I What is the current perspective on quantum interpretation?

[stevendaryl]

Entanglement does not imply action at a distance, because local relativistic QFT is an example for a QT, which of course allows entanglement but is at the same time a theory where you have only local interactions by construction.

There are two aspects to quantum theory (including QFT). One is the recipe for the evolution of the quantum state (Schrodinger's equation for QM and the differential equations governing the evolution of the field operators in QFT). The second is the recipe that says that when one performs a measurement, one gets an eigenvalue of the corresponding operator, with a probability calculated from the quantum state.

The issue with nonlocality is about the second aspect. When you say "you have only local interactions by construction", you're not including the measurement process itself.

 

[stevendaryl]

I agree with this, but there is nothing in nature that behaves that way.

The point of the example is to show that local is not the same thing as lack of FTL communication. Local implies no FTL but the converse isn't true.

 

[stevendaryl]

Ok, I am late for the discussion but I am still confused. Following @stevendaryl by locality we mean that to understand what is going on in a small region we only need the information from near by. But how is the usual Alice and Bob scenario any different?

If Bob has already measured his particle's spin along the z-axis, then Alice's result for a measurement along the z-axis is already determined, even if she hasn't performed the measurement yet. It's not probabilistic in the sense of a stochastic process. It's still probabilistic in the sense of lack of knowledge---anybody who is unaware of Bob's result will assign a 50/50 subjective probability to Alice's result.

So the strangeness of EPR (and maybe QM in general) is how probability shifts from being intrinsic probability to being subjective probability. If you believe that intrinsic probability is an objective property, then such a shift seems to be a change in Alice's situation that happens FTL when Bob performs his measurement.

If you don't believe that probability is intrinsic, but is always subjective, then that seems to be suggesting either hidden variables or superdeterminism. If there was no change for Alice due to Bob's measurement, and after Bob's measurement it is 100% certain that Alice will measure spin-up, then it seems that it must have been 100% certain before Bob's measurement.

 

[vanhees71]

Yes, but the question is should one still say that space-like events can be cause/effect related? I am not convinced that EPR, Bell and all that implies that one should.

No, space-like events cannot be cause-effect related by construction (microcausality constraint on local operators).

 

[vanhees71]

The point of the example is to show that local is not the same thing as lack of FTL communication. Local implies no FTL but the converse isn't true.

You still did no clearly define, what you mean by "local". The usual definition is that the Hamilton density is a local operator and that all local operators commute with it at space-like distances (micro-causality constraint). This excludes the possibility of space-like separated cause-effect related events, i.e., it excludes faster-than-light signal propagation. Other conclusions follow from that: the unitarity of and the cluster-decomposition principle of the S-matrix, the CPT symmetry, and the usual spin-statistics relation.

 

[stevendaryl]

No, space-like events cannot be cause-effect related by construction (microcausality constraint on local operators).

You keep saying that, but it completely misses the point. The microcausality constraint is on the field operators. But as you keep saying yourself, the only meaning to quantum theory is the predictions it makes about measurement results. Measurement results produce a SINGLE outcome from a set of possible outcomes. That is not described by the evolution of the field operators.

 

[stevendaryl]

You still did no clearly define, what you mean by "local".

Yes, I did. Local means that the conditions in one region of space evolve over a period of time $\Delta t$ in a way that is independent of the evolution of conditions in regions of space that are farther than $c \Delta t$ away (once you take into account information in the common backwards lightcone). In the EPR experiment, that is not true.

Alice in one region is measuring the z-component of spin for one particle from an anti-correlated pair.

Far, far away, Bob is measuring the z-component for the other particle.

The evolution of those two regions is not independent, and the dependency is not removed by looking at information in the common backwards lightcone.

 

[PeterDonis]

The electromagnetic field is not frame dependent.

Its value at a particular point is not. But @stevendaryl meant by "state" a global set of values for the EM field "at an instant of time". That is not invariant because "an instant of time" is not; it's frame-dependent.

 

[PeterDonis]

The evolution of those two regions is not independent, and the dependency is not removed by looking at information in the common backwards lightcone.

This is just another way of saying "the correlations violate the Bell inequalities". Which is, IMO, a much better way of saying it, since it doesn't require any questionable notion of "evolution" or "state". It's just a testable statement about the correlations.

 

[stevendaryl]

This is just another way of saying "the correlations violate the Bell inequalities".

It's not another way of saying it. There is quite a lot of work needed to go from things being local in the sense that I'm talking about to Bell's inequalities being satisfied. But it is true that every local system in my sense obeys Bell's inequalites.

 

[vanhees71]

Yes, I did. Local means that the conditions in one region of space evolve over a period of time $\Delta t$ in a way that is independent of the evolution of conditions in regions of space that are farther than $c \Delta t$ away (once you take into account information in the common backwards lightcone). In the EPR experiment, that is not true.

Alice in one region is measuring the z-component of spin for one particle from an anti-correlated pair.

Far, far away, Bob is measuring the z-component for the other particle.

The evolution of those two regions is not independent, and the dependency is not removed by looking at information in the common backwards lightcone.

This is an empty definition in the case of quantum theory. What do you mean by "conditions in one region of space"? In quantum theory what evolves are probabilities (and quantities derived from them like expectation values).

Your concrete example of course makes sense, but what you describe there is no non-locality but inseparability. There are correlations between far-distant measurements described by entanglement, but there is no non-local causal effect from one local measurement at A on the measurement at B and vice versa, because the interactions between the photons with the detectors at the places A and B are local, i.e., they fulfill the microcausality constraint such that there cannot be any instantaneous influence between the detectors at the far distant places through the measured object (in this case the two photons).

 

[stevendaryl]

This is just another way of saying "the correlations violate the Bell inequalities". Which is, IMO, a much better way of saying it, since it doesn't require any questionable notion of "evolution" or "state". It's just a testable statement about the correlations.

I realize now that what I'm sketching out is basically what Bell laid out in his "Theory of Local Beables":

https://cds.cern.ch/record/980036/files/197508125.pdf

 

[Tendex]

The point of the example is to show that local is not the same thing as lack of FTL communication. Local implies no FTL but the converse isn't true.

It would help if you clarified in what sense you are talking about FTL, in the gallilean or classical pre-Einstein-Lorentz sense where it can still be causal since there was no speed limit or in the modern (Einstein relativistic) more fundamental sense.

 

[vanhees71]

I think we should discuss within relativistic theory only, and since the only really working relativistic QT is local relativistic QFT, maybe also restrict ourselves to this.

There is also of course no problem with actions at a distance in non-relativistic physics (to the contrary it's the usual model of interactions as Newton's theory of the gravitational interaction). There's of course also no microcausality condition in non-relativistic QFT. It couldn't even be sensibly defined there.

 

[Tendex]

I think so too. That's why I wanted for stevendaryl to clarify, since he himself made a classification of qm interpretations with 2 catrgories: wrong and nonsense. My point is that talking about FTL is mathematically sound but physically wrong with the galilean mindset, and pure nonsense with the modern fundamental mindset so maybe it's better to leave it out.

 

[stevendaryl]

It would help if you clarified in what sense you are talking about FTL, in the gallilean or classical pre-Einstein-Lorentz sense where it can still be causal since there was no speed limit or in the modern (Einstein relativistic) more fundamental sense.

The issue is whether QM obeys Einsteinian relativity, when one takes into account the measurement process. As I explained in a different post, quantum mechanics has two pieces: (1) the smooth deterministic evolution of the quantum state, and (2) the nondeterministic production of a measurement result. In QFT, it's pretty clear that (1) is relativistically invariant, but it's much less clear whether (2) is. According to some interpretations of QM, relativity is only approximately true, for macroscopic measurements, but fails at a microscopic level.

In any case, I don't agree with you that you need to assume Einsteinian or Galilean relativity in order to say what FTL means. It is enough that there is a single frame in which light travels at its characteristic speed. Then relative to that choice of a frame, you can ask whether or not there are FTL influences.

 

[Tendex]

The issue is whether QM obeys Einsteinian relativity.

That's not so hard, Non-relativistic QM doesn't obey it, while its generalization RQFT does. This doesn't necessarily mean the whole construct of QFT as a mathematical theory is correct, if that's what you are suggesting.

In any case, I don't agree with you that you need to assume Einsteinian or Galilean relativity in order to say what FTL means. It is enough that there is a single frame in which light travels at its characteristic speed. Then relative to that choice of a frame, you can ask whether or not there are FTL influences.

I didn't talk about assuming anything, I asked you a simple question that you seem to refuse to answer and that's your privilege. But when you say there is a single frame in which light travel at it's characteristic speed and you mention FTL in relation with it, it is relevant to know if you are referring to frames in Galilean or Einsteinian relativity because it is not the same thing and it helps to clarify which to judge the argument adequately.

 

[stevendaryl]

That's not so hard, Non-relativistic QM doesn't obey it, while its generalization RQFT does.

The latter is just not clear.

 

[stevendaryl]

I didn't talk about assuming anything, I asked you a simple question that you seem to refuse to answer and that's your privilege.

I don't understand what your question means. Whether something is FTL doesn't depend on whether we're talking galilean or einsteinian relativity. It doesn't depend on relativity (galilean or einsteinian), since I'm only talking about a single frame. Relativity is about the relationship between two DIFFERENT frames. Within a single frame, relativity is not relevant.

Maybe it would be better to say "coordinate system" rather than frame?

 

[Tendex]

The latter is just not clear.

As I said its formulation has serious mathematical issues, but the "FTL influence" is a physical problem it does not have, it makes perfectly causal predictions.

 

[Tendex]

I don't understand what your question means. Whether something is FTL doesn't depend on whether we're talking galilean or einsteinian relativity. It doesn't depend on relativity (galilean or einsteinian), since I'm only talking about a single frame. Relativity is about the relationship between two DIFFERENT frames. Within a single frame, relativity is not relevant.

Maybe it would be better to say "coordinate system" rather than frame?

FTL might not depend on it, that's what I'm saying in a way, but you must have some idea of what you intend for your single frame when it relates with other frame. Single frame physics is not really so useful as in NRQM for quantum fields theories.

 

[stevendaryl]

As I said its formulation has serious mathematical issues, but the "FTL influence" is a physical problem it does not have, it makes perfectly causal predictions.

Oh, my god. I HATE Physics Forums discussions.

I agree that

1. In relativistic quantum field theory, the quantum state evolves according to local equations.
2. RQFT does not allow FTL signal transmission.

The issue is whether 1&2 imply that it is local. 1 by itself is not enough, because the smooth evolution does not include the measurement process by which a single outcome is selected out of a set of several possibilities. 2 is not enough, either, because although Local implies no FTL, no FTL does not imply local.

 

[atyy]

You still did no clearly define, what you mean by "local". The usual definition is that the Hamilton density is a local operator and that all local operators commute with it at space-like distances (micro-causality constraint). This excludes the possibility of space-like separated cause-effect related events, i.e., it excludes faster-than-light signal propagation. Other conclusions follow from that: the unitarity of and the cluster-decomposition principle of the S-matrix, the CPT symmetry, and the usual spin-statistics relation.

If one uses your definition of "local" in post #90, then relativistic quantum field theory is not local. So we don't have to use another person's definition, we just use yours.

If you accept that collapse interpreted as physical is "local" since it does not allow "faster-than-light signal propagation", then relativistic quantum field theory is "local".

If you reject that collapse interpreted as physical is "local" since allows "faster-than-light signal propagation", then relativistic quantum field theory is "nonlocal".

 

[stevendaryl]

If one uses your definition of "local" in post #90, then relativistic quantum field theory is not local. So we don't have to use another person's definition, we just use yours.

If you accept that collapse interpreted as physical is "local" since it does not allow "faster-than-light signal propagation", then relativistic quantum field theory is "local".

If you reject that collapse interpreted as physical is "local" since allows "faster-than-light signal propagation", then relativistic quantum field theory is "nonlocal".

Truly, I don't understand how what some people say about "collapse" isn't complete nonsense.

In an EPR experiment, with anti-correlated spin-1/2 particles, Alice and Bob agree to measure spin along the z-axis. Immediately before Alice's measurement she would say that there is a 50/50 chance that Bob will measure spin-up. Immediately after Alice's measurement, if she gets spin-up, the probability that Bob will measure spin-up changes to 0. How can you account for that sudden change? It seems to me that there are several possibilities:

1. Alice's measurement caused a physical change that affected Bob's measurement. Well, we have to rule this out because it involves FTL influences.
2. Alice's measurement, like a classical measurement, only changed Alice's knowledge about Bob. It didn't change anything physical on Bob's end. But let's think about this. If after Alice's measurement, she knows that it is 100% certain that Bob will measure spin-down, and if Alice's measurement didn't change anything about Bob, then it means that before Alice's measurement, it must have already been true that Bob would measure spin-down (only Alice didn't know it). Well, we have to reject this, too, because it involves hidden variables (the value of Bob's measurement before he makes it).
3. Umm, maybe there are two versions of the universe: One in which Alice measures spin-up and Bob measures spin-down, and another version in which Alice measures spin-down and Bob measures spin-up. Well, this would be a many-worlds interpretation, which is fanciful nonsense, and should be ruled out.

Any of those three seem like weird, unappealing choices: Unobservable FTL influences, unobservable hidden variables, unobservable alternate universes. But to reject all three seems like nonsense. Or at best, it amounts to saying: "I have no idea".

 

[RUTA]

Here is a paper that might be of interest:

M. Schlosshauer, J. Kofler, and A. Zeilinger, ``A snapshot of foundational attitudes toward quantum mechanics,'' Studies in the History and Philosophy of Modern Physics 44, 222–230 (2013).

https://arxiv.org/abs/1301.1069

 

[RUTA]

Any of those three seem like weird, unappealing choices: Unobservable FTL influences, unobservable hidden variables, unobservable alternate universes. But to reject all three seems like nonsense. Or at best, it amounts to saying: "I have no idea".

Most physicists accept that time dilation and length contraction follow from the relativity principle applied to the measurement of c, even though we have no constructive counterpart (causal mechanism) for that principle account. Indeed, this is the standard presentation of special relativity in the introductory physics textbooks. The mystery of quantum entanglement can be resolved in similar fashion via the relativity principle applied to the measurement of h. Would that principle account of quantum entanglement suffice here, or would you still require a constructive counterpart?

 

[dextercioby]

The only sensible way I know of, in which relativistic quantum field theory is defined is through Wightman's axioms which, unlike the axioms of Dirac-von Neumann for non-relativistic QM, are silent about observables, about measurements or whether measurements on a QFT observable are in agreement with Einstein' SR. And this set of axioms only work(s) for the free spinless scalar (electrically charged or not) fields. Interactions between these fields (or self-interactions) and other Poincare covariant ones, not to mention interactions with any measurement device, are already outside the axioms.

So the whole discussion of interpretation issues in SR context seems futile to me.

P.S. Sure, there is the Haag-Kastler operator algebra axiomatization, but I do not know if it is superior to Wightman's from interpretation-relevant considerations (such as measurements and interactions between measured and measurer).

 

[stevendaryl]

The only sensible way I know of, in which relativistic quantum field theory is defined is through Wightman's axioms which, unlike the axioms of Dirac-von Neumann for non-relativistic QM, are silent about observables, about measurements or whether measurements on a QFT observable are in agreement with Einstein' SR. And this set of axioms only work(s) for the free spinless scalar (electrically charged or not) fields. Interactions between these fields (or self-interactions) and other Poincare covariant ones, not to mention interactions with any measurement device, are already outside the axioms.

So the whole discussion of interpretation issues in SR context seems futile to me.

P.S. Sure, there is the Haag-Kastler operator algebra axiomatization, but I do not know if it is superior to Wightman's from interpretation-relevant considerations (such as measurements and interactions between measured and measurer).

I thought that in QFT that the measurements and observables at least included detection of particles. People certainly use QFT to compute things like cross sections, which are empirically measurable, right?

 

[atyy]

The only sensible way I know of, in which relativistic quantum field theory is defined is through Wightman's axioms which, unlike the axioms of Dirac-von Neumann for non-relativistic QM, are silent about observables, about measurements or whether measurements on a QFT observable are in agreement with Einstein' SR. And this set of axioms only work(s) for the free spinless scalar (electrically charged or not) fields. Interactions between these fields (or self-interactions) and other Poincare covariant ones, not to mention interactions with any measurement device, are already outside the axioms.

So the whole discussion of interpretation issues in SR context seems futile to me.

P.S. Sure, there is the Haag-Kastler operator algebra axiomatization, but I do not know if it is superior to Wightman's from interpretation-relevant considerations (such as measurements and interactions between measured and measurer).

I think we can consider relativity in 2+1D where there are rigorously constructed interacting relativistic field theories. Haag's own book (Local Quantum Physics, 1996) does recognize the discontinuous state update (p301), at least informally, and discusses interpretation in a way that can be understood if the principles of non-relativistic QM also apply to relativistic QFT.

 

[Fra]

Truly, I don't understand how what some people say about "collapse" isn't complete nonsense.

In an EPR experiment, with anti-correlated spin-1/2 particles, Alice and Bob agree to measure spin along the z-axis. Immediately before Alice's measurement she would say that there is a 50/50 chance that Bob will measure spin-up. Immediately after Alice's measurement, if she gets spin-up, the probability that Bob will measure spin-up changes to 0. How can you account for that sudden change? It seems to me that there are several possibilities:

1. Alice's measurement caused a physical change that affected Bob's measurement. Well, we have to rule this out because it involves FTL influences.
2. Alice's measurement, like a classical measurement, only changed Alice's knowledge about Bob. It didn't change anything physical on Bob's end. But let's think about this. If after Alice's measurement, she knows that it is 100% certain that Bob will measure spin-down, and if Alice's measurement didn't change anything about Bob, then it means that before Alice's measurement, it must have already been true that Bob would measure spin-down (only Alice didn't know it). Well, we have to reject this, too, because it involves hidden variables (the value of Bob's measurement before he makes it).
3. Umm, maybe there are two versions of the universe: One in which Alice measures spin-up and Bob measures spin-down, and another version in which Alice measures spin-down and Bob measures spin-up. Well, this would be a many-worlds interpretation, which is fanciful nonsense, and should be ruled out.

Any of those three seem like weird, unappealing choices: Unobservable FTL influences, unobservable hidden variables, unobservable alternate universes. But to reject all three seems like nonsense. Or at best, it amounts to saying: "I have no idea".

From my perspective, 2 is possibly reasonable and appealing. I will try to just explain how this is not a problem for Bell.

(To understand my line of reasoning, you need some Qbism on steroids)

Alice's measurement, like a classical measurement, only changed Alice's knowledge about Bob. It didn't change anything physical on Bob's end. But let's think about this. If after Alice's measurement, she knows that it is 100% certain that Bob will measure spin-down, and if Alice's measurement didn't change anything about Bob, then it means that before Alice's measurement, it must have already been true that Bob would measure spin-down (only Alice didn't know it). Well, we have to reject this, too, because it involves hidden variables (the value of Bob's measurement before he makes it).

To understand why you can have hidden variables, and still escape Bells theoreem. There is, and always was a problem with the premise of "realism" and $\lambda$ in Bells theorem.

Bells idea of hidden variable
$\lambda$ - suppsedly responsible for the predeterminded correlation - is "real" and in principle "known" to whole measurement device and environment, except the ignorant observer (say physicists). Ie. the hidden variable in Bells theorem represents the experimenters ignorance only! The outcome follows the single-lambda logic, without "self-interference" (Bell inequality). Ie. the interactions follow as if $\lambda$ was predetermined, and then averaged.

But, let's think about it. Does this make sense? If a bell pair is created, and isolated, would it make sense that all of the universe EXCEPT the physicists, would be informed about this?

Insted consider this.

Then let's consider the case where $\lambda$ - suppsedly responsible for the predetermined correlation - is hidden not only to the ignorant physicists, but also to the whole measurement device and environment. Then what?

Lets think about it. Does this make sense? If a bell pair is created, and really isolated until measurement, would it make sense that all no other "agents" in the universe EXCEPT the physicists, also cannot be informed about this? I think so.

To understand the point here, let's look at Qbism, where the subjective probability does not reflect correlations on observerations, but instead causally determines the action of the agent.

"An agent’s probabilities are defined by her willingness to place or accept any bets she believes
to be favorable to her on the basis of those probabilities. "

The steroid part is this: Now let's consider any other classical part fo the system an agent, say the measurement device, and all of the environment are a set of "Qbist agents". Now if these agents are ALSO ignorant about $\lambda$ as per the second case, it will follow that the actions of these agent will reflect the uncertainty even if $\lambda$ is set. This is because we assume that the action is causally related to the subjective probability, while observations are merely "correlated".

Ie. the hidden variable explains the non-local correlation of observations, but it does nto imply non-local causation. Its simply two different things. Bells original realism does not make the discintion as I see it.

I hope this does not offend anyone, I'm just trying to add a new reflection. This basic lineout is simple and selfcontained enough to hopefully not require a "reference"? Qbism is a minor but still widely discussed interpretation. The steroid addition is not standard however.

(Would WOULD required a reference, is to take this another step, and INFER physical interations (hamiltonian) from the agent-updated-mechanis. But not that I am not making that claims here. I am just saying it seems a possibilit, which is trivial enough.)

An Introduction to QBism with an Application to the Locality of Quantum Mechanics

- Christopher A. Fuchs, N. David Mermin, Ruediger Schack, https://arxiv.org/pdf/quant-ph/0703192.pdf

To associate further to Demystifiers Solipsist HV, this is the way I imagine meaning to HV. It can potentially restore some "realism" buy not in the Bell sense, the bell realism is IMO outdated. This is my problem, rather than locality. To me locality is never questioned. Non-local correlations(observations) yes, but not non-local causation (actions)

Solipsistic hidden variables
-- H. Nikolic, https://arxiv.org/abs/1112.2034

/Fredrik

 

[andrew s 1905]

Most physicists accept that time dilation and length contraction follow from the relativity principle applied to the measurement of c, even though we have no constructive counterpart (causal mechanism) for that principle account. Indeed, this is the standard presentation of special relativity in the introductory physics textbooks. The mystery of quantum entanglement can be resolved in similar fashion via the relativity principle applied to the measurement of h. Would that principle account of quantum entanglement suffice here, or would you still require a constructive counterpart?

I have tried Google to follow this up but failed to find a way of phrasing it to get sensible hits. Would you kindly point me to where I might follow it up please. Thanks Andrew

 

[Tendex]

Oh, my god. I HATE Physics Forums discussions.

I agree that

1. In relativistic quantum field theory, the quantum state evolves according to local equations.
2. RQFT does not allow FTL signal transmission.

The issue is whether 1&2 imply that it is local. 1 by itself is not enough, because the smooth evolution does not include the measurement process by which a single outcome is selected out of a set of several possibilities. 2 is not enough, either, because although Local implies no FTL, no FTL does not imply local.

Again, if your goal is to make sure about Einstein relativity in quantum theories a definition of locality related to FTL is never going to help you because it can't make that distinction for quantum theories(you should be able to see this because all quantum theories share the Schrodinger evolution and the irreversible measurement evolution that you have pointed out several times), just like it did not distinguish classical electrodynamics from QM theories. So you are always going in circles, no wonder you hate discussions.
But it seems you are aware of this when you explain the issue about 1&2 in the quote above so I can't understand why you are not drawing conclusions from it regarding trying to find a definition of locality/nonlocality/FTL that helps any regarding compatibility with Einstein relativity for quantum theories if you know of its uselessness.

 

[Tendex]

Of course I also think for the same reasons that insisting on QFT locality to stress its being relativistically compatible is pretty useless. Using mathematical arguments referring to the differential operators used doesn't seem to help much either since QFT also uses nonlocal operators like Fourier transforms.

 

[martinbn]

Here is my take on locality in the usual two particle entangled situation. Consider the two events $A$ - Alice makes a measurement and $B$ - Bob does. Take a space-like hypersurface $H$ such that both events are to the future and let $S$ be the portion of it in the past null cone of event $A$. Then for locality, to predict anything about $A$ it should be enough to know information specified on $S$ alone. If on the other hand information on the $H$ that is not in $S$ makes it possible to make better predictions about $A$, then we have non-locality. In the Alice and Bob case we do have locality according to the above definition.

As said many times before, you can consider $H'$ so that $B$ belongs to it and $A$ is to the future. In other words Bob has already made the measurement and Alice is about to. Then it seems that knowing the result of Bob Alice can make a better prediction. Except that on this $H'$ the state is factorizable. Before $H'$ it may have been $|u,d\rangle - |d,u\rangle$, but on $H'$ it is one of the two summands. So given this information on $S'$ Alice can make the same prediction. Hence there is no non-locality. You can argue that this information is not available to Alice, but so what(?), lack of information is not a good reason to talk about non-locality.

I know that this is not going to change anyone's opinion (especially BMists), I am just expressing my view on the terminology. If you prefer you could say that QM is dynamically local and correlation non-local (or constrain non-local). But in my opinion there is no point in this unless you are trying to be unclear or you are a BMist and trying to push the BM agenda.

 

[Morbert]

Alice's measurement, like a classical measurement, only changed Alice's knowledge about Bob. It didn't change anything physical on Bob's end. But let's think about this. If after Alice's measurement, she knows that it is 100% certain that Bob will measure spin-down, and if Alice's measurement didn't change anything about Bob, then it means that before Alice's measurement, it must have already been true that Bob would measure spin-down (only Alice didn't know it). Well, we have to reject this, too, because it involves hidden variables (the value of Bob's measurement before he makes it).

If Alice is careful with her algebra and only includes propositions about spin-z, she can avoid hidden variables too. Upon observing e.g. spin-up, she can assume that her measurement revealed the pre-existing property spin-up of her particle, and that bob's particle has the pre-existing property spin-down which he will observe if he measures it. The logic based on these assumptions will be reliable.

 

[vanhees71]

The issue is whether QM obeys Einsteinian relativity, when one takes into account the measurement process. As I explained in a different post, quantum mechanics has two pieces: (1) the smooth deterministic evolution of the quantum state, and (2) the nondeterministic production of a measurement result. In QFT, it's pretty clear that (1) is relativistically invariant, but it's much less clear whether (2) is. According to some interpretations of QM, relativity is only approximately true, for macroscopic measurements, but fails at a microscopic level.

In any case, I don't agree with you that you need to assume Einsteinian or Galilean relativity in order to say what FTL means. It is enough that there is a single frame in which light travels at its characteristic speed. Then relative to that choice of a frame, you can ask whether or not there are FTL influences.

Of course QM cannot obey Einstein relativity, because it's a non-relativistic theory. It cannot be relativistically invariant to begin with.

There is no other known type of relativistic QT describing real-world phenomena then local relativistic QFTs (and the Standard Model). There's not the slightest hint that relativity is not true at the microscopic level.

In non-relativistic physics there's no problem to have relative velocities larger than the speed of light. There is simply no limiting speed in Newtonian physics, and there's also no causality constraint, because there's absolute time, and time is defined as an oriented 1D manifold, whose orientation defines which events can be causally connected, namely only such events can be the cause of another event, if the time of the former is less than the time of the latter.

 

[vanhees71]

Truly, I don't understand how what some people say about "collapse" isn't complete nonsense.

In an EPR experiment, with anti-correlated spin-1/2 particles, Alice and Bob agree to measure spin along the z-axis. Immediately before Alice's measurement she would say that there is a 50/50 chance that Bob will measure spin-up. Immediately after Alice's measurement, if she gets spin-up, the probability that Bob will measure spin-up changes to 0. How can you account for that sudden change?

When Alice has done her measurement she knows also what Bob will measure or has already measured. In no way can Bob know from her measurement than by just getting a message from Alice (which can only be transmitted by a signal that's not faster propagating than the speed of light). Without the exchange of information about the measurement results all what Alice and Bob get is a random result with probabilities 50% for the one or the other outcome. The 100% correlation due to entanglement (as well as the confirmation of the predicted probabilities for the measurement outcomes) can only be revealed when doing the experiment, and it doesn't depend in which temporal order Alice and Bob perform their measurements. Within local relativistic QFT there cannot be a causal effect of one measurement on the other if these measurements are space-like separated, and such an effect is not necessary to explain the correlations, which are a property of the state of the measured system prepared in the beginning before any measurement has been done on the particles.

 

[atyy]

When Alice has done her measurement she knows also what Bob will measure or has already measured. In no way can Bob know from her measurement than by just getting a message from Alice (which can only be transmitted by a signal that's not faster propagating than the speed of light). Without the exchange of information about the measurement results all what Alice and Bob get is a random result with probabilities 50% for the one or the other outcome. The 100% correlation due to entanglement (as well as the confirmation of the predicted probabilities for the measurement outcomes) can only be revealed when doing the experiment, and it doesn't depend in which temporal order Alice and Bob perform their measurements. Within local relativistic QFT there cannot be a causal effect of one measurement on the other if these measurements are space-like separated, and such an effect is not necessary to explain the correlations, which are a property of the state of the measured system prepared in the beginning before any measurement has been done on the particles.

Indeed, only the correlations exist ahead of time, but not the things that are correlated.
https://arxiv.org/abs/quant-ph/9801057v2

 

[Fra]

Immediately before Alice's measurement she would say that there is a 50/50 chance that Bob will measure spin-up. Immediately after Alice's measurement, if she gets spin-up, the probability that Bob will measure spin-up changes to 0.

This conclusion illustrates the "problem" of not explicitly acqknowleding that probabilities are conditional or JUST thinking of them as objective outcomes.

As Vanesh wrote, after Alices measurement, the probability as per Bob is still 50%. Obviously the way Bob (and possibly his measurement device) interacts with the incoming particle is independent of of Alices probability.

IMO, Qbism brings an insight that a probability has two sides.
1. It describes correlations or observations
2. It causally determines the action of the agent

/Fredrik

 

[RUTA]

I have tried Google to follow this up but failed to find a way of phrasing it to get sensible hits. Would you kindly point me to where I might follow it up please. Thanks Andrew

We're working on a more direct pedagogical version now, that's why I asked, but here are two longer versions:

Beyond Causal Explanation: Einstein’s Principle Not Reichenbach’s
https://www.mdpi.com/1099-4300/23/1/114/htm

Answering Mermin’s challenge with conservation per no preferred reference frame
https://www.nature.com/articles/s41598-020-72817-7

 

[Tendex]

I don't agree with you that you need to assume Einsteinian or Galilean relativity in order to say what FTL means. It is enough that there is a single frame in which light travels at its characteristic speed. Then relative to that choice of a frame, you can ask whether or not there are FTL influences.

Whether something is FTL doesn't depend on whether we're talking galilean or einsteinian relativity. It doesn't depend on relativity (galilean or einsteinian), since I'm only talking about a single frame. Relativity is about the relationship between two DIFFERENT frames. Within a single frame, relativity is not relevant.

Probably off-topic in a quantum subforum but I still don't know how anyone can say that talking about possible FTL influences doesn't have anything to do with relativity, but I would ask you how you are determining that light travels at its characteristic speed, using a one-way or a round-trip measure of lightspeed to campare it with the FTL influences.

 

[PeterDonis]

I would ask you how you are determining that light travels at its characteristic speed, using a one-way or a round-trip measure of lightspeed to campare it with the FTL influences

You don't have to know what speed light is traveling to test for FTL influences. You just have to emit a light signal at the same time you emit whatever supposed FTL influence you are testing, and see which one arrives first at the destination.

 

[Tendex]

You just have to emit a light signal at the same time you emit whatever supposed FTL influence you are testing, and see which one arrives first at the destination.

My point is that this procedure only makes sense in the non special relativistic scenario, since assuming SR the supposed FTL communication would arrive its destination before emiting it, so you cannot emit both a light signal and an FTL signal at the same time, it can only be done in Galilean relativity since there the speed of light is not a maximum and FTL doesn't have problems with causality , you have classical addition of velocities also at light speed.
End of the line for the subthread discussion, stevendaryl is assuming nrqm even if he is not aware and therefore can't solve anything about relativistic QFT from that standpoint.

 

[vanhees71]

Probably off-topic in a quantum subforum but I still don't know how anyone can say that talking about possible FTL influences doesn't have anything to do with relativity, but I would ask you how you are determining that light travels at its characteristic speed, using a one-way or a round-trip measure of lightspeed to campare it with the FTL influences.

Of course it all has to do with relativity. In Newtonian mechanics nothing prevents causal effects propagating ftl. It's even the usual way interactions are model, namely as actions at a distance.

 

[RUTA]

When Alice has done her measurement she knows also what Bob will measure or has already measured. In no way can Bob know from her measurement than by just getting a message from Alice (which can only be transmitted by a signal that's not faster propagating than the speed of light). Without the exchange of information about the measurement results all what Alice and Bob get is a random result with probabilities 50% for the one or the other outcome. The 100% correlation due to entanglement (as well as the confirmation of the predicted probabilities for the measurement outcomes) can only be revealed when doing the experiment, and it doesn't depend in which temporal order Alice and Bob perform their measurements. Within local relativistic QFT there cannot be a causal effect of one measurement on the other if these measurements are space-like separated, and such an effect is not necessary to explain the correlations, which are a property of the state of the measured system prepared in the beginning before any measurement has been done on the particles.

The quantum correlations are indeed realized just as predicted by non-relativistic QM (which equals QFT for low energy phenomena as in this case). So, the mystery here has nothing to do with QM vs. QFT vs. SR. All these theories are compatible with the observed results (more so than you might suspect, see this article for example). The mystery resides in the desire for a causal mechanism responsible for those correlations and that is assumed to function per the causal structure of M4 per "local realism." No one has that answer.

 

[Tendex]

The quantum correlations are indeed realized just as predicted by non-relativistic QM (which equals QFT for low energy phenomena as in this case). So, the mystery here has nothing to do with QM vs. QFT vs. SR. All these theories are compatible with the observed results (more so than you might suspect, see this article for example). The mystery resides in the desire for a causal mechanism responsible for those correlations and that is assumed to function per the causal structure of M4 per "local realism." No one has that answer.

But causality is a precondition for scientific predictions checked using measurements. What you are calling the mystery is just a departure from the classical physics understanding of how causation works probabistically.

 

[RUTA]

But causality is a precondition for scientific predictions checked using measurements. What you are calling the mystery is just a departure from the classical physics understanding of how causation works probabistically.

Supply the causal mechanism in accord with "local realism" and you will have solved the mystery. Simply stating that the formalism maps to the experimental outcomes fails to answer the mystery.

 

[PeterDonis]

assuming SR the supposed FTL communication would arrive its destination before emiting it

Only in some frames, not in others. In any case, SR in no way prohibits using spacelike curves, which is what "FTL" actually means in SR. It is perfectly possible to model objects that move "FTL" in SR using spacelike curves; Google "tachyons". That's not to say there are no issues involved in such models, but they do exist, so your claim that the procedure I described only makes sense in non-relativistic physics is wrong.

End of the line for the subthread discussion, stevendaryl is assuming nrqm even if he is not aware and therefore can't solve anything about relativistic QFT from that standpoint.

Your dogmatic declaration that no more can be said in this subthread is wholly premature. See above.

 

[Tendex]

Supply the causal mechanism in accord with "local realism" and you will have solved the mystery. Simply stating that the formalism maps to the experimental outcomes fails to answer the mystery.

But why in accord with "local realism"?, the correlations show it to be the wrong classical recipe.

 

[Tendex]

Only in some frames, not in others. In any case, SR in no way prohibits using spacelike curves, which is what "FTL" actually means in SR. It is perfectly possible to model objects that move "FTL" in SR using spacelike curves; Google "tachyons". That's not to say there are no issues involved in such models, but they do exist, so your claim that the procedure I described only makes sense in non-relativistic physics is wrong.

I'm obviously disregarding objects like "Tachyons" since they are not compatible with causation as understood in modern science. But I am not dogmatic at all so if you want to pursue that line I'll be delighted.