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

I Wheeler's delayed choice doesn't change the past

  1. Dec 27, 2017 #1
    I'm delightfully reading the Grand Design by Hawking-Mlodinow. After many years, I became persuaded from it that Feynman's interpretation as multiple paths of interference patterns in the double slit experiment and, more generally, his path integral approach to quantum mechanics, are superior to other mathematical settings because they capture some element of reality of "unobserved motion & causality" better than fickle behaviors interchanging between particle and wave. In other words, what is strange is the nature of motion or causal link between two consecutive observations-sources of information, not the fact that what is moving is a countable, detectable unit or "solid" and permanent object as a buckyball with a precise structure C₆₀.
    What I'm going to claim and underscore here is that multiple (and simultaneous) stories in Feynman's calculation technique are not "true" stories in time; not in the sense of causative chain or sequence of events in time. I think everybody, included Feynman, was aware of this tenet. Multiple paths "capture" something that is inherent to the reality of motion, as a sniffling of proximal paths which explains randomness and relation between subsequent observations, but we don't know exactly what is it. Wheeler delayed choice experiment (WDCE)changes the future only (in the intercepted particles, not the others), does not affect the "true" past (in which the interference "was" actually generated) in those cases in which we are authorized to consider that as a real past event (in the cosmological thought experiment).
    We must distinguish two cases. First, the real WDCE experiment in which we (erroneously) deduce a time delay from the spatial location in which we choose to observe/don't observe each particle. The Feynman's multiple paths do not correspond to actual single stories. So it is a nonsense to affirm that if the site of observation-choice is "subsequent" (actually to the right) of the two (open) slits, then it follows we are observing something which is occurring "after" in the Feynman's paths. We are actually making an observation set which is alternative to the observation on the screen. We can thus compare two alternative experiments, but not to consider one experiment as the continuation of the other. In conclusion, there is no difference between the choice to observe some of the particles just in front of the slits, behind the slits or close to the detector. In all these cases we are forcing the "possible" (Feynman's) stories of the observed particles with our present action and changing the future true story of those observed particles. But we can't say anything about the unobserved part of the single particle history (in between the source and the decision probe or the screen-detector).
    Second, if we have a quasar light arriving at our planet after some billions of years, forming an interference due to the passage of photons through the two right and left ends of a big lens-galaxy, and if today the quasar and the galaxy are not anymore aligned nor existing, then each single photon arriving upon our interference screen today does have its true story. There is a true past: the emission from the quasar. Then we have only simultaneous possible paths passages through one possible side or the other of the lens galaxy and gravitational deflection & interaction with any other possible path. Eventually we have two different possible "present times" in the true history of the photon, depending on our decision: one in which the photon has been intercepted by the probe spaceship that was sent midway towards the galaxy, so that that photon won't go to form the interference pattern in case it will reach our planet (we have changed the future of that photon). The other "present time" is in case we are choosing not to observe the photon stream for a while. After that the unobserved photons conserve the interference path that was generated during the past history and will arrive on the interference screen on Earth, plotting the interference pattern. In this case, the true story ends with the impact of the photon on the screen-detector on Earth. Our decision not-to-observe has not changed in any way the true story of the single photon.
    They are not our "decisions" which change the history of events, but only facts, actual events can do that: if we actually interact with-observe the single photon, that interaction will determine a really new future chain of events for that photon. There is no way to make any interaction - or decision that won't cause a true interaction - which would determine a change in some true past event of a true history, as the emission of photon from the quasar or as the same existence of a certain photon we haven't observed yet, with or without interference.
     
    Last edited: Dec 27, 2017
  2. jcsd
  3. Dec 27, 2017 #2

    vanhees71

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    Well, Feynman's path-integral formulation of QT is equivalent to the uniquely defined QT. There's no more qualitative insights than with any other formulation of QT (including relativistic QFT, where the correct path integral is over field configurations, not paths of particles in phase space as is possible in non-relativistic QT).

    Of course, the delayed choice is not changing the past but it's rather postselecting which aspect of the situation you want to observe. Last week, I defended my habilitation about Walborn et al's realization of Wheelers ideas, which I find most intuitive. In the discussion, one of the professors stated: "But the photons is already absorbed! How can one then "postselect" whether to observe wave-like or particle-like properties after this?" He was satisfied by my answer that you can do the postselection due to the preparation of the photon pair in the polarization-entangled pair (Bell state). Here are my slides (although in German, maybe you can understand it):

    https://th.physik.uni-frankfurt.de/~hees/publ/habil-coll-talk.pdf
     
  4. Dec 27, 2017 #3

    Demystifier

    User Avatar
    Science Advisor

    Delayed choice, indeed, does not change the past, as I explain concisely in https://arxiv.org/abs/1703.08341 Sec. 5.1.
     
  5. Dec 27, 2017 #4
    Thank you Hendrick and Hrvoje. I couldn't fully understand your papers either they were both in Italian or English, but - after your replies - I believe Hawking is exaggerating in telling us (page 83):
    «...the Universe doesn't have just a single history, but every possible history, each with its own probability; and our observations of the current state affect its past and determine the different histories of the Universe, just as the observations of the particles in the double slit experiment affect the particles' past.»
    Do you agree?
     
  6. Dec 27, 2017 #5

    vanhees71

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    Hm, the only book by Hawking I like is Hawking, Ellis, The Large Scale Structure of Spacetime...
     
  7. Dec 27, 2017 #6
    I am astonished, that this has been written by Hawking.

    Regarding delayed choice experiments: To my mind, Wheeler merely suggests that it doesn't make sense to talk of "reality" before a measurement is made:

    The "past" is theory. The past has no existence except as it is recorded in the present. By deciding what questions our quantum registering equipment shall put in the present we have an undeniable choice in what we have the right to say about the past.“ [Underline, LJ]

    Wheeler in „Quantum Theory and Measurement“ (edited by John Archibald Wheeler and Wojciech Hubert Zurek), Princeton, New Jersey 1983, page 194
     
    Last edited by a moderator: Dec 27, 2017
  8. Dec 27, 2017 #7
  9. Dec 27, 2017 #8
    Maybe Mlodinow wrote that. Here is the book, check pages 82-83 https://goo.gl/images/Z6N9gr.
     
  10. Dec 27, 2017 #9

    DrChinese

    User Avatar
    Science Advisor
    Gold Member

    Naturally, almost anyone would find something to quibble with in Hawking's statement. After all, it is sweeping in the assumptions inherent in it.

    For example: I don't agree that the Universe has all possible histories, nor all possible futures. The most it could contain are histories and futures consistent with what we observe.

    On the other hand: there is no evidence that otherwise refutes Hawking's statement. And one possible deduction is that past (and future) histories are "pruned" as a result of measurement decisions made now. That implies a weak form of retrocausation. I say "weak" because you are not selecting/pruning specific histories with a measurement choice.
     
  11. Dec 27, 2017 #10
    Good point! There is no strict necessity in Quantum Theory for the Uniqueness of quantum phenomena, it is only that separate individual observers will not see mutually inconsistent outcomes from the observed events that other observers may see. Phenomena exist depending on who and how we observe them, in fact the observation call them into existence by "fiat", so there may be more than one self consistent outcome which may turn out to be mutually inconsistent from the POV of another observer, so there may be more than one Everett World in a series of overlapping event spaces like a number of criss-crossing roads along which a single observer may "travel". The "roads" may be traveled many times and in many ways... a form of event based eternalism based on a Page & Wootters Mechanism for time in a Two-State Vector Formulation. We as observers don't notice it because we are causally driven creatures.

    I am partial to the Many Interacting Worlds Interpretation of Hall, Deckert and Wiseman. We live in a Block Universe. The superposition of all quantum states (a closed set of a large number of mutually overlapping results, but not an infinite number) is an influence in this world from the "quantum overlap" of "classical" parallel Universes. A specific measurement "selects" from a background of many outcomes which local bit of the Universe will happen as a series of following consistent classical states. Each Universe is classical and bears a resemblance to our own Universe with differences, but the many quantum states in superposition are individually real and legitimate alternate choices.
    A New Hypothesis Suggests That Parallel Universes Might Interact after All - Big Think - December 2, 2017
    or the paper...
    Quantum Phenomena Modeled by Interactions between Many Classical Worlds - Michael J. W. Hall, Dirk-André Deckert, and Howard M. Wiseman - Phys. Rev. X 4, 041013 – Published 23 October 2014
    Still... a fruitful developing idea in recent developments (2017). It is version consistent with some of the more interesting counterfactual experimental outcomes that can be actually observed.
     
    Last edited by a moderator: Dec 27, 2017
  12. Dec 27, 2017 #11

    vanhees71

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    After reading Hawking's "Short History of Time" concerning "imaginary time" I was also shocked. As I said, the only book (co-)authored by him I like is the Large Scale Structure of Spacetime. It's of course much more difficult to write popular-science books that are correct and at the same time understandable to the addressed non-expert audience than to write a good textbook addressed to students of theoretical physics who are supposed to have the necessary technical prerequisites to understand the subject. That's the only excuse I have for scientists writing misleadin popular-science books. That you can do better has been shown by the (in fact very few) examples of good popular-science books (Weinberg, The first Three Minutes; Ledermann+Teresi, The God Particle (title given by a stupid publisher not by Lederman himself, but that's the only drawback of this one); Born, Einstein's Theory of Relativity (which in the 1920ies helped to keep Frankfurt university's physics department running)).
     
  13. Dec 27, 2017 #12
    Seems there is some interest in this topic. So I would like to add a comment or two to what I have already stated above regarding the "Block Universe". Assuming the Universe is indeed a "Block Universe" and time is related to quantum state entanglement via an experimentally demonstrated Page & Wootters Mechanism indicated by the experimental quantum entanglement technique referenced in this paper: Time from quantum entanglement_an experimental illustration - LMaccone&MGenovese - 1710-2013
    A Two-State Vector Formulation, as espoused by Vaidman/Feynman/Kastner/Cramer and others, would allow a very flexible view of time, almost "mechanical" in content, and leads to the possibility of a local reversal of events at the quantum state level... virtually unscrambling the egg (as it were).
    More recently this article in Science Alert and also in Nature News have these arguable examples which are highly supportive of "an Eternal Block Universe" which might have the network of quantum events being repeatedly traversed through a "reusable" criss-crossed network of interconnected Everett Worlds, yet individually reversible events (at least at the quantum electrodynamic level): Physicists Have Created a Set of Conditions in Which Time Seems to Run in Reverse - Science Alert - 2712-2017, and The new thermodynamics: how quantum physics is bending the rules - Nature News - 0111-2017 and these experiments seems to be just the prelude to the new paper:
    In no way is this "time travel" but phenomena occurring in time might be retrospectively changed applying the Wheeler Quantum Eraser Paradigm, if some quantum event cause some disastrous following chain of events in a single Everett World, it may be possible through the exercise of a strong guiding quantum correlation to "dodge the quantum bullet". The fact that we are using only "strong quantum correlations" rather than even stronger quantum entanglement, it might be possible to accomplish an outcome by "postselecting which aspect of the situation you want to observe" as Vanhees71 suggested, an aspect different from the first erroneous and disasterous sequence of events, by first invoking the NMR rollback technique followed by rolling forward and correlating an alternative classical outcome through a forcing "observation", or by simply preventing the incorrect "observation" happening by "editing it from this temporal script". It is science fiction ... but it is not speculation and is based on desktop experiment. A quantum Schrodinger's Cat could be brought back from the dead if we wanted to if we accepted the Many Interacting World's Interpretation of Hall, Deckert & Wiseman. Or Gort can resurrect Klaatu, breaking the seeming law of the irreversibly of death. I imagine there will be a practical time limit involved where such chain of event reversibility would become practically possible. But for now, we don't know what we don't know. Any other ideas out there? Doctor Strange eat your heart out.
     
    Last edited by a moderator: Dec 27, 2017
  14. Dec 28, 2017 #13
    I am of the same opinion. It is an unobserved past that gets defined into a definite past, when the present measurement takes place.
     
  15. Dec 28, 2017 #14

    Demystifier

    User Avatar
    Science Advisor

    Hawking wrote a lot of nonsense in last 10 years or so.
     
  16. Dec 28, 2017 #15
    If there is retrocausality, then the fact of object A retrocausing event X is fixed from the occurence of event X up to the retrocausation by A taking place. That is, when X occurs, the retrocausation by A is fixed. But I see it a little differently. Lets call the retrocausation by A "Y". So Y retrocauses X. Then X causes Y. So we have a causal loop. Put differently: neither of the events X or Y is causing each other, however their outcomes depend on each other.

    That is my opinion; I will not motivate it for I am not qualified enough and it would be speculation.
     
    Last edited: Dec 28, 2017
  17. Dec 28, 2017 #16

    vanhees71

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    Well, within the established and well-working theories of physics, among them local, microcausal relativistic QFT (leading to the Standard Model that up to now withstands all hard work to disprove it with bravour) there is no retrocausation possible, and there's not a single observation today that hints at its existence at all. What is an established fact are indeed the correlations described by quantum entanglement, i.e., correlations that are stronger than any that can be described by local deterministic hidden-variable theories.
     
  18. Dec 28, 2017 #17
    I think you are adressing me, @vanhees71 : do you mean that correlations with the past are possible? In my eyes correlation could be formulated in terms of two-way causality, which is something different than one-way causality.

    I fully respect QT as a correct theory, and if I am right, what I claim should be part of QT and a result of it. Probably I am saying something that, in other words, has been addressed a long time ago by people who are knowledgable and competent to do so (like you! :biggrin: ).
     
    Last edited: Dec 28, 2017
  19. Dec 28, 2017 #18

    vanhees71

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    I don't know what you mean by "correlations with the past", but of course the correlations, described by entangled states, are due to preparation of the system under investigation in the past, e.g., the nowadays widely used polarization-entangled photon pairs created by parametric downconversion: The Bell-inequality violating correlations are due to the preparation of the photon pair before any measurement is done with them, and this correlation also enables the postselection of "wave properties" or "particle properties" as described in quantum-eraser experiments (which are one particularly nice example for the possibility of Wheeler's delayed-choice idea). With a detailed enough measurement protocol you can do this postselection long after the entire experiment, including all photons, is long gone, and there is for sure no mystic retrocausation involved. It's just selecting subensembles of measured sequences of photon detections.
     
  20. Dec 28, 2017 #19
    I, personally, would render Wheeler’s statement more precisely and would say: By deciding what questions our quantum registering equipment shall put in the present we have an undeniable choice in what we have the right to say about the past in classical terms.
     
  21. Dec 28, 2017 #20

    DrChinese

    User Avatar
    Science Advisor
    Gold Member

    That's not a fair statement. It is more fair to say that such experiments can be interpreted in other ways. For example: the below referenced experiment is dismissed by you (as we have discussed previously). However, I present to thread readers an example (from a top team) of entangling particles via swapping. The swapping occurs AFTER the entangled particles have already been detected. That is certainly an observation that hints at retrocausation.

    https://arxiv.org/pdf/quant-ph/0201134.pdf
    Middle of page 5 is the entangle-swap after detection: "... Alice’s measurement projects photons 0 and 3 into an entangled state after they have been measured."

    Of course, there is no signalling possible with this scheme, as the projection of the specific entangled state is itself random.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook




Similar Discussions: Wheeler's delayed choice doesn't change the past
Loading...