Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Relative time in light of quantum entanglement experiments

  1. Apr 12, 2014 #1
    I have a question on the tension between special relativity and quantum mechanics, so please correct the category if this question is in the wrong location.

    I was looking at the write-up of an experiment: “Causality, relativity and quantum correlation experiments with moving reference frames” by H ZBINDEN, J BRENDEL, W TITTEL and N GISIN, Group of Applied Physics, University of Geneva, 1211 Geneva 4, Switzerland in 2001.


    The abstract summary is as follows:
    Abstract. Entanglement, one of the most important features of quantum mechanics, is at the core of the famous Einstein–Bohr philosophical debate [1] and is the principal resource for quantum information processing [2]. We report on new experimental investigations of the properties of entangled photon pairs with emphasis on the tension between quantum mechanics and relativity [3,4]. Entangled photons are sent via an optical fiber network to two villages near Geneva, separated by more than 10 km where they are analyzed by interferometers [5]. The photon pair source is set as precisely as possible in the center so that the two photons arrive at the detectors within a time interval of less than 5 ps (corresponding to a path length difference of less than 1 mm). This sets a lower bound on the ‘speed of quantum information’ to times the speed of light. Next, one detector is set in motion [6] so that both detectors, each in its own inertial reference frame, are first to do the measurement! The data always reproduces the quantum correlations.

    My question does not relate to the standard tensions between quantum mechanics and special relativity explored as a consequence of the EPR paradox, but rather relates to the question of temporal conflicts between SR and QM.

    According to Einstein, time is relative. In different frames of reference, time is measured at different rates and as a consequence there is no such thing as a universal present moment. The present moment is a human concept, significant psychologically only.

    However, if I am interpreting the data from the above experiment correctly, there would indeed appear to be a present moment. The present moment being defined as the moment of correlation between the entangled partners.

    If I did a thought experiment and had millions of entangled particles and placed them all over the universe in different temporal locations (i.e. gravity wells, fast spaceships etc) their moment of correlation would appear to be impervious to their relative time.

    It would seem to me that a universe that has relative moments and a universe that has a universal present moment are mutually exclusive conditions.

    I had a look in the SR section and the experimental test maintained by Tom Roberts at http://www.edu-observatory.org/physics-faq/Relativity/SR/experiments.html
    But I could find no mention of tests that test out relative time versus present time.

    Quantum entanglement experiments of particles in temporally separated frames of reference would suggest there is such a thing as a universal present moment. On the other hand Einstein’s concept of relative time emphatically rejects such a concept. Unlike the basic EPR paradox there does not appear anyway around this mutually exclusive/antagonistic set of conditions.

    This apparent temporal conflict between QM and SR appears to get no airplay.
    My question is what am I missing?
  2. jcsd
  3. Apr 12, 2014 #2
    I guess some interpretations of quantum mechanics require absolute simultaneity. If we think that some hidden message is transmitted behind the scenes during the time of measurement, this would imply that it must be absolute which of the entangled particles is measured first. By relativity of simultaneity it's frame-dependent which particle is measured first, but it makes no sense (if that argument is of any value in modern physics) that the direction of message would also be frame-dependent. One must be sender, the other is receiver.

    Bell experiments have quite strongly ruled out local hidden variable theories, but thinking a faster-than-light hidden message from one measured particle to other could be another wrong way to look at it. So the information is not carried with particles, but it's not transmitted either, but there still is a mysterious correlation between the results. I don't know, I'm just confused.
  4. Apr 12, 2014 #3
    Thanks very much for the reply Ookke.

    I am familiar with the possibilities regarding information sharing and hidden variables. What I find intriguing is that the correlation happens in what we would classify as absolute time rather than in relative time. The experiment I reference clearly shows that correlation is independent of relative time. While we may be confused as to the mechanism of transmission and while we may accept that there is no violation regarding information sharing between Alice and Bob, what seems inescapable is that the concept of relative time is breached by such experiments. The authors actually title their associated press communique as “Quantum correlations that are not sensitive to space and time” which seems a bizarre mystical sort of statement. What they meant to say (temporally speaking) I think is “Quantum correlations that are not sensitive to relative time but are sensitive to absolute time”.

    While the mechanism of correlation might cause confusion, there seems to be no confusion regarding temporal alignment of correlation. This appears to be a black and white refutation of relative time as the predominant descriptor of the temporal nature of our universe.

    So again I think there is either a fundamental problem with the concept of the universe being described fully by relative time with no place for absolute time and therefore a universal present moment, or I am missing something in my analysis of this experiment.
  5. Apr 12, 2014 #4


    Staff: Mentor

    No, there's nothing in these experimental results that requires that interpretation. Remember that the correlations are not actually known until the data from the two detectors is compared, and that comparison always happens slower than light (i.e., both detection events are in the past light cone of the comparison event).

    Also, the fact that the experiment was re-done with the detectors in relative motion, and the correlations were the same, rules out the possibility that any single "present moment" can be determining the correlations, because the two detectors do not have a common "present moment" when they are in relative motion.
  6. Apr 12, 2014 #5
    Thanks for your response Peter. I appreciate that the comparisons are done post-event and consequently rely on analysis of past events from the perspective of the observer.

    However, I disagree with your thinking and therefore your conclusion in the second paragraph of your response.
    What I think you are saying here is that because SR says that if we have things in relative motion there can be no common present moment therefore that rules out a common present moment. If I have got the crux of your argument wrong please correct me.

    Let’s forget for the moment what the underlying mystery mechanism is that causes the correlation and just focus on the timing of the correlation.

    As the detectors are set into relativistic motion we should see the correlation when measured become random as the relative times of the detectors moves out of alignment. What in fact we find is that correlations are maintained between the two detectors. If time is purely relative as SR suggests the results should be random. They even set up the experiment to effectively measure the particles in a before / before (according to SR rules) arrangement. Obviously in such a circumstance if we rely on relative time, the results should be random.
  7. Apr 12, 2014 #6


    User Avatar

    Staff: Mentor

    (addition in bold is mine of course)

    Maybe the results "should be" random, but they aren't (also confirmed by experiment many times over). Thus, we have to question the line of thinking that says that the results "should be" random... Why are you so sure that they should be?

    There's nothing in the mathematical formalism of quantum mechanics that says that my spin-up measurement of A caused B to be spin-down, or vice versa. There's nothing in the formalism that cares about whether the measurement of A preceded the measurement of B, or the other way around.

    Instead QM says that the situation in which A is measured first is indistinguishable from the situation in which B is measured first and from the situation in which they are both measured at the same time. That's not an argument for a universal present time, it's more of an argument that the concept isn't needed.
  8. Apr 12, 2014 #7


    Staff: Mentor

    What you say here is correct, but it's not quite the crux of my argument. See below.

    Can you demonstrate that this follows from quantum mechanics and SR? Because if you can't, then your claim here carries no weight. And I'll be *very* surprised if you can, because it's a pretty basic theorem in relativistic QM (meaning, the sort of thing that undergraduates get assigned as a homework problem) that field operators commute for spacelike separated events, from which it follows immediately that the correlations in experiments like the one under discussion should be the same whenever the detection events are spacelike separated, regardless of the relative motion of the detectors.
  9. Apr 12, 2014 #8


    Staff: Mentor

    More precisely, it says that all these situations are indistinguishable as long as the "A measured first" and "B measured first" cases still have the detection events spacelike separated.
  10. Apr 12, 2014 #9


    User Avatar

    Staff: Mentor

    Does it matter as far as the quantum mechanical measurements of entangled particles are concerned? That's a question, not an argument.

    We get the same correlations even when they're time-like separated and one measurement had to come first; I can't tell whether I am looking at timelike-separated or spacelike-separated measurement events just by looking at the correlated results. (Of course there are about 83 bazillion other ways that the distinction between space-like and time-like separation matters - I'm just not sure that this is one of them).

    It's worth noting that if I do my entanglement experiment with photons emitted in opposite directions, the detection events must necessarily be spacelike-separated. But if I use spin-entangled particles with non-zero rest mass, the detection events can be timelike-separated. I can't see that entanglement cares about ordering at all, even when ordering is present.
    Last edited: Apr 12, 2014
  11. Apr 12, 2014 #10


    Staff: Mentor

    In this particular scenario, it may not matter, yes. But in general quantum field operators do not commute at timelike separations; if the particular operators of interest in this experiment do, in fact, commute even if the detection events are timelike separated, that's something that would have to be established by looking at the particular operators. Whereas the proof that field operators must commute at spacelike separations is completely general; you don't even need to know what operators they are.

    In general, it certainly does. Events in a causal sequence are entangled; for example, the event of my throwing a baseball and the event of the baseball breaking a window are entangled. But the operators associated with those events certainly don't commute.
  12. Apr 13, 2014 #11
    Hi Peter and Nurgatory,

    Thank you both for your kind and educational responses.

    Just to be clear I am not trying to argue with SR experimental results, nor am I trying to argue with quantum mechanical results. I am just trying to understand the implications of entanglement on our understanding of time.

    I accept from a QM perspective that the order of measurement makes no difference to the outcome. It is effectively looking at either side of the same coin. It is the robustness of QM with respect to this phenomenon that prompts my question. From a QM perspective the measurement can be regarded as time agnostic. As noted in the experiment:

    ‘The conclusion is that correlations in the quantum world have their own causes, which cannot be reduced to those of the events, and are insensitive to space and time.’

    But where I seek to be enlightened is how SR can claim the same privilege with respect to time as QM. Is not relativity of time intrinsic to the fabric of SR? Therefore does not a classical SR explanation not only fail in providing an explanatory mechanism for quantum entanglement (as per Bell’s theorem), but does it not in fact fail from a temporal perspective in aligning with the results?

    Without reaching out to QM as an explanatory saviour, if we look at the experimental results from a classical perspective, are they not just as inconsistent with the concept of relative time in the same way they are inconsistent with trying to rely on a classical mechanism to explain the correlation of spatially separated components of a composite system as captured in Bell's Theorem?

    And Peter I am sure I could not commute your shoelaces. Excuse me if my previous response sounded impertinent, it was not meant to be.
  13. Apr 13, 2014 #12


    Staff: Mentor

    I think that you are interpreting the data completely backwards. The data shows that the temporal ordering of the measurements doesn't matter. Either can be first and the experiment turns up the same. As predicted by the relativistic theory of quantum mechanics.

    You are missing Quantum Field Theory. The "apparent temporal conflict" as you put it got plenty of airtime many decades ago, but then it was resolved in the 1950's or so. Most conflicts from the 50's get little airtime today. Modern QM is completely compatible with SR, as this experiment shows.
  14. Apr 13, 2014 #13


    Staff: Mentor

    This seems to me to be an overstatement. As I noted in my responses to Nugatory, the correlations are only "insensitive to space and time" in general for spacelike separated events, i.e., events which cannot, in classical terms, causally influence each other. That makes a big difference; see below.

    No, because the condition for the correlations to be "insensitive to space and time" in QM is precisely the *same* condition as it is in SR: that the events be spacelike separated.

    More precisely: QM says the correlations are insensitive to the order of events for spacelike separated events; SR says that all physical observables are insensitive to the order of events for spacelike separated events. The correlations are physical observables, so QM and SR are in complete agreement.

    No. Looking at the results "from a classical perspective" just means using SR without trying to bring in QM. And, as above, SR says that physics is independent of the ordering of events for spacelike separated events. So SR says that, whatever the mechanism is that explains the correlations, it must be insensitive to event ordering, and that's exactly what the experiments show to be the case.

    No worries, I didn't think it was impertinent. :smile:
  15. Apr 13, 2014 #14
    Ok, but aren't spacelike events associated with superluminal entities?
  16. Apr 13, 2014 #15


    Staff: Mentor

    Right now (in any inertial frame) my right index finger is spacelike separated from my left. Neither finger is moving superluminally.
  17. Apr 14, 2014 #16
    Thanks very much Dale and Peter for your responses.

    Dale I am ignoring QFT for the moment as I missed the 50's. :-) While, I appreciate how it extends SR into QM realm, I want to concentrate on SR purely in its classical form for this line of questioning.

    Peter, I have not quite given in to your persuasive argument. While I understand SR insensitivity with respect to space-like separated events, I still think there is a temporal flaw in SR exposed by such experiments as the one I described.

    From a QM perspective time is irrelevant with respect to qe. If you collapse one particle, you have acted on the system, hence sequence is neither here nor there, and I understand how the same applies to SR.

    What experiments such as the one I described confirms, are that this system collapse occurs instantly over space-like distances (as close to instantly as it is possible to determine) and that the collapse is not influenced in anyway by concepts such as relative time.

    Where I see SR failing is that there is no such thing as instantly. Instantly does not make sense in SR across separate frames. Instantly does not exist.

    Say we had particle A positioned here on earth and we had B on Voyager 1, I know that when I measure A, B will have a known state immediately (whether or not I go and measure it). I am not looking at sharing information or anything like that (in fact I don't even care about the information), but I know that right at that moment in time I have acted on B by acting on A.

    Now you might say that is because the collapse is insensitive to time. While that makes sense from a QM perspective I can't see how it makes sense from an SR perspective. From an SR perspective when is now? when is when?

    So the problem is not information sharing or communication (superluminal or otherwise), the problem is that the timing of the collapse happens at a system-level instantaneously across space-like separations and inertial frames.

    While SR can handle the collapse formulaically because ultimately there is no breach in its rules regarding space-like separation communication I am not convinced it handles (through omission) the temporal implications of the collapse.

    We can say that like QM, SR is agnostic to the timing and therefore SR still works. However, what do we call that shared moment in time where the entangled particles collapse. With QM we are from a system perspective coming out of the quantum realm into the classical realm through measurement so it is an irrelevant question. But we do not have that luxury in classical SR, there is no quantum realm, there is only the classical realm where time is relative.

    We have no descriptor it seems in SR for that shared moment across inertial frames of the system collapse. This common 'present' moment is missing in SR, and therefore I would say that SR as a full descriptor of time from a classical perspective would appear to fail as highlighted by temporal-style QE experiments.

    For me, if I judge classical SR purely from a temporal viewpoint it looks like Einstein painted the trees and the grass, but left out the sky.

    What am I missing?

    I really appreciate the time, energy and intelligence that has gone into responding to my questions to date and hope that I am not wasting people's time here with this line of questioning.
  18. Apr 14, 2014 #17


    Staff: Mentor

    But you can't if you're going to combine SR with QM, which you have to do in order to analyze the experiments you're describing. Most of your post is simply trying to apply non-relativistic QM reasoning to a relativistic scenario, which doesn't work.

    Meaning, the correlations are the same even if the collapse does not occur "instantly", but takes some time to happen, because you are analyzing things using a different frame. But that means your use of the word "instantly" is inappropriate, since the collapse only occurs "instantly" in a particular frame; there is no relativistic invariant that corresponds to "instant" collapse.

    Also, all this talk about collapse only makes sense if you are using an interpretation of QM that includes collapse. A no-collapse interpretation, such as the many-worlds interpretation, doesn't include anything corresponding to what you're talking about here.

    The fact that you can't combine QM and SR in the same scenario, and then try to analyze it using "classical SR". That doesn't work.
  19. Apr 14, 2014 #18


    Staff: Mentor

    This is wrong on several levels.

    First, on this forum you don't get to just ignore a mainstream scientific theory by fiat, particularly not when it is the theory which directly answers your questions.

    Second, by deliberately ignoring a theory which answers your question you make yourself look like you are simply seeking a baseless argument for the sake of argument rather than honestly trying to learn.

    Third, there is no entanglement in classical physics, so your question cannot even be asked classically.

    Fourth, QFT does not change SR in any way, it modifies QM to be compatible with SR, not the other way around. So even if your question made sense classically the SR part of the answer is the same as the QFT answer; the framework of spacetime used in QFT is "classical" SR.

    Thread closed.
    Last edited: Apr 14, 2014
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook