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

Observational data and wavefunction collapse

  1. Jul 1, 2010 #1
    Okay, let me begin by saying that I do NOT have a good foundation in quantum mechanics, but I have been completely captivated by the wavefunction collapse.

    I have looked over many explanations of the double-slit experiment over the course of a few months, and I cannot find that one single detail that I am looking for.

    I see that the eraser experiment seems to indicate that collapse has nothing to do with the detector itself, but the act of observing... and I am seriously thinking that this is so bizzare that it has no place in our objective physical universe.

    Now, I understand that no one really knows what is happening here, but how solid is the idea that the detector itself is not causing this??? If the detector were hooked up to a bell which rang when it absorbed a particle, what happens if we disconnect the bell?

    If the data registers on a screen only in real time, will the wavefunction present itself if we kill power to the monitor?

    I suppose my question is very general about the relationship between the particle and the data. I mean, are we talking about interaction which changes this particle/wave, or is just simply KNOWING "which path" enough to alter the state of the particle/wave?

    If this is the implication, why have ridiculous experiments such as destroying the data not been performed? Am I oversimplifying?
  2. jcsd
  3. Jul 1, 2010 #2


    User Avatar
    Science Advisor

    No, it does not. Delayed choice quantum eraser experiments do not need any human observer or something like that to be explained correctly and I am afraid you got your information from some crackpot sites like the one unfortunately linked on wikipedia. Delayed choice quantum eraser experiments work by picking a certain subset of all experimental detection events. That is what you need the coincidence counting for. And in this way you can of course choose a subset where you see an interference pattern or you can choose a subset where you have which-way information. However this choice is more like a filtering process and does not change the detection events which happened earlier.

    Nature does not really care about what we know. It just matters what we could know in principle. Or in other words: If you hooked up your detector to a bell and disconnected your bell, the results would not change. The interaction happened and it does not matter whether we see the result or not.
  4. Jul 1, 2010 #3
    So... the detector itself really does provide the interaction? If that is the case, this is simply understood. The interaction with the detector simply alters the probability densities in the system. Easy. We can all go home. :confused:

    Why all the hype about observers, mysterious implications, and impossible problems?

    Why not simply say that detecting a particle is like poking it with a stick? We are changing what it does when we poke it. Only, they don't say it that way. They talk about it as if looking at it changes it... Schrodinger's cat for instance. Are people just blowing this up into something it really isn't? Did Hawking really say that there are profound implications from this?
  5. Jul 1, 2010 #4
    Fortunately or unfortunately, collapse interpretations of QM have collapsed :) after the discovery of Quantum Decoherence.
  6. Jul 1, 2010 #5


    User Avatar
    Science Advisor

    Well, the quantum eraser experiments are still very good basic experiments demonstrating complementarity and showing the importance of entanglement as you have a two-photon state and not just one particle. But basically it is just that. Most interactions spoil the well-defined phase of your state and therefore ruin interference patterns.

    I do not know. All serious papers on that topic avoid this hype. However, what the media or "second-hand" sites make out of it is often far from what the original papers were actually saying.

    Well let me rephrase your first sentence in your last post: every (irreversible) interaction works as a detector. Therefore any way of looking at it must necessarily include such an interaction. Again, all serious resources on Schrödinger's cat do not attribute this to some strange effect consciousness has on our world. Unfortunately, some simplifying popular explanations do say that.

    The one interesting question which remains is more of ontological quality. Is Schrödinger's cat really a superposition of dead and alive at the same time (so none of these attributes is real) and an interaction causes it to change into one of these real states (that would be wave function collapse or decoherence) or is it either really dead or alive, but both just with some probability, but we do not know which it is until we check (which is allowed in some realistic interpretations of QM)? I think this is indeed an interesting question.

    The one quote from Hawking about Schrödinger's cat I know is:
    "My attitude is that when I hear of Schrödinger's cat, I reach for my gun."

    I suppose you mean a different one, I do not know of. ;)
  7. Jul 1, 2010 #6
    To make the wavefunction issue more visualizable, consider this analogy.

    Drop a stone in water and a ringlike wave will travel out in all directions. If a barrier with 2 holes is placed at some distance, the waves that go through the two holes will interfere. Normal classical wave stuff. However, photons and all mass particles can act just like this, even when there is only 1. Yet we never see the wave, we only see points where the photon hit. It's provable that the path that photon took depends on the whole wave though. So blocking a hole the photon presumably didn't go through has effects on where the photon gets detected (double slit experiment).

    So, if the photon really is a wave and goes through both holes, what happens to the rest of the wave when we detect the photon only at a single point? Whenever we detect a photon, then the wave disappears, and if the photon continues the wave will expand out there, until the photon is detected again. Everything the photon wave sees the photon sees and responds to. But the wave is never there when we look at where the photon is, nor does the photon respond to any part of the wave that existed before we looked.

    Edit: Note that looking only requires some way to figure out in principle where the photon is or was. It doesn't even mean you actually looked. It doesn't require an intelligent observer or other goofy literal interpretations of 'observer'.
  8. Jul 1, 2010 #7
    Wouldn't it be more accurate to say that observer based theories of wave-function collapse (a la Von Neumann) have been replaced by Decoherence.

    Dechorence still doesn't tell you how a particular state is selected in a quantum system, we still don't know if it is a random or deterministic process (for instance).

    To get round that issue you can postulate that all states exist in different universes (or something similar), but this doesn't appeal to many people.

    But in the case of these lab experiments, the "measurement" process completes when the particles become entangled with a macroscopic measuring device, be it a screen, photomultiplier, or human eye...
  9. Jul 1, 2010 #8
    Yes, Schrodinger's equation itself is just the formulation of an analogy. When I learned that, I was terribly disappointed at first.

    Hmmm.... I'd like to see that proof. For, if I trace wave propagation -- the interference nearest the center does not depend on the "whole" wave function but only on a particular portion of crests. The same is true for the next interference fringe away from the center -- I would (naiively?) say that any given experiment can't be proven to depend on the whole wave function,( one could prove that it is *possible* the results did -- but, OTOH, that proof allows that it does not *have* to depend on the whole wave function eg: results might only depend on specific parts for each run with a definite result, but different parts depending on what result was achieved.)....
    I think the truth is oversimplified in many statements about QM, and simplified statements can make one think the problem is more difficult than it is.

    Aye. Although, I am not so sure the wave (or probability pilot, or alt interpretation....) has to go "through" the hole as a classical wave would -- for a complimentary image might be formed from matter (the wall) diffracting a wave and distorting the shape of an EM field -- perhaps a reflected one.

    R. Feynman did a study where he showed the reduced rate of travel of a photon in a crystal can be viewed as the superposition of the original wave -- and many reflected waves from the lattice. I am not ruling out that a "wave" goes through the holes -- but am merely cautioning that such a field (if EM in nature) might not be identical with the only cause of interference at the destination/target.

    Yes; with more alternatives to *if* now being proffered.

    In other words, a mathematical model is not the same as Heisenberg's uncertainty principle -- just close. This is also where arguments not removing the issue you speak of, but attempting to caution people about possible subtleties come in -- for example, stochastic mechanics. If a detector is only so efficient -- and a wave is made up of varying intensity wave fronts (statistically varying like a sound wave does based on varying masses and positions of air molecules -- though transverse in nature) -- then the wave might really exist everywhere but have insufficient energy to be detected except at one point ... etc.

    A little more thought elucidates the difficulty in accepting the plausibility of the particular hypothetical example "solving" the problem altogether *by itself* -- (why only at one point instead of a few? How does the detected particle know how good the detector is in order to prevent it's friends from finding a more efficent detector... etc. ) but, at the same time, one ought to keep in mind that Just because one doesn't know HOW something can be possible -- doesn't mean it *isn't* possible. A disproof is far more demanding than mere ignorance of a proof.

  10. Jul 1, 2010 #9
    I don't wish to overly defend the analogy I used. It was an analogy for visualization of the core conceptual difficulties, not a claim of what's going on in reality. As such, the whole thing is a big IF, with some correspondence with the empirical facts. It wasn't meant to be literal.

    The proof is only empirical consistency, not a specific claim of reality, but is directly observed by blocking the hole in the double slit experiment the photon doesn't go through. The crest in QM corresponds to the most probable locations to find the particle. That why it's considered a 'probability' wave. Without the path it presumably didn't go through existing or open, there is no interference and no series of crests for the interference pattern to build up over many detections.
    Here's a classic youtube video showing the classic double slit experiment:

    I don't necessarily take the conclusions so literally as presented in the video, but the claims are perfectly valid mathematically. I keep a whole range of alternative descriptions, but individually they are completely worthless formally. The independent research section of this forum has a cute one.

    It is indeed dependent on the information available to the whole probability wave. In most situations opposite paths available destructively interfere with them having any meaning, so that effective allows you to ignore much of the formal wavefunction. This is where sum over histories and similar approaches come from. How this is physically interpreted nobody can say definitively. All we say is that formally this leads to the broadest and empirically most successful theory in the history of science. The only holdout being gravity, which requires another theory.

    Of course, I was making no attempt at strict validity. Only a picture to illustrate the conceptual issues didn't go away by realizing any device can be an 'observer'.

    Oh, I see. I search for different interpretations and models myself. But if the purpose is merely to provide some interpretation that sooths some philosophical predisposition not many people are going to care, including me. The game is to expand the predictive capacity, not simply interpret the meaning.
  11. Jul 1, 2010 #10
    Let me see if I understand this.

    Decoherence is just a fancy way of saying that interaction occurs... and that's it. Some irreversible event (or many) takes place and the wave aspect disappears. This still does not seem to imply that the wavefunction did not 'collapse,' only that an observer (as defined in some ways) did not cause it.

    From what I can gather...

    Putting all of this together, observer-dependent states in the system is simply a result. It simply means that anything we DO to observe will cause the interference to disappear. So it really is not the fact that we can see it at a given point, but rather that we are incapable of seeing anything unless we disrupt the system first.

    Could this be analogous to a blind man poking around at things with a cane? He has no other way of observing, so he has to actually touch an object and thus change its properties... No mystery, no philosophical rubbish, just a fundamental truth about the situation. ??

    I suppose you could say that the state of things would not change without the blind man's observation, but the statement is oversimplified, correct?

    If this is so, it really makes me wonder why so many people are spreading such hogwash about this experiment. One such example is Thomas Campbell who stated that experimenters unplugged the data recorders and the wavefunction reappeared.
  12. Jul 1, 2010 #11
    Yes, quantum measurements can be described as a blind man pocking around with a cane. Yes, there is a lot of hogwash observer claims being thrown around. However, thinking the conceptual issues can be brushed away so simply because such claims are so outlandish oversimplifies things quiet a bit.

    Not even in principle has any description of QM been able to decompose the theory into distinct objects with properties, or anything even resembling classical theory. It's good that you learned how ridiculous many claims can be. Now it's time to learn what really is so weird about QM, which takes a lot more patience and effort. These real conceptual issues are why people get away with throwing around the outlandish claims they sometimes do.
  13. Jul 2, 2010 #12


    User Avatar

    Quantum decoherence gives a partial view only, it does not solve the key problem.

    It just solves the real observer-problem by considering a non-real birds-view observer which never never interacts with the world, excepct to get the initial conditions. But this is a step away from the founding ideal of QM - to be a MEASUREMENT THEORY - ie we should speak about what we can say about nature, not what nature IS - to start to speak about what a bids-observer or GOD could say about nature is IMO a step in the totally wrong direction.

    I think the "hype" is mostly due to people who insist to keep various forms of realism, are trying to make sense out of the collapse and the observer. The other part of the hype is due to consistently keep confusing "observer" in the interaction-party sense and HUMANS.

    If you instead see the collapse as a plain information update, then it's far less mysterious.
    If you see the wavefunction as a specific observers/systems knowledge or expectations of some subsystem, then we see that the collapsing wavefunction of a subsystem merely is an change of state of the system in the environment encoding information about the subsystem - the implication of this is that the action of the environment changes.

    The "knowledge" one system has about another system, can be seen without human simply as the "microstate" of the system encodes information, like a memory, and how this state influences the action of the system. Like the physical state of a photographic plate encodes information about when it was exposed to light.

    So IMO it's quite possible that one observer "sees a collapse" while another observer does not. I do not think there are any sense in fundamental "objective collapses", mainly because there is no objective way to establish this objectivity. Each observer /subsystem makes their only incomplete assessment and acts accordingly.

  14. Jul 2, 2010 #13


    User Avatar
    Gold Member

    The idea about traveling wave seems to cause contradiction with quantized photons.
    That's because this wave can cause some effect say interference pattern in double slit experiment is manifestation of this effect. On the other hand only quantized particles (photons) can travel and cause something.

    So my view is that this wave is not a traveling wave but rather a standing wave along the path of traveling photon and it travels nowhere. But because it travels nowhere it should be there before photon goes there. So the last part is that previous photons from ensemble that traveled the same path created this standing wave.
  15. Jul 2, 2010 #14
    Yes zonde, that hits the fundamental issue pretty much on the head.

    The way you worded your view just triggered a connection I failed to make before, even though I've considered such models before. This would also be quiet useful in modeling certain metamaterial effects also. I have some re-reviewing of some material to do.
  16. Jul 2, 2010 #15

    I didn't know prof. Brian Greene, who is one the most prominent figures working on the foundational issues, was a crackpot. :bugeye:

    In "The Fabric of the Cosmos" he states that the erasure of the information about the which-path restores the interference pattern.

    "The Fabric of the Cosmos", p. 259

    What do you think the term 'information' is releated to? If it comes down solely to his word against yours, he clearly wins.
    Last edited: Jul 2, 2010
  17. Jul 2, 2010 #16


    User Avatar
    Homework Helper
    Gold Member

  18. Jul 2, 2010 #17
    Yes, good point. And when scientist write publications for a general audience, the distinction between comparing a relationship between a weak and strong observer effect, and the claim of a strong observer effect often gets lost. Similar conflations occur in a wide variety thought experiments that sometimes gets embedded to some extent in the scientific community as well.
  19. Jul 2, 2010 #18


    User Avatar
    Science Advisor

    When referring to the crackpot site linked on Wikipedia I referred to the link to bottomlayer. Greene is linked as a valid reference, if I remember correctly. As far as I know Greene never claimed that consciousness or human observers are needed to explain quantum eraser experiments.

    Information erasure in quantum eraser experiments refers to setting some reversible which-way marker which can be undone. For example you can take a double slit and put one polarizer at the position of each slit and set one to horizontal and one to vertical. By measuring the polarization you could now know which slit a detected photon passed through. To restore the interference pattern and destroy which-way information all you need to do is to filter the light from both slits using a polarizer which is set to 45° between horizontal and vertical. It will let 50% of the emission from both pass through and the photons from both slits are indistinguishable again. The interference pattern will reemerge and which-way information will be lost. However, this does not depend on whether some person actually observed this or not.

    Information is somewhat ill defined in this context. There are two basic possibilities:

    a) Information corresponds to "leaving a trace" or in other words it corresponds to an irreversible interaction. That seems logical to me as an irreversible interaction causes changes in entropy which can be described using the same formalism as used for Shannon's definition of entropy in information theory.

    b) You also count reversible interactions as information. In the above example this would be the situation behind the polarizers placed at the slits. If you measured now, you would get the which-way information, but if you do not measure and make the photons originating from the two slits indistinguishable again, there is no way to ever find out which slit one certain photon went through. In this case already the possibility to get some kind of information at some time is referred to as information. I dislike this usage.

    However, I do not know the exact wording Greene used in his book, but I would not consider what happens in DCQE experiments really as erasure of information itself, but rather as erasure of the possibility to get some information.
  20. Jul 2, 2010 #19
    "There's no such thing as wave-function collapse"

    Ok, explain how measuring one particle in a two particle entangled system causes the other particle to select a state (before it is measured)

    I don't think decoherence explains wave-function "collapse", rather it just explains why we don't get superpositions of states in macroscopic objects in usual environments.

    That's not to say that quantum state selection need be anything too mysterious, just that we don't have the physics model for it yet, eg if we're making observations restricted to a 3-brane universe and the "wave-function" is due to effects including higher dimensions (so we only have a projection, cross-section or other subset of it in our universe) it's no surprise that it appears mysterious for now.

    (Surprisingly, a recent Scientic American feature ( 12 events that will change the everything ) gives extra dimensions only a 50-50 chance. God, let's hope it doesn't turn out to be one of Alain Connes' weird unfathomable constructions ;) )
    Last edited: Jul 2, 2010
  21. Jul 2, 2010 #20
    Pardon my ignorance, but what other way can you figure out in principle where the photon is with out an eventual intelligent deductive interpretation?:confused:

    Again, forgive me if this is a silly question, but how can you predict an event without interpreting the meaning of the processes that led to that event? Doesn't predictive capacity only have meaning under some formal framework of reasoning/rationalizing phenomena in which interpretations must strictly adhere to? Like mathematical logic for instance. And even then don't mathematical formalisms require some physical interpretation to have any meaning ?
    I'm just asking.

    This is what I mean @ the highlighted above. Is there anyway of communicating knowledge without interpretation? Perhaps i'm delving into philosophical domains now. I'm just curious.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook