non-negative just as in Newtoni-
an mechanics. Thus a stochastic mechanical system never actually "tunnels" through a potential barrier; rather, it is "kicked" over the barrier by a large enough positive energy fluctuation resulting from a configuration fluctuation."

Presumably, similar would be the case if it could be formulated in Barandes' theory. Or maybe not. It seems reasonable to me that the Barandes model would do the same thing if it could be formulated. But there is no evidence for that other than stochastic mechanics and Barandes' model fall broadly under the same interpretational hat and they both claim to derive or prove some correspondence between quantum mechanics and stochastic systems.

 

[DrChinese]

I'm not trying to say anything about how it might work. I'm only trying to describe the conditions that appear to me to be necessary to have entanglement, without making any hypotheses at all about why those are the necessary conditions. I understand that nobody has a mechanism that could explain why. But that doesn't change the fact that, for example, if we take away the "interaction" node at the BSM (in my graph description), entanglement is no longer present--and as you have said, whether or not there is an "interaction" node at the BSM is under the experimenter's control. Nobody understands how it could be that the experimenter can control whether photons 1 & 4 are entangled by choosing whether or not to let photons 2 & 3 interact at the BSM. But that is, factually, the case, as the experiments show. I'm just trying to describe that factual condition in a way that might help to make it clearer exactly what is required.

My bad, I wasn't trying to actually ask about the underlying mechanism for entanglement swapping - but I see why it looks that way.

What I meant was: Yes, there is one of your nodes at the BSM when swapping is active. No disagreement there. And you punted on the question of whether it was an "interaction" or a (joint) measurement. But for anyone wanting to draw Einsteinian causality/locality into the equation (by assumption of course), how would that serve as an "out"? It won't. If there is no opportunity for an interaction at the 2&3 beam splitter, there is no swap. That is true regardless of where or when the BSM is located/executed.

As I re-read your post, I don't think there is a disagreement. It is a true statement that as far as I know, there must be something like the nodes you describe to enable entanglement. However, they don't need to respect locality and/or causality between the end nodes.

Ken Wharton refers to your diagram as a "W" entanglement type, for the entanglement swapping setups. For the more common entanglement from a single PDC crystal, he calls that a "V".

 

[DrChinese]

This is no answer, but what is wrong with this conceptual outline that ias how I understand it?

1&2 share a "hidden variable"
3&4 share another "hidden variable"

noone else can get they without breaking the entanglement, ie only 1 and 2 themselves, "know" the keys. That is the premise. same with 3 & 4.

If we can escape bells inequality, it could explain bother experiments and correlations. By "matching" the keys of 2&3; we know 1&4 keys correlate as well but without know what they are. - no spacetime region needed.

But the tentative escape of bells ansatz is that it is not possible to make a markov division of the process indexed by these hidden keys; because the interaction itself is influenced by the fact that the keys are hidden. So bell theorem does not apply to such KIND of hidden variable.

Baranade has no such theory but argues for its possibility? That one has trouble to even imagine what such as theory would be like, is a separate problem.

/Fredrik

OK, this is the point I keep making - and apparently I am not explaining well enough.

So there are 2 initial "keys" (1&2 and 3&4) according to your hypothetical explanation. You think that if you know ("match") those keys, you can predict the 1&4 entanglement. There are a lot of reasons* what that cannot be true. But I will focus on the one related to the BSM and whether swapping is enabled or not.

When entanglement is enabled: we read those keys - and make a prediction that is certain (correlation=1). When entanglement is not enabled: we read the exact same keys! But the outcome is random (correlation=0). This is actually performed in the Ma paper ("Experimental delayed-choice entanglement swapping"). Obviously, there are no keys/hidden variables as you want to describe them.

To repeat: everything registers the same at the BSM regardless of whether there is a swap or not. But there is only a swap if the 2 and 3 photons overlap within a narrow time window - so they are allowed to interact (and they must be indistinguishable). The experimenters delay the "2" (or "3") photon if they don't want a swap to occur. But you assert the photons already possessed whatever property value is being measured regardless of timing. And you assert there is no physical interaction capable of affecting the 1&4 photons in any way due to their overlapping in the beam splitter.

So whether they overlap or not shouldn't even matter from your perspective. But experiment says otherwise. And the results are pretty much back or white.


*Violation of a Bell Inequality in swapping experiments also disproves the "key" idea, which is just another attempt at a local hidden variable model with a different label.

 

[iste]

This is the experiment, and saying his "proof not is faulty" does not make it cover this experiment as it might otherwise for basic PDC entanglement (without a swap).

If entanglement swapping is a process of a unitarily evolving quantum system then I don't understand why what you say would be the case  if Barandes was correct. He would claim that he has proved that he could convert that description into a stochastic system.

irreversible interaction at the BSM creates a swap

Collapse may not be real. It is not a foregone conclusion that it is and several functioning formulations of quantum mechanics do not have a physical collapse. On a formal level, Barandes explicitly identifies the collapse as simply statistical conditioning with no physical content.

You are the one who says: Entanglement is always "directly attributed to an initial local interaction". Either Barandes claims this, or he doesn't. Either way, it is factually incorrect (experimentally falsified).

My thoughts is that the entanglement part of entanglement swapping is quantum; we then have two different entangling scenarios going on, both respectively caused by initial local interactions.

The swapping part, even though clearly is occuring in quantum systems is not really quantum at all. You can make sense of this with any kind of set of correlations that: if A and B are correlated, and then C and D are correlated, then if you correlate B and C, A and D will be correlated. This kind of reasoning can be applied to literally anything.

You can correlate B and C by simply conditioning on two of their respective outcomes. If you do not condition on two of their respective outcomes, there will clearly be no correlation if all the outcomes occur because clearly all of the pairs - B1C1, B1C2, B2C1, B2C2 - can occur equally. If outcomes from A and D have a one-to-one mapping with those of B and C, then there will clearly be no correlation between A and D either because there is none between B and C. But if you condition on B1C1, you are going to get the same pair of outcomes for A and D all the time so they are now correlated.

You don't need quantum mechanics to justify these correlations beyond the regular entanglements.

2. This too is factually incorrect: There is no statistical conditioning. ALL cases in which photons 2 & 3 arrive within the specified narrow time window (indistinguishably) lead to a swap, and are counted - and they show perfect correlations (1). All cases in which photons 2 & 3 arrive at the BSM but are made distinguishable fail to lead to a swap, and are counted. They show only random correlations. The only difference is whether the experimenter chooses to make them distinguishable or not

A Bell state measurement is equivalent to statistical conditioning in the Barandes framework. It is an experimental exercise in statistical conditioning.

My thoughts are that indistinguishability is a requisite for the coherence required for the Bell state. In Barandes' formulation, an intrusive interaction with an entangled state would cause it to lose its correlation and coherencr, due to something called a division event. Presumably, distinguishability for the Bell state is like that, like the way a measurement causes decoherence or which-way information ruins interference. Those are all covered by division events which would ruin the kind of entangled states, the kind of correlations that allow the conditioning exercise I have just given.

3. My point being: how can the interaction at the BSM *not* be physically real

Only physical collapse isn't real, but if you allow (i.e. construct the experimental conditions conducive to allow) the two different entangled systems to correlate and pretend there is a collapse (i.e. just statistical conditioning maybe without even knowing it), you will get the correlations naturally just by assuming regular conventional entanglement.

 

[PeterDonis]

here is one of your nodes at the BSM when swapping is active. No disagreement there.

Ok, good.

And you punted on the question of whether it was an "interaction" or a (joint) measurement.

For my formulation it doesn't matter which it is. All three kinds of nodes serve the same function as far as determining whether there is a connected subgraph.

for anyone wanting to draw Einsteinian causality/locality into the equation (by assumption of course), how would that serve as an "out"? It won't.

I wasn't trying to address such questions; to me they are part of the "how" that I wasn't discussing. I don't think my formulation gives any particular help to any particular attempt at interpretation of the "how". It's just a way of describing exactly what conditions are required for entanglement according to the experimental facts.

However, they don't need to respect locality and/or causality between the end nodes.

Yes, of course.

Ken Wharton refers to your diagram as a "W" entanglement type, for the entanglement swapping setups. For the more common entanglement from a single PDC crystal, he calls that a "V".

Ah, ok. I figured someone had come up with something similar already.

 

[Fra]

OK, this is the point I keep making - and apparently I am not explaining well enough.

So there are 2 initial "keys" (1&2 and 3&4) according to your hypothetical explanation.

Or i was thinking in terms of 2 pairs or correlates keys. The bsm serves to find the new 2&3 pairs that "match" then we know corresponding 1&4 do.

You think that if you know ("match") those keys, you can predict the 1&4 entanglement. There are a lot of reasons* what that cannot be true. But I will focus on the one related to the BSM and whether swapping is enabled or not.

When entanglement is enabled: we read those keys - and make a prediction that is certain (correlation=1). When entanglement is not enabled: we read the exact same keys! But the outcome is random (correlation=0). This is actually performed in the Ma paper ("Experimental delayed-choice entanglement swapping"). Obviously, there are no keys/hidden variables as you want to describe them.

I dont follow your issue as i see no problem here.

i am getting convinced that you think the HV necessarily defines the states, as they do in bells thinking. This the difference between ignorance HV and HV that are "subjective" ie subjective to 1&2 and 3&4 only. They are exposed only upon physicaö interaction

But you assert the photons already possessed whatever property value is being measured regardless of timing.

No, i didnt say that. I said the photons are correlated; they are not determined one by one.

And you assert there is no physical interaction capable of affecting the 1&4 photons in any way due to their overlapping in the beam splitter.

The 2&3 interaction is physical and required to identify the 1&4 that are correlated. info from bsm are classically transmitted.

So whether they overlap or not shouldn't even matter from your perspective.

Of course it matters. The DATA from bsm is the key of the protocol. you probably assume the photon states are predtermined relative to environmebt from HV. That is not what i suggest.

*Violation of a Bell Inequality in swapping experiments also disproves the "key" idea, which is just another attempt at a local hidden variable model with a different label.

No, what i lined out is not a bell style HV theory.

Hidden key pair does NOT predetermine the photons, the only define their relation. This keys have no relation to the environment. This is how they are hidden. So not same as "ignorance". Which is imo related to the lack of markov indivisibility.

/Fredrik

 

[pines-demon]

There are definitely parts of his latest interview where he implies this has fully interpretational consequences, parts where he talks about no need for observers I remember distinctly, as well as other parts. Its very general but it is unambiguous that the theory, if true, implies particles are always in one configuration at a time. That 100% implies some kind of interpretation that is different to what the great majority of people currently think, and would completely deflate things like the measurement problem. There is no way he could talk about this deflating quantum mechanics or making it more boring if there wasn't some tangible interpretational element here, even if not everything is explained about why quantum mechanics would be that way. The wavefunction not being real is in and of itself part of interpretation. True, what he would be pushing here is a distinct formulation of quantum mechanics, but it directly implies an interpretation too - the kind it shares with stochastic mechanics.

You said before that one could be agnostic given the state of the theory. Now you are eager to defend there is some interpretation. I do not think it is worth it unless he provides more content about it. I mean Barandes wants to provide us right away with interpretational conclusions but gives no intuitive mechanism on how to understand an indivisible stochastic process. That's what we need to understand here, because even if equivalent to quantum mechanics it is equally mysterious and nonclassical. He is proposing some stocastic force that permeates space and allows to bypass Bell's theorem. In most cases that would mean that his theory is nonlocal.

Bohmian mechanics also has one configuration at a time, superdeterminism does to.

If it was just like Wick rotation, Barandes wouldn't be going on about how the wave function isn't real or quantum mechanics becomes more boring or observers are deflated.

Well it can be both a mathematical duality and an interpretation. However for me it is not deflational in the sense that we have other similar interpretations and Barandes interpretation barely provides a way to think about it. Also the duality seems to be between quantum mechanics and indivisible non-Markovian processes (INMP). The latter is not well studied at all, so quantum mechanics (QM) can help INMP more than INMP can help QM.

And if you look at the bottom of that paper at the conditions required to satisfy classical stochastic processes in terms of a two-time correlation function - I am pretty sure that at the core of Barandes' idea is that this kind of thing is violated. The classical two-time correlation function implies divisibility. So Barandes doesn't have to avoid the result because the indivisible stochastic process is not of the kind of stochastic process the paper is using to compare to quantum mechanics and saying is incompatible. The Barandes process is compatible.

People have tried to break the axioms of probability before to get quantum mechanics without amplitudes but that does not mean it is easier to comprehend or to interpret. Is Barandes proposal even a stochastic process at this point? We really need a simple example, hopefully a classical one.

 

[Fra]

3. My point being: how can the interaction at the BSM *not* be physically real, given the facts per 2.?

For me the BSM is certainly real, but i do not see it as a measurement of keys of 2 & 3 it measure of the relation. Knowing the relation between 2 keys, is the output in my conceptual world.

/Fredrik

 

[Fra]

There are definitely parts of his latest interview where he implies this has fully interpretational consequences, parts where he talks about no need for observers

If this is his main idea forward I am more sceptical.

He also mentions in one talk that his take on "observers" is that they are suffiently complex to qualify as a macroscopic system - ie just like in copenhagen interpretation. So this os where the transistion probabilites would then have to be encoded.

I personally think barandes view might mate well with interacting neural network models.

https://arxiv.org/html/2209.07577v3

https://www.nature.com/articles/s41534-018-0081-3

A key here are the hidden layers in neural networks. They are real but hidden. See paper. This mode of thinking is different than the hilbert/wave thing and might offer new perspectives to old issues.

/Fredrik

 

[DrChinese]

1. Collapse may not be real. It is not a foregone conclusion that it is and several functioning formulations of quantum mechanics do not have a physical collapse. On a formal level, Barandes explicitly identifies the collapse as simply statistical conditioning with no physical content.

2. ... we then have two different entangling scenarios going on, both respectively caused by initial local interactions.

3. The swapping part, even though clearly is occurring in quantum systems is not really quantum at all. You can make sense of this with any kind of set of correlations that: if A and B are correlated, and then C and D are correlated, then if you correlate B and C, A and D will be correlated. This kind of reasoning can be applied to literally anything.

You can correlate B and C by simply conditioning on two of their respective outcomes. If you do not condition on two of their respective outcomes, there will clearly be no correlation if all the outcomes occur because clearly all of the pairs - B1C1, B1C2, B2C1, B2C2 - can occur equally. If outcomes from A and D have a one-to-one mapping with those of B and C, then there will clearly be no correlation between A and D either because there is none between B and C. But if you condition on B1C1, you are going to get the same pair of outcomes for A and D all the time so they are now correlated.

4. Only physical collapse isn't real, but if you allow (i.e. construct the experimental conditions conducive to allow) the two different entangled systems to correlate and pretend there is a collapse (i.e. just statistical conditioning maybe without even knowing it), you will get the correlations naturally just by assuming regular conventional entanglement.

1. Collapse may not be physical, but it may be proven so in experiments. Interpretations don't prove this one way or another, even those that appear otherwise consistent.


2. No, we initially have 2 separate and independent systems. They have no connection or correlation to each other whatsoever, on any polarization measurement basis. They need not be local to each other.

This should be obvious, as you could have just as easily have >2 separate and independent quantum systems (of N=2 photons) to start with. None of those have any particular correlation whatsoever to the others either. In other words: a swap could in principle be made to occur between any of dozens of initially entangled pairs (1&2, 3&4, 5&6, 7&8, ...) created anywhere and at any time. It is the near simultaneous and indistinguishable arrival of one photon of 2 of those systems that creates the swapped entanglement and binds them into a single quantum system.


3. Surprised you would bring this up, because there is definitely no similarity to classical correlation. I can only assume you still do not understand important details about the operation of the BSM. Maybe I can help clarify a few points.

Yes, 1&2 are correlated. And yes, 3&4 are correlated. But that correlation is quantum, completely different than any other type of classical correlation (you should already know this, since you understand Bell). The 1&2 pairs are correlated on EVERY SAME polarization basis, and we know that relationship. That is actually an infinite number of possible bases. Same for 3&4 pairs. But... the number of possible outcomes at the BSM is only 4 (there are only 4 Bell states). Further, not all DIFFERENT polarization bases are correlated for 1&2 pairs. For example, if you measure photon 1 at 0 degrees and photon 2 at 45 degrees, there is precisely zero correlation. Ditto if you measure photon 1 at 20 degrees and photon 2 at 65 degrees, right?

At the BSM, you get 2 bits of information (and I am streamlining a bit here for simplicity, the nuances do not change our conclusion): Bit 1 is whether both photons emerge from different ports, or they exit from the same port (Transmit or Reflect itself does not matter). Bit 2 is whether they are polarized the same or different at any same angle (H or V itself does not matter). Importantly: the angle for testing Bit 2 need not be the angle used for the measurements on photons 1 & 4. And in fact, the 1 & 4 photons can be measured on a basis that does not commute with the 2 & 3 measurements. So here's where your thinking falls apart:

i) We get the values for Bit 1 and Bit 2 regardless of whether the 2 & 3 photons are distinguishable or not. But if they are distinguishable, there is no swap. According to your thinking, they should swap - regardless of overlap in the BSM - as long as we have the key. After all, you think the "key" (Bits 1 and 2) is being revealed to unlocking the hidden correlation of the 1 & 4 photons - and that there is no physical aspect to the overlapping in the BSM within the time window. But experiment shows your hypothesis does not occur. A swap is solely dependent on indistinguishable overlap in the beamsplitter, in addition to decoding the 2 bits to determine the Bell state. This is entirely in keeping with theory and experiment.

ii) You talk about "correlating" B and C (though it might be more helpful if you used the labeling of the papers, 1/2/3/4 rather than A/B/C/D). I hate to be the one to break this to you, but there is absolutely NO correlation between measurement outcomes of A and B (or C and D) in relevant versions of these experiments. Why not, when A and B are in a known entangled state? Answer: A and B are normally never measured on the same basis, and neither are C and D. But they can be, and if they are, the result is in fact as you predict!

This is easily seen in the Ma reference I provided, see Figure 3 and check it closely. The a) side is where the experimenter decides to execute the swap, and the b) side is where the swap is not executed (due to induced distinguishability). When the choice of measurement basis is |𝐻〉/|𝑉〉 (i.e. the same for all 4 photons), there is strong correlation regardless of whether a swap is executed or not. Not surprising, it's exactly for the reason you mention. So we agree here.
But otherwise... no. If we instead choose to measure the outer photons (1 & 4) on the |𝑅〉/|𝐿〉 or |+〉/|−〉 bases while measuring the inner photons on the |𝐻〉/|𝑉〉 basis, you only get correlation when there is a swap. Read this past sentence again if you are not sure: because there is only a random relationship between the |𝐻〉/|𝑉〉 basis and the |𝑅〉/|𝐿〉 or |+〉/|−〉 bases in the initial entangled pairs. But there is perfect correlation (within experimental limits) in the final entangled pair IFF a swap is executed. You can easily see that there is NO correlation in the b) side [no swap] for the |𝑅〉/|𝐿〉 and |+〉/|−〉 results, completely different than for the a) side [swap executed].


4. There are interpretations that deny the physical nature of collapse at all levels. But the cited experiment(s) make clear that the overall statistics change according to the choice of the experimenter as to whether to execute a swap or not. This choice can be made after the fact (in the cited Ma delayed choice version) and nonlocally (more easily seen in Field test of entanglement swapping over 100-km optical fiber with independent 1-GHz-clock sequential time-bin entangled photon-pair sources).

---

If these experiments are not a direct demonstration of spatial nonlocality and/or a violation of temporal causality, it makes for a pretty good disguise. Combine this result with Bell's Theorem (and related experiments), and both locality and realism (noncontextuality) take a pretty good hit.

 

[DrChinese]

1. Or i was thinking in terms of 2 pairs or correlates keys. The bsm serves to find the new 2&3 pairs that "match" then we know corresponding 1&4 do.

2. The 2&3 interaction is physical and required to identify the 1&4 that are correlated. info from bsm are classically transmitted.

3. Hidden key pair does NOT predetermine the photons, the only define their relation. [You specify"2 pairs or correlates keys" corresponding to 1&2 and 3&4]

/Fredrik

1. This is a common flaw in understanding these experiments. All indistinguishable 2 & 3 pairs within the time window cause swaps. Distinguishable 2 & 3 pairs within the time window do not lead to a swap, even though the same "keys" are present and recorded.


2. Well, is it physical or not? If it is, then you are acknowledging nonlocal action. And then we are in agreement.

And why mention the fact that results from different components of a entangled quantum system must be reported and brought together with a classical signal? This is a circular comment with no significance. ANY possible experiment showing nonlocality inherently requires signaling to bring together recorded results. So what?


3. Guess what? ALL 1&2 pairs have the same fixed and known relationship. And ALL 3&4 pairs have the exact same fixed and known relationship. Because they are all created in the same Bell state. After a swap, and only after a swap, they find themselves in 1 of 4 Bell states that randomly appear.

But according to your thinking, I guess that means all 1&4 pairs must have the same fixed relationship since the initial pairs have known fixed relationships. And yet... only swapped pairs do. The 1&4 relationship is purely random otherwise.

 

[Fra]

Perhaps "match" was the wrong word but..

All indistinguishable 2 & 3 pairs within the time window cause swaps.

Yes, this is what I meant by "match". Thus corresponding 1&4 are entangled; NOT the others.

2. Well, is it physical or not? If it is, then you are acknowledging nonlocal action. And then we are in agreement.

Well if you mean the bell meaning of "nonlocal" then sure it's "nonlocal". But this is not a problem.

And why mention the fact that results from different components of a entangled quantum system must be reported and brought together with a classical signal?

Without the "match" signal; you can't select the entangled 1&4 pairs.

3. Guess what? ALL 1&2 pairs have the same fixed and known relationship. And ALL 3&4 pairs have the exact same fixed and known relationship. Because they are all created in the same Bell state. After a swap, and only after a swap, they find themselves in 1 of 4 Bell states that randomly appear.

Yes, but the problem is to find the 2&3 pairs that by chance are related. All of those are NOT related; thus we need the BSM results to "pick" the corresponding 1&4.

But according to your thinking, I guess that means all 1&4 pairs must have the same fixed relationship since the initial pairs have known fixed relationships.

No, not sure how you conclude that I think this?

And yet... only swapped pairs do. The 1&4 relationship is purely random otherwise.

Agreed.

/Fredrik

 

[DrChinese]

1. Well if you mean the bell meaning of "nonlocal" then sure it's "nonlocal". But this is not a problem.

2. Without the "match" signal; you can't select the entangled 1&4 pairs.

3. Yes, but the problem is to find the 2&3 pairs that by chance are related. All of those are NOT related; thus we need the BSM results to "pick" the corresponding 1&4.

4. No, not sure how you conclude that I think this? [DrC: "But according to your thinking, I guess that means all 1&4 pairs must have the same fixed relationship since the initial pairs have known fixed relationships."]

5. Agreed.

/Fredrik

1. The question I asked was: "Is it physical or not?" If you believe it is physical, then you accept there is action at a distance.

The other option is to deny physical action or effects that transcend the usual Einsteinian limits. Instead, the BSM becomes an act of discovery, and does not in any way affect the compiled results by its presence or absence.


2. Without a laser or a photon detector, you also can't perform the experiment. Yes, you need the results of 4 photon measurement, specifically a four-fold coincidence. So what?


3. Not sure what you are saying here. The experiment includes all four-fold coincidences, and they are all factored in to the results.


4. Because you imply that there is some "hidden" relationship between the 1&2 and 3&4 streams that can be "unlocked" or otherwise discovered. There is no such initial relationship, and there is no experimental evidence (or theory for that matter) that there is. The relationship between the 1 & 4 pairs must be created by an action, it cannot be "revealed" by any known method otherwise.


5. Yay!

---

I'm not specifically trying to trash anyone's preferred Interpretation. I ask only that these swapping experiments be compared to the assumption(s) inherent in an Interpretation. Every single Interpretation includes at least 1 additional assumption over canonical QM*. If that assumption cannot withstand/survive the evidence provided by experiment, then perhaps a re-evaluation of that Interpretation is needed. I personally have no idea how nature performs these incredible feats. But I want to know the boundaries provided by these incredible tests.



* For example: the Bohmian interpretation assumes there is an exact position for quantum particles at all times, and although unknowable it is capable of instantaneous action at a distance. MWI assumes there are many worlds that are forever hidden from our perception. A lot of interpretations deny hidden variables but also deny action at a distance. Some feature retrocausal action/communication. Etc.

 

[PeterDonis]

the problem is to find the 2&3 pairs that by chance are related

It's not by chance. Whether or not photons 2&3 create a swap at the BSM is under the experimenter's control. The experimenter cannot control which of the four possible swap operations takes place, but they can control whether some swap operation takes place, or none does.

 

[iste]

1. Collapse may not be physical, but it may be proven so in experiments. Interpretations don't prove this one way or another, even those that appear otherwise consistent.

Yes, I just made that statement because sometimes you put forward arguments that imply that seem to imply that collapse must be physically real.

2. No, we initially have 2 separate and independent systems. They have no connection or correlation to each other whatsoever, on any polarization measurement basis. They need not be local to each other.

Yes, this is what I said. When I said "respectively", I meant that each entangled system is initiated by its own initial local interaction which is separate from the other system's initial local.

Yes, 1&2 are correlated. And yes, 3&4 are correlated. But that correlation is quantum

Yes, I said this. I said the entanglement part is quantum. The entanglement part corresponds to what you are talking about in the rest of the paragraph I took your quote from.

i) We get the values for Bit 1 and Bit 2 regardless of whether the 2 & 3 photons are distinguishable or not. But if they are distinguishable, there is no swap. According to your thinking, they should swap - regardless of overlap in the BSM - as long as we have the key.

No, I acknowledged that distinguishability as a requisite for the Bell state and suggested that this is reflected in Barandes' model. I quote myself:

"In Barandes' formulation, an intrusive interaction with an entangled state would cause it to lose its correlation and coherence, due to something called a division event. Presumably, distinguishability for the Bell state is like that, like the way a measurement causes decoherence or which-way information ruins interference. Those are all covered by division events which would ruin the kind of entangled states, [ruin] the kind of correlations that allow the conditioning exercise I have just given."

A swap is solely dependent on indistinguishable overlap in the beamsplitter

This is what I imply in the part I quoted of myself - indistinguishability comes with the coherent Bell state and similar appears to be the case in the Barandes model where entangled states lose their correlations and lose their nonseparability when subject to division events.

ii) You talk about "correlating" B and C (though it might be more helpful if you used the labeling of the papers, 1/2/3/4 rather than A/B/C/D) [... etc, next two paragraphs.]

What you say here doesn't seem problematic to me at all. It makes complete sense that the non-separable correlations would turn out like this because you are only correlating one part of the systems 2 & 3, so you wouldn't expect to get correlations for the additional parts of the systems in 1 & 4. The entangled state gets you correlations for more than one part and so you get more than just one set of correlations in 1 & 4. Coherence and indistinguishability may be requisites for the totality of the correlations, and those kinds of constructs feature in Barandes' model. In principle, I don't see why Barandes' model can't explicitly feature the distinction between separable and non-separable correlations like in this paper since non-separable correlations already explicitly feature in the Barandes model. The non-separable correlations and notion of division events applied to them also would explain why the entanglement swaps in the Barandes' picture so I guess the swapping actually does require quantum parts.

So maybe the swapping part is a bit more quantum than I seemed to imply, but it the correlations in 1 & 4 I still think are jist from transitivity, just transitivity with the right kind of correlations for 2 & 3. Obviously, the description I made, which I will repeat:

"You can correlate B and C by simply conditioning on two of their respective outcomes. If you do not condition on two of their respective outcomes, there will clearly be no correlation if all the outcomes occur because clearly all of the pairs - B1C1, B1C2, B2C1, B2C2 - can occur equally. If outcomes from A and D have a one-to-one mapping with those of B and C, then there will clearly be no correlation between A and D either because there is none between B and C. But if you condition on B1C1, you are going to get the same pair of outcomes for A and D all the time so they are now correlated."

... is overly simplistic but I don't see why it couldn't be extended to a case with more multi-faceted correlations in different bases like in the paper you cite.

So, I would argue that the mechanism still is at its core about a transitivity of correlations, which isn't especially quantum albeit in a quantum setting. Now it is more explicitly something like: if you want quantum correlations between 1 & 4, you need quantum correlations between 2 & 3. But this is still at heart about the transitivity of correlations from within each pair of 1 & 2, 2 & 3 and 3 & 4.

4. There are interpretations that deny the physical nature of collapse at all levels. But the cited experiment(s) make clear that the overall statistics change according to the choice of the experimenter as to whether to execute a swap or not. This choice can be made after the fact (in the cited Ma delayed choice version) and nonlocally (more easily seen in Field test of entanglement swapping over 100-km optical fiber with independent 1-GHz-clock sequential time-bin entangled photon-pair sources).

Yes, these are consistent with a statistical conditioning account because conditioning is effectively what the Bell state measurement is and this conditioning is required for the swapping correlations. Time would be irrelevant when it comes to statistical conditioning. Non-separable measurements might be seen as another form of conditioning, so you're effectively just choosing different forms of statistical conditioning, one more multi-faceted than the other.

The Ma paper actually supports this I think in a section near the end in the supplementary information for the "Bipartite State Analyzer..." section (and the authors appear to endorse a statistical conditioning kind of view too).

"When Victor performs a BSM, photons 1 and 4 are only entangled if there exists the information necessary for Victor to specify into which subensembles the data are to be sorted. In our case the subensembles correspond to |Φ−〉23 or |Φ+〉23. Without the ability for this specification, he would have to assign a mixture of these two Bell states to his output state which is separable, and thus he could not correctly sort Alice's and Bob's data into subensembles. This is confirmed by evaluating the experimental data obtained in a BSM but without discriminating between |Φ−〉23 and |Φ+〉23. Then there exists a correlation only in the |𝐻〉/|𝑉〉 basis (0.55 ± 0.06) and no correlations in the |+〉/|−〉 (0.02 ± 0.05) and |𝑅〉/|𝐿〉 (0.01 ± 0.05) bases, similar to the situation when Victor performs a separable-state measurement."

Suggesting that the Bell state results can plausibly be found as subsets of separable subensemble results.

 

[Fra]

I think are making some small incremental progress to understanding each other... But I see again that some of this is terminology; we loose focus and keep getting stuck on terms.

I will omitt commenting on what i think are related issue, to keep the discussion from diverging.

1. The question I asked was: "Is it physical or not?" If you believe it is physical, then you accept there is action at a distance.

I consider the interactions and readout of the 2&3 BSM to be physical interaction - meaning it's part of the "quntum system" - not just happning on the "observer side".

The measurements of ALL 1&4 correlations are also physical interactions, but the process of post-selecting the records that correspond to the BSM-tag that makes 1&4 correlate, is not a "physical interaction", as it is simply a "processing" of records on the observer side. But communicating the BSM-data to the processing site is a physical process! This is why time ordering is irrelevant; but no sorted results appears until the tags are communicated.

It is also why there is no need to action at a disstance, ie signal nonlocality.

But if you speak about action at a distance as "bell nonlocality" then yes its there, but this is not a problem; this is why i don't use this term; as it SOUNDS like a problem; but it's not. So for me it adds confusion to discussion imo.

Bells notion locality and realism are all blurred up in this ansatz - which i think is part of the problem.

4. Because you imply that there is some "hidden" relationship between the 1&2 and 3&4 streams that can be "unlocked" or otherwise discovered. There is no such initial relationship, and there is no experimental evidence (or theory for that matter) that there is. The relationship between the 1 & 4 pairs must be created by an action, it cannot be "revealed" by any known method otherwise.

100% agreed!

The problem is that i do not see; why you infer this from what I write?

Ineed the entangled STREAM of remote entangled pairs; after post-selection; are "engineered". It is a choice and combination of technology.

But why would this contradict what I said?

I said the relationship exists between 1&2, and between 3&4. And to MAKE/CREATE the correlated 1&4 pair; the BSM data and 2&3 interactions are required.

There is no "physical connection" between the two single photons/particles 1&4. Thus no need to action at distance. The CORRELATION of the pairs from the filtered STREAM, are engineered created by the choice of an experimenter, yes.

Do we agree here? If not, why?

If so, then what is our disagreement about?

5. Yay!

Yes progress :)

---

I'm not specifically trying to trash anyone's preferred Interpretation. I ask only that these swapping experiments be compared to the assumption(s) inherent in an Interpretation.

I fully agree that any interpretation must be able to conceptually "handle" experiment. And I do not see a problem with this; and the "escape" boils down to the problem in bells ansatz that baranders points out.

I think that not all HV theory "that explain correlation", satisfy premises of bells theorem. And noone here fleshed out a mature theory, but some "interpretations" holds the possibility for such theories, and this is the premise for my argumenation.

The problem is not lack of determinism, the problem is to understand the nature of causal relations between parts of the system. How can the "interference" patterns be while we do have the correlations? QM describes it, but are we conceptually satisifed with its explanatory value?

/Fredrik

 

[Fra]

It's not by chance. Whether or not photons 2&3 create a swap at the BSM is under the experimenter's control. The experimenter cannot control which of the four possible swap operations takes place, but they can control whether some swap operation takes place, or none does.

Fully agreed. By chance i referred to this.

/Fredrik

 

[DrChinese]

1. Yes, I just made that statement because sometimes you put forward arguments that imply that seem to imply that collapse must be physically real.

2. Yes, this is what I said. When I said "respectively", I meant that each entangled system is initiated by its own initial local interaction which is separate from the other system's initial local.

3. No, I acknowledged that distinguishability as a requisite for the Bell state and suggested that this is reflected in Barandes' model. I quote myself:

"In Barandes' formulation, an intrusive interaction with an entangled state would cause it to lose its correlation and coherence, due to something called a division event. Presumably, distinguishability for the Bell state is like that, like the way a measurement causes decoherence or which-way information ruins interference. Those are all covered by division events which would ruin the kind of entangled states, [ruin] the kind of correlations that allow the conditioning exercise I have just given."

This is what I imply in the part I quoted of myself - indistinguishability comes with the coherent Bell state and similar appears to be the case in the Barandes model where entangled states lose their correlations and lose their nonseparability when subject to division events.

4. ... "You can correlate B and C by simply conditioning on two of their respective outcomes. If you do not condition on two of their respective outcomes, there will clearly be no correlation if all the outcomes occur because clearly all of the pairs - B1C1, B1C2, B2C1, B2C2 - can occur equally. If outcomes from A and D have a one-to-one mapping with those of B and C, then there will clearly be no correlation between A and D either because there is none between B and C. But if you condition on B1C1, you are going to get the same pair of outcomes for A and D all the time so they are now correlated."
...

So, I would argue that the mechanism still is at its core about a transitivity of correlations, which isn't especially quantum albeit in a quantum setting. Now it is more explicitly something like: if you want quantum correlations between 1 & 4, you need quantum correlations between 2 & 3. But this is still at heart about the transitivity of correlations from within each pair of 1 & 2, 2 & 3 and 3 & 4.

Yes, these are consistent with a statistical conditioning account because conditioning is effectively what the Bell state measurement is and this conditioning is required for the swapping correlations. Time would be irrelevant when it comes to statistical conditioning. Non-separable measurements might be seen as another form of conditioning, so you're effectively just choosing different forms of statistical conditioning, one more multi-faceted than the other.

5. The Ma paper actually supports this I think in a section near the end in the supplementary information for the "Bipartite State Analyzer..." section (and the authors appear to endorse a statistical conditioning kind of view too).

"When Victor performs a BSM, photons 1 and 4 are only entangled if there exists the information necessary for Victor to specify into which subensembles the data are to be sorted. In our case the subensembles correspond to |Φ−〉23 or |Φ+〉23. Without the ability for this specification, he would have to assign a mixture of these two Bell states to his output state which is separable, and thus he could not correctly sort Alice's and Bob's data into subensembles. This is confirmed by evaluating the experimental data obtained in a BSM but without discriminating between |Φ−〉23 and |Φ+〉23. Then there exists a correlation only in the |𝐻〉/|𝑉〉 basis (0.55 ± 0.06) and no correlations in the |+〉/|−〉 (0.02 ± 0.05) and |𝑅〉/|𝐿〉 (0.01 ± 0.05) bases, similar to the situation when Victor performs a separable-state measurement."

Suggesting that the Bell state results can plausibly be found as subsets of separable subensemble results.

1. Yes, I think that experiment shows that what happens at the BSM is a form of physical action. I find it difficult to see any way out of that conclusion. Collapse in that view occurs at different times and places, and as it cannot be said to occur at the locations of the individual component measurements by Alice, Bob, Victor, etc.


2. Agreed, but not sure why you don't simply call them "distant sources" or similar. Using the word "local" tends to mislead.


3. I know you meant "indistinguishability" rather than "distinguishability".

But what you are saying is again factually incorrect. Prior to the action at the BSM, the 2 and 3 photons are distinguishable and always have been. Their interaction at the BSM makes them indistinguishable. There is no point where that are "made" to be coherent. The sources merely start them off in phase, a basic requirement no different than requiring them to be of the same wavelength. So your statement that an "intrusive interaction with an entangled state would cause it to lose its correlation" never happens and has no relevance here.


4. This too is factually incorrect. There is no correlating whatsoever occurring at the BSM. When a swap occurs, the beamsplitter causes the 2 & 3 photons to become indistinguishable - we agree this is a requirement for a swap. Next, the polarizers tell us whether the 2 & 3 photons are orthogonal and whether they exited the beam splitter from the same or different ports. These are the 2 bits that tell us which of the 4 Bell State will result. This in no way amounts to correlating anything.

If the 2 & 3 photons do not overlap (indistinguishably), no swap occurs. The B1C1 conditioning you mention:

a) Can be therefore be discovered regardless of whether that overlap occurs, but the swap only occurs when there is overlap; that needs addressing.
b) There is never, I repeat never a correlation between the measurement outcome of photon 1 and the measurement outcome for photon 2 in the cases where they are measured on different bases. No correlation! That is true regardless of whether a swap occurs, since those bases do not commute. This is basic, and has nothing to do with swapping.


5. You misunderstand what the authors are saying. What they mean: If you ignore the 2&3 outcomes, there is no particular correlation between 1&4 after a swap. You need the BSM result to know which Bell state occurs.

So no, there are no Bell states in separable situations.

 

[DrChinese]

1. The measurements of ALL 1&4 correlations are also physical interactions, but the process of post-selecting the records that correspond to the BSM-tag that makes 1&4 correlate, is not a "physical interaction", as it is simply a "processing" of records on the observer side.

2. But communicating the BSM-data to the processing site is a physical process! This is why time ordering is irrelevant; but no sorted results appears until the tags are communicated.
... Indeed the entangled STREAM of remote entangled pairs; after post-selection; are "engineered". It is a choice and combination of technology.
...I said the relationship exists between 1&2, and between 3&4. And to MAKE/CREATE the correlated 1&4 pair; the BSM data and 2&3 interactions are required.

3. There is no "physical connection" between the two single photons/particles 1&4. Thus no need to action at distance. The CORRELATION of the pairs from the filtered STREAM, are engineered created by the choice of an experimenter, yes.

Do we agree here? If not, why? ... If so, then what is our disagreement about?

1. There is no post-selection. All 4 fold detections with the time window are considered. There is a need to identify the resulting Bell state. But you can read the same 2&3 information regardless of whether a swap occurs. Using your logic, reading that information should "reveal" the 1&4 Bell state (which will be the same as the 2&3 Bell state). But no Bell state occurs UNLESS there is physical overlap of the 2&3 photons in the beam splitter. That is precisely the opposite of what you assert. IFF there is overlap does a swap occur. How do you explain that nugget?

2. It is trivially true that results of observations that are spacelike separated must be communicated to a central location for compilation, and that communication cannot be done faster than light.

So I ask this simple question: if there is no FTL signaling mechanism (which there is none such known), does that require also that there can be no FTL action at a distance? Because such a conclusion is not a logical deduction from the premise. Most scientists would say that it is at least possible that there could be FTL action even if there is no FTL signaling.

3. These statements talk past each other. I can't say precisely what the *direct* connection between 1&4 are after a swap, I only know they are in fact (as you state, and I agree) engineered by the physical choice of the experimenter. In my mind, all 4 photons are connected as a single quantum system after a swap. (Note that the term "after" does not really mean anything other than after the 4 fold detections occur.)

To summarize: There is no post selection occurring when there is a swap. The communication to bring all observed results together does occur at light speed or less, a fact that has no relevance whatsoever to this discussion as this is true of all experiments. A swap requires physical overlap of the distant delayed 2&3 photons in a beam splitter in addition to reading 2 bits of information to indicate the resulting Bell state.

 

[PeterDonis]

There is no post-selection.

Remember that you are (implicitly) adopting an interpretation in which the quantum state describes individual runs of the experiment. But there are also statistical interpretations (e.g., the interpretation used in Ballentine) in which the quantum state does not describe individual runs. In a statistical interpretation, "post-selection" is required, since you have to do it in order to get the statistics you need from the experimental data.

FTL action

This, again, implicitly assumes that the quantum state describes individual experimental runs, instead of just making predictions about the statistics over large enough numbers of runs.

 

[martinbn]

So I ask this simple question: if there is no FTL signaling mechanism (which there is none such known), does that require also that there can be no FTL action at a distance? Because such a conclusion is not a logical deduction from the premise. Most scientists would say that it is at least possible that there could be FTL action even if there is no FTL signaling.

The action you say might exist is not going to be FTL, but it will act back in time, the effect will be in the past of the cause. I am not sure if this makes any sense!

 

[JC_Silver]

I'm trying to follow the discussion, but as a high school teacher I'm not very well versed on QM, so forgive me if i get everything wrong.

I'm not sure I'm following why Barandes reaches his conclusion and why Dr Chinese's experiment shouldn't work, so if I may expose what I gathered so someone might clarify things.

I'm going to compare what I understand of QM and what I understand of Barande's paper and hope anyone helps me understand a little better.

Regular QM:
- Data from a particle can be described by a wavefunction, which is a function that behaves as a wave but has a complex number, it is this complex number's rotation that dictates the interference pattern seen in experiments with particles such as the electron
- We know that squareing the module of the wavefunction gives us the probability density of finding the particle in a certain position (or whatever it is that you are calculating for).

So the standard interpretation is that the wavefunction is a complex function that helps calculate the probability density of a state.

Barandes:
- The particle follows a completely random path, so random it cannot be known until it interacts with some system large enough to cause a "division event" when the particle takes a definite value. This can describe as a matrix. Non-Markovian processes forbid certain movements, so particles cannot take paths that would break the indivisibility of the process, these paths generate a interference pattern seen in experiments, but are made of pure probability rather than a wavefunction.
- The Born Rule comes from the math of non-Markovian stochastic processes naturaly and has no need for explanation.

So the "Barandian" interpretation would be that the wavefunction is a tool that helps calculate the probability density of a state.

-----

If we cannot know what the path of a particle is between division events, then is the particle even real? Or local? Why is this atributed to the particle?
The only conclusion I can see is that instead of the particle being in a "superposition" following a "wavefunction" until "collapse" is that the particle is in a "undescribable stochastic state" following "every path allowed by its stochastic nature" until "the division event".

Shouldn't this be just regular Copenhagen Interpretation with the "wave" from the wavefunction swapped for statistical processes? Am I completely lost?

Again, sorry for the probably stupid question.

 

[iste]

You said before that one could be agnostic given the state of the theory. Now you are eager to defend there is some interpretation.

There are layers to interpretation and this goes for all interpretations not just Barandes' formulation. For instance, Bohmian mechanics and Many Worlds are both distinct interpretations defined by certain interpretational features. But within these interpretations there are disagreements like whether the Pilot wave is a physical object or not, or you could be agnostic. Barandes can be agnostic about the underlying mechanisms behind an indivisible stochastic process whilst still endorsing the interpretation of an indivisible stochastic process where particles are always in one position at any time and the wavefunction is real; these latter two features are unambiguously in the area of interpretation but we can still be ambiguous about the specific mechanisms for why the system behaves as it does. You can be a Bohmian and believe particles go along definite determinate trajectories but be agnostic about whether the pilot wave is a physical object. You can advocate Many Worlds and be agnostic about their ontological status (bizarrely, but I read a philosophy of science encyclopedia article that suggests there are different ways of looking at Many Worlds). There are layers to all interpretations.

Bohmian mechanics also has one configuration at a time, superdeterminism does to.

Yes, but they are also unambiguously not identical to Barandes' interpretation mathematically or interpetationally.

I do not think it is worth it unless he provides more content about it.

That's completely fair.

I just disagree; even ignoring other stochastic formulations, the idea of particles being in definite positions still has attractive implications even if you don't know exactly why they behave the way they do - definite positions gets rid of all of the measurement problem issues and it complements the way we view the world, not just at an everyday level but in almost every other field in science outside of quantum theory.

The quantum interpretation problems came about because we couldn't figure out how to reconcile quantum theory with the pre-quantum, classical-like view of the world. So I would argue that if a formulation appears that reconciles the quantum and pre-quantum, it kind of takes away the very reason people started looking at different interpretations. Barandes' formulation claims it is at least consistent to describe quantum systems in a pre-quantum way, giving a route to the aforementioned reconciliation even if not all the interpretational details are there.

Obviously, you could argue that Bohmian mechanics does exactly the same thing as said here. But I think there is a subtle difference in that Bohmian mechanics effectively postulates quantum mechanics then adds particles separately on top while Barandes' idea can be seen as deriving quantum mechanics from the behavior of a more general kind of system that already has definite configurations and so does not require the additional Bohmian step.

He is proposing some stocastic force that permeates space and allows to bypass Bell's theorem. In most cases that would mean that his theory is nonlocal.

It depends what you mean by nonlocal I guess. Barandes' theory would presumably be Bell nonlocal if it establishes a direct correspondence between quantum and stochastic behavior. But quantum mechanics is already nonlocal anyway. Quantum mechanics is already weird so even if you cannot rule out a weird underlying explanation to Barandes' theory, I still don't think this is necessarily a strong argument against it since all quantum interpretations so far have been a bit weird in certain senses.

At least in the area of entanglement, I think the Barandes theory does suggest memory of local interactions (as opposed to overt communications like say in the Bohmian theory) is a sufficient mechanism which additionally explains many other aspects of the system's evolution, whether in the case of single or multiplr systems. So adding an explicit non-local communication to this doesn't seem parsimonious, at least in the case of entanglement, imo. Obviously this is all dependent on whether Barandes is actually correct or not.

People have tried to break the axioms of probability before to get quantum mechanics without amplitudes but that does not mean it is easier to comprehend or to interpret. Is Barandes proposal even a stochastic process at this point? We really need a simple example, hopefully a classical one.

I think there are plausible ways of viewing these things that aren't completely unintuitive. In post #282, I talk about indivisibility in terms of joint probability violations (and link pages from a source). The system has no unique joint probability distribution. But this doesn't necessarily mean that the statistics don't exist at all, just that they exist on different mutually exclusive probability spaces - mutually exclusive because of context-dependence. Various authors I have read talking about this include Dzhafarov, Khrennikov, Abramsky, Pitowsky and other people analyzing Fine's theorem.

I cite the second paper in #282 by Sokolovski as being a nice picture of indivisibility because it is clearly talking about the indivisibility of path integral "trajectories" in a similar kind of sense - trajectories evince different statistical probability spaces that are mutually incompatible and depend on measurement... or rather - to put it in Barandes' terms - depending on the statistical coupling to other systems, because the measurement disturbance properties in Barandes' formulation (and I believe in quantum mechanics generally) are properties any kind of physical system can have via interaction. Measurement is just a special case.

Can ignore below in red since based on misreading of passage on page 8 which I talk about in future post #380

You get something similar in the first paper I link in post #282, by Milz and Modi. I feel like I have to read between the lines a bit here but they state elsewhere that if processes are indivisible then they cannot be Markovian. So this should apply to indivisibility. (but non-Markovian processes can actually be divisible) ** (comment at end)

on page 8:

"It is always possible to write down a family of stochastic matrices for any non-Markovian process. Given the current state and history, we make use of the appropriate stochastic matrix to get the correct future state of the system. In general, for Markov order m, there are at most dm distinct histories, i.e., μ ∈ {0, ... , dm−1 − 1}; each such history (prior to the current outcome) then requires a distinct stochastic matrix to correctly predict future probabilities ... On the other hand, such a collection of stochastic matrices for a process of Markov order m could equivalently be combined into one d × dm matrix"

So it seems to be talking about non-Markovianity in the sense that different histories have different statistics - like different context-dependent probability spaces or ensembles, like in the Sokolovski paper. When you have a unique joint probability distribution satisfying consistency conditions in the form of Chapman-Kolmogorov Markov property, all the statistics can be fit onto one probability space (with unique joint probabilities) which determines the evolution through time in a history-independent way.

So maybe indivisibility just means the statistics are context-dependent on the history and can be disturbed or changed by interactions with other systems or contexts, but the statistics of the system's behavior do exist. This doesn't seem that bizarre or unintuitive to me even if the reason for the indivisibility may be unknown or even strange.


** (This following note might actually be irrelevant but I put in anyway because the idea of a non-markovian divisible process actually seems weirder to me than an indivisible process. Clearly the different kinds of stochastic systems that can exist are quite complicated.)

On page 15 of that Milz / Modi source they mention that divisible systems can be non-Markovian (my impression that such systems are a bit unusual as they have even been described as 'pathological' before, quoted in arXiv:2401.12715v1) and then give the following note as a citation [43]:

"Due to this inequivalence of divisibility and Markovianity, the maps t : s in Eq. (38) cannot always be considered as matrices containing conditional probabilities P(Rt|Rs)—as these conditional probabilities might depend on prior measurement outcomes—but rather as mapping from a probability distribution at time s to a probability distribution at time t [36,398]. This breakdown of interpretation also occurs in quantum mechanics [190]. In the Markovian case, t : s indeed contains conditional probabilities."

So to me they are suggesting this is kind of like a special case of a divisible system but where history-dependence again precludes context-independent probabilities unless the system is Markovian.

When they say quantum case in that quote I believe they are talking about quantum stochastic processes (based on the source) which is a bit of a different topic despite the name.

 

[DrChinese]

1. Remember that you are (implicitly) adopting an interpretation in which the quantum state describes individual runs of the experiment. But there are also statistical interpretations (e.g., the interpretation used in Ballentine) in which the quantum state does not describe individual runs. In a statistical interpretation, "post-selection" is required, since you have to do it in order to get the statistics you need from the experimental data.


2. This, again, implicitly assumes that the quantum state describes individual experimental runs, instead of just making predictions about the statistics over large enough numbers of runs.

1. I am reporting an experiment, not implicitly assuming an interpretation.

Please tell me what post selection is occurring. I place 2 orthogonal polarizers at the outputs of the beam splitter. Yes, I require 2 clicks within a time window. But each and every time these go off, the related 1 and 4 photons will certainly* indicate opposite results for any same polarization measurement (linear or circular).

2. It doesn’t matter whether the quantum state describes averages of runs, or individual runs. There is a certain prediction for each and every run, and for their average. Much like GHZ, if you follow the analogy.

The controlled variable is whether or not the 2&3 photons are allowed to interact at the beam splitter. The statistical interpretations predict there is no physicality to the overlap that changes the stats - it’s performed too late to make a difference. But there is a difference.


*within experimental efficiency of course

 

[DrChinese]

The action you say might exist is not going to be FTL, but it will act back in time, the effect will be in the past of the cause. I am not sure if this makes any sense!

These various experiments (together and individually) show both an FTL effect (no common past) and a retrocausal (delayed choice) effect. Some call it nonlocality in spacetime. You can label it as you like. All I am asserting is what it “looks like” and not what it actually is.

 

[gentzen]

If we cannot know what the path of a particle is between division events, then is the particle even real? Or local? Why is this atributed to the particle?
The only conclusion I can see is that instead of the particle being in a "superposition" following a "wavefunction" until "collapse" is that the particle is in a "undescribable stochastic state" following "every path allowed by its stochastic nature" until "the division event".

Shouldn't this be just regular Copenhagen Interpretation with the "wave" from the wavefunction swapped for statistical processes? Am I completely lost?

You understood pretty well why Barandes' formulation is not necessarily progress. However, he does untangle the "undescribable stochastic state" a bit: There are finitely many states, systems are composed by the cartesian product, a subsystem is one or more component of such a product, where the other components are "ignored". Note that this is different from a subset of the finitely many states, it is more like an equivalence relation of the set of states.

You can also see that the probabilities for the different states are supposed to vary continuously with the time parameter. The state itself on the other hand cannot really vary continuously, because the states are discrete. And you can see that a division event is assumed at t=0. You could probably avoid the need for it, if you would follow the strategy of the QBists more closely. But I guess Barandes prefers simplicity here.

In the end, you could simply diagonalize the density matrix at t=0, use a basis compatible with it (or a suitable coarse graining) at t=0, and then "rotate as fast as you wish" to the basis which you really want to work in (i.e. the one giving the "true ontology").

 

[PeterDonis]

Please tell me what post selection is occurring.

It occurs when you compare the experimental results to the predictions of QM. Those predictions are statistical. In order to do the statistics right, you have to separate the runs into buckets based on the BSM results, and test the observed statistics against QM predictions for each bucket separately.

There is a certain prediction for each and every run.

No, there isn't, because there is no way to predict which of the four possible Bell states will be produced by the BSM interaction for each individual run. You can only post-select the runs into buckets after the fact, when you know the BSM results. That's not prediction, that's post-selection.

 

[iste]

...

Apologies, just realized I completely misread the following section I quoted in my last post, shown below.

"It is always possible to write down a family of stochastic matrices for any non-Markovian process. Given the current state and history, we make use of the appropriate stochastic matrix to get the correct future state of the system. In general, for Markov order m, there are at most dm distinct histories, i.e., μ ∈ {0, ... , dm−1 − 1}; each such history (prior to the current outcome) then requires a distinct stochastic matrix to correctly predict future probabilities ... On the other hand, such a collection of stochastic matrices for a process of Markov order m could equivalently be combined into one d × dm matrix"

I had the impression it was saying you could only combine them as when you have a Markov process - but it clearly it isn't saying this! I guess, bias of reading what you want to read.

The context-dependence is still more or less still my interpretation on the indivisibility due to a joint probability violation regarding transition probabilities as expressed by Chapman-Kolmogorov equation, and this looks like what the Sokolovski paper I cited is talking about. I guess you will have to read and see for yourself in the Milz / Modi paper:

https://scholar.google.co.uk/schola...3713567&hl=en&as_sdt=0,5&as_ylo=2020&as_vis=1

Pages 12 - 15, 35 - 38. Divisibility specifically mentioned page 15.

Page 35 they say more explicitly:

"As we mention in Sec. III B, one of the fundamental theorems for the theory of classical stochastic processes, and the starting point of most books on them, is the Kolmogorov extension theorem. It hinges on the fact that joint probability distributions of a random variable S pertaining to a classical stochastic process satisfy consistency conditions amongst each other, like, for example, s2 P(S3, S2 = s2, S1) = P(S3, S1); a joint distribution on a set of times can always be obtained by marginalization from one on a larger set of times. Fundamentally, this is a requirement of noninvasiveness, as it implies that not performing a measurement at a time is the same as performing a measurement but forgetting the outcomes."

"Put differently, the validity of the KET is based on the fundamental assumption that the interrogation of a system does not, on average, influence its state. This assumption generally fails to hold in quantum mechanics," (page 13).

Clearly though it isn't just about measurements, like the context-dependence of the double-slit experiment behavior for instance. And the measurement-related disturbance is a special case of disturbances any physical system can impose by interaction.

 

[iste]

But what you are saying is again factually incorrect. Prior to the action at the BSM, the 2 and 3 photons are distinguishable and always have been. Their interaction at the BSM makes them indistinguishable

That's all that I implied. I'm just making a point that analogues of this kind of thing exist in the Barandes theory.

So your statement that an "intrusive interaction with an entangled state would cause it to lose its correlation" never happens and has no relevance here.

That's just me implying that if they were distinguishable they couldn't be correlated in terms of the Bell states.

This too is factually incorrect.

Its not if one thinks that a BSM is formally equivalent to statistical conditioninh. This is what the Barandes paper would imply and I have read other papers that would imply similarly. Its not factually incorrect. And I'm am pretty sure a Bell state implies a correlation so I would say it is correlating. You are correlating by picking out ensembles where you always have the same two outcomes.

a) Can be therefore be discovered regardless of whether that overlap occurs, but the swap only occurs when there is overlap; that needs addressing.

Again, Barandes' paper seems to account for this. It contains concepts of non-separable correlations, coherence, decoherence.

b) There is never, I repeat never a correlation between the measurement outcome of photon 1 and the measurement outcome for photon 2 in the cases where they are measured on different bases. No correlation! That is true regardless of whether a swap occurs, since those bases do not commute. This is basic, and has nothing to do with swapping.

But they would have a correlation in the same basis right? They are entangled... which isn't even something I am contesting.

5. You misunderstand what the authors are saying. What they mean: If you ignore the 2&3 outcomes, there is no particular correlation between 1&4 after a swap. You need the BSM result to know which Bell state occurs.

So no, there are no Bell states in separable situations.

"When Victor performs a BSM, photons 1 and 4 are only entangled if there exists the information necessary for Victor to specify into which subensembles the data are to be sorted. In our case the subensembles correspond to |Φ−〉23 or |Φ+〉23. Without the ability for this specification, he would have to assign a mixture of these two Bell states to his output state which is separable, and thus he could not correctly sort Alice's and Bob's data into subensembles. This is confirmed by evaluating the experimental data obtained in a BSM but without discriminating between |Φ−〉23 and |Φ+〉23. Then there exists a correlation only in the |𝐻〉/|𝑉〉 basis (0.55 ± 0.06) and no correlations in the |+〉/|−〉 (0.02 ± 0.05) and |𝑅〉/|𝐿〉 (0.01 ± 0.05) bases, similar to the situation when Victor performs a separable-state measurement."

They are saying that correlations in other bases exist in Bell state subensembles and they disappear when you mix up the data so you cannot tell the difference - but the relevant basis correlations are preserved. Similar, to my artificial example where the correlations exist if you isolate a subensemble and disappear when you mix up the data... only in this case in the experiment, it is a more complicated quantum correlation, not the simple classical one. But again, if you were to mix up all data from all Bell states then there would be no correlations anywhere. So its like levels of mixing. Also similar to how the interference fringes on idler photons in delayed choice eraser disappear when you just mix up the data. All that is happening is conditioning on different subensembles of data.

 

[iste]

If we cannot know what the path of a particle is between division events, then is the particle even real? Or local? Why is this atributed to the particle?
The only conclusion I can see is that instead of the particle being in a "superposition" following a "wavefunction" until "collapse" is that the particle is in a "undescribable stochastic state" following "every path allowed by its stochastic nature" until "the division event".

Shouldn't this be just regular Copenhagen Interpretation with the "wave" from the wavefunction swapped for statistical processes? Am I completely lost?

Again, sorry for the probably stupid question.

The particles always have definite configurations, i.e. definite positions, even during indivisibile evolutions. Indivisibility just means you cannot construct a certain kind of law that describes probabilities for moving between intermediate parts of a trajectory. But everywhere along the trajectory the system has definite outcomes, even when unmeasured. If the particle is one stochastic system, measurement just means adding an additional system that interacts with the particle. Both the particle and the measuring device are always in definite configurations.

 

[PeterDonis]

When Victor performs a BSM, photons 1 and 4 are only entangled if there exists the information necessary for Victor to specify into which subensembles the data are to be sorted.

Note that this description assumes a statistical interpretation, in which "sorting into subensembles" happens. An interpretation in which the quantum state describes individual runs of the experiment would describe this differently, along the lines of the descriptions @DrChinese has been giving.

 

[JC_Silver]

The particles always have definite configurations, i.e. definite positions, even during indivisibile evolutions. Indivisibility just means you cannot construct a certain kind of law that describes probabilities for moving between intermediate parts of a trajectory. But everywhere along the trajectory the system has definite outcomes, even when unmeasured. If the particle is one stochastic system, measurement just means adding an additional system that interacts with the particle. Both the particle and the measuring device are always in definite configurations.

Is that demanded by the math? Because I understand that this is the conclusion given on the paper, but when I look at the math used the only thing I gather is that in between division 1 and division 2 the particle is unknowable and trying to predict its properties gives us all of QM weirdness.

I don't see why a regular particle with definite positions wouldn't be able to be describe by regular statistics.

By saying it can't be known by any means tells me this is different from classical physics, the particle is doing something weird where classical statistics break down and we are not allowed to look.

The math only says that between two divisions the particle's states are fundamentally unknowable, that is the superposition.

This is what confuses me, to me the math doesn't seem to demand that the particle have definite values, it demands that its state be unknowable.

 

[Lord Jestocost]

There is no QM “weirdness”. Some people, even some physicists, are just completely "confused" because classical physics and its concepts fail to account for certain physical phenomena.

 

[jbergman]

Is that demanded by the math? Because I understand that this is the conclusion given on the paper, but when I look at the math used the only thing I gather is that in between division 1 and division 2 the particle is unknowable and trying to predict its properties gives us all of QM weirdness.

I don't see why a regular particle with definite positions wouldn't be able to be describe by regular statistics.

By saying it can't be known by any means tells me this is different from classical physics, the particle is doing something weird where classical statistics break down and we are not allowed to look.

The math only says that between two divisions the particle's states are fundamentally unknowable, that is the superposition.

This is what confuses me, to me the math doesn't seem to demand that the particle have definite values, it demands that its state be unknowable.

I agree. Which is why I made my emperor's new clothes comment vis a vis interpretations. This "stochastic approach" looks to me like nothing more than dressing up the typical QM evolution.

 

[iste]

Note that this description assumes a statistical interpretation, in which "sorting into subensembles" happens. An interpretation in which the quantum state describes individual runs of the experiment would describe this differently, along the lines of the descriptions @DrChinese has been giving.

I don't think it is. They are just describing their data.

 

[PeterDonis]

They are just describing their data.

They are describing their data in a particular way, in terms of "subensembles". The only reason for doing that is to perform statistical analysis. If we are discussing interpretations, such an analysis is only relevant for statistical interpretations.

 

[Fra]

So I ask this simple question: if there is no FTL signaling mechanism (which there is none such known), does that require also that there can be no FTL action at a distance? Because such a conclusion is not a logical deduction from the premise. Most scientists would say that it is at least possible that there could be FTL action even if there is no FTL signaling.

How do you define "FTL action"?

/Fredrik

 

[iste]

They are describing their data in a particular way, in terms of "subensembles". The only reason for doing that is to perform statistical analysis. If we are discussing interpretations, such an analysis is only relevant for statistical interpretations.

I don't follow. If interpretations are independent of the predictions of quantum mechanics then they should be independent of the statisticsl analyses experimenters perform.

 

[martinbn]

They are describing their data in a particular way, in terms of "subensembles". The only reason for doing that is to perform statistical analysis. If we are discussing interpretations, such an analysis is only relevant for statistical interpretations.

I think they only say to look at the sets of results of Alice and Bob, and the subsets of for which e specific outcome happened at Victor's measurment on photons 2&3. No interpretation is needed here.

 

[iste]

Is that demanded by the math? Because I understand that this is the conclusion given on the paper, but when I look at the math used the only thing I gather is that in between division 1 and division 2 the particle is unknowable and trying to predict its properties gives us all of QM weirdness.

Yes, it is implied by the fact that you are talking about a stochastic system. If you sample the system, it has definite configurations at every point in time. Furthermore, you can always assign probabilities to any point in time, the system having started at an initial time. What you cannot do is take the system moving from initial time to another point in time and then construct conditional probabilities regarding the intermediate times. Particles are in definite configurations between division events.

I don't see why a regular particle with definite positions wouldn't be able to be describe by regular statistics.

There is a distinction between the statistics that describe particle behavior and the fact that it exists in definite configurations. Similar to how there is a distinction between the probabilities you assign to dice rolls and the fact that you can roll dice and get a specific outcome. Superpositions reflect the statistical description and so there is nothing inconsistent about the particle being in definite configurations during superposition.

 

[PeterDonis]

If interpretations are independent of the predictions of quantum mechanics then they should be independent of the statisticsl analyses experimenters perform.

To the extent that the statistical analysis is to compare predictions with experiment, of course it is independent of any interpretation, yes.

But that very fact means that you cannot use anything about the above to argue for or against any particular interpretation. In the post of yours that I responded to in post #383, you were trying to use the statistical subensembles to argue against the interpretation @DrChinese is using. At least, that's the only reading of your post I can come up with that makes it relevant at all to this thread.

 

[pines-demon]

In the end, you could simply diagonalize the density matrix at t=0, use a basis compatible with it (or a suitable coarse graining) at t=0, and then "rotate as fast as you wish" to the basis which you really want to work in (i.e. the one giving the "true ontology").

What do you mean here?

 

[gentzen]

In the end, you could simply diagonalize the density matrix at t=0, ...

What do you mean here?

You mean on a high level? Basically the same as he in https://arxiv.org/abs/2309.03085 on page 2:

Without any important loss of generality, the set of times $\mathcal{T}$ will be assumed to include an element denoted by $0$ and called the initial time. It will be further assumed that at the initial time $0$, the time-dependent dynamical map $f_t$ trivializes to the identity map $\operatorname{id}_\mathcal{X}$ on $\mathcal{X}$:
$f_0 = \operatorname{id}_X ,\quad\text{ or }\quad f_0(i) = i\quad[\text{for all i }\in \mathcal{X}]. (6)$

I mean: you can finesse this point in various ways if you want. But it is more a question of aesthetics and simplicity than a real limitation of generality.

You want to have a time dependent quadratic stochastic matrix as your starting point. Using a division event at $t=0$ is a simple way to achieve this, therefore Barandes did this. You could also be "more general" and introduce a linear dependence on an abstract probability vector (of same dimension as the phase space), without a connection to any specific time $t_0$. But it might feel a bit more complicated and less well motivated. But you could motivate this abstract probability vector as the eigenvalues of the density matrix.

What I described was how to stay with the division event at $t=0$, and still use the eigenvalues of the density matrix as the abstract probability vector.

 

[iste]

To the extent that the statistical analysis is to compare predictions with experiment, of course it is independent of any interpretation, yes.

But that very fact means that you cannot use anything about the above to argue for or against any particular interpretation. In the post of yours that I responded to in post #383, you were trying to use the statistical subensembles to argue against the interpretation @DrChinese is using. At least, that's the only reading of your post I can come up with that makes it relevant at all to this thread.

I don't think a statistical conditioning interpetation of an experiment has to necessarily be tied to a specific interpretation: e.g. arguing against retrocausality in delayed choice.

https://scholar.google.co.uk/scholar?cluster=3918555048962557101&hl=en&as_sdt=0,5&as_vis=1

I was initially responding to a point where Dr. Chinese invoked delayed choice.

 

[PeterDonis]

I don't think a statistical conditioning interpetation of an experiment has to necessarily be tied to a specific interpretation

If "interpretation of an experiment" just means "comparing QM predictions to the results", then of course not.

But again, that means you can't use "statistical conditioning" to argue against any QM interpretation.

 

[JC_Silver]

Yes, it is implied by the fact that you are talking about a stochastic system. If you sample the system, it has definite configurations at every point in time. Furthermore, you can always assign probabilities to any point in time, the system having started at an initial time. What you cannot do is take the system moving from initial time to another point in time and then construct conditional probabilities regarding the intermediate times. Particles are in definite configurations between division events.



There is a distinction between the statistics that describe particle behavior and the fact that it exists in definite configurations. Similar to how there is a distinction between the probabilities you assign to dice rolls and the fact that you can roll dice and get a specific outcome. Superpositions reflect the statistical description and so there is nothing inconsistent about the particle being in definite configurations during superposition.

I'm not gonna lie, this is what gets me about the non-Markovian process, because the dice rolls isn't a non-Markovian process, it's a regular stochastic process. Since I'm not well versed in stochastic processes of any kind and I can't find good sources on non-Markovian processes online that are not by Barandes (as we used to say in the long past, my Google-fu isn't strong enough), I'm left not understanding exactly why the particle has definite positions between division events, because as far as I understand, the dice roll also doesn't exist between one roll and the next.

Again, sorry for bothering >.<

 

[DrChinese]

It occurs when you compare the experimental results to the predictions of QM. Those predictions are statistical. In order to do the statistics right, you have to separate the runs into buckets based on the BSM results, and test the observed statistics against QM predictions for each bucket separately.

No, there isn't, because there is no way to predict which of the four possible Bell states will be produced by the BSM interaction for each individual run. You can only post-select the runs into buckets after the fact, when you know the BSM results. That's not prediction, that's post-selection.

Yes, technically there is separation of the results into the 2 discernible buckets (Bell states) of 4 fold coincidences. Calling it "post-selection" may be correct at some level, but it is also misleading to readers. We combine the results of ALL 4 near-simultaneous detections per specification, which is normal for all experiments. Nothing is being "thrown away" because it doesn't fit our ideas of what is being studied. As I described, it is also possible to use polarizers instead of polarizing beam splitters; in that case, there is only a single bucket: would you call that "post-selection" too?

And calling the results in each bucket as being statistical is also misleading to some readers. The prediction is certain in each and every case. (And for this statement I am referring to the Ma experiment specifically, as it does not feature a CHSH-like inequality.)

 

[pines-demon]

I can't find good sources on non-Markovian processes online that are not by Barandes (as we used to say in the long past,

It is normal, Barandes himself says that there is almost no literature on non-Markovian processes, and most of the literature out there is about non-Markovian divisible processes, so Barandes stuff is very niche.