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

Superposition and retrocausality

  1. Oct 30, 2014 #1
    Hello guys,

    I made this analogy for double slit experiment. I would like to share this idea and I would like to hear your opinions and ideas. Thank you. I apologize for any typos or bad grammar, I am not native english speaker.

    Here it is:

    Analogy for wave collapse

    I am using term KNOWER instead of OBSERVER because delayed choice quantum eraser has shown us that not act of observing change the outcome, but very knowledge of which path information does.

    Imagine we ask a random person a question we know the answer for, but he does not. We will give that person two options (A or B), both equally like to be the right answer.

    If the person doesn’t know the right answer, it means that both the answers (A and B) can be the right one. In other words, THE RIGHT ANSWER is in superposition of two states (A and B).

    When we tell to the person which answer is the right one (let’s say its answer A), it immediately causes collapse and leaves to the person only the one right answer. Answer B is no longer right, it collapsed to category WRONG ANSWER.

    We represent a particle

    A person represents knower

    Two options represent double slit

    Not knowing the right answer represents wave


    As we know which answer is the right one right from the beginning, is is obvious that superposition does only exist in relation to knower. The right answer has been always only one. Having more than only one option to answer creates possibilities, but it doesn’t mean that there is more than one right answer at the same time.

    This might also explain retro causality. It’s exactly the opposite. It is not that person who finds out the right answer immediately tells us which one is it, we had known the right answer beforehand. Particles had known the outcome before which path information was obtained.

    IN CASE OF QUESTIONS ABOUT DELAYED QUANTUM ERASER: this way we can imagine we use a machine to record the right answer and pass it to the person without yet hearing it. The outcome is already set. If the person destroys the record machine, he has still a wave of possibilities A or B to be the right answer. If the person hears the recorded answer, it gives him only the one right answer.

    As to explain how the particles can know the WHICH PATH information before they are actually sent into slit, I dont have an answer for that. Especially difference between quantum mechanics and classical physics. As if in a quantum world, a person hearing a question always either know or not know the right answer, whether in macro world hearing the same question means you know the answer immediatelly.
    Last edited: Oct 30, 2014
  2. jcsd
  3. Oct 30, 2014 #2


    User Avatar

    Staff: Mentor

    You are basically suggesting that there's a right answer to the question (and "knower" knows what it is) and that the appearance of superposition is the result of the other person not yet knowing this answer.

    That feels like a sensible interpretation, and if you had suggested it in 1935 Einstein would have agreed with you - google for "EPR quantum" to find the classic EPR paper.

    However, thirty years later John Bell (google for "Bell's theorem" and pay particular attention to DrChinese) proved that some quantum mechanical phenomena cannot be explained in this way. Under certain conditions the quantum mechanical prediction and observed behavior cannot be reconciled with the idea that the superposition had a definite value that we just didn't know.
  4. Oct 30, 2014 #3


    Staff: Mentor

    Cant understand why you don't think knowing which path it took is not an observation. In fact the delayed choice experiment is simply an experiment showing in simple cases decoherence can be unscrambled:

    Knower is a poor choice because it implies conciousness which most definitely is not required. Even observation has that semantic difficulty implying observer - but it common terminology so nothing can really be done about it, except to be 100% clear on its meaning in QM.

    QM cant be modelled by classical probability theory for all sorts of reasons. The following paper explains the continuous transformation reason:

  5. Oct 31, 2014 #4


    User Avatar
    Gold Member

  6. Oct 31, 2014 #5


    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    In my opinion Qbism is very misleading. It suggests that the quantum state implies a non-obejectiveness (subjectiveness) in our very possibility what to know about Nature. That's misleading, as was explained by Nugatory above.

    The mathematical description of quantum theory and the connection with observables due to Born's Rule is the core of quantum theory, and it describes objective knowledge about objects. The quantum state is formally represented by a statstical operator, and physically it represents (an equivalence class of) preparation procedures on the objects described by the formalism. E.g., it is a clear procedure how to prepare a particle beam of some sort (say protons at the LHC) with a quite well defined momentum distribution. Nothing of this is subjective. You can measure the momentum of the particle with clearly defined procedures too and verify objectively, whether you succeeded with preparing a beam of particles with the desired momentum.

    Qbism is just another interpretation of quantum mechanics on the level of metaphysics. It uses the socalled Bayesian viewpoint to interpret the probabilities predicted by quantum theory. The Bayesian idea is to make sense of the metaphysical meaning of probabilities for single events (or small ensembles). In practice this is a rather empty approach, because physics is about making predictions about the outcome of real experiments in the lab, and since these predictions in the case of quantum theory are probabilistic, you must "collect statistics" to make (objective!) statistical tests whether the predicted probabilities are consistent with the observations in nature. This naturally (in the literal sense) brings you back to the frequentist interpretation about the meaning of the mathematical concept of probabilities in terms of statistical analysis of data. I don't see any merit of the Qbism approach for physics practicioners (theorists or experimentalists alike). It rather obscures the meaning of the quantum theoretical formalism than brings any further insight! As I stressed above, this in itself is my personal point of view, and in some sense also metaphysics ;-).
  7. Oct 31, 2014 #6
    How can consiousness not be required? How do you KNOW something( or observe something) without realizing you do so?
    And the reason why I chose KNOWER instead of OBSERVER is because observing doesn’t necessarily leads to knowledge. Imagine there is a camera that can record trajectory of fired photon through double slit. Now imagine we set this camera to record this experiment from the perfect side view( 90degrees), that way we create 2D illusion to observer. This way we can see the whole process of photon being fired, passing through slit and hitting the wall behind. But we cannot see which slit it went through, therefore I think it would create interference pattern( although this is thought experiment, I cant prove it). But it would show difference between observation and knowledge.
  8. Oct 31, 2014 #7


    User Avatar

    Staff: Mentor

    Neither knowledge nor observation are necessary to eliminate the interference pattern. All that is necessary is that the incoming particle interact with something else in any way that is affected by which slit it goes through.

    The words "observer" and "observation" are widely used, but that's a historical accident. In the earliest experiments, the "something else" was always a detector that someone would eventually look at, so people fell into the habit of saying "observation" when "interaction" would have been better.

    How would a camera (or any other device) record the passing of a photon without somehow interacting with it? You will have to specify how the photon interacts with the camera before we we can say whether the interaction is such that an interference pattern will be formed or not. But either way, it is completely irrelevant whether anyone ever looks at the results gathered by the camera.
  9. Oct 31, 2014 #8
    Well i thought that quantum eraser is about avoiding interaction, and that knowledge of which path is enough to eliminate the interference pattern.

    I meant is as pure hypothetical device. But let´s consider this instead: I do the double slit experiment with measuring apparatus. Let´s say result with WHICH PATH information goes into one envelope and the result with PATTERN information goes to another. If I open WHICH PATH envelope first, then I know which slit photon went through, and thus PATTERN result in the other envelope must be sum pattern. But if I burn the envelope with WHICH PATH information first, and then take a look at the result, will I see interference pattern?

    Or am I missing something here? Is what are you trying to say that even burning the envelope won´t make any difference, because the interaction has already happened?
  10. Oct 31, 2014 #9
    What you are talking about is pretty much "hidden variables", and as noted by Nugatory it can not reproduce QM predictions, but you need a different setup to see why, not dual-slit. For a simple Bell-test setup you can search these forums for "DrChinese challenge".

    Even for your dual-slit musing though, I do not understand what is the point or conclusion you are making... superposition only exists in relation to knower, and then what? The end result, presence or absence of interference pattern is pretty objective, so what are you actually saying? How do "we" (particle) decide where to hit the screen and whether to draw a pattern?
  11. Oct 31, 2014 #10
    Indeed, discarding the which-path measurement on a macroscopic level will not restore the interference. Only very precise undoing of the measurement (reversing it, as in literally doing the whole interaction in reverse) on the quantum level does.
  12. Oct 31, 2014 #11


    User Avatar

    Staff: Mentor

    [QUOTE="Matush, post: 4898404, member: 528930]
    Or am I missing something here? Is what are you trying to say that even burning the envelope won´t make any difference, because the interaction has already happened?[/QUOTE]

    yes, that is how it works.
  13. Oct 31, 2014 #12


    Staff: Mentor

    The same way its not required in the everyday common-sense view of the world. That world exists independent of conciousness, observers etc etc.

    QM is a theory about observations that appear in such a world.

    There are issues - but its not the stuff pop-sci books prattle on about such as conciousness causes collapse.

    Weinberg gives a pretty good overview:

    And so does Lubos:

    Last edited: Oct 31, 2014
  14. Oct 31, 2014 #13


    Staff: Mentor

    It's about decoherence being reversible in some simple cases.

    And indeed you dont get an interference pattern if you know which path - but there is only one way to do that - observe it.

    Of course.

    But I understand how pop-sci accounts can confuse the issue. Forget them. You will get the real deal here.

    Last edited: Oct 31, 2014
  15. Oct 31, 2014 #14


    User Avatar
    Science Advisor
    Gold Member

    A couple of elements to the puzzle which may assist you:

    1. It is not the knowledge itself, but the potential to know, that determines if there is interference. The setup does not have to produce the answer, only that the answer *could* have been learned in principle. In such case, no interference.

    2. The "interaction" need not be "physical" in the usual sense. You do not need to see it go through one slit or the other and by the act of placing something at the slit, eliminate the interference. You can place polarizers at BOTH slits and get interference or NO interference according to their relative settings. So it is not the polarizer per se that eliminates the interference. It is the relative setting, which is not physical in the usual sense - as either way a polarizer must be traversed.
  16. Nov 5, 2014 #15
    Interpretations of quantum mechanics are made somewhat more difficult by the assumption that something explainable happens outside of observation. But it's only ever observables that really need an explanation (to the extent they do). Explanations themselves are not really in need of any explanation - other than one might be curious as how they came about. The simplest explanation for explanations are that observables inspire them.

    Now by observation is meant not just what is observed in a given experiment, but also that which is, in principle, observable in an experiment. Anything that can't be observed (such as which slit a particle passes through) throws into question the usefulness of the concept that such presupposes. What does it mean to propose a slit through which a particle passes? What is a particle in this situation? Obviously the concept exists (since we can imagine it) but to what observable (in principle or practice) does the concept apply? It doesn't. What does it explain? It doesn't.

    Prior to QM, such a concept was a reasonable way of explaining an observation. For example, I might explain the appearance of a horse in terms of that same horse having been previously hidden in the stable, from which it then emerged. Becoming visible. Likewise the particle component of a particle detection, could be conceived as a particle travelling through space, unobserved, and then hitting a detection screen, and that could explain the particle detection (the observable).

    But to the extent that the concept of such a particle, travelling in such a way, has failed to fully explain a particle detection (or rather the pattern of particle detections), the concept itself must be considered as not being a good enough explanation. The concept itself, of a particle localised in space, prior to detection, does not itself need an explanation. It is, itself, the explanation, albeit not a very good one.

    This is why it's somewhat silly to ask questions as to what slit a particle went through, or how an unobserved cat can be both alive and dead. Just as with the particle/slit, the concept of a cat in such a situation (superposition) is an answer - not a question. And a better answer in terms of the cat might be that the cat is not alive and dead, but alive or dead - not just prior to observation, but more importantly: after observation.

    After observation?

    Yes. When you open the box, in what state is the cat?

    The unavoidable answer, perversely enough, is that the cat is alive or dead.

    Last edited: Nov 5, 2014
  17. Nov 6, 2014 #16


    User Avatar
    Science Advisor
    Gold Member

    Of course, you can ask and answer that question. But if you do, you won't see any interference.
  18. Nov 6, 2014 #17
    It is not really a question in the first place. It is a concept - it is the concept of a particle passing through a particular slit. We could repose this concept as a question, such as which slit, but there will be no answer. Because the concept itself is already an answer (to a much older question). If there are any questions here, it is whether the concept of a particle passing through a particular slit is a good answer.
  19. Nov 6, 2014 #18


    Staff: Mentor

    Hmmmmm. Observation in QM is a subtle issue. As mentioned previously there are variations of the double slit such as the quantum eraser experiment that show its really decoherence at work and it can sometimes be unscrambled, so even the question of observation in a usual sense is problematical:

    I personally prefer the view in QM observation and decoherence are synonymous - but that's just me.

    I think it reveals what a poor tool our usual concept of terms like observation, passing through a slit etc are in QM and we need to be 100% clear on what they mean in that context.

    Last edited: Nov 6, 2014
  20. Nov 6, 2014 #19
    Which-way experiments can be understood as a way of challenging the concept of a particle pursuing some specific path. While posed as a question (which-way) the experiments can be understood as a way of throwing into question the concept presupposed by the which-way question. The concept itself is being challenged. By the experiment.

    But this is often not obvious. There will be a certain attachment to the concept of a particle pursuing a definite path. And to the extent that an experiment agrees with such a concept, the concept will find a home (so to speak). A measurement made close to a slit, for example, can easily provoke the concept of a particle having passed through that slit, and not the other.

    But this is an expression of the concept rather than the experiment.

    The alternative concept of a particle travelling through both slits (or the concept of a particle as a wave function) is not in any way affected by a measurement being made close to one of the slits. We will still see the distribution of detections (some close to the slit and others further afield) that remain perfectly consistent with the concept of wave functions (or similar concepts).

    The fact that interference patterns are interrupted can be understood in the same way we understand that a view of the moon is interrupted when we close our eyes.

    Last edited: Nov 6, 2014
  21. Nov 6, 2014 #20


    Staff: Mentor

    Again I don't disagree.

    But what is gained?

    Why not simply associate observation with decoherence and all such issues disappear? Even the tricky delayed choice experiment is then easy to understand.

    Why is the moon there when we are not looking - answer - its being looked at by the environment all the time.

Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Similar Discussions: Superposition and retrocausality
  1. What is retrocausality? (Replies: 16)