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

Entanglement, causality and local realism

  1. Jun 7, 2009 #1
    I have been walking around with the notion that "quantum entanglement implies a sort
    causal relationship due to the states having a shared history". I am now not so
    sure that this is correct.
    When I talk about entanglement, I use the argument that the preparation of the entangled state places information in the entangled state at the time of preparation and this "influences" future measurements. Where "this influences future measurements" in the last sentence means that the entangled states share a Hilbert space containing the eigenvectors of the joint Hamiltonian of the entangled objects. I'm not sure if I am way off course in my comprehension or if I have a problem with my language.
  2. jcsd
  3. Jun 7, 2009 #2
    In QM, an entangled system of particles is treated as a single entity. Observables calculated for the system CAN NOT and SHOULD NOT be attributed to parts of the system independently. Doing so leads us to silly paradoxes such as spokey action at a distance. Non-locality is inherent in the decision to consider distant particles as a single entity. There is nothing mysterious here.

    Just like if the system is "The sun, and an observer with a series of colored goggles", an observable could be the color seen by the observer. This color can not be attributed to the observer only or to the sun only without leading to paradoxes. For example, if you say the color seen by the observer belongs to the sun, then the observer will be able to change the "color of the sun" superluminally by changing the goggles they are wearing.
  4. Jun 13, 2009 #3
    Thank you for your help. I'm still working on getting the concepts clear. I appreciate the help. I have found that the short answer is NO, entanglement doesn't imply causality. I was referred to a paper http://arxiv.org/abs/quant-ph/9904042 that goes to the question of causality and entanglement.
  5. Jun 13, 2009 #4
    Just to add a little to mn4j's comments...When you say "...that preparation of the entangled state places information in the entangled state at the time of preparation..." your argument sounds remarkably similar QM arguments involving "hidden variables". This was essentially what Einstein,etal had stated in the EPR paper, so you are in good company. However, Aspect's experimental violation of Bell's Inequality has pretty much ruled out any explanation of non-locality (spooky action at a distance) that might be attributed to hidden variables.
  6. Jun 13, 2009 #5
    I strongly disagree with the above. Comparing Aspect's experiment to Bell's inequality is apples to rocks. The assumptions used by Bell in deriving his inequalities are so handicapped, that they amount to a straw man. Consequently, violation of said inequalities by Aspect type experiments, should not be surprising.

    Note that it can be argued (and more reasonably so) that violation of Bell's inequalities by Aspect type experiments demonstrate the non-validity of Bell's inequality. Thus the issue of non-locality and causality does not even begin to come into the picture until Bell proponents can prove that the inequalities are accurate representations of all possible deterministic local and causal systems. This has not been done. However, there are numerous demonstrations of the errors of Bell's inequalities.

    I will draw your attention to a recent one:
    Where the author shows convincingly that Bell blundered right from his first equation by assuming the wrong topology for EPR experiment.
    Last edited: Jun 13, 2009
  7. Jun 14, 2009 #6
    In deference to the author that originated the post...might we agree that the intention of the Aspect experiment was to test Bell's Theorem and that within the parameters of that test and at that time, it was accepted that Bell's Inequality was violated. Hence it is the preavailing belief (at this time) that Bell's Inequality was violated and that non-locality is a characteristic of the Quantum World...not to say that this is the final word on the subject.

    I will review the archive in depth when I get a chance...I note that it seems to have been updated on nunmerous occasions due to challenges.
    Last edited: Jun 14, 2009
  8. Jun 14, 2009 #7
    I appreciate all the attention given to my question.
    I am not trying to suggest there are hidden variables. That's where I run run into a problem with my language.
    What I am grasping at is how to express Bell or CHSH violations. Causality in my mind has to figure somewhere in the framework of entanglement. Can you just make measurements on arbitrary systems and then decide after the fact to communicate your results and do the statistics and test for violation? That wouldn't make any sense. There has to be some prior agreement or knowledge or belief and a way to meaningfully correlate the measurements. This process of agreement prior to future measurements "feels" like the notion of causality. I understand that with or without any prior agreement the measurements inherently are unaffected. Admittedly, I am getting into epistemology here.
  9. Jun 14, 2009 #8
    That Bell's inequality was violated by experiment is the only fact here. The rest is conjecture, trying to answer the question "Why was Bell's inequality violated?". The issue I and many others have raised is the fact that the so-called "prevailing belief" is misguided and unjustified.

    Think about the following silly example for a second:

    1) Canoe comes up with a set of inequalities which we shall term "Canoe's inequality". It states that for all real numbers, n + n < 2n-1
    2) Experiments are performed on real numbers showing that n + n = 2n
    3) Canoe concludes that experimental violation of his inequalities prove that numbers are not real.

    As you can see, the conclusion (3) neglects completely to discuss the validity of (1). It just so happens that in this case, the non-validity of (1) is obvious. However in the case of Bell's inequality, it is not so obvious to the untrained eye that Bell's inequalities are not valid. It is not surprising that nobody has EVER performed a macroscopic locally real experiment which proved the validity of Bell's inequalities.

    As I explained above, violations of the Bell of CHSH can only mean the inequalities do not correctly model the system being measured. The only way causality or locality come into the picture is IF and only IF, Bell or CHSH inequalities have been proven to EXHAUSTIVELY and ACCURATELY model all POSSIBLE locally causal systems. This has not been done. In almost half a century, since Bell's work, nobody has experimentally verified that Bell's inequalities are correct. So it is mind-boggling that Bell proponents purport to say how nature must be, based on an unverified mathematical theory such as Bell's or CHSH inequalities.

    It seems to me you are trying hard to explain how causality comes into the picture. But all of this is irrelevant to the issue once you understand what I have explained above, because the idea that all possible hidden-variable theorems have been forbidden is just fantasy. In any case, you are right that is possible to have epistemological dependence (logical dependence) in a situation in which there is no ontological causal relationship.

    For example consider two clocks located at opposite ends of the universe. They will be correlated by virtue of their being clocks, even if they run at different speeds. If time-tagged data of each clock were collected during the same time period and analysed together, a correlation will be observed even if it was impossible for information to travel from one to the other during the measurement.
  10. Jun 14, 2009 #9
    I appreciate all the responses. And I probably don't understand all of what is being said.

    I am aware of the open question related to CHSH. My intuition also tells me that there are no hidden variables.
    Your comment about clocks implies there has to be a notion of a clock and time and have these notions before make any observations. It also implies you know of the existence of both of these before hand and know to bring information together related to both of these. I don't see any hidden variables there. I do see a big belief system being brought to borne (sp?)

    Maybe I don't really understand what causal means?
  11. Jun 14, 2009 #10


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    That's an exaggeration at best, and crazy talk at worst. I'm still not sure where on that scale your comments are.

    That doesn't make any sense.

    Wrong. Locality is still an issue, because when you derive the inequalities, you assume that the value of the quantity you're going to measure doesn't change when you rotate your measuring device.

    Thanks for posting that. It looks interesting. Interesting enough to deserve its own thread actually. I think I'm going to start one.

    Edit: I did, but it got deleted, and I got a formal warning for violating the forum rules. :smile:

    That's not what he's saying. He's just saying that the set S0 appears (implicitly) in one step in Bell's argument, where e.g. S3 would have been an equally valid choice. (I still haven't understood why it is, but I'm working on it).

    Come on, that's not even an argument. Do you think straw man arguments are valid when you use them?

    First of all, it isn't possible even in principle to "prove the validity of Bell's inequalities" by performing experiments. Second, I agree that it's not surprising, but for a very different reason. My reason is that QM predicts Bell inequality violations. If QM can kick CM's *** in every other way, we have no reason to expect the opposite when we do these experiments.

    Again...not possible even in principle.
    Last edited: Jun 15, 2009
  12. Jun 14, 2009 #11
    Again thanks for all the replys.

    Does anyone have any references on Quantum Mechanics, Probability and the notion of causality? (Preferably something outside the Copenhagen Interpretation.)

    Does anyone know if the Quantum Bayesians have touched on the notion of causality yet?
    I find their work really exciting.

    I'm getting a very wild picture of the Universe the likes of which may knock my hair back. I haven't had that experience in a long time :) (Sorry, I had to mention that I really do get off on this stuff).
  13. Jun 14, 2009 #12
    Bell's theorem assumes that any local hidden variable formulation of a GLOBAL experimental situation MUST be analyzable into its parts (expressable in factorable form) in order to be considered a local hidden variable formulation.

    In effect, the data sets accumulating at A and B in the archetypal optical Bell test are to be treated as independent by any lhv formulation.

    But of course this is, obviously, wrong. The data sets at A and B are not independent. Entanglement experiments are designed to produce dependent, spatially separated data sets -- the preparation of which entails some sort of common cause parameter/property which is the basis for the entanglement formulation (nonfactorable) and the data correlations.

    So, what does an experimental violation of a Bell inequality tell us? Not much beyond the fact of the violation (though it can be used as a rough indicator of the presence of engtanglement). It doesn't mean that anything FTL is happening, or that nothing exists in the channels connecting emitters and detectors (the hidden variable thing). Obviously, something is going on between emission and detection -- it's just that nobody knows how to talk about it any more precisely than the standard QM formalism (which pretty much ignores the question of what's happening between emission and detection).

    So, our intuitive notion of the meaning and cause of quantum entanglement is alive and well.
  14. Jun 14, 2009 #13


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    The books by Ballentine and Isham both have good discussions about probability. Ballentine argues (convincingly in my opinion) that we shouldn't think of a state vector as representing an objective property of the system, and instead as a representation of an objective (statistical) property of an ensemble of identically prepared systems. (I prefer to say that a different way: QM doesn't describe the world. It's just an algorithm that tells us how to compute the probabilities of possible results of experiments, given the results of other experiments). Isham seems to agree with that, but he still makes an effort to always explain both how an "anti-realist" (like Ballentine) and a "realist" thinks. It's interesting to get both perspectives.

    I don't even know what that is.

    That's why we're here. :smile:
  15. Jun 14, 2009 #14
    A model of the Universe that could describe the causality of non-locality (among other things) would be very "wild" indeed. Some basic requirements would seem to have to be:

    1) It would have to be Relativistic.
    2) It would have to be Indeterministic.
    3) It would have to be expanding at a rate that was ever-increasing.

    Good luck with your search.

    and...kudos to Fredrik for speaking up while I was absent. He is correct. We shouldn't be surprised by the violation of Bell's Inequality. It is what is predicted by QT..and QT is arguably the most successful theory ever (IMHO - not looking for a debate here).
  16. Jun 14, 2009 #15
    Fredrick thanks for the references. I'll check them out. What is causality I now is a bizarre question but, it's those simple ones that can be a real journey :)

    The Bayesian are C.A. Fuchs, Carleton Caves, Asher Peres and others. The group is working on a new interpretation of QM. If you want to feel like a rebel follow their work :) (Admittedly, I'm a big fan of Fuchs' ideas, and the Bayesians. I've never been comfortable with the Copenhagen Interpretation, I don't know why but it always seemed inelegant Gasp! to me. It felt almost like the physics was so weird that there really wasn't language to express the ideas. That it seems to be beyond the reach of allot of people, even the smart ones :))

    Thomas, thanks for the reply. And again, I do understand that CHSH hasn't been shown to work for all dimension Hilbert spaces. I agree with your views and I found your exposition very clear.
    I do have an issue with talking about what happens during emission and detection. I don't think it makes sense to talk about what cannot be observed and observation counts as detection. To talk about what can't be observed and to go further an try to ascribe anything to such a situation really is more metaphysics than physics.
  17. Jun 14, 2009 #16
    Canoe thanks for the reply.

    To connecting causality and non-locality, your reply about connecting the two being non-deterministic is intriguing. Why do you say that? I can see stochastic but not non-deterministic? Maybe this is semantics on my part?
    In the end, I fear that even talking about causality and non-locality is very dangerous. I'm not thinking there is a direct connection between causality and entanglement like in hidden variables. I don't even think they communicate out side what the experimenters chose to communicate. And maybe that's all there is to it. Causality in this framework would be what is believed will happen or has happened. It's still constrained and not chaos. Then again I'm still wrestling with all this :)

    I don't know why, other than physics has to fit experiment and be able to predict, you say the other things.
  18. Jun 14, 2009 #17

    I know I read somewhere (I can't remember quite where but I suspect i could locate it if need be) that Bohr went off at Heisenberg when the latter espoused the idea that the act of measuring caused uncertainty. In otherwords, there is an objective reality but we are just too clumsy to precisely measure it without causing a disturbance to the system. That would be stochastic.

    However, conventional QT, whichever interpretation you favor, prescribes an underlying indeterminism to nature. Hence my very general comment that any succesful model of the universe would seem to need to explain "Uncertainty".

    As to the other things I mentioned:

    A model of the universe that somehow would incorporate both indeterminism and relativity would, at some level, unify physics.

    And as to my comments about the universe expanding at an ever-increasing rate, this has actually been observed and verified by astronomers in the last 10, or so, years. I think it would be a fair statement to say that astrophysicists are somewhat stumped by this discovery.
  19. Jun 15, 2009 #18
    Canoe, thanks for the reply. I don't know if I agree with the notion of an objective reality. I do believe in a consensus on reality.
  20. Jun 15, 2009 #19
    I think one of the terms bandied around in this type of discussion is "counterfactual definite", which is a mouthful that basically says while observing something changes the outcome, there would have been a defined outcome if you had not observed it as well.

    Essentially it can be summed up as realism, so the options are non-local but real (counterfactual definite), or local but without a definitely real state.

    Either could fit the observations, but experiments, particularly Bell Inequality tests, show that the Universe does not fit BOTH of those parameters. One must be wrong, and it just so happens that modern quantum theory "makes more sense" under the non-local but real assumption, which also allows it to avoid hidden variable postulates.
  21. Jun 15, 2009 #20
    Jaynes, E. T., 1990, `Probability in Quantum Theory,' in Complexity, Entropy, and the Physics of Information, W. H. Zurek (ed.), Addison-Wesley, Redwood City, CA, p. 381; http://bayes.wustl.edu/etj/articles/prob.in.qm.pdf
  22. Jun 15, 2009 #21
    I know, it is not fashionable to point out the problems with Bell's inequalities. If you can't tell already, I'm not fazed by however you may choose to characterize my comments.

    Then you must be talented at ignoring the obvious. Violation of Bell's inequality by real experiments means ONLY that Bell's inequality does not accurately model said experiments. What else did you think it meant?

    Wrong. Locality is still an issue, because when you derive the inequalities, you assume that the value of the quantity you're going to measure doesn't change when you rotate your measuring device.[/quote]
    Look, you need to study some basic logic.

    Your argument goes like follows:
    1) Bell's inequalities are valid for all local realist hidden variable theorems.
    2) Violation of Bell's inequalities by experiments show that local realist hidden variable theorems are impossible.

    I say claim (1) is unproven. If you think it has been proven, show me where. And until you can prove (1), discussion of (2) is a red-herring and completely irrelevant. On the contrary, there are published examples of local realist hidden variable theorems which violate the inequalities. You only need one such example to crush Bell's dreams. I will give you three:
    * Clifford Algebra Valued Local Variables violate Bell's inequalities: Adv. Studies Theor. Phys., Vol. 1, 2007, no. 12, 603 - 610
    * Event-by-Event Simulation of Einstein-Podolsky-Rosen-Bohm Experiments: Foundations of Physics, Vol. 38, Issue 4, p.433, 2008.
    * Exclusion of time in the theorem of Bell. K. Hess et al 2002 Europhys. Lett. 57 775-781 ; "Breakdown of Bell's theorem for certain objective local parameter spaces", PNAS February 17, 2004 vol. 101 no. 7 1799-1805
    No! He is saying S0 is not valid. Read it when you get a chance.

    Interesting that you said that because I was showing how straw-man arguments are not valid. And Bell's theorem is a giant straw-man.

    If Bell's inequalities can not be proven, then we never get to claim (2) and Bell's theorem therefore fails.

    Hehe, I like the way you phrased this one. I will say rather that Bell's inequalities FAIL to account for the predictions of QM, and the results of real experiments. Isn't it obvious that the only odd man out is Bell's inequalities. Everything else agrees with each other. (QM agrees with experiments and vice versa). Bell's inequalities are beginning to look like a trojan horse.

    Now I begin to understand where you are coming from. It's probably all about kicking CM's *** to you. Experimentalists can not only measure individual events in a double slit diffraction with electrons, they can also measure very accurately the time-order in which they arrive on the screen. I will only entertain such boasting and fanboyism when QM is able to predict those details which have been known for decades. For the moment though, the logic of the argument in favor of Bell's theorem is seriously lacking.
  23. Jun 16, 2009 #22


    User Avatar
    Science Advisor
    Gold Member

    Bell designed his assumptions to match normal, reasonable definitions of locality and realism. That is why you are in the minority. Since Bell, virtually every physically possible test of entanglement has told us the exact same thing. Namely, that nature is not local realistic.

    Like you, I have references for "disproofs" of Bell. Those would be dwarfed by the repeatable experiments supporting Bell, which are easily over a thousand at this point. And please note this: there is not a single experiment yielding a value for entangled correlations that is NOT in substantial agreement with the predictions of QM and in violation of the predictions of ANY proposed local realistic theory. That includes all of the references you provided.
  24. Jun 16, 2009 #23


    User Avatar
    Gold Member

    virtually every physically possible test of entanglement - would be quite a stretch.
    Haven't heard of entanglement experiments that had taken detection loophole into consideration.
  25. Jun 16, 2009 #24
    Prove it. Show me a single experiment with locally real macroscopic objects which validates this claim, failing which it is a mass of hot air. And this is the crucial point. Unless you can prove that Bell's assumptions are exhaustive and inclusive of all possible local hidden variable theorems, you can't even begin to attribute any validity to Bell's theorem.

    At one time, the majority also thought the earth was flat. Like I thought, you have not even grasped the very basics of my argument you quoted above. Yes, experiments have validated the predictions of QM. So what? The issue before us is to show an experiment that has validated BELL's INEQUALITIES!

    Hehe, you need to read my argument again and make sure you understand it before you purport to provide experiments (which all violated Bell's inequalities btw), as proofs of Bell's inequalities.

    1) The agreement between QM and experiment has NEVER been an issue, so stop diverting attention to it as if it ever was.
    2) All three references I provided show that local realist theories which agree with QM and experiments and VIOLATE Bell's inequalities are possible. Don't make claims about articles you have not read. If you want to discuss them in a non-hand-waving manner, let's do so.
    3) Note this: There is not a single experiment confirming the validity of Bell's inequalities.
  26. Jun 16, 2009 #25


    User Avatar
    Science Advisor

    Rowe et al, "Experimental violation of a Bell's inequality with efficient detection", Nature 409, 791-794 (15 February 2001)
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook