Thursday, 7 December 2006

Quantum Physics Debate, 2: The Many Worlds Interpretation rebutted

Further to our debate on the Many Worlds Interpretation of Quantum Physics begun here at Not PC by Brian Scurfield, doctoral student Mark Sadgrove offered this response which has been lightly edited to appear here as a post (thanks, Mark):

I probably shouldn't wade in here, but it is nice to see this topic being discussed on a New Zealand oriented site, simply because, academically speaking, there are some fine minds in this country when it comes to research into fundamental quantum physics, and it would be nice if that translated into wider public interest.

Disclaimer: I'm a Quantum Physics PhD who has published work with Rainer Leonhardt and Scott Parkins mentioned in comments in the original post. Specifically, my own work could be said to err more towards supporting the notion that interaction with the outside world is the only thing that's needed to explain the quantum measurement problem away... Having said that, the Many Worlds Interpretation is as far as I'm aware still an interpretation rather than a theory with new, testable predictions, so I'm really no more qualified than anyone else or for that matter any more biased by my use of established quantum theory when it comes to commenting on this stuff.

I have a few points to make:

1) The "Many worlds interpretation" is still called an interpretation. Is anyone aware of any specific predictions that Many Worlds theory makes which are outside of standard quantum theory? In this case there might be hope for confirming it one way or another...

I seem to recall that some (perhaps prominent?) physicists have said that a successful quantum computer would force people to believe the many worlds theory because, well, that super-classical computing power must be coming from somewhere , the idea being that we share the possible computing power of a huge number of somehow existent universes when we use a quantum computer. Personally I still don't find this convincing. From a very utilitarian point of view, that computing power arises because nature, as embodied in the laws of quantum mechanics, allows it to occur. Quantum mechanics just works that way and that's that.

The reason people are groping for interpretations is because they find it difficult to form a clear, intuitive picture of the natural process that is occurring [in observations of Quantum phenomena], unlike in classical mechanics where little balls colliding with each other is to a crude degree all the mental imagery you'll ever need. But an interpretation is just window dressing until it actually leads to intuition which makes NEW predictions. Smarter people than me are backing 'many worlds,' but I'd like to know if they've made any progress on the prediction front.

2) Occam's Razor. I always thought that the most compelling reason for being suspicious of the Many Worlds Interpretation was Occam's Razor which states that given the infinite possible explanations that one can give for a physical phenomenon, the simplest possible explanation (that is the one with the fewest parameters) should be preferred. [N.B., it can be simply stated as the law of succinctness, in Latin: "entia non sunt multiplicanda praeter necessitatem," or, "entities should not be multiplied beyond necessity."

'Many Worlds' theory multiplies Universes ad infinitum just so we can have a cosy mental picture of the natural processes described by QM. A lot of people think that this is a dear price to pay! This is a capitalist-oriented site right. [Oh yes - Ed.] So do you really think that Nature would be this wasteful of resources when cleaner, more efficient single-universe models can explain things just as well? (Of course, the point is that some people think that ONLY Many Worlds theory can explain all the observed phenomena in which case it doesn't matter how "wasteful" it seems, because it's the only game in town.)

Another way to look at it is through the testability lens which a lot of people have also brought up. If you claim that there are multitudinous other universes which are created by quantum "splitting" or "differentiation" events but, oh hang on, you can't ever reach them or even feel their influence except via a rather unspectacular interference experiment, then why should I believe you? People who do believe such things are, for example, good candidates for believing that there is a very specific God in Heaven - a bearded man in the Judeao-Christian tradition for example, exactly as described in the bible [sorry: The Bible].

While we have no real evidence of such specifics, believers might say, if exactly such a god did exist it explains a few things about the world. Well perhaps, but why believe all these things that you can't verify along with the very general idea of a creator which does not imply any specific form for god.

==> Do we really have to swallow a multiverse teeming with infinite slightly differentiated universes all equally "real" just to explain the quantum measurement problem? Are there leaner versions of many worlds without infinite versions of each of us floating around in them?

(3) Is QM an interpretation? Remember that Physicists don't have to buy into any interpretation of quantum mechanics to use it effectively. In a comment on the previous post, Brian S took Scott Parkins' to task about a quantum jump "interpretation". But I don't think Scott was using any interpretation. The jump issue in quantum mechanics, or the collapse of the wave function, ra, ra, ra, whatever you wish to call it, is I think the bare minimum theory needed to predict naturally observable events. I don't agree that there is any interpretation going on here.

The facts are that when you look at the distribution of photons from a slit experiment on a detector screen, each one makes a "dot" at just one point - that's your measurement. To describe the distribution of points you use quantum mechanics. Quantum mechanics describes the photon as a "delocalised" or non-point-like entity - a wave as such which gives the requisite interference pattern. It also provides a probabilistic rule telling you how likely a given measurement is given the form of this wave. What you do to connect the wave model of the pre-measurement photon to the measurement you actually make - your interpretation of the observed physical phenomenon - is up to you, but it doesn't change the basic description of what happens which is standard quantum mechanics.

Some rather vocal proponents of the de-coherence project think that they've got the quantum measurement problem - jumps and all - ironed out by just considering the interaction of a closed quantum system with it's environment. I favour de-coherence myself (remember I'm biased - I invoke it a little in my thesis) because it seems clean. The idea is that the idea is that the nice probability waves that don't have any particular "position" are in reality rapidly converted into particular "classical" states in any real system because there is interaction with the environment. I don't think many people believe that de-coherence theory is enough to explain the measurement problem away completely, but it IS a theory and it does help make sense of certain experimental observations to some degree.

Personally, de-coherence looks to me like Quantum Mechanics without the difficult jump, but with the Born Rule explained (see W. H. Zurek, Probabilities from Envariance, 2004 - still controversial!!!) and no multiple Universes required.

Reading the recent Wheeler preprint ('100 Years of the Quantum' -Tegmark & Wheeler [pdf]) referenced in the previous post, the De-coherence theory is in fact presented as being an important addition to Everett's Many Worlds Theory and not an explanation in competition with it. I am not convinced after a cursory review of that paper that there is really anything left to explain if dec-oherence works as well as Tegmark and Wheeler imply in this preprint, but even so the "Many Worlds Theory" they describe does not seem as grandiose as the usual idea of infinite simultaneously existing Universes being almost identical to our own. Rather, a far more subtle "Many Minds" idea is suggested which seems rather different to me, or at least is far less suggestive in a sci-fi sense.
* * *
Okay sorry to go on. Maybe my overall point is that Many Worlds Interpretation seems to excite a lot of people when it's really a subtle idea, for which refinement might remove a lot of the sci-fi elements -- this is even if it survives as an attempt to provide a picture of the undoubtedly successful quantum mechanical laws. It seems to me that the successful implementation of Shor's algorithm alone is not enough to irrefutably prove the existence of multiple "me's" in Universes of which I can never otherwise observe the effect.

As a final note, I attended the Quantum Physics of Nature last year and saw Zeilinger's labs in Vienna. Very nice! I also heard debates about precisely this stuff and I can assure you that the physics community is nowhere near a consensus...

LINKS: Quantum Physics Debate, 1: The Many Worlds Interpretation - Brian Scurfield at Not PC
Introduction to Quantum Physics - Wikipedia
Many Worlds Interpretetation - Wikipedia
Born Rule - Wikipedia
Probabilities from Envariance - Arxiv.Org
Copenhagen Interpretation - Wikipedia
The Quantum Aristotle - Peter Cresswell, SOLO

RELATED: Philosophy, Science



    Mark Sadgrove said...
    [I'm a Quantum Physics PhD who has published work with Rainer Leonhardt and Scott Parkins mentioned in comments above.]

    Mark , Welcome. Perhaps, that your long comment be re-posted by PC (owner of the blog) as Part 2 [Done - Ed.].

    I suspect that PC, will come out with an article of his own, in terms of philosophical implications of Quantum Mechanics, perhaps part 2, if your very nice post is not to be made as part 2, then PC can post his one as part 3.

    I was at the Department as well in the mid-1990s, and late Professor Dan Walls & Prof. Mathew Collete were the lecturers for Quantum Mechanics. Prof. Leonhardt & Prof. Harvey were the ones teaching 'Opto-Electronics'. Dr. Bold's son (Geoff) was in the same class as me.

    I guess that I will see you at the 'inaugural symposium of the Dodd-Walls Centre for Photonics and Ultra Cold Atoms' , next Saturday week , that is 9 December.

    I might try & spot you for a chat.



    Why should I believe you are conscious? After all, the only consciousness I ever experience is my own and I cannot experience your consciousness. Isn't believing in billions of other consciousnesses just to explain the reality of my own perceptions wasteful? Surely nature is much more parsimonious?


    I'm well aware of Frank Tipler's stranger theories, but he would hardly be the first scientist to believe in a creator. I cite him because I think that his point about non-locality is correct. Arguments need to be taken on a case-by-case basis, for a man who is wrong about many things can be right about some things.

    BTW - The multiverse considered as a whole is a much much simpler thing than any particular universe and in fact contains zero net information.

  3. Mark Saidgrove said...
    [...the Many Worlds Interpretation is as far as I'm aware still an interpretation rather than a theory with new, testable predictions.]

    I wouldn't want to add more to Mark's comment, because it has been brought up many times in this debate, but I just want to stress these FACTS:

    #1) MWI is non-testable after 50 years when it first proposed by Everret.

    #2) MWI has not proposed any single PREDICTIONS yet despite after 50 years to be worth a serious look at, except that it is a fascinating topic to start talking about at parties or newsgroup.

    I fired some emails recently to proponents of MWI such as David Deutsch & Frank Tipler to list PREDICTIONS of MWI to date just in case I have missed that myself from research publications. There has not been any reply from Tipler, despite sending 2 emails. Deutsch replied to my query and said, that MWI is pretty difficult to test. Well I say no more than that. In my Sunday School days when I was young, I recalled that the Sunday School teacher started every Sunday by asking the class 2 questions, which were more like indoctrination:

    #1) Who made you?

    Some kids said, "It was mummy & daddy" and some said, "It was the Almighty God". The teacher corrected those who answered "mummy & daddy" to say that the correct answer was "God Almighty" where this entity cannot reveal itself to humans. In other words, it is physically impossible to communicate with such entity or to test if such entity does exist. I suspect that if such entity exist then it must reside in one of those other multiple universes.

    #2) Where is God?

    Every kid knew that the answer was "He is everywhere". Now, the MWI said that other parallel universes are everywhere around us and possibly occupying the same space as our universe. Again MWI fits in with this "God is everywhere".

    Now Professor Frank Tipler (an MWI proponent) has just stopped short of declaring that the "Laws of Quantum Mechanics" must involve an extra parameter to represent GOD. He has quoted on internet interviews/articles that he wouldn't rule out an "Intelligent Designer".

    Until , MWI comes up with some useful predictions for humanity to be used in practical technology applications in our universe, it should be remained as an interesting topic of discussion at parties so that to impress girls, but not a useful topic for researchers to get bogged down in debating, since it hasn't made any useful contribution to technology. Quantum computers will be built with or without MWI, and this is fact not an arbitrary assertion. Los Alamos (US Nuclear facility research center) scientists had built some primitive experimental proto-type machines over recent years, and their effort had none to do with MWI.

  4. Brian S said...
    [Arguments need to be taken on a case-by-case basis, for a man who is wrong about many things can be right about some things.]

    I hope you're right here Brian, regarding Frank Tipler's interpretation of Physics. Lets see, when he comes up in the future with the modification of the standard quantum mechanics equations to perhaps include a parameter for existence of GOD. I predict that Professor Tipler will call this new parameter the "Almighty Constant", with such modification to standard quantum mechanics equations, it would be possible to solve the new formulation to find out which of the many infinite universes that GOD resides in.

    Yes, Brian , watch out for Tipler in the coming years, where I suspect that he will incorporate the "Intelligent Grand Designer" idea in to standard quantum mechanics equations with the inclusion of the "Almighty Constant" parameter.

  5. Brian S said...
    [I have explained the point several times now. You believe in an instantaneous, undetectable, and unpredictable collapse. That ... violates relativity, ...]

    No, no, no Brian, you have not. You have explained several times, but all obfuscation, not clear on what you were describing. You merely gave general description and not specific.

    OK, to be fair to you, I will re-state my question for a specific example.

    Explain how the instantaneity observed in Anton Zeilinger's , et al, experiment at Innsbruck?

    Brian S said...
    [That not only violates relativity, it violates common sense.]

    So, you're implying that Non-locality violates relativity but not the existence infinite multiple universes which violates conservation principles (energy & matter)? I take it that you favour one violation and simply ignore another violation?

  6. Kirk, McCoy, Scotty and Uhura (and didn't she look good in her new outfit) got propelled into a parallel universe on the weekend so it must be true, right?


  7. Insider said...
    [Kirk, McCoy, Scotty and Uhura (and didn't she look good in her new outfit) got propelled into a parallel universe on the weekend so it must be true, right?]

    I think that they all did. When they came back from one of those parallel universes into our universe, their names had been interfered with while they were there, which they are now known as Ian Wishart, Nicky Hagger, John Key, Helen Clark.

  8. Brian,

    The following is a response from (Dr) Gregor Weihs, who is a co-author of the paper for the "Innsbruck University Double-delay choice" experiment together with Anton Zeilinger. Gregor Weighs now teaches at Institute for Quantum Computing of the University of Waterloo, Canada.

    Gregor Weihs said...
    [Thank you for your interest in my work. The interpretation you choose
    (many worlds or multi-verses) doesn't change anything about the physics. Therefore I agree with you that it isn't testable. I am also convinced that (local) hidden variable models can't explain the entanglement we see in the lab. But hidden variables and many worlds are not the same. Many worlds were invented precisely to feel better about entanglement. They don't explain anything as far as I am concerned, but as before, this is more a matter of taste.]

    As Mark was saying in his comment, that there is no consensus amongst Physicist, there is one clear barrier to accepting MWI by a majority is that the 'testability' of MWI is physically impossible, whether it is 10, 50, 100 , 200 or 1,000 years from now, it would still be unachievable.

  9. You've got the Star Trek thing wrong FF! What really happened is that Kirk got himself entangled with Helen Clark's brain after the Enterprise had a bizarre time-travelling accident. He and Clark now share some kind of weird telepathic link.

    Said Kirk: "It's spooky action at a distance. Sort of like a mind-meld with the Borg, but without their sense of humour. I've found that I now have the ability to retrospectively change the laws of physics at will. Normally Scotty would be quite upset about that except that he's busy doing strange things with Tribbles after becoming entangled with David Benson Pope's left testical."

  10. FF,

    Where did you get the idea that the MWI is untestable from Deutsch's reply? According to you, he said that the MWI is pretty difficult to test. "Pretty difficult" is not "impossible". Bose Einstein Condensates were "pretty difficult" to test at the time of Einstein's prediction.

  11. Right, didn't see that this had been reposted, sorry, just made another long comment under Brian's original post.

    Just to clear things up, I actually have my PhD, although it's pretty fresh (alf a year old) and you can download my thesis off the UOA Physics Dept. website - follow the publications link and look for the thesis link, from memory. Also, this discussion is definitely concerning the philosophy of quantum mechanics, so I'd like to point out that my qualification doesn't make me an authority or anything, although it's fair enough to expect me to have a (well informed?) opinion on this stuff.


    P.S. Did you guys know that one of the largest quantum computer candidates to date ( Technical paper, "Deterministic Quantum Teleportation with Atoms," Nature, June 17, 2004; Technical paper, "Deterministic Quantum Teleportation of Atomic Qubits", Nature, June 17, 2004. ) was built with the help of a New Zealander? Murray Barret now works at Otago but his QC research was performed in America in the Wineland group. Not sure if he'll be keen on being bombarded with questions though!

  12. Mark,

    I came across a publication of yours here:

    "Observation of Robust Quantum Resonance Peaks in an Atom Optics Kicked Rotor with Amplitude Noise"

    I don't understand the detail of your publication, however I do understand the main concepts. Would you share your knowledge with us here in this forum about the technology implication of the research you published in that paper quoted above?

  13. Mark,

    I'll quote from your post on the other thread.

    Deutsch is obviously a REALLY smart guy and I find his arguments stimulating and his criticism of the utilitarian approach to physics rather trenchant (I guess I effectively promoted this approach in my first post). But damned if I still don't feel CONVINCED by his argument about the power of quantum computers proving many worlds. I won't feel convinced ever, in fact, unless someone can provide me MATERIAL proof (or similar) of those worlds. The existence of some large calculating resource outside of our universe is very different (to me at least) to the idea of multiple universes populated with multiple you-and-is. I feel compelled to be cynical about such an idea from the core of my scientific cynicism... but then why doesn't a smart guy like David Deutsch I wonder?

    Deutsch's argument does not rest just on the existence of quantum computers. In his book, Deutsch builds the argument up from a number of different directions. Many of these I touched on in my original post. You say you feel compelled to be cynical about the idea that there are other versions of yourself. This is understandable but - as I'm sure I don't have to tell you - sheer incredulity is not a reason to dismiss a theory. It is not cynicism that the MWI needs, but our best criticism.

    Brian S, as for your comment about the consciousness of others, well perhaps we can agree that, apart from all the emotional stuff that makes me regard others as conscious like myself, from a rational viewpoint, the consciousness of others will always remain a (very useful) hypothesis which we will never be able to verify directly.

    Of course, that is an assertion. We don't know enough about consciousness to say that for sure. Similarly, people say that the MWI is not testable like it is an established fact. It is not. And I would say that our prospects of directly testing the MWI are far better than our prospects of directly testing another person's consciousness.

    I don't really think that's solipsism just a brutal fact. But to boil down your argument to it's point, perhaps it's more illuminating to ask why nature bothers to be anything at all... I mean apart from doubting the existence of others, I should perhaps doubt the existence of my physical body (clearly unecessary and wasteful merely to create the perception of a physical body which I experience) etc etc, and in fact even the existence of a coherent, enduring "I" seems doubtable. But I think that you interpret my Occam's Razor argument as a solipsist sort of argument.

    Yes, that is how I interpreted it.

    My point was that given what we observe in nature, the Many Worlds Interpretation seems to require a lot more resources to explain the observations than interpretations such as Copenhagen or (pushing it) Decoherence.

    However you look at it, quantum mechanics requires far more resources than we observe. The question is: to what extent should we regard those additional resources as real? In the Copenhagen Interpretation, nothing is real until collapse. So what we observe is the outcome of an interaction of the real with the possible. But what does that mean? We're on a slippery philosophical slope, and our explanation is rapidly going to require far more additional machinery than one that just says those additional resources are real.

    Just like the solipsism argument.

    Now when we regard those additional resources as real, and then ask about the structure of those resources, we arrive at the conclusion that reality is partitioned into layers down which information flows easily and that those layers are in fact universes.


    You mentioned decoherence. As you pointed out in your original comment, decoherence is an important part of the MWI. However, I don't think decoherence can be regarded as an interpretation in itself. It does not give a mechanism for wave-function collapse, as you seem to be saying it does. Rather it just explains why the observables of macroscopic objects *appear* to be single-valued.

  14. Mark S. Here again, long after the debate ended.
    I've been reading a little bit here and there about "Many Worlds" over the past months (It was the Nth anniversary of Everett's paper recently!), and I thought that I should mention something I discovered: being a fan of decoherence I should (to at least some degree) admit to being a Many Worlds advocate! A closet MW theorist perhaps?

    I felt rather surprised to learn this myself, but the key point is that since I subscribe to the theory of decoherence, I don't believe that the wave function collapses. Instead, according to decoherence, the appearance of a collapse is created by the extremely rapid interaction of the an initially pure quantum state with its environment, leading to the "wavelike" nature of the original state (i.e. it's existence in many states at once) being "distributed" over the state + environment (if my terminology is vague, it's because I'm not an expert in the general field of decoherence, I've just used it in very specific instances once or twice). Then when you go and look at just the objects comprising the state itself (i.e. an atom) rather than the whole quantum system (i.e. the atoms plus an impinging light field, plus other atoms nearby) you don't see the quantum wave nature of the atom any-more. That's because the interactions with the environment have lead to superpositions of the atom and environment states rather than just of the atomic states (it's actually a form of entanglement, if you're fond of that buzzword).
    THE POINT IS there's no collapse. If there's no collapse, then you're a many worlder... but the description above sure doesn't sound like a universe full of alternative "me"s does it. That's what I just can't get out of Everett's theory - a vast meta-Universe of unknowable parallel worlds... I just don't think that Many Worlds leads enevitably to such a picture, or, if it does, how we're ever supposed to find out if it's true. Everett himself didn't use the phrase "Many worlds". That's the main thing I want to be careful about.

    Hope everyone enjoyed the MWT anniversary!

    Oh yeah one more thing. Brian S. makes a lot of Deutch's assertion that the power of quantum computing has to come from SOMEWHERE (obviously from a near infinite set of Universes containing copies of us all right?! ;)) How then do you explain the fact that at least some of the power of Quantum computing comes from the wave nature of matter alone, and can therefore be reproduced by CLASSICAL wave theory (no many worlds needed there guys.)
    To support these assertions, I offer to references to Physical Review:
    i) No entanglement needed to do a classical beating quantum search (just wave properties)
    S. Loyd, Phs Rev. A 61 010301(R)

    ii) An OPTICAL ONLY implementation of a quantum search algorithm (using just the wave properties of light, which can be described by classical equations).
    N. Bhattacharya,et al. Phys. Rev. Lett. 88, 137901 (2002)

    Also if anyone's interested, I've just published some work with the group here in Tokyo on using Quantum interference to make a ratchet.

    M. Sadgrove, et al. Phys. Rev. Lett. 99, 043002 (2007).

    Mark Sadgrove

  15. Gah, well I guess if you adopt a smarty-pants tone online, you're setting yourself up for a fall. I just remembered that, although what I wrote about speed ups using classical waves is correct, it missed the point.

    THE BIG THING about quantum computing is that you get the speed up free of charge in an important sense: there is no nasty growth in the resources you require as you make the size of your computer bigger.

    You can do things fast with classical waves, it's true, but you'll always be forced to use some physical resource or parameter in a way that gets exponentially demanding as you increase the computer size. For example, perhaps you'll have to specify the frequency of the waves with exp(N) accuracy where N is the size of your system.

    Okay, I have to admit, that's a pretty important point and I have to hand it to Brian S that he's been saying that all along (or at least reporting the observations of the very clued of Deutsch).

    Despite the fact that I knew this before I posted my last message, I have to admit my little screw up just now makes me appreciate all the more that it IS strange that Quantum Mechanics lets you get exponential speed-up without exponential resources. Of course, duh, that's why everyone is so keen on it.

    It's a little embarassing working out one's philosophy in public. I'll go back to lurking now. I will, however, admit that this discussion has made me readjust my thinking about many worlds, although my fundamental objections are still the same.

    Mark S.


1. Commenters are welcome and invited.
2. All comments are moderated. Off-topic grandstanding, spam, and gibberish will be ignored. Tu quoque will be moderated.
3. Read the post before you comment. Challenge facts, but don't simply ignore them.
4. Use a name. If it's important enough to say, it's important enough to put a name to.
5. Above all: Act with honour. Say what you mean, and mean what you say.