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

Question on Uncertainty Principle

  1. Sep 1, 2011 #1
    If Uncertainty Principle is a result of 'fiddling' of an apparatus with what is being observed (for example light particle/wave of a detector fiddling with particles/waves being observed), then wouldn't uncertainty principle cease to apply if in the future a 'smaller' or 'less interfering' element of nature is discovered for use for observation? Say, a via the use of a "string powered detector" or whatnot.

    Also, I notice a lot of disagreement between websites on it. The most 'zany' appear to imply the principle is about an abstract concept of "the observer can not know" while others take it down to earth and talk about interference of the apparatus observing and the particles. I tend to think the 2nd camp is the most rational, hence my question: why do they take it so strongly and not consider it a "weak" principle, since, they don't seem to have proven they will "never" be able to measure them without interfering with them [since it appears in the future such a method/technology/science may be discovered], unless I miss something?

    I have been offered advice that there are experimental proofs, e.g. involving zero point energy:

    I'm not convinced any of this isn't a result of interfering equipment or incompletely isolated experimental setups. Whenever I click a link about those proofs I read about controversies, objections, and scientists suggesting other reasons explaining them.

    Most importantly many of the physicists I see on videos lecturing on it, when asked why it happened they offer as a personal opinion the "light through a dark corridor" example[/analogy], or a variation. This is: you may only see if light goes through the corridor, by putting your hand between the light and its target, abstracting its path (the observer is in a room perpendicular to the corridor with no way to see the light other than obstructing it; no air dust is present).

    If they see clearly a relation of the uncertainty principle with apparatuses obstructing the initial behavior of particles then I keep asking: why not treat it as a "weak" principle since in the future science+technology may appear that can see particles without obstructing their initial behavior?
  2. jcsd
  3. Sep 1, 2011 #2


    User Avatar
    Science Advisor
    Gold Member

    Good question! The Heisenberg Uncertainty Principle (HUP) is a fundamental aspect of quantum theory. It really has nothing to do with the resolution of observational devices, and that is why it would NOT be considered a "weak" principle. Here are a few reasons your idea is incorrect (and I hope you don't mind my bluntness):

    i) Experimentally, it only applies to non-commuting observables. Commuting observables are not constrained by the HUP. If it had something to do with observational issues, then all observables would be affected. So for an electron the x, y and z spin components do not commute, and so an exact reading of the x component (which is accurately accomplished using an S-G apparatus) renders y and z completely uncertain. Meaning, that if you knew the state of y or z previously, that state is now gone.

    ii) You can measure 2 non-commuting observables on a pair of entangled particles, separated by some distance (entangled particles will essentially have identical properties after suitable basis adjustment). By your idea, the HUP would be violated. But it isn't! The results are no different than if you performed 2 measurements on the same particle. In other words, the entangled particles serve as the perfect means of checking whether the act of observation "disturbs" the particle so that its state is changed. It turns out that the HUP is the guiding rule here.

    iii) Lastly, you can "erase" the results of an observation and return the particle to its earlier state, something which should NOT be possible according to your idea.
  4. Sep 2, 2011 #3
    I think the eraser experiment may help me to understand it better. I prefer an intuitive method rather than going through math.

    thanks for the pointer.
  5. Sep 2, 2011 #4
    Doc Chinese perhaps I can ask you to expand a little for a dabbling philosopher.

    I came to the forum looking for a clear understanding of whether quantum indeterminacy is (in my lingo) epistemic or ontological uncertainty. By epistemic I mean 'what we can know' and by ontological I mean 'what there really is'. So the result of a fair roullette wheel may be epistemically uncertain (we don't know where the ball will land and hence bet) but that does not mean that it is ontologically uncertain - God-like knowledge of starting conditions and understanding of mechanics will tell us where the ball will land (let us assume).

    So, with quantum indeterminacy is the indeterminacy ontological or epistemic? In other words, could God know the x, y and z spins simultaneously? (I am not religious but the notion helps my question)

    Well, from what I read elsewhere it looks like the question is confusing to physicists for a couple of reasons:

    Broad reason - "why think that there is an ontological realm at all, if we can't EVER know, then what does ontology amount to but unverifiable speculation? I am not interested in that!"

    Narrow reason - "knowledge of the initial conditions is impossible in the quantum realm and so no amount of understanding of the subsequent mechanics will help determine a future state - it is not just difficult to predict (like roulette) but impossible!"

    I am not concerned with the broad reason here but rather the narrow one. Couldn't God know the x, y and z spins simultaneously? Couldn't God know the initial conditions?

    Now, I think what your answer above said was that the problem was ontological not just epistemic since the equipment being used was not at issue - in other words, God couldn't know. What I don't understand is how this claim is established - I cannot imagine a proof that would establish this. I didn't entirely follow your answer. For example:

    "Meaning, that if you knew the state of y or z previously, that state is now gone."

    Again, if we imagine a God-like non-interventionist measurement of x spin, would that not leave y or z in the original state? Surely the "knowing" of the x-spin isn't the problem, it is HOW we come to know the x-spin that is the problem.

    Sorry if this is simple-minded but, as with so many people, I find the idea baffling. It would make much more sense to me if it were epistemic indeterminacy that we were dealing with rather than ontological. But the issues are discussed as if the problem is fundamental in the world and not just fundamental in how we come to know the world.

    In advance, thanks.
  6. Sep 2, 2011 #5


    User Avatar
    Science Advisor
    Gold Member

    Welcome to PhysicsForums! These are great questions. I don't pretend to have absolute answers, but I think I can provide some useful information on the matter.

    In QM, it is generally considered that not only are x, y and z simultaneously unknowable, they do not simultaneously exist. (Please be aware that strictly speaking, there are interpretations in which this is not the case, such as the non-local Bohmian interpretation, but I don't think this should effect things for our purposes.)

    Let's review a couple of items.

    First: If I place a series of identically oriented x polarizing filters in the path of a particle already oriented to pass (say x+), then each one leaves the particle in the same x+ state – no change. Each one also “spins the roulette wheel” on y and z, leaving them indeterminate until measured. This is true because any previous y or z state is gone, and this is experimentally verified. And follows QM as well.

    Second: This leaves open the question of whether God could know the new y and z orientations. But then we have Bell’s Theorem. Bell demonstrated that there were no combinations of y and z that match the statistical predictions of QM. Bell did not use the angle settings x, y and z per se, but rather other angle settings for which the principle could be demonstrated. For simplicity, I would recommend reading a page of mine which demonstrates this issue:

    Bell’s Theorem with Easy Math

    I believe that these two points should be understood before going much further, as these will be cornerstones of the argument.
  7. Sep 4, 2011 #6
    Dr Chinese,

    Thanks greatly for the steer - I had looked for your much-lauded page on Bell's theorem but couldn't find it before!

    Now, I think I am starting to get the idea a little better. I have some thoughts on the maths which may fall out of the simplification you provide or it may be inherent in the Bell proof:

    - If the angles were 100,200,300 degrees then the probability of any AB or BC matching would be (cos100)squared - 0.03 (approx). If the angles were 2, 4 and 6 degrees then the probability of AB or BC matching would be .9987. If the angles were 90, 180 and 270 then the probability of them matching would be 0. Yet in all cases the "classical" probability would be the same - .333. This is beyond weird. Surely if the rotation of angles can shift the probability from 0 (totally impossible) to .9987 (quite near certainty) then there must be something in the angles that is dictating the matching. The classical probabilities are bluntly assigning equal probability to each of the outcomes - surely it is just missing some understanding of the impact of the angles.

    - If the photon passes through 120 and then, after an interval, is tested against 120 again, did you say that the probability of it passing through is now 1? If so what of two entangled photons both of which are tested simultaneously at 120, then at 120 again and then at 240 (three tests on each single photon). Would the third reading be expected to match in each photon's cases?

    Thanks for this - immeasurably helpful (hehe).
  8. Sep 6, 2011 #7


    User Avatar
    Science Advisor
    Gold Member

    1. Your math is slightly off but you have part of it correct. The 33% minimum does apply across the board as you state.

    However, for the angles you must average 3: AB, BC and AC. When you do that for 90, 180 and 270, the result is 33% (average of 0, 0 and 100 as 90 and 180 always match). For 2, 4, 6 degrees it is a bit less than 100% (which is still above 33%, so no obvious violation).

    2. No, they would not be expected to match but about 25% of the time.

Share this great discussion with others via Reddit, Google+, Twitter, or Facebook