Tuesday, August 09, 2005

Interpretation of QM and progress

One of the recorded discussions in Toronto was between Lee Smolin and Martin Roček. They debated the achievements of the research focusing on the interpretational issues of quantum mechanics.

Martin Roček argued that almost no progress occured in this field while Lee Smolin claimed that quantum computing is one of the counterexamples. I tend to agree with Martin.

Quantum mechanics looks weird. As Sidney Coleman once said, if thousands of philosophers were trying to find the weirdest possible thing for thousands of years, they would have never come up with something as strange as quantum mechanics.

But quantum mechanics is true and successfully tested. All those great physicists who could not swallow it and who were proposing various "completions" were wrong. This list includes de Broglie, Schrödinger, Einstein, Podolsky, Rosen, Bohm, Bell, and many others.

We should certainly admit that the questions raised by these Gentlemen were instrumental in deepening our understanding of quantum mechanics. The physicists have understood the EPR effect, entanglement, quantum teleportation, and some mechanisms useful for quantum computing. But it is even more important to say that their opinion that quantum mechanics had to be modified has been falsified as much as you can get.

There have been some unsatisfactory features of the Copenhagen interpretation - such as its unjustified separation of the world into the microscopic and macroscopic realm - but these features do not really affect the predictions for any experiment that has been done (or even realistically planned). Moreover, these features have been more or less clarified by the concept of decoherence from the 1980s that has become a part of satisfactory, modern neo-Copenhagen interpretations of quantum mechanics such as the Consistent Histories.

There also remain several open questions about the interpretation of quantum mechanics in the context of cosmology and quantum gravity (what are the right observables near the Big Bang?), but these questions can hardly be answered before we properly understand dynamics of quantum gravity. Traditionally, dynamics and interpretation are two separated parts of the intellectual structure called quantum theory, and it is only dynamics with which most serious physicists may be spending hundreds of hours and about which they can write quantitative papers. While it is fair to say that quantum gravity has the capacity to smear the boundaries between dynamics and interpretation, it is not fair to say that the people who are simply questioning the foundations of quantum mechanics are contributing to the research of quantum gravity.

Concerning quantum computing, I believe that one of the important boosts for its development came from the 1982 article by Richard Feynman about simulating the quantum world by computers. And Feynman was no philosopher who would like to think about subtle philosophical principles and feel unsatisfied with a working theory of physics because of some abstract prejudices. On the contrary, Feynman thought that the philosophers were politely speaking dopes, and his favorite interpretation of quantum mechanics was summarized by the dictum "Shut up and calculate!".

I personally don't see any significant future for the enterprise of "improving the foundations of quantum mechanics" because the foundations work; all major recipes how to modify them drastically have been shown inconsistent with reality (which showed that the "opponents" of quantum mechanics do not understand an important part of physics well); and new progress in identifying the complete interpretation of quantum mechanics will require more concepts that we will learn from dynamics of quantum gravity.


  1. "Shut up and calculate" is a quote most likely by Mermin and meant as a derrision against Copenhagen.


    >>But perhaps Feynman was offering such advice to others who were searching for a better understanding of the quantum mechanical formalism. I can't believe that. He said that he "always had a great deal of difficulty understanding the world view that quantum mechanics represents," and added, "I still get nervous with it."2 Nobody who felt that way would ever respond with "shut up and calculate" to conceptual inquiries from the perplexed.<<

    Not that it matters to anyone but the hhistoricians of the field.

    To me there is a simple physical question in this: Does Quantum MLechanics describe meassurement with the right probabilities? Meassurement is a physical process occuring constantly, independent of human observation and metaphysical questions.
    It appears that QM requires ad hoc hypothesis going beyond unitary evolution to describe it (the classical being collapse, in MWT it's the insertion of the Born rule by hand). These are the "interpretations", often tied to or inspired by philosophical thoughts.

    Decoherence also tells us why it's so hard to devise experiments testing the different ad hoc hypothesis, it does not tell us wether or not QM does describe meassurement.

    There are several reasons that have lead several accomplished physicists to suspect that there might be something still lurking there and that we might find it in Quantum Gravity, one particular thing being the role plays in normal Quantum Mechanics, now this though is an issue of actually doing the calculations discussing them and seeing what they tell us.


  2. Something that troubles me in terms of "quantum computing" is the understanding that along with the Window on the Universe, this might be seen in relation to Glast. In this sense, the Calorimetric view, might have revealled the deeper interplay of such a computerized processes?

    I think this might be Smolin's point, and the assessment you gave in terms of how Smolin is approaching this?

    But while this view is being established in relation to this method, one more avenue raises the question of LIGO information and how it would be defined in quantum computer aspects, and the rebulding process this information might reveal of those cosmological events?

    Any points of clarification for sure.

  3. Dear Frank,

    I personally find it a bit strange if Dr. Mermin tries to steal a piece of Feynman's fame by pointing out that he has said the famous quote, too.

    I don't know whether Dr. Mermin said it or not, but it is almost definitely less important than that Feynman did say it. And the fact that Feynman replied it - first as an answer to a student who asked about the interpretation of QM - is a well-known fact that you can even find in some of the Feynman's books.

    Feynman thought that QM was very strange, much like Coleman or me. He said that while only "12" people used to understand GR, it's safe to say that none understands QM. But these were mostly jokes; Feynman himself was definitely convinced that he knew how to apply QM to make predictions, and this is what mattered to him.


  4. Expecting that almost for certain what I can expect from Peter Woit, some one who enjoys ridicule hard working string theoretists for failures, but would not be able to come up with his own ideas, and some one who would absolutely NOT tolerate any discussion of any successful ideas, and who seeks to quarantine any of such dangerous ideas, I thus use the space of Lubos, who at least has some intelligent and can torelate the discussion of ideas, in order to preserve something I just posted.


    All those proposed ways of "modifying" quantum theory are just technicality details that does not address one fundamental question physicists should have asked long time ago, but still have not asked. Does it make a difference how QM is described, by Schrodinger equation, operator dynamics, path integral, or other method. It's only a difference in mathematical form, not a difference in physics.

    The whole QM is built up upon a few fundamental principles. Actually I think there's only one and a half principles: The uncertainty principle, and then there's the Pauli exclusion principle, applicable only to SOME particles. You can add none-linear terms to schrodinger equation when spacetime discreteness kicks in. But the bottom line question is still "has the uncertainty principle be change in any way" and the answer can only be yes or no. Very likely, adding none-linear terms to schrodinger equation is merely a technical necessity in order to preserver the uncertainty principle and preserve a constant hbar. You really can not say you have modified QM if there is only technical modification but the fundamental principle remain untouched. And there is good reason to believe that uncertainty principle WILL be preserved even at a scale of discrete space and time.

    The fundamental question to ask is why the nature needs to observe an uncertainty principle to start with. You can say "it is just the way nature is" but that is hardly a satisfactory answer. You need to ask what is the more fundamental reason that causes uncertainty principle to become a necessary certainty. Presume you are the creator and you can start over re-creating a new universe from scratch, disregard any none essential physics laws we know and set a brand new set of laws. Is it possible, then, for you, to construct a self-consistent universe, where uncertainty principle doesn't exist, every things is classical, and further more, masses do not attract each other except for through interaction of electrical charges.

    Can such a universe be designed, excluding both QM and GR principles, and still be self-consistent, and operational? If not, why? If yes, why?

    I am going to just ask the question, without discussing any of my own thoughts, lest Peter would though I am discussing my "pet ideas". It is never wrong to just ask questions, and if you do ask the right question you may make progress.


  5. Lubos says:
    "Feynman thought that QM was very strange, much like Coleman or me. He said that while only "12" people used to understand GR, it's safe to say that none understands QM. But these were mostly jokes; Feynman himself was definitely convinced that he knew how to apply QM to make predictions, and this is what mattered to him."

    Surely Feymann understand QM better than any of us, but I guess he is right in saying no one understands QM. He is better than the rest of us in that he realized he didn't know. And the rest of us don't even know enough to know that we don't know.

    Lubos, being able to carry out QM calculations and predict things doesn't mean you know QM. It only means you know the math formulations of QM. A properly programmed computer can do the same QM derivations and make predictions, and even do it better and faster than we do.

    But a computer doesn't under understand why its code needs to be written that way, instead of some other ways. We the human know the nature observes the uncertainty principle and all that, but really doesn't understand why the nature HAS to observer UP, and why there can't be an alternatives. Until we understand why UP is necessary, we really don't understand QM.

    That's what Feymann means that QM is strange.


  6. Just wanted to add the link by Gerard t' Hooft, on the Future of Quantum Mechanics.

  7. Does Quantum mechanics look weird? For a good bunch of philosophers, Classical Mechanics was weird. Quantum Mechanics is sweeter to swallow, because it forbids you to speak about finite momentum within an infinitesimal distance. What is weird, or at least unexpected, is that in order to implement this prohibition QM uses complex probabilities to average trajectories. We have some understanding of why it is so; after all also we get imaginary exponentials when we regularise a dirac delta, but while classical mechanics has a kind of "first principles" approach, QM has not got an author for its "Principia".

  8. I was under the impression that the pressure being exerted by the quantum gravity camp "opposite" string theory was being exerted, not on a modification of QM, but rather, on its reinterpretation (such as people like Carlo Rovelli have been advocating).

  9. Frank and Lubos,

    Lubos wrote:
    >Feynman thought that the philosophers were politely speaking dopes, and his favorite interpretation of quantum mechanics was summarized by the dictum "Shut up and calculate!".

    Frank wrote:
    >"Shut up and calculate" is a quote most likely by Mermin and meant as a derrision against Copenhagen.

    Lubos replied:
    >…Feynman did say it. And the fact that Feynman replied it - first as an answer to a student who asked about the interpretation of QM - is a well-known fact that you can even find in some of the Feynman's books.

    As it happens, I’m actually one of the students (not the first, I’m sure!) who asked Feynman about the interpretation of QM, so I know from personal experience what his reaction was. I took QM from Feynman my junior year at Caltech (‘74-’75) and Intro to High Energy Physics my senior year (‘75-’76).

    He never actually told me “Shut up and calculate!” He was too nice a guy for that: despite the image he created of himself as a socially boorish bull-in-a-china-shop, he was actually a quite considerate person, at least in dealing with undergrads (I had an oral final with him the end of my senior year which I was dreading – it had been a pretty tough course – but the oral final actually ended up being a pleasant experience).

    However, while Feynman was too polite to say it to any of us explicitly, “Shut up and calculate!” was clearly his attitude. I tried to get him to explain why he felt that way -- he was not willing to discuss it (though he was happy to discuss other matters at length).

    My impression was that he had simply concluded that, after geniuses such as Einstein, Bohr, Heisenberg, etc. had struggled with the problem, none of the rest of us was likely to make any progress and we should not waste our time. I think he was probably agnostic as to whether any future progress in physics would radically change the structure of QM and would have thought speculating about the same also to be a waste of time (i.e., “Shut up and calculate!”).

    Incidentally, at Caltech, I also had John Schwarz for a freshman course, and Polchinski and I were in the same dorm, so I had a chance to know a couple of the eminences in string theory when they were still “nobodies.”

    All the best,


  10. To me “Shut up and calculate!” is basically lowering the research to an engineering project. Don't ask why, just follow the instructions in the user manual step by step and you can get the result. That's engineering, not science exploration.

  11. About QM, I wrote time ago
    suggesting how near a path integral is of a regularised minimum law, ie of a set of regularised delta primes. This is a way to understand why the imaginary weights, and it could be interesting to check if other regularisations of delta prime are equivalent to quantum mechanics. On other hand, the approach does not make clear the role of angular momentum, and angular momentum is the key of mechanics, either classical or quantum. Even liuville theorem is mostly a way to account it at infinitesimal level.

  12. Speaking as a philosopher, not a scientist, it seems to me that quantum theorists are guilty of a mjor non-sequitor. They infer non-causation from unpredictability. Or, in other words, from our uncertainty regarding the predicted behavior of quantum particles, they draw the invalid conclusion that this quantum behavior is uncaused. Am I wrong about this?

    I would argue, to the contrary, that the principle of causation (that every contingent event has a cause)is an a priori truth which no amount of empirical observation can overturn--and I have seen nothing in QM that would overturn it other than the above non-sequitor.

  13. Dear Dr. Cowan,

    as a physicist, I would say that if a conjecture is described as an "a priori truth", then this conjecture is not a scientific hypothesis but rather a religious dogma that a scientist is not interested in (and moreover, scientists tend to believe that most such dogmas are false). In science, every conjecture may be questioned and every conjecture needs a positive body of observational, experimental, and mathematical evidence to be trusted.

    Physicists only use the word "cause" - in analogy with other words - for those causes whose existence has direct physical consequences or indirect physical consequences that may be derived via mathematical reasoning.

    For example, one may conjecture that the precise result of a quantum measurement depends on (or "is caused by") some a priori pre-existing hidden variables that obey local natural laws. Such an assumption implies that the so-called Bell's inequalities must be universally satisfied.

    One may show experimentally that Bell's inequalities may be easily violated by doable experiments. This proves that the result of the measurements in quantum mechanics cannot be determined or caused by any pre-existing physical degrees of freedom as long as their dynamics is local.

    One may speculate about some non-local contrived untestable influence that is causing a particular result to be chosen, but physicists are usually not interested in angels on the tip of a needle if their existence is untestable (and if their existence seems to break the Lorentz invariance and other principles).

    The violation of Bell's inequalities is one of the examples how physics can show that philosophers' preconceptions may be heavily incorrect. In fact, it is likely that 99% of all "a priori truths" that the philosophers have had about the nature have already been falsified. The existence of "cause" behind every quantum microphenomenon is a newer example; the flat Earth is an older one.

    And this fact is also one of the main motivation for Coleman's (and perhaps originally Feynman's, as David Goss told me) statement that philosophers could not invent something working and as strange as quantum mechanics even if they spent thousands of years with it.

    A group of people can decide that no amount of experimental evidence can overturn a certain dogma as you said; but these people should not be surprised if they are viewed by this Reference Frame, much like other scientifically thinking people, as equivalents of the Inquisition, religious bigots, or perhaps monkeys jumping in between palms rather than cutting-edge thinkers. Thanks for appreciating my clear formulation.

    All the best

  14. Lubos:

    I am not interested in Steve Cowan's religous view but I have to defend his philosophy view point, because his is the accepted and established philosophy view point of most scientists, except for those who does not even have an idea what it is all about.

    You have demonstrate you have zero training in philosophy, no wonder you have no idea what Steve was talking about when he says "a priori truth".

    There are SOME truth that you simply have to assume they are true, but there is no way you can prove it by experiment or logic, because they are just too fundamental to be proveable.

    Have you heard about the 5th postulate?

    Or try to prove to me that 1+1 = 2.

    Or try to prove to me that the universe exists.

    Or try to prove to me a number of other things that are painfullly obviously true, but you simply can't show why they are true. Those are the type of things that you simply HAVE to unconditionally accept as truth, like the universe must exists and every thing we see and touch must be real.

    What Steve meantioned, the principle of causation (that every contingent event MUST has a cause)is an a priori truth. And the most important one that the whole fundation of science is based on!!! If you don't accept that one unconditionally, then there is no science at all. It's that simple.

    It is ONLY because we CAN rationalize things we see and observer, and we are capable of finding connections between things, and that such connections exist, that we can find true of the nature by doing experimental science. If everything just happen randomly on their own and there is no calsual relationship whatsoever between things, and everything is explained away as coincidence, then physics laws do not exist and nor can we study science.

    And science only developed when the principle of causation is widely accepted, and it is today still the very fundation that science exists at all. Too bad some science researcher totally have no idea of this important fundation. Those people are just enginners since they do not have the wisdom of a scientists.


  15. Lubos said:

    "This proves that the result of the measurements in quantum mechanics cannot be determined or caused by any pre-existing physical degrees of freedom as long as their dynamics is local."

    I am very glad that you are careful enough to not forget to add the qualifier at the end of your sentence, "as long as their dynamics is local"

    That is exactly the key to understand the Aspect experiments. You can either interpret it as demonstrating lack of causuality, or as lack of locality. Either way the experiment can be explained satisfactorily. But a philosopher would rather prefer to keep the causuality, and give up the locality, than to preserve locality and lost causuality. And that is in line with the firm believe of the principle of causation.

    And No I do not think a philosopher can develope any science truth. But that is NOT because their philosophy is wrong, rather it is because they do NOT do experiments. Same is true for physicists. Without doing experiment no physics truth can ever be found.


  16. Dear Quantoken,

    one can assume whatever he wants, but in science it makes sense to ask whether a meaningful statement is correct in nature or not and the answer is never obvious, especially if the question is a very difficult one.

    The statement that there always exists a material reason - a cause - why a certain result of a quantum measurement is chosen by nature is a wrong statement.

    The fifth postulate is, as shown by general relativity, not generally valid either. On the other hand, according to all of our working theories and counting systems, 1+1=2 and "the universe exists" are valid statements.

    Let me repeat that there are no statements in science that one must unconditionally accept as true - especially not any statements that you have ever been capable to make - and whoever thinks that there are such statements is a dogmatic bigot.

    Thousands Quantokens can repeat millions of times that there must exist a material cause for any particular outcome of every quantum measurement - or even that this misconception is an inevitable dogma for every scientist. But even if they repeat it millions of times, the statement will remain untrue and every person with IQ above 70 and elementary knowledge of the status of quantum theory knows why these Quantokenish statements are wrong and why Quantokens are crackpots (and this example is definitely not the only reason).

    Quantum mechanics does not say that *everything* is a coincidence, as you incorrectly argue. But it definitely says that a particular result of a particular experiment is a result of coincidence and only the probabilities of different outcomes may be predicted. If you're unsatisfied with this fundamental fact about our Universe, I recommend you to apply for asylum in a different one. Maybe, in this other Universe they may think that you are a scientist as opposed to a complete moron. Good luck.


  17. Lubos:
    "Let me repeat that there are no statements in science that one must unconditionally accept as true - especially not any statements that you have ever been capable to make - and whoever thinks that there are such statements is a dogmatic bigot."

    Lubos, do you realize that the very statement you just made, as highlighted above, is one such statement that you just have to accept as valid, without prove. Once you try to rationalize it, then it contradicts itself by proving itself as a counter example, hence invalidate itself!!!

    Do you know Godel's theorem?

    The 5th postulate may not be a good example. What about the first 4 postulates, Lubos?

    You have NOT shown me how you can prove "1+1=2". You know it's true, I know it's true. We all agree it's true. But no one can prove it. Nor is it necessary to prove it. We just accept it as so.

    I feel it is as tough as proving 1+1=2 to try to teach you what is a priori truth. You need to grab a few philosophy books and read them. Note you came from a former iron curtain country, you probably know nothing more than Maxism or Leninism in terms of philosophy.


  18. Lubos said:
    "Quantum mechanics does not say that *everything* is a coincidence, as you incorrectly argue. But it definitely says that a particular result of a particular experiment is a result of coincidence and only the probabilities of different outcomes may be predicted."

    I did not make that argument. I have not even meantioned quantum mechanics so far. I have not argued what QM says and doesn't say. So don't put words in my mouth.

    I original argument is that once you remove principle of causation as an a priori truth, then you are left to resort to pure coincidences to explain everything, because you have denied the causation principle. On the other hand, if you do not want to attribute things to coincidence, you have to accept causality and accept the principle of causation.

    Now, by refering to QM, you are trying to make a case that maybe something are causual, and something are coincidential. You want a half and half situation where the principle of causation is right half of the time, and wrong the other half of time.

    That doesn't work. You either accept the principle of causation in whole piece, or reject it in one whole piece. There is no half way compromise, Just as you can NOT apply one set of physics laws to half of the universe and the other half with a different set.

    Now, certain aspects of outcomes of quantum processes are un-predictable, as proven by experiments. I do not dispute that. However, un-predictable does NOT equal to no causation. Every thing still has a cause behind it. It's just there is no way we can observe that cause without disturbing the system. So, since the cause can not be precisedly determined, the outcome is not predictable. But the whole principle of causation is fully preserved.

    It's just like when a computer generate a pseudo random number or a cryptographic key, it is un-predictable to an outside observer what the key could be. But causation is fully preserved on the computer. Every little step of computation is completely deterministic and completely spelled out in the code. It's just you can not detect the input, so you can not predict the output. The same thing works for QM.


  19. Hi There!

    I have been following the arguement regarding causation and I must admit that Quantoken has a point!

    I believe that Lubos has mixed up the induction problem with the a priori causation issue! Quantum Mechanics is NON DETERMINISTIC and thereby violates the induction hypothesis! But one can certainly set up QM to be able to follow the chain of causation in QM! Try out the Heisenberg version of QM rather than Schrodinger's version!

    Interesting enough, chains of causations are essentially a description of SpaceTime! So, some new a priori statements maybe needed to describe Quantum Gravity, rather than Causation! However, in flat spacetimes, QM certainly obeys causality a priori!

    An amateur mathematician.

  20. Philosophically, QM seems to have a few weak areas. One is the assertion that Schrodingiers Wave Function is a probablility function - is there any actual evidence to support this?

    The second is the concept of Measurement. This seems to be a classical concept, yet it is being applied to quantum interactions. Classical physics allows observations which do not affect the objects being observed - a one way flow of information. Quatum interactions involve information exchange - you can only observe something by bouncing something else off it.
    This is true in classical physics but can be ignored because the objects being observed are so much larger than the objects being used to make the observaton(e.g. billiard balls vs photons), or by observing the effect of a large number of small objects (e.g. a n electric current is the sumation of the movements of a very large number of electrons).

    While a measurement is an interaction between two objects, the Copenhagen Interpretation seems to focus on the observer instead of the interaction.
    Some of the difficulties in understanding QM may be due to asking the wrong question, or asking the right question the wrong way. If you do an experiment to measure light as a wave you get a wave answer and if you do one to measure photons you get photon answers.
    Therefore asking a QM question in terms of a classic observer doing a measurement rather than as a interaction involving two way flow of information (energy, momentum, etc) is likely to produce an answer which is a confusing mixture of Quantum and Classical.