Quantum entanglement: left or right first?

[gzhpcu]

If you consider Aspect's classical quantum entanglement experiment, i.e., highly energized calcium atoms, when returning to their normal state, emit back-to-back two photons with identical spin:
- there is a detector on the left and one on the right separated by a very large distance
- when a measurement is taken on one of the two photons, the Schrödinger wave function collapses, and a definite spin state is taken, immediately setting the spin state of the other photon.

If you add two observers: for one observer it might appear that the wave function collapse was simultaneous, for the other observer it will not appear so. You might even have a situation that depending on their position/speed, one observer sees the measurement being taken on the one on the photon to left first (ergo setting the spin state of the photon on the right) and the other observer sees the measurement on the photon to the right as taking place first, setting the spin state of the photon to the left.

So which one sets the state of the other?

[korjik]

Whichever observation occurs first in the frame of the entangled photons. The collapse is simultaneous too. This 'spooky action at a distance' is one of the things that bugged Einstien about QM

[Fortis]

I had a bit of a think about this one a while ago. I suspect that if we constrain ourselves to those things that are observable then we will probably find that the two cases (left or right) are indistinguishable, i.e. there is no test that we can perform to determine which of the two detectors "collapsed the wavefunction."

[gzhpcu]

Whichever observation occurs first in the frame of the entangled photons. The collapse is simultaneous too. This 'spooky action at a distance' is one of the things that bugged Einstien about QM

What do you define as "the frame of the entangled protons"?

[Grey]

So which one sets the state of the other?Neither. Even framing the question suggests that there has to be some kind of absolute sense of which event precedes the other. For spacelike separated events, that's not a valid assumption.

[Bad jcsd]

Why's this in against the main stream. It's a good question, even if vanilla quantum mechanics isn't Lorentz covaraint it's not difficult to make it Lorentz covariant and thta doesn't make the problem go away (also photons don't have reference frames in relativity).

I'd guess the answer depends on which interpretation you use, in the Copenhagen inetrpretation the wavefunction has no physical reality so physical actions causing what is purely a mathematical conceit to do something is moot. When you take an approach which assigns a reality to wavefunction I'd guess you'd argue that there is no true wavefunction collapse only that both observers decohere with the quantum system with no absolute time order.

No causality violation is necessary (though if you do go for interpretaions such as Bohmian mechanics then you do get causality violations) as it cannot be used to send information FTL, so Fotris is correct there is no experiment we can perform to determine which came first.

[gzhpcu]

Neither. Even framing the question suggests that there has to be some kind of absolute sense of which event precedes the other. For spacelike separated events, that's not a valid assumption.

Does this imply that the event is simultaneous, independent of time?

[Grey]

Does this imply that the event is simultaneous, independent of time?I wouldn't say that, either. :) I'd say that the question of whether the event is simultaneous, or one or the other of the observations happened first, is a meaningless question. Two events that are spacelike separated simply do not have any unambiguous temporal relationship to each other, even if it turns out as a result of quantum mechanics that the two events are correlated in a manner that appears to imply causation.

[gzhpcu]

I wouldn't say that, either. :) I'd say that the question of whether the event is simultaneous, or one or the other of the observations happened first, is a meaningless question. Two events that are spacelike separated simply do not have any unambiguous temporal relationship to each other, even if it turns out as a result of quantum mechanics that the two events are correlated in a manner that appears to imply causation.

Hmmmm, calling it a "meaningless question" just seems to imply to me that we don't have a clue... :-)

[Bad jcsd]

Hmmmm, calling it a "meaningless question" just seems to imply to me that we don't have a clue... :-)

It's like asking of two arbitary points A and B which one is on the right? There is no absolute answer as it depends very much from which direction we are loooking (though events with timelike sepration have a definite temporal order).

the fialure of simulatenoity at distance is something thta occurs in all theories which incorparate special relativity.

[Grey]

Hmmmm, calling it a "meaningless question" just seems to imply to me that we don't have a clue... :-):) Maybe, but the universe doesn't know either. Asking which one happened first is assuming that our common-sense ideas of time (that if there are two events, either one or the other happens first or they happen at the same time) are actually true. But though we certainly don't understand time completely, there are some things we know about it. One of those things we know for sure is that our common-sense ideas about it are dead wrong!

[turbo-1]

http://streamer.perimeterinstitute.ca:81/mediasite/viewer/

Penrose's presentation at the Perimeter Institute gives his take on this conundrum.

[peteshimmon]

If its a meaningless question then it was
a meaningless experiment:)

[gzhpcu]

So maybe Feymann's "sum over histories" approach is a better way of looking at this problem?

Thanks for link turbo-1.

[gzhpcu]

:) Maybe, but the universe doesn't know either. Asking which one happened first is assuming that our common-sense ideas of time (that if there are two events, either one or the other happens first or they happen at the same time) are actually true. But though we certainly don't understand time completely, there are some things we know about it. One of those things we know for sure is that our common-sense ideas about it are dead wrong!

OK. I but we do know is that we have a situation of nonlocality, and that measuring the spin on one photon immediately sets the spin on the other one.

[korjik]

What do you define as "the frame of the entangled protons"?

The same way you define every other frame. We have to assume an inertial frame if we are going to reduce the problem to what you described.

Grey is right tho. the question is a bit meaningless since all the different observers are going to disagree on who saw what first. That is a basic property of relativity that I should have remembered.

[gzhpcu]

However, I still have a problem: Events happen and (for example) light transmits the image of the event to the observer. Just as when I look up at the sky, I am looking at various points in time. In the case of quantum entanglement, taking a measurement of one of the two photons, results in its spin being known out of all possible spin states, and thereby the spin state of the entangled photon is predetermined immediately, irregardless of the distance separating the two. When the measurement of the other photon's spin is taken, it will correspond to this preset state.

If a third observer, positioned exactly in the middle of the two photons, waits for a light signal transmitting the result of the measurement, it would seem to me that this observer would be in a unique position to say which of the two photons was measured first and thereby set the state of the other. That other observers will have a different view is to be expected, but not in this case. This would seem to me to be the priveleged observer, the only one able to answer the question I originally posed.

[turbo-1]

How about if causality was taken out of the picture? What if the observation of the spin state of either particle was of no consequence to the state of the other, but that both particles are destined to have the same spin state no matter where or when the wave-form is collapsed by observation?

[gzhpcu]

I believe that Einstein had argued originally something similar, however, the standard model of QM states that until observed, the spin states are undetermined. They both are described by Schroedinger's probability wave function (could it be that these are both identical in form?). He thought that they already had their spin states set in reality.

[turbo-1]

I do not have an adequate understanding of the value of indeterminance and entanglement in QT. I hope, though, that as real-world measurements of subatomic particles progress, we might see a clarification of that dichotomy. I have an expectation that wave/particle duality can be explained with a better understanding of this dichotomy. As an optician, I am quite comfortable with the wave nature of EM. I have a sneaking suspicion, though, that the particular aspects are more related to the mechanics of observation, in which observations seem to be quantized in an otherwise anolog universe.

[Grey]

In the case of quantum entanglement, taking a measurement of one of the two photons, results in its spin being known out of all possible spin states, and thereby the spin state of the entangled photon is predetermined immediately, irregardless of the distance separating the two. When the measurement of the other photon's spin is taken, it will correspond to this preset state.Your claim that this happens immediately (i.e., at the same time) presumes that there is such a thing as "at the same time" for two events that occur at distant locations. According to relativity, there is no such thing. No, really, simultaneity for distant events does not exist independently of a reference frame. And since all reference frames are equally valid, there's no universal meaning to simultaneity.

If a third observer, positioned exactly in the middle of the two photons, waits for a light signal transmitting the result of the measurement, it would seem to me that this observer would be in a unique position to say which of the two photons was measured first and thereby set the state of the other. That other observers will have a different view is to be expected, but not in this case. This would seem to me to be the priveleged observer, the only one able to answer the question I originally posed.In the middle of the two photons and moving how? At rest? If so, at rest with respect to what?

We make a measurement on a photon. We know that the measurement of the other photon will be correlated (actually in most actual cases, it's not that they'll have the same spin, it's that they'll have opposite spin). Does that mean it changes "at the same time" as the measurement of the other photon. No. It only means what it says, that the two are correlated in a way that will require a superluminal connection between the two. Note that the fact that the two particles have some correlation does not necessarily mean that the property in question has to be well-defined ahead of time.

[gzhpcu]

Your claim that this happens immediately (i.e., at the same time) presumes that there is such a thing as "at the same time" for two events that occur at distant locations. According to relativity, there is no such thing. No, really, simultaneity for distant events does not exist independently of a reference frame. And since all reference frames are equally valid, there's no universal meaning to simultaneity.

In the middle of the two photons and moving how? At rest? If so, at rest with respect to what?

We make a measurement on a photon. We know that the measurement of the other photon will be correlated (actually in most actual cases, it's not that they'll have the same spin, it's that they'll have opposite spin). Does that mean it changes "at the same time" as the measurement of the other photon. No. It only means what it says, that the two are correlated in a way that will require a superluminal connection between the two. Note that the fact that the two particles have some correlation does not necessarily mean that the property in question has to be well-defined ahead of time.

I never meant (as Einstein maintained) that the property in question has to be well-defined ahead of time. The correlation between the two just means they will always display the same result. What the result is, however, remains undetermined until measurement.

Is space time then really like a comparison made to a loaf of bread, which exists, and can be sliced at different angles (according to the perspective of different observers) and thereby delivering different results as to the sequence of things?

Trying to understand this better: when a single event occurs, it seems to me that the time of reference of the stationary observer right where the event occurs is privileged in the sense that no other observer can see the event any "sooner" due to distance and/or relative motion. Yes, I know there are problems with the word "sooner"... :-(

I see what you mean about simultaneity, however. If you can't agree on distances and time, then there is not much to agree on...

[Grey]

Is space time then really like a comparison made to a loaf of bread, which exists, and can be sliced at different angles (according to the perspective of different observers) and thereby delivering different results as to the sequence of things?Although I find this pretty strange, too, it does indeed seem taht our universe works very much like this.

Trying to understand this better: when a single event occurs, it seems to me that the time of reference of the stationary observer right where the event occurs is privileged in the sense that no other observer can see the event any "sooner" due to distance and/or relative motion. Yes, I know there are problems with the word "sooner"... :-(An observer located right at the site of an event when it occurs certainly sees the event with no delays due to propagation. What what do you mean by "stationary", though? Stationary with respect to what?

[gzhpcu]

"Stationary" in respect to the location of the event.

[Grey]

"Stationary" in respect to the location of the event.And how are we to decide what that is? It's harder than you might think. An event is a single point in space and time, and determining it's "location" at a later time requires establishing a reference frame, which isn't universal.

Let's look at an example. Out in distant space, there are two observers moving relative to each other. Both are inertial observers and consider themselves to be at rest. At the moment that they meet, they both witness an event right at the meeting point. The first observer thinks that the event happened right where he has been standing, and continues to stand, so that he is stationary with respect to the event. The second observer thinks that she has not moved at all, so she must be stationary with respect to he event. Each observer thinks that the other approached the location of the event, reached its location just as the event took place, and then moved away afterward. How do we decide who is right?

[gzhpcu]

Let's be a bit more specific: the explosion of a space station. The observer is in a spaceship very close by since a week. His/her distance from the space station has been fixed all that time. Would this be "stationary"?

[Grey]

Let's be a bit more specific: the explosion of a space station. The observer is in a spaceship very close by since a week. His/her distance from the space station has been fixed all that time. Would this be "stationary"?Depends. Probably from the point of view of the person in the spaceship. But what about the person who is aware that the space station (with the space ship along with it) is in orbit and moving constantly? Or, if it would be better to avoid gravity and general relativity, what if the spaceship and space station are en route to the Andromeda galaxy, moving (relative to the Milky Way) at a substantial velocity. Are they stationary or not? They certainly feel no effects of movement, though that galaxy has a significant blueshift (and that other one behind us is rather redshifted). If so, what about the point of view of the person en route from Andromeda to the Milky Way? Or do we have to be stationary relative to the Galaxy to be "really" stationary. What about our movement toward the Great Attractor, then?

As for an observer who thinks that the space station is rushing by, and explodes as it passes, the fact that it was the space station that exploded doesn't mean that we necessarily have to consider it to be the "stationary location" of the event. After all, if there's a traffic accident at Fifth and Maple, we say that the accident happened at that fixed point relative to the Earth, even though both cars were moving. If someone asked a few days later where the accident you saw happened, you wouldn't say something like "right where the car is!", because we tend to consider Earth to be our fixed reference frame. So you'd be justified in saying that this space station came flying past, and just as it got to that spot right there, it exploded, and the debris has since continued on roughly the same trajectory. The person who has been staying a fixed distance from the spaceship is equally justified in saying that the spaceship was at rest when it suddenly exploded, though there was this other spaceship that came flying past at the same time.

It may sound like I'm being flip, but this is at the heart of relativity. Under relativity, there are no preferred reference frames. Sticking to special relativity, any inertial observer is perfectly entitled to consider herself at rest, and everything else moving around her. A second inertial observer is free to do likewise, even if he is moving relative to the first.

[gzhpcu]

My apologies if I seem a bit thick, but by "stationary" I intended to mean in respect to the spacestation (which also happens to be where the event - explosion - takes place), where the spaceship "parked" near the space station is in the same intertial system as the the space station.

In respect to the loaf of bread analogy: There can not be anyway to slice the loaf in a way to see the event "sooner" than the slice from the point of view of the spaceship "parked" next to the space station. The loaf of bread is growing (arrow of time) allowing the present and past to be sliced in various fashions, but not arbitrarily in the future from the point of view of any "stationary"-to-the-event observer.

[Grey]

My apologies if I seem a bit thick, but by "stationary" I intended to mean in respect to the spacestation (which also happens to be where the event - explosion - takes place), where the spaceship "parked" near the space station is in the same intertial system as the the space station.I'm going to harp on this. I know that's what you mean, but why is the spaceship automatically the preferred reference frame in which to judge the location of the explosion? If we judge the reference frame of a car accident, we tend to do so in the reference frame of the Earth, not the car. The point I'm making is that deciding that the space station is fixed (and therefore that it is stationary relative to the location of the event) is an arbitrary choice of reference frame. In relativity, we don't get to do that. Or rather, we can do that, but we have to accept that any other choice of reference frame would be equally valid.

[gzhpcu]

I did not say it was the preferred reference frame in which to judge the location of the explosion, but rather the preferred reference frame in experiencing the exposion first. But I see what you mean: being "stationary" is irrelevant, a space ship flying by would make the original space ship no longer special...

[Grey]

I did not say it was the preferred reference frame in which to judge the location of the explosion, but rather the preferred reference frame in experiencing the exposion first. But I see what you mean: being "stationary" is irrelevant, a space ship flying by would make the original space ship no longer special...Exactly. So there are any number of reference frames that will be right at the explosion when it happens. They'll all be right there for it, but they'll also all have different ideas about when other events happened and for events that have spacelike separation from the explosion, they'll disagree about the time ordering.

So applying this to the wave function collapse thing, if we wait until the entangled particles are spacelike separated before making any measurements, then there will be valid reference frames where the left one was measured first, valid reference frames where the right one was measured first, and valid ones where they happened at the same time. What does that mean for the "collapse of the wave function"? Well, if it's physical (it's just as common to consider it a mathematical abstraction), then it's not bound by the normal rules of causality. An event that causes something else doesn't need to specifically happen before that something else. The universe is weird. :)

[gzhpcu]

Thanks for straightening me out again, Grey... ;-)

[Grey]

Thanks for straightening me out again, Grey... ;-)It's a fun discussion. And trying to describe it always helps remind just how strange the laws of the universe really seem to be! ;)

[John Dlugosz]

Two photons entangled to have identical spin, and observers with spacelike separation:

In reference frame A, I see L make a measurement first. Then I see R make a measurement, and it is consistant with what L found. "consistant" can be complex and involve statistical probabilities, if they used different axis of measurement, to prove entanglement as opposed to preset values that existed all along.

In reference frame B, he sees R make a measurement first, and then sees L make a measurement, which is consistant with R's result. B sends his results to me.

I see that he got the same results I did for what L and R measured.

The measurement events at L and R do not happen in some order. They both exist in spacetime, period.

Whatever unnamed deity is responsible for such things chose a random but consistant pair of outcomes to assign to each observer, and "posted" them as a single accounting transaction, with a view of the Universe as a 4-D slab of spacetime, with all events past and future layed out. Time only exists when moving through the slab on a defined time axis, observing the changes in the other 3 dimentions as you move.

--John