Copenhegen interpretation or Many World Interpretation?

  1. Can there be an experiment (even a thought experiment), which could settle the debate CI vs MWI? Now a days many scientists seem to favour MWI. I find it interesting because it perhaps allows time travel (the possibility of an observer going back to his/her own past) while avoiding the grandfather paradox, since, having visited your past, you can come back to a "different present" in a parallel universe. This is what I read in one of the popular science books.
     
  2. jcsd
  3. Fredrik

    Fredrik 10,303
    Staff Emeritus
    Science Advisor
    Gold Member

    No, there's no such experiment.

    We would need a quantum theory of gravity before we can even try to make an argument for that time travel thing. So that popular science book is doing some wild speculation.
     
  4. In principle... probably, yes. Every interpretation of a model has implications along with it, and it is just a matter of time before someone clever enough comes along and sees a way to measure the reality of one set of implications verse another.

    I am not that someone, however. All I can add is we have no mathematical understanding of the collapse of a probability into a specific outcome, which means if our quantum model reflects "actual reality", then it follows that either our model is incomplete, or else the other information produced by our model (the other outcomes and their probabilities) should also be realized somehow.

    The amazing accuracy of quantum predictions is a fair vote in the "actual reality" column, however the failure to accomodate gravity is a contradictory vote in the "incomplete model" column. That is more or less the juncture on which this debate stands.
     
  5. If you mean we don't have the technology for such an experiment, then that's true. If you mean such an experiment is impossible in principle, that's not true. There may or may not be an experiment that will distinguish MWI from other no-collapse interpretations; however, there are theoretically measurable difference between MWI and collapse interpretations like Copenhagen. In particular, measurement is theoretically reversible in MWI while it is not in Copenhagen. Quantum decoherence makes it questionable whether such experiments will ever be possible in practice—but that is different from there being no such experiment, full stop. See here for a more detailed overview.
     
  6. Fredrik

    Fredrik 10,303
    Staff Emeritus
    Science Advisor
    Gold Member

    I mean that it's impossible in principle. It is so by definition of "interpretation". If a statement about the real world has falsifiable consequences, then it's a theory, not an interpretation.

    I would say that this is a misunderstanding of the term "Copenhagen". I almost never use that term myself, because you can't find two people who define it the same way. But I have a strong opinion about one aspect of it at least: The idea that measuring devices "are classical" is a serious misunderstanding of Bohr's original idea, which was more along the lines of: Science requires theories to be falsifiable. To be falsifiable, they must make predictions about results of experiments. So the definition of science requires indicator components of measuring devices to be for all practical purposes indistinguishable from classical.

    That's how I interpret Bohr anyway. I know that there are lots of people who interpret his statements as saying that the laws of nature are such that measuring devices are fundamentally classical. But I find it very unlikely that Bohr believed anything that silly.

    What am I supposed to do with a dictionary entry on the MWI?
     
  7. Which is why, in many ways, many worlds "interpretation" is a misnomer. The fact that people call it an interpretation doesn't rule out the possibility for experimental tests. This is bizarre, backwards reasoning: I named specific tests for non-collapse QM like MW, and instead of saying, 'If there are tests, we shouldn't call it an interpetation,' you seem to say, 'That can't be right, it's called an interpretation so there are no tests.' Language is descriptive, not prescriptive, on reality. Sometimes those descriptions are wrong.

    I never said Copenhagen was about measuring devices "being classical", so I have no idea what your objections have to do with anything I've said. Wavefunction collapse is a component of the Copenhagen interpretation (at least, the modern take on it), and is the entire reason the latter has led to the so-called measurement problem. Since collapse is a one-way, non-unitary process, it leads to theoretically distinguishable predictions of collapse vs. no-collapse QM. I have never seen the Copenhagen interpretation not defined with, in part, wavefunction collapse so while you're perfectly entitled to your own private definition, I think it's a very non-standard one and certainly not what the OP had in mind with the question.

    Since I specifically linked you to a section entitled "Tests of the MWI", I would think "read it" would be a cracking start.
     
    Last edited: Jun 2, 2013
    Derek Potter likes this.
  8. Fredrik

    Fredrik 10,303
    Staff Emeritus
    Science Advisor
    Gold Member

    I'm saying that if it is an interpretation, there are no tests. I don't think I've heard the claim that it's not an interpretation before.

    The idea that wavefunction collapse is a non-unitary physical process that projects the wavefunction exactly onto an eigenstate, is almost certainly based on those misunderstandings of Bohr that I mentioned, not on Bohr's actual beliefs.

    Now that we know about decoherence, there's no reason to think that there's such a thing as a non-unitary collapse. The decoherence approach is to apply QM to a larger system that includes an environment. The result is an apparent collapse. Interactions with the environment put the system in a state that's practically indistinguishable from a collapsed state.

    This doesn't contradict Copenhagen, because we're just using the same theory (and if we want to, the same interpretation) to a larger system.

    I didn't see that you linked to that section. The page is displayed from the top until it has finished loading, and then the browser scrolls down to that section. This takes a couple of seconds. I had already closed the tab before that. I might take a look at it later.
     
    Last edited: Jun 2, 2013
  9. I'm still don't understand why you're so hung up on the word interpretation. Many worlds makes a number of claims. Those claims are testable, in principle. If you think that means it shouldn't be called an interpretation, then it's not an interpretation by your reckoning. The fact that people call it an interpretation does not say whether it may or may not be testable. It just means people might be using an inappropriate word for it.

    Whether it's a historical misunderstanding or not, that is what people currently mean by the term. Between this and the last quote, I think you are getting unnecessarily hung up on language. By "Copenhagen interpretation", it's reasonable to assume the OP meant what nearly everyone mean by the term today. Instead of considering whether it's possible to distinguish between MW and "the usual framework people mean by 'Copenhagen'", you're arguing that the term is inappropriate for the concept. While a perfectly reasonable discussion to have for it's own sake, it's largely irrelevant here. Your focusing on the word and not it's intended referent.

    Fine, we won't say the C-word since it bothers you. We'll say that the answer to, "Can Copenhagen and Many Worlds be empirically distinguished?" is "no" on the grounds that "Copenhagen" is not a well-defined model of QM, being subject to disagreement on its definition. Instead, we'll ask, "Can Many Worlds be empirically distinguished from other well-defined models of QM?" The answer to that is yes. MW is empirically distinguishable from (and has been experimentally preferred over some) variants of QM of involving wavefunction collapse. Whether it can be empirically distinguished from other non-collapse variants is not yet clear.

    Actually, you just described Many Worlds pretty well, so apparently you don't need empirical convincing of it. All Many Worlds quantum mechanics, at its most basic, consists of is the claim that wavefunctions don't collapse and so the universe's wavefunction stays in an overall superposition. However, for subsystems, quantum decoherence gives the appearance of collapse since entanglement allows individual subsystems entangled with others to follow the statistics of mixtures instead of superpositions. A "world" is just one of the states of a macroscopic system that composes the (incoherent) statistical mixture after quantum decoherence. I've always thought deWitt did Everett a major disservice by branding the latter's theory "Many Worlds", since it seems to conjure up mistaken thoughts of some kind of multiverse for people. All it is, at it's simplest, is quantum mechanics with the collapse postulate removed and no new ones (like Bohm's) to replace it.

    If you think this view is compatible with Copenhagen, then I'll reiterate that I think your use of the word is very non-standard. However, the point remains: if we put aside "Copenhagen interpretation" as a not-very-helpful term, then various purported models of quantum mechanics—models that are usually, though perhaps inaccurately, called "interpretations"—may still be experimentally distinguished from one another. Many Worlds—which you are apparently already on board with, even if you haven't realized it—can be distinguished from other mainstream models.

    No worries. I do recommend reading it though, and I hope you aren't put off by the source. I think physicists can often be overly dismissive of the work done by philosophers of physics, who are generally extremely well versed with the formalism and do a lot of things that would probably be better classified as 'physics' than 'philosophy'.
     
  10. tom.stoer

    tom.stoer 5,489
    Science Advisor

    I guess you will agree to the following remark: decoherence does not cause a collaps to one pure state but only an effective diagonalization of the reduced density matrix (which still dos not correspond to a pure state). So the generalized quantum state "dead and alive" decoheres to "either dead or alive". But for a single cat still both possibilities are contained in the reduced density matrix. So decoherence
    a) does explain why we observe a classical state of the cat, but
    b) it does not tell us which state (either "dead" or "alive") will be observed.
    So - even taking decoherence into account - in some sense there is still a collapse of the density matrix "dead or alive" to one pure state, e.g. "dead".
     
  11. Which is why what Fredrik is describing is precisely the Many Worlds interpretation as it's now understood. Without genuine collapse (and thanks to decoherence), the result of me doing a Schrodinger's cat experiment is that the reduced density matrix of the cat and me together (which decoherence has effectively diagonalized) is an incoherent mixture of "cat is dead and I saw a dead cat in the box" and "cat is alive and I saw a live cat in the box". Since in our subsystem that superposition is now incoherent, the two macroscopic pure states in the mixture ("worlds") continue evolving completely independently of each other.
     
  12. tom.stoer

    tom.stoer 5,489
    Science Advisor

    Yes, this is decoherence as an application of the QM formalism.

    And now we talk about an interpretation. Instead of interpreting the incoherent superposition as independent branches in MWI, one can refute MWI and insist on the collapse of the incoherent superposition to a pure state. The question is whether there could be an experiment to test these two interpretations.
     
  13. And the answer to that question, which was the point of my very first post before getting side tracked by linguistic hair-splitting, is, again, "Yes, in principle". I linked to this overview which discusses empirical means of distinguishing between no-collapse and collapse quantum mechanics and cites some specific proposals and already performed preliminary experiments. It's not a settled issue by any means, but neither is it an "In principle impossible to settle" issue.
     
  14. Fredrik

    Fredrik 10,303
    Staff Emeritus
    Science Advisor
    Gold Member

    I find it difficult to do that, because I think the "standard" version of the CI is pretty silly, and possibly also logically inconsistent.

    I would consider those "variants" different theories. I don't doubt that they, or at least some of them, are distinguishable from QM in principle, because whoever came up with these alternatives probably made sure that they would be.

    To me, this is just quantum mechanics...except maybe for the idea that there's a wavefunction of the universe. That should probably be considered an MWI idea. "Copenhagenish" interpretations don't say that it's fundamentally wrong to apply QM to large systems, but they put a lot of emphasis on the measurement, and this suggests that it may not make sense to include yourself in the wavefunction.

    You're probably right about that. This reminds me of this quote by Asher Peres:
    There seems to be at least as many different Copenhagen interpretations as people who use that term, probably there are more. For example, in two classic articles on the foundations of quantum mechanics, Ballentine (1970) and Stapp (1972) give diametrically opposite definitions of “Copenhagen.”​
    Source: http://arxiv.org/abs/quant-ph/9910078
     
  15. Fredrik

    Fredrik 10,303
    Staff Emeritus
    Science Advisor
    Gold Member

    Yes, I agree, there's still some sort of "collapse", but now we're talking about a kind of "collapse" that doesn't contradict the MWI.
     
  16. The term "wavefunction collapse" does not appear in books published before 1975, so people like Bohr, Feynman, Einstein had probably never heard of it. (I know many people use the term, but what it means to them in terms of observable phenomena I have no idea.)
     
  17. Fredrik

    Fredrik 10,303
    Staff Emeritus
    Science Advisor
    Gold Member

    I thought it was mentioned in von Neumann's book from 1932, which I haven't read myself. Maybe he called it something else. "Reduction of the state vector" is another term for it.
     
  18. stevendaryl

    stevendaryl 3,090
    Science Advisor

    The way I think of it, Copenhagen is not so much a precise physical theory as it is a recipe for getting testable predictions out of quantum mechanics. Wave function collapse is really part of the recipe, and not part of the ontology. You can interpret as saying: After a measurement, you can act AS IF the wave function has collapsed to the eigenstate of the observable that was measured corresponding to the eigenvalue obtained. I don't think Copenhagen treats this collapse as a physical event.

    You can certainly make a theory of wave function collapse so that it's interpreted as a literal event, but I don't think Copenhagen does that.

    Actually, it seems to me that the intuitive idea of a wave function collapsing as a physical event has serious difficulties if you have more than one particle. With more than one particle, the wave function is a function on configuration space, rather than physical space, and so it's not possible in any straight-forward way to interpret "collapse" as some rapid or instantaneous change of a field.
     
  19. Well, insofar as a "Copenhagenish" interpretation suggests that, such an interpretation/variant/whatever you wish to call it is empirically distinguishable from MWI. Now we are getting extremely deep into Gedanken experiment territory as the technology required would be truly fantastic (and the experiments possibly unethical); however, in MWI, not only does it make perfect sense (and is required) to include the experimenter in the quantum state, sufficient control (i.e. the ability to isolate from the environment to such a degree that decoherence does not occur) of the combined experimenter+experiment system allows coherent superpositions. Whether it is because of objective collapse or because of whatever reason a "Copenhagenish" interpretation disallows it, macroscopic objects (including people) cannot, in principle, be in coherent superposition in some variants, and they can in MWI.

    Perhaps it is possible to work out a less extreme experimental difference between MWI and your Copenhagenish interpretations, but the existence of one in principle is sufficient to answer the OP's question in the affirmative. In the meantime, I will amuse myself by imagining physicists being hurled at high velocity at a double slit in the hopes of observing an interference pattern:rofl:
     
    Last edited: Jun 2, 2013
  20. stevendaryl

    stevendaryl 3,090
    Science Advisor

    Von Neumann, in his treatment of quantum mechanics, describes two types of processes associated with a wave function: (1) Smooth evolution according to the Schrodinger equation, and (2) projection of the wave function onto an eigenstate following a measurement. This second type process is what people mean by "wave function collapse", I think, whether or not the term was used.
     
  21. That's a perfectly reasonable framework; however, it's getting rather close to bare instrumentalism. Instrumentalism, of course, is a reasonable position to take, but you're essentially removing yourself from the "interpretation game" when you take it. I think this is fine as a working theory, since it's entirely sufficient for anyone who just uses QM. However, it shouldn't be mistaken as a genuine interpretation of QM's formalism, nor should it be taken as eliminating the need for one—for, as we are now discussing, different wavefunction ontologies are, in principle, distinguishable and so the practical instrumentalist/agnostic position will (hopefully) not be tenable forever.

    The point I'm making here is that we need to compare apples to apples. There are variants of quantum theory that arise from interpreting (and/or extending) the formalism of QM in different ways. MWI is one, Bohmian mechanics is another, as are various objective collapse models, and so on. Some of these are in principle testable, and it's important that that be understood. On the other hand, it doesn't make sense to compare MW with your Copenhagen/instrumentalism viewpoint, since all it is is the neutral position you adopt in the absence of sufficient evidence in favour of a particular QM interpretation. Which is fine, but it shouldn't stop us from working on devising experiments that will yield such evidence.

    Of course, some people go a ways further than you do and adopt the position that "a recipe for getting testable predictions" is all QM is, that there is no independent wavefunction, no epistemic framework, nothing. That position—ontological instrumentalism, instead of just practical instrumentalism—is an interpretation of QM, insofar as it denies an independent existence to all the referents of QM's formalism. Whether such an interpretation is also experimentally testable, I'm less sure. Maybe evidence for objective wavefunction collapse would count as evidence against it. In any case, it doesn't sound like this is your position anyways so this is beside the point.
     
Know someone interested in this topic? Share a link to this question via email, Google+, Twitter, or Facebook

Have something to add?

0
Draft saved Draft deleted