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

Quantum collapse

  1. Mar 7, 2009 #1
    this is a question that has been bugging me from the begining of my learning of quantum mechanics. why does the the wavefunction collapse when we do a measurement?
    since we have no concrete idea about collapse, we use probability theory to understand quantum phenomenon. but at least from classical framework, we use probability when we dont know the full picture. in some sense we can call probability the 'mathematics of our ignorance'.
    now bell has proved that no hidden variable theory of quantum mechanics is not possible. and because of that it seems that the it is impossible to understand the dynamics of collapse.
    but why not? i may not be able to predict that at a particular measurement which eigenvalue shall i get, but cant i predict that given a specific type of measurement, that is understanding of all the parameters involved, we should be able to tell for which parameters which eigenvalue shall i get. i may not be able to control the parameters and hence we shall still need the framework of quantum mechanics, but is this proposition impossible?
    in this respect i remember the famous comment of einstein - 'god doesnt play dice with the world'. but then even dice has some rules, some mechanics, some dynamics.
  2. jcsd
  3. Mar 7, 2009 #2


    User Avatar
    Science Advisor

    Well that's not quantum mechanics per se, it's the Copenhagen interpretation.

    Personally, I regard it as a bit of a non-issue. The 'problem' or 'paradox' lies mostly in how the question is framed. (Zeno's paradox is another example). You're regarding the 'measuring' system and the 'measured' system as separate, when in reality, they are not, since they interact.

    This all goes back to the early quantum physicists who were a bit confused, or rather, conservative. They were trying to stick with old 'classical' ideas, such as the idea that a system always has a definite value for its measurable properties, and that measurements, at least ideally, can be performed in a way that does not interfere.

    If you drop all that and simply regard the wave function as the fundamental description of the system, rather than some mathematical device that could be used to get measurable values, then the metaphysics of it all becomes much simpler.

    It's not an unusual thing. I recall reading something by Stephen Weinberg where he noted that it's not easy at all to follow the reasoning in Newton's "Principia", even to someone well-versed in Newtonian mechanics. Because, Newton wasn't really the first 'Newtonian' physicist. He was the last non-Newtonian physicist. I think the same type of confusion applies in this case.
  4. Mar 8, 2009 #3
    I think that the collapse of the wave function occurs only in the mind of the observer.In the old Schrodinger cat paradox,the cat (at least to my way of thinking) was never in a two state system.It was always in one state or the other,until a conscious mind knew which one.The whole measurement problem surely has to take account of the connection between consciousness and quantum phenomena.
  5. Mar 8, 2009 #4
    The user alxm is right. The quantum events are the primary. The classical notions arise from quantum ones as the inclusive picture - all different events are piled up and considered as relating to one "free", for example, particle. Classical picture is a kind of cinematographic illusion. The primary are rare black points on each frame, the secondary or illusory is the "image" obtained with superimposing different frames.

  6. Mar 8, 2009 #5
    can you just explain to me what is meant by this statement?
  7. Mar 9, 2009 #6
    The idea that there is a collapse, or that an intelligence is required to collapse the waveform, or that conciousness affects quantum phenomena is rubbish.
    If it were so you could, for example, record the result automatically and look at it a week later. What happens then ? does the waveform collapse a week after the event ?
    Does a flea know if the cat is dead or alive ?

    As I have said before on this forum the Cat was Schrodinger's parable showing the stupidity of the view prevelant at the time that nothing is there if no one is looking.

    See the cartoon on Nearing Zero, where the attractive lab assistant is just a waveform when the Prof. is not looking at her.

    Read about the deBroglie - Bohm version, beaten to death at the Solvay Conference in 1927. Read The Undivided Universe by Bohm.
  8. Mar 9, 2009 #7
    You say "the idea that there is a collapse--------is rubbish.Do you think then that the two forms of the wavefunction continue to exist independently? ;do you really mean that!!?.If so,how do you explain why we only see one outcome when we make a measurement ?. I agree with your view about "the stupidity of the view----that nothing is there if no one is looking",and I would say that even when no one is looking,a quantum system will have come down on one side or the other of the quantum fence.
  9. Mar 9, 2009 #8
    When someone has a car accident an observer going to help the person inside thinks: He might be dead or alive(Wave function), when that person finally sees the driver (Measurement) one of the 2 possibility is not possible anymore (Collapse of the Wave function).

    Is this a good example?
  10. Mar 9, 2009 #9
    Yes, and it is reading then start again. Not COLLAPSE all, only CHANGE. What we want knowing happening when two wavfunction strike together do know they it?
  11. Mar 9, 2009 #10


    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor

    No, it is not a good example.

    There have been so many threads on this here on PF that I feel silly repeating all of this again, but here goes.

    There is a difference between classical and quantum system. This is no big surprise. Let's say you toss a coin and it falls into a box before you can see how it lands. So you don't know if it is head or tail. Same situation as the accident. Does this mean that the coin system is now in a superposition of states of "head" and "tail"? No, it doesn't! The coin is a classical system. It doesn't work that way. We detect no superposition effects of |head> + |tail> wavefunction.

    What you have is that the coin is already in a definite state. It is just that you do not know what it is. So you say that there is a 50% probability that the coin produces head, and 50% probability that it is tail. It is either OR.

    This classical statistics is clearly different than the superposition of a bipartite quantum system. A quantum system, before it is measured, is really in a superposition of ALL the possible states that it can have. It means that there is no definite out YET before it is measured, and before you take a peak at it. This is clearly different than the classical (coin) system, where in that situation, the coin already has a definite state, and it is just our ignorance that is forcing us to consider all the possibilities.

    Now, how can one tell the difference between the two if, upon measurement, all we see is just ONE value? This is where knowing what is meant by "non-commuting" and "non-contextual" observablein quantum mechanics is crucial. One can make a measurement, say, or a non-commuting observable value, which will preserve the superposition of the the other non-commuting observable. For example, say [A,B] != 0, where A and B are observables. If I measure B, the wavefuction collapses for B, but NOT for A! If I can find some way of making such a measurement on the identical system repeatedly, I may be able to detect that A is in a superposition of states even though I'm only measuring B.

    This is naively what is done when we detect bonding-antibonding states in chemistry, the coherence gap in the SQUID experiments out of Delft/Stony Brook, etc.. etc. These phenomena cannot occur if the systems are already in a definite state. This is why we invoke the superposition principle as vital in QM, because it fits into numerous experimental observations, and why it is VERY different than classical systems.

    Edit: I've posted references in this post:


    Last edited: Mar 9, 2009
  12. Mar 10, 2009 #11
    The Bohm - DeBroglie view of quantum processes is much better in my view. Bohm notes that although the treatment of statistical processes is satisfactory in the Copenhagen view, the individual quantum processes use unsatisfactory assumptions - like the collapse of the wave function. The mysterious wave-particle duality. The inability to give a clear notion of what the reality (ontological interpretation) of a quantum system should be.
    Then there's nonlocality.
    (I've somewhat paraphrased Bohm)
    Read: - The Undivided Universe by Bohm.

    From Cambridge University Cavendish Lab. http://www.tcm.phy.cam.ac.uk/~mdt26/pilot_waves.html
    Last edited: Mar 10, 2009
  13. Mar 10, 2009 #12
    Nice observation.
  14. Mar 11, 2009 #13
    This sentence doesn't make grammatical sense.What do you really mean?
  15. Mar 11, 2009 #14
    But what if the observable here commutes with every other physical observable? That is to say, measuring whether the coin is heads or tails won't change any other physical observable? Is that right? So, I could look at the date of the coin and then look whether it's heads or tails and the observation of the date won't affect or interfere with the heads or tails state of the coin. This seems reasonable to me. All these macroscopic measurements seem to commute to me and I'm just back to good old classical probability. But that's not to say that the quantum theory should fail to describe this system and I fail to see any reason that quantum mechanics doesn't work for large systems as well as small. It's just that if there is any non-commutation, the effects are generally minuscule.

    I have a feeling I must be wrong here, because you've emphasised an enormous difference between this macroscopic (not classical) case and the microscopic case generally treated with quantum mechanics.

    Edit: If there is a genuine difference between macroscopic and microscopic systems, what use is schroedinger's cat, the atypical macroscopic system described as a superposition of two base states, |dead> and |alive>
  16. Mar 11, 2009 #15


    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor

    I don't know exactly what you are asking here.

    If A and B commutes, then for a non-degenerate case, measurement of A will also give you B. For example, H and p commutes for a free particle. If you know momentum, then the energy is also determined. There's no mystery here.

    What is invoked in the Schrodinger Cat-type experiment is the non-commuting/non-contextual observables. These DO exist!

    Furthermore, the "size", as in microscopic versus microscopic isn't the issue. It is how well and to what extent can you maintain coherence of the system. In the Delft/Stony Brook experiments, they showed a superposition in a system as large at 10^11 particles! This is because superconductivity can show coherence at the "macroscopic" scale! There's nothing that say that we can't have it larger. The big problem is how to isolate the system from coupling to the environment that will destroy the coherence.

  17. Mar 11, 2009 #16
    I don't really understand this line. Coherence of what?

    Yes, it looks like I completely forgot what commuting means. Swap every time I said "commuting" to "non-commuting". I might even have meant anti-commuting, but I'm not sure. Obviously measuring the heads or tails state of the coin won't tell you what date the coin is. Although, looking at the coin, you could see both the state of the coin and its date (assuming the date is printed on both sides, which it isn't).

    What I meant to ask was "why is there any difference between the case of a coin and the case of a quantum system, with regards to the mathematical description?". I see that there is a difference in interpretation and in that regard I'm agreed that in one case we do not know, although the state is defined and in the other case, we do not know until we make the measurement. It seems like you're assuming the Copenhagen interpretation implicitly.
  18. Mar 11, 2009 #17

    The fact that we don't detect superpositions like |head> + |tail> is simply because due to very fast decoherence the rest of the universe becomes entangled with the state of the coin. So, in general, we have a superposition like:

    |head>|universe_1> + |tail>|universe_2>

    There is then no way to do interference experiments to demonstrate the difference between a classical state in which the coin has either tail up or head up. But the formalism of quantum mechanics still demands that, generically, you end up in a superposition of two very different states.

    The only way to get rid of the superposition is to assume new physics in which non-unitary effects lead to a real collapse of the wavefunction. But there isn't a shred of evidence for such non-unitary effects.
  19. Mar 11, 2009 #18


    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor

    Then how do you explain all those observations that I've cited? Obviously, the presence of the coherence gap in the SQUID measurement is a result of such superposition, and it was measured! No analogous effect can be demonstrated for the classical state.

  20. Mar 11, 2009 #19


    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor

    Do you not understand the usage of "coherence" in what I posted, or do you not understand "coherence" as used in QM, i.e. the onset of classical phase with decoherence?

    In the classical case, the description is never a "superposition" of the actual state, but rather in the state of YOUR knowledge of the system. This is why the absence of "realism" in QM, i.e. the absence of an already-established state that hasn't been measured, is such a major issue with QM. I think I've highlighted a paper in the "Noteworthy papers" sticky that showed that even when one invokes non-locality, one still cannot save realism. It means that these states in QM are not yet determined until they are measured. This is clearly different than the classical case where "either or" cases occurs. You will get either head, or tail, but not a mixture/superposition of both, even before you toss the coin or take a look after the coin lands. Not only that, you cannot perform ANY experiment (there's none so far) equivalent to the experiments that I've mentioned.

  21. Mar 11, 2009 #20
    The observations are consistent with quantum mechanics. Even the measured decoherence time is consistent with theoretical predictions.

    After the system has decohered you cannot experimentally demonstrate a difference between the following two possibilities:

    a) The system undergoes a "real collapse" in the non-unitary sense (requires new physics)

    b) The system does not undergo any non-unitary evolution, rather it has become completely entangled with the environment.

    The reduced density matrix for a) and b) are the same and there is no way to distinguish between the two alternatives.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook