Results 1 to 24 of 24

Thread: Information symmetry in a double slit experiment

  1. #1
    Join Date
    Jun 2009
    Posts
    1,915

    Information symmetry in a double slit experiment

    In a double slit experiment, it is said that if the experimenter observes which slit each particle passes through, this destroys the interference pattern on the screen. Imagine each particle has an observer riding on it who can record which slit the particle passes through, but the experimenter does not have access to this information Is there still an interference pattern?

    For a more dignified way of putting it (getting rid of human observers): Is it possible to have some "internal" change in each particle that records which slit it passed through, without a corresponding change in the experimental apparatus that causes it to also get that information? By a "change" , I mean some observable taking on a definite value from its set of possible values.

  2. #2
    Join Date
    Dec 2006
    Posts
    1,496
    I think it is the non disturbing action at each slit that is important. If the particle could contain some L or R identifier that was passively introduced at the slits then the interference pattern would I think remain. And if that identifier could be decoded at the screen, presumably we could map the L and R paths. An observer on each particle could do that I suppose and each observer could tell us which slit their particle went through and we coukd map the paths without having any effect on the interference pattern. And If the observer were to shout out the slit as they passed through it, I would still expect the interference pattern to be produced. But the moment any disturbance occures with the particle, we lose the interference pattern. So perhaps the question should really be about disturbance - is it actually the case that if we could detect the particle at the L or R slit with 100% non disturbance, would the interference pattern remain? I think it would, but I am not at all sure that any such non disturbing measurement will ever be possible. So we will never know.

    That at least is my understanding of the single shot two slit experiment.

  3. #3
    Join Date
    Mar 2010
    Location
    United Kingdom
    Posts
    7,311
    It's not so much about disturbing the photon as extracting information (disturbing the state of the system the photon is part of). If information about the photon never leaves the system then the pattern will appear. If you extract information by any method it won't. You see a good example of this in the quantum eraser experiment (https://en.wikipedia.org/wiki/Quantum_eraser_experiment) - here the photon is measured internally to the system but then that information is erased. And the interference patterns remain.

    So to your original scenario - as long as your photon riding observers never report home then the pattern appears. If they do then it doesn't.

  4. #4
    Join Date
    Dec 2006
    Posts
    1,496
    I suppose that if each observer on their respective photons report their L or R passing well after the interference pattern is displayed then no disturbance of the system takes place and the interference pattern occurs. Whereas if they reported their L or R status in transit the system is disturbed in the manner you describe.

  5. #5
    Join Date
    Apr 2011
    Location
    Norfolk UK and some of me is in Northern France
    Posts
    9,228
    It is like maxwell’s demon, who can by information push all the fast particles to one end of the box or the pilot wave idea that Bell’s inequality tests, if you can use the information, even in a single photon, you lose the superposition pattern. That leaves the nature of unmeasured states as weird as ever. (Ken G was good on this interpretation)
    sicut vis videre esto
    When we realize that patterns don't exist in the universe, they are a template that we hold to the universe to make sense of it, it all makes a lot more sense.
    Originally Posted by Ken G

  6. #6
    Join Date
    Jun 2009
    Posts
    1,915
    Quote Originally Posted by profloater View Post
    It is like maxwell’s demon, who can by information push all the fast particles to one end of the box or the pilot wave idea that Bell’s inequality tests, if you can use the information, even in a single photon, you lose the superposition pattern. That leaves the nature of unmeasured states as weird as ever. (Ken G was good on this interpretation)
    I don't see that the situation closely resembles Maxwell's demon since the observer's on the particles are not making any decisions about where the particles go.

    As to saying the interference pattern depends on whether the observers on the particles reveal their information to the experimenter, what happens if the observers survive the particles impact with the screen. After the double slit part of the experiment is over, the experimenter flips a coin to determine whether he will interrogate the observers about which slit they passed through. Will there be an interference pattern?

    Can the experimenter predict the outcome of the coin flip by seeing whether an interference pattern appears or doesn't ?

    To me, the least paradoxical answer is to say that if something happens to a particle that makes it possible to store which-slit information in the particle then, at the same time, something must happen to the experimental apparatus to store a record of which-slit information in it. So the interference pattern does not appear - whether a human experimenter is aware of the which-slit information or not.

    However, tidy as that answer is, I don't know whether it is correct!

  7. #7
    Join Date
    Jun 2009
    Posts
    1,915
    Quote Originally Posted by Shaula View Post
    You see a good example of this in the quantum eraser experiment (https://en.wikipedia.org/wiki/Quantum_eraser_experiment) - here the photon is measured internally to the system but then that information is erased. And the interference patterns remain.
    That answers the question if the information is collected and then definitely erased. What happens (as mentioned in my previous post) if we stochastically decide whether to destroy the information or not? (I'm not thinking exclusively about using photons in a double slit experiment. People are doing experiments where "quantum effects" are observed in aggregations of molecules - although I don't know if "quantum effects' yet involves double slit experiments.)

  8. #8
    Join Date
    Mar 2010
    Location
    United Kingdom
    Posts
    7,311
    Quote Originally Posted by tashirosgt View Post
    That answers the question if the information is collected and then definitely erased. What happens (as mentioned in my previous post) if we stochastically decide whether to destroy the information or not? (I'm not thinking exclusively about using photons in a double slit experiment. People are doing experiments where "quantum effects" are observed in aggregations of molecules - although I don't know if "quantum effects' yet involves double slit experiments.)
    The article mentions the delayed choice eraser (https://en.wikipedia.org/wiki/Delaye...quantum_eraser) as well which I think sort covers that. In that the choice is made after the photon is detected but before you look at the pattern overall (as looking at the pattern effectively extracts information from the system). If you have kept the path data in any form and not erased it then the photon detection doesn't contribute to the interference pattern. If you erase it after the detection is recorded then the detections do contribute to the pattern.

    The practical difficulty is keeping the apparatus in a mixed state. If we could somehow isolate the equipment perfectly such that information didn't flow out of it at all then QM says that you could record the whole experiment, then erase the path data and see a diffraction pattern. Or not erase it and then see no pattern. The critical part is the leakage of information from the isolated system to the environment.

  9. #9
    Join Date
    Jun 2009
    Posts
    1,915
    Quote Originally Posted by Shaula View Post
    The article mentions the delayed choice eraser (https://en.wikipedia.org/wiki/Delaye...quantum_eraser) as well which I think sort covers that. In that the choice is made after the photon is detected but before you look at the pattern overall (as looking at the pattern effectively extracts information from the system).
    I see that the delayed-choice experiment covers the case of when to measure something, but I don't see that it involves a stochastic procedure to determine whether to measure something. Am I not understanding the wikipedia article?

  10. #10
    Join Date
    Mar 2010
    Location
    United Kingdom
    Posts
    7,311
    Quote Originally Posted by tashirosgt View Post
    I see that the delayed-choice experiment covers the case of when to measure something, but I don't see that it involves a stochastic procedure to determine whether to measure something. Am I not understanding the wikipedia article?
    OK, sorry, I didn't see the stochastic part as the important bit. It doesn't matter how you select what you erase or measure - what matters is whether the information gets out of the isolated system or not. You could do that via manual selection, based on the time of day, coin flips, radioactive decay - it doesn't really make a difference as far as I know.

  11. #11
    Join Date
    Jun 2009
    Posts
    1,915
    Quote Originally Posted by Shaula View Post
    what matters is whether the information gets out of the isolated system or not. You could do that via manual selection, based on the time of day, coin flips, radioactive decay - it doesn't really make a difference as far as I know.
    It's disturbing to me that the experimenter could predict the outcome of a coin flip based on whether an interference pattern appears before the coin is flipped!

  12. #12
    Join Date
    Mar 2010
    Location
    United Kingdom
    Posts
    7,311
    Quote Originally Posted by tashirosgt View Post
    It's disturbing to me that the experimenter could predict the outcome of a coin flip based on whether an interference pattern appears before the coin is flipped!
    But they can't. Once you have measured the pattern you have effectively already made the call as to whether the path information is available or not. Erasing the information after you have made a measurement has no effect on the measurement. The delayed choice bit relies on the information about the detection remaining inside the isolated system and not available to the observer until after the choice is made.

  13. #13
    Join Date
    Jun 2009
    Posts
    1,915
    Quote Originally Posted by Shaula View Post
    But they can't. Once you have measured the pattern you have effectively already made the call as to whether the path information is available or not. Erasing the information after you have made a measurement has no effect on the measurement. The delayed choice bit relies on the information about the detection remaining inside the isolated system and not available to the observer until after the choice is made.
    No doubt, I'm thinking of this in a too simplistic manner, but here goes:

    In my scenario, the particles contain the which-slit information "inside" them and the particle and information survive the impact with the screen. (I imagine the particle sticking to the screen.) After the particles hit the screen, the experimenter makes the coin-flip decision whether to extract the which-slit information from the particles or destroy it.

    As I mentioned before, my solution is that the experimental apparatus gets the which-slit information at the same time as the particle gets it, so the particles hit the screen in the same way they hit it when the experimental apparatus uses which-slit detectors.

  14. #14
    Join Date
    Aug 2008
    Location
    Wellington, New Zealand
    Posts
    4,434
    Hi tashirosgt. Unfortunately your scenario is not the delayed choice or quantum eraser experiments. It is observers who have the information, not particles. The choice is made while the particle is in flight. A particle hitting a screen is a measurement and whatever the observer then does, does not change that measurement. The solution is to a non-problem.

    A good delayed choice experiment: Experimental Realization of Wheeler's Delayed-Choice Gedanken Experiment (PDF)
    Wave-particle duality is strikingly illustrated by Wheeler's delayed-choice gedanken experiment, where the configuration of a two-path interferometer is chosen after a single-photon pulse has entered it: Either the interferometer is closed (that is, the two paths are recombined) and the interference is observed, or the interferometer remains open and the path followed by the photon is measured. We report an almost ideal realization of that gedanken experiment with single photons allowing unambiguous which-way measurements. The choice between open and closed configurations, made by a quantum random number generator, is relativistically separated from the entry of the photon into the interferometer.

  15. #15
    Join Date
    Mar 2010
    Location
    United Kingdom
    Posts
    7,311
    Quote Originally Posted by tashirosgt View Post
    No doubt, I'm thinking of this in a too simplistic manner, but here goes:

    In my scenario, the particles contain the which-slit information "inside" them and the particle and information survive the impact with the screen. (I imagine the particle sticking to the screen.) After the particles hit the screen, the experimenter makes the coin-flip decision whether to extract the which-slit information from the particles or destroy it.

    As I mentioned before, my solution is that the experimental apparatus gets the which-slit information at the same time as the particle gets it, so the particles hit the screen in the same way they hit it when the experimental apparatus uses which-slit detectors.
    As I understand it then if they were able to keep the whole system in a mixed state then the outcome would depend on whether the information is erased. But you can't check the pattern and leave it in the mixed state, as soon as you pull some information out then the possible states of the system are constrained. So if you leave the information on the path available and observe the pattern you won't see interference and it doesn't matter what you do afterwards to the path information - you've measured the system in a way that means you can infer that the information is there. So it can't predict a coin flip because in order to make a prediction you have to have made a measurement that takes it out of a mixed state and so into a state where the coin flip doesn't matter.

    If you made the coin as part of the apparatus then you could infer what the result was after the flip had happened by looking at the pattern, or you could infer the pattern by looking at the coin result. But the coin has to be an unobserved part of the apparatus and you can't separate them out in a way that creates these macroscopic retrocausal effects you are talking about.

    In general all the quantum strangeness is only apparent if you leave your apparatus isolated. As soon as you start poking it to take measurements then these effects start getting suppressed.

  16. #16
    Join Date
    Jun 2009
    Posts
    1,915
    Quote Originally Posted by Shaula View Post
    But you can't check the pattern and leave it in the mixed state, as soon as you pull some information out then the possible states of the system are constrained.
    Let me see if I understand that. (It is disturbing in the same way that "The moon isn't there unless you look at it" is disturbing! )

    Suppose the experiment forbids observing whether an interference pattern exists until after the coin flip. Then the experimenter can use the coin flip (and the required actions of destroying versus examining the which-slit info in particles) to accurately predict whether the pattern exists. In repetitions of that experiment, there is a 0.5 probability that the pattern does exist. In a different experiment, the experimenter observes whether an interference pattern exists before the coin flip. In this experiment there is zero probability of observing an interference pattern. The outcome of the coin has no effect.

    So if you leave the information on the path available and observe the pattern you won't see interference and it doesn't matter what you do afterwards to the path information - you've measured the system in a way that means you can infer that the information is there.
    The latter experiment is consistent with my idea that the creation of which-slit information is symmetrical between an observer "inside" the particle and the experimental apparatus. Nevertheless, my idea is vague. Also statements about the information being "on the path available" are vague. It's hard to define "information" in a technical way. If we make a technical definition, then does our technically defined information have a spatial location? Can it be "on a path"?

  17. #17
    Join Date
    Jun 2009
    Posts
    1,915
    Quote Originally Posted by Reality Check View Post
    Hi tashirosgt. Unfortunately your scenario is not the delayed choice or quantum eraser experiments. It is observers who have the information, not particles.
    However, I think it's generally agreed that "observers" need not be human beings. In a macroscopic particle, I can imagine some apparatus inside it that detects and records the which-slit information.

    The choice is made while the particle is in flight.
    Well, we shouldn't say "choice" unless we specify a chooser.

    A particle hitting a screen is a measurement and whatever the observer then does, does not change that measurement. The solution is to a non-problem.
    I agree that, in a classical way of thinking, a particle hitting a screen creates a "physical record", which may or may not be examined. If we are thinking of experiments where physical records are created and then destroyed without being examined, I don't know that thinking of them in the classical way works.

  18. #18
    Join Date
    Jun 2009
    Posts
    1,915
    Quote Originally Posted by Len Moran View Post
    I think it is the non disturbing action at each slit that is important. If the particle could contain some L or R identifier that was passively introduced at the slits then the interference pattern would I think remain. And if that identifier could be decoded at the screen, presumably we could map the L and R paths. An observer on each particle could do that I suppose and each observer could tell us which slit their particle went through and we coukd map the paths without having any effect on the interference pattern. And If the observer were to shout out the slit as they passed through it, I would still expect the interference pattern to be produced. But the moment any disturbance occures with the particle, we lose the interference pattern.
    I agree that the idea that measurement must cause a disturbance isn't accepted in most modern discussions of QM. (e.g. the uncertainty principle is not explained by that idea, in contrast to what some early founders of QM thought.) So I think it's fair to think of a non-disturbing measurement of the which-slit information. (Although the idea of "disturbance" hasn't been defined in a technical way.) My question is whether the creation of which-slit information must be symmetrical between a particle and the experimental apparatus. Imagining that the particle can contain a physical record of the which-slit info at the same time that the (other) experimental apparatus does not leads to imagining some strange scenarios.

  19. #19
    Join Date
    Mar 2010
    Location
    United Kingdom
    Posts
    7,311
    Quote Originally Posted by tashirosgt View Post
    Suppose the experiment forbids observing whether an interference pattern exists until after the coin flip. Then the experimenter can use the coin flip (and the required actions of destroying versus examining the which-slit info in particles) to accurately predict whether the pattern exists. In repetitions of that experiment, there is a 0.5 probability that the pattern does exist.
    I think you need to be very careful about the word exist here. Rephrase the experiment. The apparatus is put into a mixed state where the outcome of the observation is not defined. What you are doing with the coin flip is deciding which outcome you are going to force the system into. The way you have phrased it kind of sounds like you have a 'hidden variables' interpretation going on where the existence of the pattern is somehow secretly there when the system is in a mixed state - it isn't. Pattern and non-pattern both 'exist'. The coin flip is simply you deciding what sort of measurement you are going to make and hence what you will see. In all of the experiments it is more correct to say that the pattern/no-pattern are both states that are part of the mixed state. The coin flip doesn't affect that at all - all it does is select for you how you are going to measure the system and hence what states are possible to observe.

    Quote Originally Posted by tashirosgt View Post
    In a different experiment, the experimenter observes whether an interference pattern exists before the coin flip. In this experiment there is zero probability of observing an interference pattern. The outcome of the coin has no effect.
    Yus, because you have extracted the information by choosing what kind of measurement you are going to make before you flip the coin.

    Quote Originally Posted by tashirosgt View Post
    The latter experiment is consistent with my idea that the creation of which-slit information is symmetrical between an observer "inside" the particle and the experimental apparatus. Nevertheless, my idea is vague. Also statements about the information being "on the path available" are vague. It's hard to define "information" in a technical way. If we make a technical definition, then does our technically defined information have a spatial location? Can it be "on a path"?
    You don't need to really localise information like that in this experiment. The information is in the system and not leaked to the environment. It doesn't matter which part of the system you believe it to be associated with. The path info could be 'on' the particle, the slit, a separate detector or written down by tiny gnomes. The key thing is that it has not breached the boundary of the isolated system that is in a mixed state.

    Information does have a technical definition. In fact it has several. But the most commonly used definition of quantum information associates it with a quantum system. In this example the system is the entire apparatus (because it is an isolated system in a mixed state) so that is what the information is calculated for. It doesn't really make sense to try to point to somewhere and say "it is there". Information in this example is a system property, not a physical thing. It is a bit like asking where the temperature is stored in a block of ice.

  20. #20
    Join Date
    Aug 2008
    Location
    Wellington, New Zealand
    Posts
    4,434
    Quote Originally Posted by tashirosgt View Post
    However, I think it's generally agreed that "observers" need not be human beings. ...
    Correct - but what you imagine inside a macroscopic particle is the cat or detector in Schrodinger's cat thought experiment. Your observer is in a superposition of states until we can measure their state. Your observer has measured and not measured the which way information.
    The word choice means anything that chooses, e.g. in the experiment I cited, the chooser is a quantum random number generator.
    A particle hitting a screen creates a physical record (measurement) classically and in some interpretations of quantum mechanics. A collapsing wave function (see the table near the bottom of the article) is the "essence of an measurement".

  21. #21
    Join Date
    Dec 2006
    Posts
    1,496
    This quote from Anton Zeilinger I think (if I read it correctly) presents this issue of path information quite starkly in the sense that any semblance of which path information, no matter how tenuous, will prevent the occurrence of interference. In this sense, even a completely passive physical detection of "something" passing through the slit constitutes an informational disturbance, even if the results of that passive detection are never looked at. The fact that the information is "somewhere" to be looked at is enough to destroy the interference.


    The superposition of amplitudes ... is only valid if there is no way to know, even in principle, which path the particle took. It is important to realize that this does not imply that an observer actually takes note of what happens. It is sufficient to destroy the interference pattern, if the path information is accessible in principle from the experiment or even if it is dispersed in the environment and beyond any technical possibility to be recovered, but in principle still ‘‘out there.’’ The absence of any such information is the essential criterion for quantum interference to appear.

  22. #22
    Join Date
    Jun 2009
    Posts
    1,915
    Quote Originally Posted by Len Moran View Post
    This quote from Anton Zeilinger I think (if I read it correctly) presents this issue of path information quite starkly in the sense that any semblance of which path information, no matter how tenuous, will prevent the occurrence of interference.
    I find discussions of "information" vague. Can there be information that nobody knows? If we create a thought experiment about information, it usually done from the viewpoint of an omniscient narrator who postulates that some information exists and some information does not. The existence of information is apparently different that the existence of some entity that knows the information.

  23. #23
    Join Date
    Dec 2006
    Posts
    1,496
    Quote Originally Posted by tashirosgt View Post
    Can there be information that nobody knows?
    That's an interesting point.

    Perhaps Zeilinger would say that any successful detection of which path data would in principle produce information if someone were to look at it. If no one ever did look at it, that doesn't alter the fact that the experiment has been designed to extract the data and it is only that fact that prevents the interference pattern from appearing. Presumably he would also say that if the interference pattern was not lost, then the experiment has failed to extract any which path data.

    It almost seems like he is saying that a "proof" that a detection of "something" passing through each slit (even if the detection is entirely passive and physically non disturbing in nature) has been successful is if the interference pattern is lost. The informational element of such a designed experiment is perhaps of no consequence to the actual "successful detection" in terms of the loss of the interference pattern - it is not the looking at information that destroys the pattern - it is the implementation of an experiment that in principle can extract such information that destroys the interference pattern.

  24. #24
    Join Date
    Mar 2010
    Location
    United Kingdom
    Posts
    7,311
    Quote Originally Posted by tashirosgt View Post
    I find discussions of "information" vague. Can there be information that nobody knows?...The existence of information is apparently different that the existence of some entity that knows the information.
    It doesn't matter if any one knows the information. What matters is whether the information travels outside the system that is in a mixed state. You could pipe that information straight into /dev/null and it wouldn't matter - as soon as the information about the quantum state of the system leaves the bounds of the system then the future phase space of the system is constrained. I don't think it is helpful to think in terms of observers and knowledge as this adds an unnecessary level of complexity. At the most basic level it is about flows of information. Every time information leaves the system it alters the available phase space in a way linked to the information content. So the future evolution of the system is now constrained to a smaller set of possibilities. It doesn't matter if the observer is conscious, a piece of silicon or capable of registering the information - it flowed so the system is perturbed away from the previous mixed state towards a more constrained state.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •