Simultaneity for entangled particles

[Dale]

Does the act of measuring one of the particles change the past so that both particles are retroactively put into a determined state from the time they were emitted?

Some people believe exactly that:
http://arxiv.org/abs/0706.1232

 

[michelcolman]

No, it doesn't work that way. The total pattern of photons seen by Bob never shows an interference pattern regardless of what happens to the entangled photons, entangled photons behave differently than unentangled ones in this sense. However, if Alice measures the photons in a particular way that ensures the "which-path" information is "erased" (so her measurements give no information about which slit each of Bob's photons went through), then if Bob and Alice get together and compare notes, an interference pattern will be seen in the subset of Bob's photons whose entangled twins went to either of Alice's two detectors. I really recommend looking over the [post=2908070]example[/post] involving the delayed choice quantum eraser if you want to get a better understanding.

I have heard about the "erasing" experiment, haven't quite understood the fine parts of it (for example, the addition would not appear to add up to a uniform picture at first sight), but anyway, it's a different experiment.

Mine is a lot simpler. But why can't entangled photons behave like non-entangled photons in the double slit experiment? They're just photons in an indeterminate state that happen to have a twin somewhere. Or did nature somehow understand our little trick so it decided to consider the whole setup a measurement that causes decoherence? I know QM is full of this sort of unintuitive results, so I wouldn't even be totally surprised.

 

[michelcolman]

Some people believe exactly that:
http://arxiv.org/abs/0706.1232

At least that would solve the simultaneity problem. Measuring one particle causes the pair to have been decoherent from the beginning, so there's no fighting over whose reference system is to be used.

But don't the past experiments already include some component to verify that decoherence has not taken place yet before the measurement?

 

[t_barlow]

Are we saying that entangled photons cannot be made to behave this way, ie like double slit experiment?

No they can't. Entangled particles always look like spots, in particular Bob's will always look like spots regardless of what Alice does to hers. The reason is, each particle when considered in isolation appears to be in a mixed state, which means it resembles a statistical mixture of up/down, which never show interference, unlike a coherent superposition of up/down.

The way I see it is, if you measure an entangled particle, you do not change the state of its entangled pair. What changes is your information about the other particle, which is stored in your physical state (e.g. as a dark spot on a screen), so it is local to you. That information tells you what you will get if you go and measure the other particle, but it doesn't tell you what someone else found when they measured it. The only way to meaningfully compare what two people measured on spatially separated systems is to bring those two observers together into causal contact and see what they say. Fortunately quantum mechanics guarantees us consistency, which is nice.

 

[michelcolman]

If you measure the entangled particle in ether frame of reference aren't you going to get the results that you would normally get? Can you even tell which measurement happened "first"

It doesn't matter much except if you are trying to determine when the decoherence happened. If it is possible for an observer to receive a photon and say for sure "this photon was in an indeterminate state before I looked at it, and now it took state x", then you can have a problem with two observers who can both be considered to be "first" but only the real first will have received an initially indeterminate photon.

 

[michelcolman]

The way I see it is, if you measure an entangled particle, you do not change the state of its entangled pair. What changes is your information about the other particle

Wait a minute... so you're saying that the particles were in some state all along, we just did not have the information? What's the difference with ordinary, non-entangled particles that happen to have some opposite or equal property then? What makes entanglement so special if all that was missing was our information on the state of the two particles? There's nothing spooky about that! Or did I misunderstand what you were saying?

 

[t_barlow]

If Alice and Bob always make the same measurement- e.g. same angle of a polarizing filter- then yes entanglement doesn't seem so spooky. But the state isn't determined all along, since when Alice and Bob choose the angle randomly, they find that when they chose orthogonal angles, their results are completely uncorrelated, which can't happen if the states are all predetermined. And yet they still manage to get correlated results when they choose the same angle! Actually it's a while since I've done these calculations so I might be wrong somewhere.

 

[JesseM]

Well 'respected' is a relative label. Both might well argue they are victims of scientific 'consensus politics' - genuine openness to anything radically different is largely an ideal not reality. This is the cry of climate change sceptics - peer-reviewed journals refuse to publish their papers, then they are accused of being fringe scientists because they don't get papers published in peer-reviewed journals.

It's also the cry of creationists/intelligent design advocates, relativity deniers/ether theory advocates, biologists who don't think HIV causes AIDS, and so forth. Hopefully you agree that in at least some of these cases, the real problem is the poor nature of the arguments and their failure to provide an alternate explanation for most of the evidence that is taken to support the mainstream theory (and I have never seen any evidence that anything different is true for 'climate change skeptics')

At any rate both Prikas and Christian do make quite strong arguments, from quite different perspectives.

How can you, as a nonexpert without understanding of the technical details, make the judgment that the arguments are "strong"? Aren't you just seeing them in a favorable light because you are predisposed to like the conclusions?

If you read the pretty non-technical intro in Prikas's paper, I find it hard to counter the logic - energy/momentum will 'magically' alter given a particular succession of measurements, assuming entanglement.

That just seems like an argument from incredulity. As long as the change in energy/momentum obeys some clear mathematical rules and doesn't violate any known conservation laws or other basic principles, I see no reason a priori to rule out the hypothesis that nature actually follows these mathematical laws. Who are we to tell nature how to behave?

Joy Christian seems to be particularly rigorous and beat off all critics to his earlier papers.

How can you say he "seems to be particularly rigorous" if you can't follow the math at all? Just because he uses an authoritative tone of voice and includes a lot of equations?

 

[michelcolman]

No, it doesn't work that way. The total pattern of photons seen by Bob never shows an interference pattern regardless of what happens to the entangled photons, entangled photons behave differently than unentangled ones in this sense. However, if Alice measures the photons in a particular way that ensures the "which-path" information is "erased" (so her measurements give no information about which slit each of Bob's photons went through), then if Bob and Alice get together and compare notes, an interference pattern will be seen in the subset of Bob's photons whose entangled twins went to either of Alice's two detectors. I really recommend looking over the [post=2908070]example[/post] involving the delayed choice quantum eraser if you want to get a better understanding.

OK, I now understand that experiment (I think), but something is bothering me. If you replaced D1 and D2 with a single detector (located at "BS"), you apparently would not see any interference pattern for those coincidences at all (the superposition of R01 and R02). I don't understand why that would be. Why do you get a pattern when looking at erased information that has a 50% probability of going to D1 or D2, while you don't get any pattern if they all go to a detector located at BS? In the latter case, you still don't have any information about the path, so I would expect an interference pattern. And certainly if you then replaced BSA and BSB with mirrors as well, so everything ends up in the same detector. Actually, you might as well remove all of the detectors and mirrors, and there would still be no interference. Right?

In the original double slit experiment, you got an interference pattern when you did not add any detectors, and the pattern disappeared when you merely added a detector to figure out which photons went through which slit. This seems to be in complete contradiction with what I understood from the erased information experiment.

In fact, the only explanation seems to be that, if you set up an experiment that allows you to delay the decision whether or not to measure stuff (or if you try something like my Alice and Bob FTL experiment), nature will go "nice try, kiddo" and remove all interference patterns. Is that what would happen?

 

[JesseM]

OK, I now understand that experiment (I think), but something is bothering me. If you replaced D1 and D2 with a single detector (located at "BS"), you apparently would not see any interference pattern for those coincidences at all (the superposition of R01 and R02).

You probably could still see an interference pattern if your detector at BS could measure the momentum of the incoming photons, since the photons that would have gone to D1 would be going in a different direction than the photons that would have gone to D2. It's an interesting question whether all possible ways of detecting the idlers could be separated into groups of different possible results (like idlers going to D1 vs. D2, or idlers having one momentum vs. another) such that the subset for each result would show an interference pattern if the which-path information was lost (and perhaps with some additional assumptions like the idea that each measurement must be as precise as possible given the limits of the uncertainty principle, and that the detectors must be arranged so that no idlers will avoid hitting a detector).

Actually, you might as well remove all of the detectors and mirrors, and there would still be no interference. Right?

No, as t_barlow said above, the fact that they are entangled (in a way that at least had the potential to allow you to determine the which-path information if the idlers had been measured in the right way) guarantees that the total pattern of signal photons will never show interference.

In the original double slit experiment, you got an interference pattern when you did not add any detectors, and the pattern disappeared when you merely added a detector to figure out which photons went through which slit. This seems to be in complete contradiction with what I understood from the erased information experiment.

How is it in contradiction? The key issue is whether you measured the which-path information. In the original double slit experiment, if you added detectors but their measurements weren't sufficient to determine the which-path information--if for example the uncertainty in spatial position from their measurements was larger than the distance between slits--then you would still get interference, it's not the mere presence of "detectors" which destroys interference but rather the details of what they actually detect.

In fact, the only explanation seems to be that, if you set up an experiment that allows you to delay the decision whether or not to measure stuff (or if you try something like my Alice and Bob FTL experiment), nature will go "nice try, kiddo" and remove all interference patterns. Is that what would happen?

Interference is always lost in the total pattern of signal photons (if the photons were entangled in such a way that it would have been possible in principle to determine the which-path information by measuring the idlers in the right way), but as I said I'm not sure about the question of whether it's always possible to recover an interference pattern in some subset if you perform sufficiently detailed which-path-erasing measurements on the idlers.

 

[kfx]

I don't know. When you measure a particle, how do you know whether or not it was still in a superposition of states one microsecond earlier?

All you know is that, if you don't measure it, it remains in a superposition of states (causing interference patterns etc.) And if you do measure it, it turns out to be in a definite state. But for all you know, it may have been in a definite state all along, knowing in advance that you were going to measure it. How can you tell?

The Kochen-Specker no-go theorem tells that particles cannot be in a definite state "all along", it must be the measurement itself that causes the particle to collapse to a definite state.

Read this for a simple introduction why: http://www.cs.auckland.ac.nz/~jas/one/freewill-theorem.html

 

[kfx]

The problem is that there's no such thing as a "stationary observer".

Dunno, if the observer is stationary with regard to the measurement devices, doesn't that count?

You seem to underastand the relativity theory issues better than me, though.

 

[kfx]

If Alice and Bob always make the same measurement- e.g. same angle of a polarizing filter- then yes entanglement doesn't seem so spooky. But the state isn't determined all along, since when Alice and Bob choose the angle randomly, they find that when they chose orthogonal angles, their results are completely uncorrelated, which can't happen if the states are all predetermined. And yet they still manage to get correlated results when they choose the same angle! Actually it's a while since I've done these calculations so I might be wrong somewhere.

Actually I'm confused. If we're measuring position of a particle, then where do angles come into this?

Does it make sense to talk about measuring position in the context of this experiment at all? I think it does. With measuring it, we are reducing uncertainty; not only of the measured particle itself, but also of the twin particle. Position is a observable, just like the spin or polarization. But does selecting different measurement basis affect the outcome here as well? Position, roughly speaking, seems to be a binary property (certain/uncertain), and independent of the basis. Does the Kochen-Specker theorem still holds? Why?

 

[darkhorror]

It doesn't matter much except if you are trying to determine when the decoherence happened. If it is possible for an observer to receive a photon and say for sure "this photon was in an indeterminate state before I looked at it, and now it took state x", then you can have a problem with two observers who can both be considered to be "first" but only the real first will have received an initially indeterminate photon.

So then let's say we have two frames of reference and in each frame they are first. Now they won't agree who did the test first but they will agree with the results. Wouldn't they both be "real"? Seems like that would make sense as what happens when depends on your frame of reference.

how is it a problem that both can be considered first? there is no overarching master frame of reference where the "real" first happens. But in reality both are real. So just depending on the frame you chose you get the results.

 

[Q-reeus]

JesseM wrote:

It's also the cry of creationists/intelligent design advocates, relativity deniers/ether theory advocates, biologists who don't think HIV causes AIDS, and so forth. Hopefully you agree that in at least some of these cases, the real problem is the poor nature of the arguments and their failure to provide an alternate explanation for most of the evidence that is taken to support the mainstream theory (and I have never seen any evidence that anything different is true for 'climate change skeptics')

Sure, in many cases, but not all, the 'outsider' deserves the tag. My point was your slant about those two being typical of a 'tiny minority' is not a valid argument in itself. Lumping creationists together with these two gents is inane - chalk and cheese. And regardless of whether ultimately right or wrong, there is plenty of evidence 'climate change skeptics' have been unfairly treated. Recall a famous resignation recently, or those infamous e-mails?

That just seems like an argument from incredulity. As long as the change in energy/momentum obeys some clear mathematical rules and doesn't violate any known conservation laws or other basic principles, I see no reason a priori to rule out the hypothesis that nature actually follows these mathematical laws. Who are we to tell nature how to behave?

Don't have to know all the math to follow Prikas's principal argument. From that paper:

"Our purpose, in the present part of the article, is to prove that two or more correlated particles, even when they are unable to interact with a certain Hamiltonian (i.e.: when they are at great distance, even with walls of Pb between them, even when every particle with its measuring devices is entrapped in rooms deep beneath the surface of the earth), they exchange energy and angular momentum, and this is what I call ”non-locality”.
We will also prove that every quantum theory, orthodox or of hidden variables, suffers from this non-locality. This holds, because neither the current theory nor the alternative ones are responsible for non-locality. We will try to prove that the idea of two, non-interacting, distant particles in zero spin state ”together” is solely responsible for the whole novelty of non-locality."

He next considers a zero-spin initial state particle, decaying into + and - spin fermions that become well separated. Performing 3 successive measurements on one particle forces the other particle to have reversed it's WELL DEFINED initial spin 'remotely'. Yes the overall energy/momentum is conserved but that's hardly the point. Energy/momentum is NECESSARILY EXCHANGED remotely, with no chance of any causal agency. Merely an 'argument from credulity'? If you are an expert here then how about actually reading through that paper and providing a proper point-by-point critique that can set us all straight. Personally I think anyone making a serious claim to point out not only the 'apparent' absurdities, but also provide a plausible resolution deserves fair consideration. You disagree?

How can you say he "seems to be particularly rigorous" if you can't follow the math at all? Just because he uses an authoritative tone of voice and includes a lot of equations?

True I don't understand a lot of his math. And maybe that's his main problem - in a league of his own and precious few other specialists have the skills to even debate his findings. The conclusions are clear enough though. As he claims 'quantum weirdness' can be reproduced in an entirely classical arrangement, hopefully soon experimental results will put it on a firmer basis. Once again, if you are the expert, give us a point-by-point critique after actually reading his paper(s). I threw the links in here as food for thought, not as an excuse for a bashing. And I have never claimed either of them must be 'right' - OK!

 

[michelcolman]

The Kochen-Specker no-go theorem tells that particles cannot be in a definite state "all along", it must be the measurement itself that causes the particle to collapse to a definite state.

Read this for a simple introduction why: http://www.cs.auckland.ac.nz/~jas/one/freewill-theorem.html

OK, but what I was thinking was that, at the time of measurement, the particle decides to "have been" in some state since it was emitted. Retroactively. But if you don't measure it, it does not. Is that still impossible?

Otherwise, we're back to the initial question, when exactly does the twin particle take its definite state? If it's simultaneous, then in which frame of reference?

 

[michelcolman]

So then let's say we have two frames of reference and in each frame they are first. Now they won't agree who did the test first but they will agree with the results. Wouldn't they both be "real"? Seems like that would make sense as what happens when depends on your frame of reference.

how is it a problem that both can be considered first? there is no overarching master frame of reference where the "real" first happens. But in reality both are real. So just depending on the frame you chose you get the results.

I understand your point: both can claim to be first, it does not matter whether A's measurement influenced B's or the other way around, they just got the same result.

But would they not be able to tell that they received an already decoherent particle? I hear people say that the twin particle "simultaneously" takes its definite position, so I assume that there must be some way of telling whether or not that happened. Otherwise, why even talk of "spooky" action if the particle might as well have been in some state all along (or at least made up its mind ahead of time)? The fact that it's considered spooky at all, means there must be some experiment to actually verify the spookiness. Verify that the particle did indeed change at some specific moment. If you have that kind of experiment, the order of measurement would be important. If no such experiment exists, the particle may have been in some predetermined (hidden) state all along.

 

[Q-reeus]

My last entry, referring in part to Athanasios Prikas's paper at http://arxiv.org/abs/0710.1008" , contained:
"Performing 3 successive measurements on one particle forces the other particle to have reversed it's WELL DEFINED initial spin 'remotely'."
As stated that is wrong. There is no well defined initial spin state, only a well defined anti-correlation of spins. What I meant to and should have said was "...WELL DEFINED previous spin state (after the first measurement had been performed on the 'twin').." The essence being that by orthodox theory 'entanglement' survives until a JOINT measurement is made. Thus a succession of measurements on one particle implies definite correlations for the other particle at the instant of each such successive measurement, and the remote correlation (entanglement) only disappears following actual measurement of the second particle. Prikas's suggested remedy is probably wrong as the proposed correlation function would seem to clash with experimentally confirmed results. His main achievement is to point out the implications re energy/momentum exchange when a certain sequence of multiple measurements are performed. Most Alice-Bob scenarios only look at a single pair of measurements, and the discussion is about 'information', which seems less of a concrete problem than mysterious changes in energy/momentum as Prikas has demonstrated.

 

[John232]

If you assume a particle traveling close to the speed of light underwent spacetime dilation then that particle would have its worldline contract to zero and its time would be undefined, so then how could it not interact simultaneosly with other particles that shared its history?

 

[Phrak]

You are trying to make sense of quantum theory by assigning a physical process to "information acquired upon reduction of uncertainty". I use the terms 'information' and 'uncertainly' not in the quantum mechanical sense but in the sense of information theory. In other words, upon observation, more knowledge is obtained about a physical state than was available before an observation was made. This should be seen as a purely subjective phenomena until demonstrated otherwise.

There are many theoretical attempts to caste this objectively. The experimental physicists seem to do a better job at making peace with the absurd.

 

[Demystifier]

Nonlocality and relativity may peacefully coexist:
http://xxx.lanl.gov/abs/1002.3226
http://xxx.lanl.gov/abs/1006.1986

 

[Phrak]

Nonlocality and relativity may peacefully coexist:
http://xxx.lanl.gov/abs/1002.3226
http://xxx.lanl.gov/abs/1006.1986

What's the jist of the argument?

 

[Demystifier]

What's the jist of the argument?

That velocities faster than light are not in contradiction with the principle of relativity.

 

[John232]

You are trying to make sense of quantum theory by assigning a physical process to "information acquired upon reduction of uncertainty". I use the terms 'information' and 'uncertainly' not in the quantum mechanical sense but in the sense of information theory. In other words, upon observation, more knowledge is obtained about a physical state than was available before an observation was made. This should be seen as a purely subjective phenomena until demonstrated otherwise.

There are many theoretical attempts to caste this objectively. The experimental physicists seem to do a better job at making peace with the absurd.

I am sure they do make better sense out of it mathmatically. But, the lorentz is a factor in the mathmatical equations of light speed traveling particles. I asked how is there not a way information couldn't be simulateos for an object traveling the speed of light because I think that would be one bug in the explanation. Somehow a particle traveling the speed of light that would have contracted worldlines would end up being able to be detected at different points at the same time.
At the start of particle physics it was said that there is no way to know for sure about the frame of reference of a particle itself so then it was wrong to assume that it did expereince spacetime dialation, but what if it did in some fanshion, but how would we ever know for sure what a particle itself perceived?
So then the question would be what would separate a particles perspective from our own that would allow it to be observed with an approx speed and location when its worldline should be contracted to zero?

 

[michelcolman]

If you assume a particle traveling close to the speed of light underwent spacetime dilation then that particle would have its worldline contract to zero and its time would be undefined, so then how could it not interact simultaneosly with other particles that shared its history?

It could be simultaneous for one observer, but then it would not be simultaneous for a different observer.

Of course, from the point of view of a particle traveling at light speed, everything is simultaneous. Those reference systems always give nonsensical results. I'm talking about real observers, who can never travel at light speed. If one of those observers finds the entanglement effect to act instantaneously, it will take a finite amount of time for a different observer, and it will actually work backwards in time for yet another one. That's just a basic result of relativity.

So, as far as I can tell, the only ways of getting around this are:
- the effect happens simultaneously in some reference frame, for example the frame of the source emitting the entangled particles
- the effect happens in such a way that it does not matter when it happened, since nobody can tell the difference. However, how can researchers claim to have created entangled particles over some distance or time, then, if they can't tell when decoherence occurred?

 

[John232]

The forward motion does seem different than the affect it would have on the spin, but if you send an electron traveling close to the speed of light down a tube that will reflect the wave at a half wavelength it will not travel down that tube. Any voltage at the start of that path will not have voltage. It is as if the particle knows that it is a half wave length before it even travels down it. So then it could have some precognition, but still be seen to travel at a defenet speed.

There is something that separates the two reference frames so that we can observe the particle to react from its own frame of reference while at the same time we are unable to see the effects it does directly at the same. Traveling the speed of light for a particle doesn't create an effect where it is seen to travel at infinite speed even though its worldline is being crunched to zero in its own frame, but we still observe the effects it has on the particle itself. It's as if spacetime only contracts for it, but the effects caused by it affects what we see it do.