Interpretation of a state in quantum entanglement

[Lord Jestocost]

Realism is basic assumption of scientific approach.

With all due respect, this statement makes no sense without a defining what "realism" means for you!

 

[edguy99]

Actually, you left out the most important one:
...But even if you insist on that distinction, we can still agree that we have it down to two possibilities for anyone who wishes to think that u is an intrinsic property of the particle:
1) It is an intrinsic property that is acquired at an unknowable point in the proper time history of the particle. (In which case we have the problem similar to an unknowable aether frame.)
2) It is an intrinsic property that is acquired when the measurement is done on the particle itself. (In which case people need to significantly change the language they use about when entangled states undergo their changes!)

In itemizing the possibilities, aren't you missing 3) It is an intrinsic property that is acquired when the particles are entangled?

 

[Ken G]

In itemizing the possibilities, aren't you missing 3) It is an intrinsic property that is acquired when the particles are entangled?

The only way to make that work, and obey Bell's theorem, is to invoke superdeterminism, where the measurement that is made is also part of the properties of the system, where now the system also includes the measuring device and the scientist who thinks they are making a free choice about what to measure. We certainly cannot rule out that possibility, but it just isn't the way science is set up. We are never looking for the equations that tell us what we are going to measure, we are looking for the equations that tell us what we will get when we make a choice to measure something. I'm not even sure one could convey any sense of "truth" in the scientifically testable sense to the concept of superdeterminism.

Indeed, since I regard the properties as mental devices we use to build correct expectations, it is as irrelevant to me if superdeterminism is "true" as it is irrelevant if some divine being is choosing the outcomes of all measurements. It's simply not what shows up when you do science. So if I object to claims that there is some preferred yet unknowable reference frame in which the properties are set simultaneously, I will equally object to some unknowable superdeterminism that allows the outcomes to be determined in advance of deciding on what will be measured. If we stick to quantum mechanics, we have that the entangled Bell state I'm talking about looks like |uu>+|dd>, and that state simply does not attribute either of the properties to the particles in advance.

 

[edguy99]

The only way to make that work, and obey Bell's theorem, is to invoke superdeterminism ...

Assuming the particle can be described by it Jones vector and that entangled particles have matching Jones vectors, why invoke superdeterminism?

 

[Ken G]

With all due respect, this statement makes no sense without a defining what "realism" means for you!

For the purposes of our discussion, it seems our views of "realism" overlap insofar as we both associate the concept of "intrinsic properties" that are native to the individual particles with "realism." We know that the realist must allow that entangled systems cannot confer properties to individual particles, instead the whole entangled system would then have properties, but after the measurements are said and done and the entanglements are broken, that's when the realist must allow that the individual particles have their own instrinsic properties. Personally I do not take that view, but I know that many people like realism and will not part with it so easily, but even for them I must challenge the idea that the properties are conveyed at any time other than when the individual measurements are done on the individual particles. My objection is only that the latter language is never used in entanglement scenarios, even when the realist picture of intrinsic properties is taken on (as it usually is).

 

[Ken G]

Assuming the particle can be described by it Jones vector and that entangled particles have matching Jones vectors, why invoke superdeterminism?

The Jones vector does not tell you the property that will be taken on by the particle when measured, so one cannot say that eventual property is encapsulated in the Jones vector. Hence, for you, I still ask the key question: when, in their own proper time, did the particles, post-entanglement, acquire their individual properties? I don't see how the Jones vector lends any insight there.

 

[edguy99]

The Jones vector does not tell you the property that will be taken on by the particle when measured, so one cannot say that eventual property is encapsulated in the Jones vector. Hence, for you, I still ask the key question: when, in their own proper time, did the particles, post-entanglement, acquire their individual properties? I don't see how the Jones vector lends any insight there.

Maybe I am missing something here. The Jones vector gives absolute numbers for u or d if measured on a basis vector, and the correct probability of an u or d measurement (cos^2 of the angle from the basis vector) when off the basis vector. Does not assuming the entangled particles have the same Jones vector match experimental results?

 

[zonde]

Ken G, our discussion goes into too many directions. If you want to stick to particular topic please point it out.

yet normally people using that language are not called on to admit they are assuming an unknowable and inaccessible preferred reference frame-- something that we always admonish people not to do in relativity, but apparently have no objection to in quantum mechanics!

In special relativity preferred reference frame is superfluous, but SR can certainly survive if I put a label "preferred" on one of the reference frames (just a proof of principle, no particular commitment).

The way you mean it is in the sense of separating the object from the scientist studying it, so you regard an intrinsic property as an objective one. But the other meaning is that "objective" means "all reasonable people can agree to it," whereas subjective means that we can have widely different opinions, but that carries no connotation that the property is intrinsic to the particle, only that reasonable people agree on the property. That latter meaning is the only one essential to science, and is preserved when saying that properties are a form of information used to make predictions. But I know that realists, like Einstein famously was, will not be easily led down that road, so that's why I offer the alternative view that still avoids the need for a preferred reference frame.

I would rather go with something similar to the other meaning of "objective": "all reasonable people can give consistent account of phenomena".
You need to exclude the case where people agree just for the sake of being in agreement.

It's nowhere in the scientific method, and that's a crucial point. The historical track record of adding requirements to the scientific method, beyond the method itself, is rather dubious!

Besides other things realism is in the idea about communication with other observers and recording of experimental results and other boring stuff like that.

I'm not sure what you mean there, Bohmians are not free to violate Bell's theorem. All they do is choose realism over locality-- they hold that the property of a system is determined as it unfolds in its own proper time. The "pilot wave" is their nonlocal device for enforcing the necessary correlations, so in a preferred foliation of spacetime the pilot wave would appear to move instantaneously, and in some frames it would appear to move backward in time. So the collapse in the measurement on one particle is not the issue, it is how the correlation is enforced that is the issue here. Since I start the whole question after the property is acquired, whether or not there is any collapse is of no consequence to me, that's a whole other issue. No-collapse does not mean the property was always there, it only means the property was deterministically established by nonlocal pilot waves, and in some preferred foliation of spacetime, the two properties were acquired simultaneously. The Bohmian does not say properties cannot be acquired, rather, they say they are acquired deterministically and at non-arbitrary times, and that's why I view the appeal to a preferred reference frame as Bohmian realism.

In Bohmian mechanics outcome of measurement is determined by in initial position of particle and pilot wave. Pilot wave is non-local but positions of particles are local. So in a sense measurement is determined by initial positions of particles and how experimentalist sets up the experiment and determines pilot wave. Nothing is random there so there is no collapse. We can speak only about apparent collapse.

 

[zonde]

Maybe I am missing something here. The Jones vector gives absolute numbers for u or d if measured on a basis vector, and the correct probability of an u or d measurement (cos^2 of the angle from the basis vector) when off the basis vector. Does not assuming the entangled particles have the same Jones vector match experimental results?

Calling Jones vector a "hidden variable" should give you the idea that your model can't violate Bell inequalities.

 

[edguy99]

Calling Jones vector a "hidden variable" should give you the idea that your model can't violate Bell inequalities.

You cannot call a Jones vector a "hidden variable" wrt Bell inequalities since it does not match the model that Bell examined. Specifically, the Jones vector does not give "absolute" numbers (only probabilities) when measured outside the basis vector. Bell assumes a "absolute" match between entangled particles even when measured off of their basis vectors.

 

[zonde]

You cannot call a Jones vector a "hidden variable" wrt Bell inequalities since it does not match the model that Bell examined. Specifically, the Jones vector does not give "absolute" numbers (only probabilities) when measured outside the basis vector. Bell assumes a "absolute" match between entangled particles even when measured off of their basis vectors.

Hmm, then how your model reproduces prediction of perfect correlations at any angle?

 

[edguy99]

Hmm, then how your model reproduces prediction of perfect correlations at any angle?

You are correct that the Jones vector does not predict perfect correlations at any angle. Do you have a reference for an experiment that has achieved "perfect correlations at any angle" without throwing out the non-correlated particles?

 

[zonde]

You are correct that the Jones vector does not predict perfect correlations at any angle. Do you have a reference for an experiment that has achieved "perfect correlations at any angle" without throwing out the non-correlated particles?

Yes, if I remember correctly this experiment does exactly that:
Loophole-free Einstein-Podolsky-Rosen experiment via quantum steering

 

[edguy99]

Yes, if I remember correctly this experiment does exactly that:
Loophole-free Einstein-Podolsky-Rosen experiment via quantum steering

From the article: "Alice’s result (the prediction which she makes for Bob’s result) is immediately sent back to Bob via coaxial cables. If Alice detects no photon, Bob counts this as an inconclusive event from Alice (0)."

They are throwing out all the "non-matches" as inconclusive. These would include the ones the Jones vector would predict as not being the same.

 

[zonde]

From the article: "Alice’s result (the prediction which she makes for Bob’s result) is immediately sent back to Bob via coaxial cables. If Alice detects no photon, Bob counts this as an inconclusive event from Alice (0)."

They are throwing out all the "non-matches" as inconclusive. These would include the ones the Jones vector would predict as not being the same.

They are not throwing them out. They are including them in calculations.
"To close the fair sampling loophole one must also account for Alice’s inconclusive (0) results when she detects no photon and include these results when calculating the steering value."

 

[Ken G]

Maybe I am missing something here. The Jones vector gives absolute numbers for u or d if measured on a basis vector, and the correct probability of an u or d measurement (cos^2 of the angle from the basis vector) when off the basis vector. Does not assuming the entangled particles have the same Jones vector match experimental results?

The experimental results in the scenario I gave are that u occurs half the time, and always for both particles. This is a prediction of the initial Bell state, |uu>+|dd>, so there is no issue with statistical predictions of the experiment. I am not asking why do we get that experimental result, I'm asking that if it comes out u for both particles, and we interpret that outcome as an intrinsic property of the particles after the entanglement is broken (I'm not interested in the entangled state, or the Jones vector, that's not the question), then when was that intrinsic property acquired? The Jones vector (classically), and the Bell state (quantum mechanically), are both moot on that question, so don't relate here.

 

[Ken G]

Ken G, our discussion goes into too many directions. If you want to stick to particular topic please point it out.

Perhaps the key issue is, if we both stick to the realistic perspective that u is an intrinsic property of each individual particle in the situation I describe after the entanglement is broken, then I argue the only proper time we could use to say when those properties were acquired were when the observations that came out u were performed for each particle. That is the only way to avoid an arbitrary foliation of the spacetime. You take a different approach where there is a preferred foliation that is unknowable, and I'm saying the unknowable is an unsound foundation for any scientific interpretation. Yet most entanglement language does pretend there is some unknowable foliation such that the entanglement is broken "instantaneously" for both particles, which would mean prior to the observation on an unknowable one or other of the particles. I don't like that language, common though it is, because it violates the same sensibilities we use to shout down absolute simultaneity language in relativity.

In special relativity preferred reference frame is superfluous, but SR can certainly survive if I put a label "preferred" on one of the reference frames (just a proof of principle, no particular commitment).

Yes, that's what Poincare and Lorentz tried to do, but their approach is widely viewed as unresponsive to the core lesson of relativity: that no such preferred frame exists, expressly because it would be unknowable if it did.

I would rather go with something similar to the other meaning of "objective": "all reasonable people can give consistent account of phenomena".

That cannot hold, because the "account" of what happened involves interpretation, and we cannot expect all reasonable people to arrive at the same interpretation (witness quantum mechanics interpretations). Indeed, you can't get much more subjective than interpretations-- so "accounts of phenomena" are highly subjective. Instead, we must say "objective" means that all reasonable people can agree on the outcome of the experiment, or the property that the experiment measures, but not the account of the phenomena.

You need to exclude the case where people agree just for the sake of being in agreement.

Yes, they should come to the agreement via independent paths, that's fine.

Besides other things realism is in the idea about communication with other observers and recording of experimental results and other boring stuff like that.

Sure, we must allow that what is communicated is objectively real, and the properties are objectively real, but we can do that without saying the properties are anything but objectively real concepts we communicate in order to form correct expectations. Still, I don't need to go to those less realist places, I' m content to stay at the level where the properties are intrinsic but set during measurements in all cases.

In Bohmian mechanics outcome of measurement is determined by in initial position of particle and pilot wave. Pilot wave is non-local but positions of particles are local. So in a sense measurement is determined by initial positions of particles and how experimentalist sets up the experiment and determines pilot wave. Nothing is random there so there is no collapse. We can speak only about apparent collapse.

I agree with all that, it is the apparent collapse to which I refer-- the Bohmian thinks the collapse is determined, but it's still a collapse in the sense that the particle is acquiring properties it did not possesses previously. One can take a deterministic approach without saying that all the properties have always existed. But we needn't concern ourselves with the Bohmian view, it suffices to wonder whether the properties are intrinsic, and when they are acquired.

 

[Boing3000]

That is the only way to avoid an arbitrary foliation of the spacetime.

Actually, I would like to read some reference that backup this claim. This seems to me a arbitrary claim, as as been pointed out be reference in this post.
There is one state that begin it's existence at entanglement (and the not big bang, like the reference bizarrely claim).
So there is at least a non arbitrary foliation which is pointed to by GR, and on which everybody agrees, including you : proper time. Actually, with massive particle, a testable hypothetical variation on when the correlation occur is perfectly defined, and probably within experimental range.

As a side note, few of the people making claims about "arbitrary" choice of frame(s) for labs, are willing to admit that they do an arbitrary choice of frame when deciding the orientation of detectors, which as far as I know is not included into the quantum state description, but nonetheless used into the lab(s) setup(s).

So it is possible to entangled batch of electron. It is possible to drive half of then atop the Himalaya, it is possible to bury the other half in a mine in Texas.
I am not bothered more about about computing the proper time of each batches, than to compute how to orient the detectors. Both question are answered by the unique implicit indisputable geometry of space time (causality included)

 

[zonde]

[Property] was acquired by the particle when the measurement was done on that particle, which at least allows properties to change via local interactions, but of course from Bell's theorem we all know that the result is going to have to satisfy some nonlocal constraint that will compromise to a significant degree the idea that the property is intrinsic to the particle. Still, we can save intrinsic properties if we allow them to be subject to nonlocal constraints that are intrinsic to the system to which the particle belongs, but not intrinsic to the particle itself. For me, if we cannot make the constraints intrinsic to the particle, then I see little advantage in making the properties intrinsic to the particle either, especially when I see that I only use the property concept to build expectations (in my mind) about the outcomes of observations, so it seems natural to allow that properties are mental constructs that require no "time of acquisition" in the particle proper time.

It is possible that two measurements of entangled particles are separated by large time. Then this nonlocal constraint would have to act (if we agree that it requires some physical mechanism) across time i.e. this hypothetical physical mechanism would have to "remember" how it should constrain the measurement. For me it is easier to model the situation in such a way that this "memory" is intrinsic to the particle.
But of course this is personal preference and without some experimental results that tells apart two approaches there is no solid reason to chose one approach over another.

I would like to add separate comment about this phrase:

I only use the property concept to build expectations (in my mind) about the outcomes of observations

Of course you use property concept to build expectations about outcomes and this happens in your mind. But later on you check these expectations against experimental facts. And these experimental facts do not happen in your mind. They happen in physical reality. If there is good agreement between your expectations and experimental facts you can start to speculate that there is something else in your model that corresponds to reality the way your expectations correspond to experimental outcomes. And that is the point where I start to think that this "property" might have counterpart in physical reality.

 

[Ken G]

Actually, I would like to read some reference that backup this claim.

I don't understand what you're asking for here, you want a reference that simultaneity between two widely spatially separated events requires an arbitrary foliation of spacetime? That would seem to be one of the more basic results of relativity theory, I would expect the reference would be required from anyone who claimed that simultaneity was an absolute property that did not require an arbitrary foliation (i.e., an arbitrary choice of coordinates).

This seems to me a arbitrary claim, as as been pointed out be reference in this post.

Could you perhaps find the place in that website that you are talking about, more specifically? That website would seem to summarize criticisms of Bohmian mechanics, and one of those criticisms is likely the fact that Bohmian mechanics must regard the property as being acquired at some particular time in the proper time history of the particle, which would require an arbitrary foliation of spacetime. So without some specific quotation of what you are talking about, I would expect that website to be largely in agreement with my claim, and I could then use that very website as the reference you seek.

There is one state that begin it's existence at entanglement (and the not big bang, like the reference bizarrely claim).

Of course, but as I said above, that one state evolves. In Bohmian mechanics, it later is deterministically converted into a property of the particle in question. Surely you don't think Bohmian mechanics requires that no changes in state or property ever occur?

So there is at least a non arbitrary foliation which is pointed to by GR, and on which everybody agrees, including you : proper time.

Proper time is not a foliation, it is just the "trunk of the tree." But more to the point, as I pointed out above, the proper time is precisely where the issue is left ambiguous-- we cannot say or know at what point in the proper time stream the property was acquired, unless we say it was acquired at measurement of each particle. I do not object to realists who hold to that latter position, as I said so many times, my objection is to the oft-repeated claim that the properties change "simultaneously" or "instantaneously," words that require an arbitrary foliation.

Actually, with massive particle, a testable hypothetical variation on when the correlation occur is perfectly defined, and probably within experimental range.

Certainly not, to do a correlation, you must have already done the observation on both particles, making the issue moot unless you adopt the approach I just stated-- again not what you find in popular descriptions.

As a side note, few of the people making claims about "arbitrary" choice of frame(s) for labs, are willing to admit that they do an arbitrary choice of frame when deciding the orientation of detectors, which as far as I know is not included into the quantum state description, but nonetheless used into the lab(s) setup(s).

Of course the detection frames are arbitrary. The issue is that this arbitrariness is acceptable because it is not intrinsic to the particle, like properties are supposed to be after the entanglement is broken.

I am not bothered more about about computing the proper time of each batches, than to compute how to orient the detectors. Both question are answered by the unique implicit indisputable geometry of space time (causality included)

I'm afraid I don't know what you are saying there, it sounded like something requiring a reference!

 

[Boing3000]

I don't understand what you're asking for here, you want a reference that simultaneity between two widely spatially separated events requires an arbitrary foliation of spacetime?

We are not speaking of event arbitrary separated. We are speaking of a single event called entanglement, whose "state" is testable along two perfectly non-arbitrary worldlines with precise and unique proper coordinates.

Could you perhaps find the place in that website that you are talking about, more specifically?

And in General Relativity, there is no natural way, even once chosen a space-like slice of space-time (that has no reason to possibly be a flat one) as a definition of the "initial time", to determine how space-time will have to be sliced into more "simultaneous" classes of events in the future. If we try the "simplest solution", that is taking the parameter of time defined as the age (the time spent since the Big Bang), so as to define simultaneity as equality of age, this cannot work because it will run into lots of singularities (especially at the centers of planets and stars).

(underline's mine). This site seems to make the same arguments than yours, except the fact that it admit BB as a valid starting point (a thing it seems to me your are not even willing to admit).

I would expect that website to be largely in agreement with my claim, and I could then use that very website as the reference you seek.

Except this site at least admit there is a non-arbitrary frame. My question to you is why do you claim that entanglement, which is single and uniquely defined event where to start foliation (with respect to the tools uses in the laboratory (clock AND polarizer) syncrhonized/aligned)

Of course, but as I said above, that one state evolves. In Bohmian mechanics, it later is deterministically converted into a property of the particle in question. Surely you don't think Bohmian mechanics requires that no changes in state or property ever occur?

A cases with an entangle state that evolve in time would be an even more interesting feature. But the only quibbling here is to say if there is at least one experimental system of coordinate that allow to define simultaneity and then TEST for it.

Proper time is not a foliation, it is just the "trunk of the tree."

That I certainly don't understand. I think that GR predict that little invisible clock "ticks" along each particle would be exactly at the same event/place whatever the frame used where it sits on the word-line "path". I am even pretty sure that if your beam of electron/photon just slingshot around a heavily rotating black-hole, the entanglement have no known reason (as per QM) to be "broken". And then the first question became how to orient the filters...

But more to the point, as I pointed out above, the proper time is precisely where the issue is left ambiguous-- we cannot say or know at what point in the proper time stream the property was acquired, unless we say it was acquired at measurement of each particle.

We can always relate the time of the both measurement to the proper time of particles. We can thus test/eliminate certain type of "influence", notably FLT ones.

I do not object to realists who hold to that latter position, as I said so many times, my objection is to the oft-repeated claim that the properties change "simultaneously" or "instantaneously," words that require an arbitrary foliation.

That's fair, except that I have no understanding of on what physics theory your argument stand on. Isn't GR and the equivalence principle a non-arbitrary foliation ? Ins't it the only possible (the BB one of the article seems dubiously remote to me) (and testable) foliation ?

Certainly not, to do a correlation, you must have already done the observation on both particles, making the issue moot unless you adopt the approach I just stated-- again not what you find in popular descriptions.

"Certaintly" ? How so ?. Don't we detect difference in proper time already with atomic clock ? I don't say it is easy, I say it is much more concrete then to speak about measurement made on particle in another galaxy.

Of course the detection frames are arbitrary. The issue is that this arbitrariness is acceptable because it is not intrinsic to the particle, like properties are supposed to be after the entanglement is broken.

This is no logical. If you take as granted that entanglement is not a property of particles (like Bell proved), but of detectors, the issue you have are even more problematic.
In such a case, just explain how do you orient detector if you are working on a beam of photon across the Atlantic (or maybe having bounce of satellites mirror...) Your position would appears logical to me if you can do this without ever taking the actual path of particles into consideration...(I say you have to do it even implicitly)

 

[Ken G]

We are not speaking of event arbitrary separated. We are speaking of a single event called entanglement, whose "state" is testable along two perfectly non-arbitrary worldlines with precise and unique proper coordinates.

Your statement is not responsive to the scenario I described, and the question I asked. Again, the scenario is that we have two entangled particles, which are then separated quite widely, perhaps all the way to alpha Centauri. Later, a measurement is done on one particle, achieving a "u" result, such that it is known the other particle will yield "u" if measured similarly. Then I posed this question: when did the second particle acquire the property "u"? You have not answered this question, nor is it true that there is some "unique proper coordinates" that provides an answer to the question I posed.

I am even pretty sure that if your beam of electron/photon just slingshot around a heavily rotating black-hole, the entanglement have no known reason (as per QM) to be "broken".

Yes, that would not break the entanglement, we need the "u" result somewhere in my scenario.

We can always relate the time of the both measurement to the proper time of particles. We can thus test/eliminate certain type of "influence", notably FLT ones.

I'm not sure what you mean here, but it seems to be subsumed in the question I posed above, that you have not answered.

That's fair, except that I have no understanding of on what physics theory your argument stand on. Isn't GR and the equivalence principle a non-arbitrary foliation ?

No, it isn't. A foliation is a way to parse the spatial and temporal coordinates separately, it is akin to choosing coordinates. Normally, we think of the "trunk" of the tree as the proper time of some observer, and we then "branch out" spatially in some way that has the flavor of being perpendicular to the proper time stream, but that perpendicularity is only local in GR.

"Certaintly" ? How so ?. Don't we detect difference in proper time already with atomic clock ?

The issue is not if we can detect proper time, it is if we can detect when the property was acquired. That can only be done via the measurement, which then cannot tell you when it was acquired it can only tell you when the measurement was done.

This is no logical. If you take as granted that entanglement is not a property of particles (like Bell proved), but of detectors, the issue you have are even more problematic.

I don't see any problems there, nor have you pointed to any. But notice that I never said entanglement was not an intrinsic property of the system, my entire question is about when the entanglement is broken, and refers to the properties of the particles themselves, not the properties while they are still entangled.

In such a case, just explain how do you orient detector if you are working on a beam of photon across the Atlantic (or maybe having bounce of satellites mirror...)

This is why you need a connectivity across spacetime (and GR tells you how to preserve orientation along a world line), so that you can meaningfully define the "u" state consistently for the two particles. It is arbitrary for either particle by itself, but has a meaning of "same orientation" for the correlation between the results.

 

[Ken G]

It is possible that two measurements of entangled particles are separated by large time. Then this nonlocal constraint would have to act (if we agree that it requires some physical mechanism) across time i.e. this hypothetical physical mechanism would have to "remember" how it should constrain the measurement.

Yes, what we know from experiment is that the constraint is applied somehow, and the issue for the discussion is more "when" is it applied, rather than "how" it is applied, as the former would seem to be easier to establish. However, my point is that it is not easier to establish, unless we say either that the "when" is "whenever the scientist used the notion", or for those with a more realist bent, the "when" is "whenever the measurement was done on the particle acquiring the property in question."

For me it is easier to model the situation in such a way that this "memory" is intrinsic to the particle.

But that's what Bell's theorem says it cannot be, in the sense that it cannot be a hidden local variable. So even if you hold that the property is intrinsic to the individual particle once acquired, it cannot be that the constraints on that acquisition are entirely intrinsic to the individual particle.

But of course this is personal preference and without some experimental results that tells apart two approaches there is no solid reason to chose one approach over another.

The freedom to choose here is in asserting when the property is acquired. You are choosing to answer that by looking for a preferred foliation of spacetime that is unknowable, and although that is common in the language found, it suffers the same problems as all preferred reference frames in relativity: they never show up when you look for them. They are like the person on the roll call whose name is always called but never says "present and accounted for."

Of course you use property concept to build expectations about outcomes and this happens in your mind. But later on you check these expectations against experimental facts.

But why does checking against facts imply the properties are intrinsic to the particle? Properties are testable elements of experiments, so can only be said confidently to be intrinsic to the experiment. The experiment demonstrably involves particles, and an apparatus, and a scientist forming expectations and testing them. So the property can be intrinsic to any of those, or their combination, and still be a perfectly self-consistent element that is true to the scientific process.

And these experimental facts do not happen in your mind.

Not entirely in the mind, no, but the mind is involved or it is not an experiment and cannot be shown to involve properties. So perhaps your objection is my claim that the properties are intrinsic to the analysis, but note that when I say that, I include in the analysis more than just the mind, there is also everything that the mind is making sense of-- they are all part of the analysis. So the analysis involves a mind that is considering data that comes from an apparatus that acts on a particle, those are all elements of the analysis so to say the property is intrinsic to the analysis means it is intrinsic to all of that. But the key point is, the "when" the property is acquired then has a clear answer: it is when the analysis is completed.

If there is good agreement between your expectations and experimental facts you can start to speculate that there is something else in your model that corresponds to reality the way your expectations correspond to experimental outcomes.

Sure, so "reality" is also part of the analysis. But when you discover that properties cannot be said to be acquired by particles at particular times other than when you need to use the property, you can also start to speculate that your thinking process is playing a role in the meaning of a property. And that is the point where I start to think that this "property" is intrinsic to the analysis, and the analysis, which includes my mind, the apparatus, and the particles, is all I ever meant by "physical reality."

But I realize that last bit is not going to be accepted by realists, which is why my actual point here is a strong objection to the common language that when property u is measured on one particle, that property is "instantaneously" or "simultaneously" acquired by the other particle. That language is arbitrary, undefinable, and pretty close to meaningless, given what we know about relativity. So the realists should instead hold that properties are always acquired at the time of measurement on the particle in question, and nonrealists can hold that the property is acquired just when the information is available in the analysis, as properties are only a form of information and nothing more. Neither of those approaches requires giving some unknowable meaning to "simultaneity" across large distances.

 

[PeterDonis]

there is at least a non arbitrary foliation which is pointed to by GR, and on which everybody agrees, including you : proper time.

Proper time does not give a foliation of a spacetime. It doesn't even make a valid time coordinate, since an event that lies on multiple worldlines can have multiple proper times.

 

[PeterDonis]

Isn't GR and the equivalence principle a non-arbitrary foliation ?

No. The equivalence principle has nothing to do with foliations at all (it only talks about small patches of spacetime). The principle of relativity (which is what I think you actually meant) says that the laws of physics take the same form in any coordinate chart. It says nothing at all about how to construct any particular coordinate chart (which would be equivalent to constructing a foliation).

 

[Boing3000]

Your statement is not responsive to the scenario I described, and the question I asked. Again, the scenario is that we have two entangled particles, which are then separated quite widely, perhaps all the way to alpha Centauri. Later, a measurement is done on one particle, achieving a "u" result, such that it is known the other particle will yield "u" if measured similarly.

You don't need to make extravagant and un-testable scenario with undefined "later it is known". You can do physics with 15km remote location for example.
In such a context the synchronization procedure is not questionable.

Then I posed this question: when did the second particle acquire the property "u"? You have not answered this question, nor is it true that there is some "unique proper coordinates" that provides an answer to the question I posed.

I did, but the simplicity of the response eludes you: proper time. If we are talking of photon, there is no way to modify the setup, because proper time is 0 anyway.
I propose we test using a perfectly non-arbitrary definition of time used along both beam. Thus you need massive particle like electron in a similar setup, and you need a beam twice as fast and twice as long, keeping both detector at equal distance from the source (yes that require some mirror)
Then we can test if the correlation still only dependent of the synchronization of filter clock, or if the proper time of the electron change the correlation.
It is possible (maybe not easy technologically) two distinguish between the time of the measurement, and the time of the measured. (in short sorting out between non-locality and realism)
I you test the "younger" particle first (frame lab) (even so lightly), is is no surprise (as per proper non-locality) that the correlation is reflected on the "older" particle
But if you test the "older" first, it is surprising that the correlation still holds. If it does hold non-locality loose some point. Does it ?

The issue is not if we can detect proper time, it is if we can detect when the property was acquired. That can only be done via the measurement, which then cannot tell you when it was acquired it can only tell you when the measurement was done.

Yes and measurement are done at some precise timing to know which pair of electron we measure, and where (in the lab frame). There is no issue to decide when the property is acquired, what "speed" that correlation is bound with, in whatever frame you see fit to analyse.
The example is more obvious with GR and gravitational time dilatation, but I think that is also quite a far-fetched experiment...

No, it isn't. A foliation is a way to parse the spatial and temporal coordinates separately, it is akin to choosing coordinates. Normally, we think of the "trunk" of the tree as the proper time of some observer, and we then "branch out" spatially in some way that has the flavor of being perpendicular to the proper time stream, but that perpendicularity is only local in GR.

You lost me there. As PeterDonis said, I probably meant the principle of relativity; where all clock ticks at one second per second, whatever your preferred coordinate choice is, there is only on time per word-line.

I would also want you to respond to how would you "align" filter detector orientation. (I understand how you would synchronize clock (you need SR/GR for that). I would like you to do so only by using QM, and not the proper geometry of beams. Because all I am saying is to take at face value that you cannot ignore SR/GR when doing any position/angle/timing measurement.

 

[Boing3000]

Proper time does not give a foliation of a spacetime. It doesn't even make a valid time coordinate, since an event that lies on multiple worldlines can have multiple proper times.

I am only interested in a hypothetical proper time of entanglement. Of testing/measuring it. I am pretty sure I will need to use detectors world line labelled with their own proper time and intersecting with it. That's the point actually.

 

[PeterDonis]

I am only interested in a hypothetical proper time of entanglement.

Proper time along the worldline of each entangled particle is well-defined, yes. But it doesn't give you a foliation of spacetime.

 

[Ken G]

You don't need to make extravagant and un-testable scenario with undefined "later it is known". You can do physics with 15km remote location for example.

I don't understand what you are claiming here, I do need the scenario I mentioned to bring up the issue that I wish the scenario to bring up. It is not relevant that other scenarios are possible that do not encounter this question.

 

[zonde]

Yes, what we know from experiment is that the constraint is applied somehow, and the issue for the discussion is more "when" is it applied, rather than "how" it is applied, as the former would seem to be easier to establish. However, my point is that it is not easier to establish, unless we say either that the "when" is "whenever the scientist used the notion", or for those with a more realist bent, the "when" is "whenever the measurement was done on the particle acquiring the property in question."

What type of improvement is it that you are arguing for? Do you mean it is better interpretation, better theory or just a better convention in communication?

But that's what Bell's theorem says it cannot be, in the sense that it cannot be a hidden local variable. So even if you hold that the property is intrinsic to the individual particle once acquired, it cannot be that the constraints on that acquisition are entirely intrinsic to the individual particle.

Yes, constraints can't be intrinsic to the particle. I completely agree with this. I believe I was not claiming anything like that.

The freedom to choose here is in asserting when the property is acquired. You are choosing to answer that by looking for a preferred foliation of spacetime that is unknowable, and although that is common in the language found, it suffers the same problems as all preferred reference frames in relativity: they never show up when you look for them. They are like the person on the roll call whose name is always called but never says "present and accounted for."

Problems like that are common for all interpretations. Non-local collapse is not an exception.

But why does checking against facts imply the properties are intrinsic to the particle? Properties are testable elements of experiments, so can only be said confidently to be intrinsic to the experiment. The experiment demonstrably involves particles, and an apparatus, and a scientist forming expectations and testing them. So the property can be intrinsic to any of those, or their combination, and still be a perfectly self-consistent element that is true to the scientific process.

I am not claiming certainty where there is none.
And no, properties are not testable elements of experiments. Only calculated expectations are.

Not entirely in the mind, no, but the mind is involved or it is not an experiment and cannot be shown to involve properties. So perhaps your objection is my claim that the properties are intrinsic to the analysis, but note that when I say that, I include in the analysis more than just the mind, there is also everything that the mind is making sense of-- they are all part of the analysis. So the analysis involves a mind that is considering data that comes from an apparatus that acts on a particle, those are all elements of the analysis so to say the property is intrinsic to the analysis means it is intrinsic to all of that. But the key point is, the "when" the property is acquired then has a clear answer: it is when the analysis is completed.

Experimentalist of course is involved in design of experiment. But beside that mind is not involved in experiment. Just like physical processes outside laboratory happen even when nobody is analyzing them with some model.

Sure, so "reality" is also part of the analysis.

No. Only correspondence to reality is analyzed. But that analysis is completely different. It's analysis about our confidence in the model.

But when you discover that properties cannot be said to be acquired by particles at particular times other than when you need to use the property, you can also start to speculate that your thinking process is playing a role in the meaning of a property. And that is the point where I start to think that this "property" is intrinsic to the analysis, and the analysis, which includes my mind, the apparatus, and the particles, is all I ever meant by "physical reality."

You can speculate whatever you want. Just check that at the end you can calculate the right expectations.

But I realize that last bit is not going to be accepted by realists, which is why my actual point here is a strong objection to the common language that when property u is measured on one particle, that property is "instantaneously" or "simultaneously" acquired by the other particle. That language is arbitrary, undefinable, and pretty close to meaningless, given what we know about relativity. So the realists should instead hold that properties are always acquired at the time of measurement on the particle in question, and nonrealists can hold that the property is acquired just when the information is available in the analysis, as properties are only a form of information and nothing more. Neither of those approaches requires giving some unknowable meaning to "simultaneity" across large distances.

That language (that a property is "simultaneously" acquired by the other particle) is meaningful within certain interpretation. You are free to not accept particular interpretation if you don't like it.
If you say that this language is common in interpretation independent descriptions of phenomena maybe you could give some example?

 

[Ken G]

What type of improvement is it that you are arguing for? Do you mean it is better interpretation, better theory or just a better convention in communication?

The improvement is the elimination of language that rests on the existence of an unknowable preferred reference frame. As such, it is precisely the same improvement that language about light propagation received when reference to an aether frame was dropped from the Lorentz transformation. So that would be a better interpretation and a better convenction in communication, though not a different theory because what is being dropped is unknowable and untestable.

And no, properties are not testable elements of experiments. Only calculated expectations are.

That's a what a property is. What else would be scientific?

Experimentalist of course is involved in design of experiment. But beside that mind is not involved in experiment. Just like physical processes outside laboratory happen even when nobody is analyzing them with some model.

This is a crucial aspect of nonrealism that a lot of people get wrong. To have a mind be involved does not require the mind be part of the apparatus, it only requires that a mind be used to say that an experiment happened and what that means. In fact, the mind is more important than the apparatus-- that's the whole concept of a "gedankenexperiment" after all! Certainly in a gedankenexperiment, a mind is not involved in the experiment, as there is no experiment, but that's not what mind "involvement" means-- one has no gedankenexperiment if there is no mind. Similarly, if we say the environment carried out an experiment when no one was around, it is we who are saying it, so our minds are demonstrably involved even when no mind is present on the scene-- as no mind is present on the scene in a gedankenexperiment either. I find it ironic that realists never object to gedankenexperiments, so they don't seem to recognize the inconsistency in allowing hypothetical apparatuses carrying out some measurement when there is no physical experiment present, while disallowing hypothetical minds doing the analysis when there is no physical scientist present!

No. Only correspondence to reality is analyzed. But that analysis is completely different. It's analysis about our confidence in the model.

You say "correspondence to reality," I just say "reality." My words are more direct, and more scientific as a result. Nonrealism is so much more pragmatic, more agnostic, more precise, and downright more realistic as a result.

You can speculate whatever you want. Just check that at the end you can calculate the right expectations.

We certainly agree that all science does is create and test expectations. That you want it to be expectations that "correspond to reality" is outside of the scientific method, as I pointed out above. What you are doing is distancing the results by forcing them to "correspond" to something, instead of just being what they are: results, period. But since I know you are going to do that, as you are a realist, I offer the alternative interpretation, that the properties are intrinsic to the particles, that the constraints on the particles are intrinsic to the system as a whole, and above all, that the properties are conveyed by the measurements on each particle individually.

That language (that a property is "simultaneously" acquired by the other particle) is meaningful within certain interpretation. You are free to not accept particular interpretation if you don't like it.

I can say precisely the same thing about the aether for light propagation. Indeed, I would, it is precisely the same attitude, and should be rejected for precisely the same reason: it never shows up when looked for. The scientist should never build their prejudices into their models-- if the prejudice never presents itself in any of the data, it is inevitable that it will eventually be dropped altogether.

If you say that this language is common in interpretation independent descriptions of phenomena maybe you could give some example?

Where did I say the language is common in "interpretation independent" descriptions? Would I say that reference to an aether is interpretation independent? I said only that it is common, which it clearly is. Here are quotes from several of the first google hits on entaglement:
https://en.wikipedia.org/wiki/Quantum_entanglement
"Recent experiments have measured entangled particles within less than one hundredth of a percent of the travel time of light between them.[7] According to the formalism of quantum theory, the effect of measurement happens instantly."
https://www.quantamagazine.org/entanglement-made-simple-20160428/
"We will, according to quantum theory, get those results even if great distances separate the two systems, and the measurements are performed nearly simultaneously."
https://www.sciencedaily.com/terms/quantum_entanglement.htm
"As a result, measurements performed on one system seem to be instantaneously influencing other systems entangled with it."

All three of those are expert articles, loaded with important insights into entanglement, and to be fair, the second one can be interpreted as saying that in some reference frame the measurements can be regarded as simultaneous, not that they are simultaneous, while the third one throws in the words "seems to", but nevertheless the casual reader will miss the significance of these subtle escape acts, and easily fall into the common trap of imagining that simultaneity is an unambiguously defined element of the entanglement phenomenon. It should instead be enough to state that the measurements are made outside each other's light cones, or that neither could send a subluminal message to the other. So why say that quantum mechanics predicts the outcomes occur "instantaneously" or "nearly simultaneously"? Those are strikingly naive remarks in contrast with the herculean efforts to describe the profound subtleties of entanglement. I understand why they are there, the nonexistence of simultaneity is simply not the point of the articles, but my point is that this kind of language is so widespread it becomes ossified into the lexicon, and pushing back against that is the purpose of this discourse.

 

[Boing3000]

I don't understand what you are claiming here, I do need the scenario I mentioned to bring up the issue that I wish the scenario to bring up. It is not relevant that other scenarios are possible that do not encounter this question.

Your question is "Then I posed this question: when did the second particle acquire the property "u"?" have been answered many times now. It have been answered for photon in the experiment referenced by synchronized clock reading on site A & B (some arbitrary choice)
In your impossible mind experiment, the easiest thing it that the electrons (I suppose) traveled to alpha centaur inside a boxes fitted with a clock.

Now let's not forget you haven't yet answer the very simple question about the same implicit rational you made in that statement:

Later, a measurement is done on one particle, achieving a "u" result

"Later" with respect to what ? And most importantly "up" with respect to what ? You don't seem to realize that "up" is as a relative notion than "time". And you don't seem to realize that relativity have a unique answer for both questions because it have a unique way to describe space-time geometry.

I take the simpler view not only to acknowledge you can define an absolute local "up" in both cases as you can define an absolute local "age" in both case. They came from the same theory.

 

[Ken G]

And most importantly "up" with respect to what ? You don't seem to realize that "up" is as a relative notion than "time".

Yes, I don't realize that-- because it isn't. There is a unique way to parallel transport the meaning of "up" along a worldline, otherwise conservation of angular momentum would be meaningless.

And you don't seem to realize that relativity have a unique answer for both questions because it have a unique way to describe space-time geometry.

Yes I don't realize that either. We have connections in relativity that allow parallel transport to be defined, but we do not have any connection that defines simultaneity in an invariant way.

I take the simpler view not only acknowledge you can define an absolute local "up" in both cases as you can define an absolute local "age" in both cases.

Of course you can define local age, that's called proper time. I've invoked that concept over and over, I invoke it to pose the question you seem to imagine you have answered but you have not: when in the proper time of each particle did they acquire the "u" property? If you think you've answered it, I'd sure like to hear what that answer was.

 

[Boing3000]

Yes, I don't realize that-- because it isn't. There is a unique way to parallel transport the meaning of "up" along a worldline, otherwise conservation of angular momentum would be meaningless.

Finally, we are making progress. Although computing it may be quite difficult, it is uniquely defined.

Yes I don't realize that either. We have connections in relativity that allow parallel transport to be defined, but we do not have any connection that defines simultaneity in an invariant way.

So you can defining alignment (which also is a identity of value) across large distance is OK. But strangely the other invariant "proper time" is not ?
What allows you to make a distinction between both ?

Of course you can define local age, that's called proper time. I've invoked that concept over and over, I invoke it to pose the question you seem to imagine you have answered but you have not: when in the proper time of each particle did they acquire the "u" property? If you think you've answered it, I'd sure like to hear what that answer was.

Well, you don't seem to be hearing the answer: The proper time of the particle at the event "interacting with the filter".
If you think you cannot use connections in relativity that allow proper time to be uniquely defined, you are just contradicting yourself.

 

[Ken G]

Finally, we are making progress. Although computing it may be quite difficult, it is uniquely defined.

Computing what? What are you talking about?

So you can defining alignment (which also is a identity of value) across large distance is OK. But strangely the other invariant "proper time" is not ?

You keep confusing the invariant proper time with the ability to mark simultaneous moments across two different proper time streams. It is the latter that does not exist, and that is what I am talking about. Obviously proper time itself is invariant, as I've said many times now.

Well, you don't seem to be hearing the answer: The proper time of the particle at the event "interacting with the filter".

This was one of the possibilities I offered above, so how could I not hear it? But unfortunately it's not what I'm talking about, I'm talking about the stress put on simultaneity in common descriptions of entanglement breaking. I don't think you have understood, simultaneity is quite different from proper time.

If you think you cannot use connections in relativity that allow proper time to be uniquely defined, you are just contradicting yourself.

I cannot contradict something I never said.