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

I Conceptual catch in EPR

  1. Aug 9, 2016 #1
    I question whether there would be a conceptual catch in EPR proposal to get both non commutative variables define with accuracy at one location simultaneously. The idea of separability that EPR introduced to questioning the completeness of QM is that whenever one measures a physical variable q at Alice location, one can predict with absolute accuracy the-q of Bob particle without disturbing it.
    My claim is that there is still uncertainty in p variable (the other non-commutative variable) of Bob`s particle simply because it is not yet measured. So the prediction of one of non-commutative variable does not mean that both variables are known before the measurement taken place on the other variable. And consequently, this does not grantee that Bob particle possess elements of reality regarding p and q.
     
  2. jcsd
  3. Aug 9, 2016 #2

    DrChinese

    User Avatar
    Science Advisor
    Gold Member

    Mostly I would agree. There is no question that EPR thought that p and q could be simultaneously measured without uncertainty. They felt that any other perspective meant that the reality of Bob was dependent on the nature of an observation by Alice (who is presumably spacelike separated). And that would be the counter view you are asserting - that we live in an observer dependent reality.

    Of course, your view is consistent with Bell. EPR did not have the benefit of his analysis of the situation. :smile:
     
  4. Aug 9, 2016 #3

    Strilanc

    User Avatar
    Science Advisor

    We have definitions for what we mean by uncertainty, and they don't distinguish between "I know what the next measurement will be because I already measured it just now" and "I know what the next measurement will be because of [complicated reasons]".

    As an analogy, suppose I flip a normal everyday coin then cover it with my hand. You don't know if the coin landed heads up or tails up. You're maximally uncertain about the outcome. But then I lift the coin and show you the bottom, which happens to be heads. Are you still maximally uncertain? After all, you still haven't seen the top of the coin.

    Of course not! You immediately infer that the coin landed tails up. The bottom of the coin and the top of the coin are related; there is mutual information present. Given one side, you can infer the other. And the quantum case with EPR pairs is no different in this respect: there's mutual information, you see one side, and then you infer the other side. (What is different and interesting is that the mutual information is somehow stronger. You get the same correspondence along multiple non-commuting axes.)
     
  5. Aug 9, 2016 #4

    stevendaryl

    User Avatar
    Staff Emeritus
    Science Advisor

    I don't remember who was the genius who thought of changing the noncommuting variables from [itex](p, q)[/itex] to different spin components, but the latter is a big improvement, in my opinion. But in any case, if you have two noncommuting variables [itex]A[/itex] and [itex]B[/itex], and you have perfect correlation (or anti-correlation) between two different particles, then:
    1. If Alice measures [itex]A[/itex] for her particle, she knows with certainty what value Bob will get for a measurement of [itex]A[/itex].
    2. If Alice measures [itex]B[/itex] for her particle, she knows with certainty what value Bob will get for a measurement of [itex]B[/itex].
    In case 1, assuming no non-local interaction, it seems that [itex]A[/itex] for Bob's particle had a definite value. But the value for [itex]B[/itex] can be uncertain. In case 2, the value for [itex]A[/itex] can be uncertain.

    But EPR reasoned that if Alice's choice to measure [itex]A[/itex] or [itex]B[/itex] is freely chosen (not predetermined), and if that choice has no influence on Bob or his particle, then Bob's particle can't go from some state with an uncertain value of [itex]A[/itex] to a state with a definite value of [itex]A[/itex], because that would be a change in the state of Bob's particle. So they reasoned that [itex]A[/itex] must have already been definite before Alice made her choice. Similarly, [itex]B[/itex] must have been definite. So they were led to the conclusion that both variables must have had definite values all along.

    I think it was impeccable reasoning, but was never-the-less proved false by Bell's inequality. How flawless reasoning can be empirically refuted is one of those big mysteries of the universe. :wink:
     
  6. Aug 9, 2016 #5

    Nugatory

    User Avatar

    Staff: Mentor

    Bohm, I believe.
    Just about everyone's opinion, I expect.
     
  7. Aug 10, 2016 #6
    But concluding that the particle must have a physical reality without being measured based on a premise that the physical reality of Alice particle comes to a definite physical value after being measured is inconsistent argument. You don`t see self reference here! One can`t prove a conclusion based on a preposition which is itself a rephrasing of that conclusion because this would be an inconsistent logical thought. (by rephrasing here, I meant that the system collapses into a particular state which in general implies that the physical quantity has no physical reality at all before the measurement and of course is the case if no measurement at all has been taken place).
    For example, if A implies B and B implies C is a theory, then if C comes to contradict the physically reality, then one should argue that A is a wrong start. But if we start with C implies B and B implies C which is not physical real then why we consider it as axiom or premise in that theory in the first place?
     
    Last edited: Aug 10, 2016
  8. Aug 10, 2016 #7

    stevendaryl

    User Avatar
    Staff Emeritus
    Science Advisor

    No, it's not. Let's assume that, according to some frame [itex]F[/itex], Alice measures variable [itex]A[/itex] slightly before Bob measures the same variable. So for a brief period of time, Alice knows what Bob's result will be. So during that period, Bob's particle is in a state of having a definite value of [itex]A[/itex], even though he hasn't measured it. Either it was always in that state, or it made the transition to that state when Alice made her measurement.

    The other way out is to say that the state of Bob's particle, as deduced by Alice, is not objective, but is a subjective state, relative to Alice's information. That's sort of the Many-Worlds approach.
     
  9. Aug 10, 2016 #8

    stevendaryl

    User Avatar
    Staff Emeritus
    Science Advisor

    I realize that I can't even parse what you're saying.
     
  10. Aug 10, 2016 #9
    Because she knows what her result comes up with, right? So EPR admits that Alice`s particle has no physical reality before the measurement, right? but this is the same conclusion they disproved in their paper, So my question how one uses a conclusion to prove or disprove itself?
     
  11. Aug 10, 2016 #10

    DrChinese

    User Avatar
    Science Advisor
    Gold Member

    They don't say that, no. They simply say that Alice can learn about Bob with 100% certainty without disturbing Bob. Once they draw that conclusion, they conclude that there is physical reality independent of the actual measurement. That conclusion is based on a couple of assumptions, one of which Bell showed to be unsupported.
     
  12. Aug 10, 2016 #11

    stevendaryl

    User Avatar
    Staff Emeritus
    Science Advisor

    That's the way a proof by contradiction works: You start off making certain assumptions (in this case, locality plus observables have no definite values until they are measured) and you derive a contradiction. That means that one of your assumptions was wrong. EPR assumed that the wrong assumption was that observables lack definite values until they are measured. (Bohm instead assumed that the wrong assumption was locality.)
     
  13. Aug 10, 2016 #12
    So, how can we use a wrong assumption to prove a true statement? How do we use "observables have no definite values until they are measured" to arrive to "observables have definite values without measurement as per EPR definition of the complete physical theory". I am not sure if this a real proof of contradiction, may be I need to learn more about the subject. The true meaning of proof of contradiction is as follow; having A in hand, prove that A implies B. To do this, assume -B (negation of B) that leads to -A by logical induction, however, this contradicts the fact that we already have A which means -B can not be true and hence A implies B.
     
  14. Aug 10, 2016 #13

    DrChinese

    User Avatar
    Science Advisor
    Gold Member

    EPR showed you could predict a definite value of any of Bob's observables in advance. The logical deduction was that Bob's observables existed independent of and prior to measurement. It did require the assumption, eminently reasonable, that measuring Alice does not in any way alter Bob.

    QM implies that Bob's non-commuting observables do NOT have simultaneously definite values (essentially your "observables have no definite values until they are measured"). So that was the basis of the contradiction.

    (Of course the "eminently reasonable" assumption was subject to attack - and it was attacked.)
     
  15. Aug 10, 2016 #14
    For the first part which regards the prediction of Bob's observables, I still do not see a valid logical deduction behind it because one can not invalidate some statement, which is "Bob's particle has no definite position without measurement", based on a premise of QM that validates it. This is not a proof by contradiction. Plus from physical point of view, QM would simply say that although Bob's particle is not measured, yet it is entangled with Alice's one and it has no separate state vector from the defintion of entangled state.
    For the second part, Einstein didn't argue about the uncertainity principle, since 1927 as he simply moved from trying to prove that QM is inconsistent to that it is incomplete.
     
  16. Aug 10, 2016 #15

    Nugatory

    User Avatar

    Staff: Mentor

    Proof by contradiction is not quite that strong. If we can prove that A and -B implies both C and -C, that is sufficient to show that A implies B. However, that is not sufficient to exclude the possibility that -A and -B are both true.

    In the EPR argument, proposition A is the laws of QM, -B is "unmeasured properties have definite values" and C is "we can determine the simultaneous values of two non-commuting observables". A and -B implies C, but also A implies -C so we have our contradiction: Either A is false and -B is true, or A is true and -B is false, or both of them are false.

    If you start with the EPR premise that -B is true you'll conclude that A is false. Bohr and company preferred the other position, that A must be correct and therefore B, no matter how distasteful, was also correct. Bell's contribution was to provide an experimental method for distinguishing the two possibilities.
     
  17. Aug 11, 2016 #16
    I admire your comment but there are two points:
    1) The logical contradiction of a form (if A implies B and if A implies –B) works only when the hypothesis, A is true. Because if A is false then there would be no contradiction. In English, if QM is true and it leads to two opposite statements, then there is a contradiction, may because of something wrong in the logical induction not in the premise that QM is false.
    2) A implies B itself is not justified by the way EPR did. The reason again is related to logical invalidity to prove or disprove some statement using the statement itself.
     
  18. Aug 11, 2016 #17

    Nugatory

    User Avatar

    Staff: Mentor

    Right, so if we find that A implies B and A implies -B we can conclude that A is false.
    You may still be missing the point of the EPR argument. QM (proposition A) does not lead to two opposite statements, C and -C, it leads only to C. However, QM and the additional assumption (proposition -B) that unmeasured quantities have a definite value does imply both C and -C. Thus, (A and -B) has to be false, which implies (-A or B).
    I don't understand this at all. A and -B are separate assumptions; the proposition whose truth was assumed to start the proof by contradiction was neither of these, but rather (A and -B). Once we've found that (A and -B) leads to a contradiction, we know that (A and -B) has to be false, and the question is whether it's A that is false, -B that is false, or both of them.
     
  19. Aug 11, 2016 #18
    I don't think so. This is just a contradiction not a proof by contradiction. We can not prove that A is false here. Suppose A is "it rains" and B is "it will turn green", then saying: if it rains, it will turn green and if it rains it will not turn green, does not prove that it is not raining, simply because may be it is raining on the sea for the second case. The proof by contradiction would be similar to: prove that because it is green, then it must have been raining. To do this, we assume it is not raining, then because of this it will never turn green, assuming the complete association between them, but this contradicts the fact that we have green already then we must conclude that it must have been raining.
     
  20. Aug 13, 2016 #19
    I do not understand why you mentioned that (A and -B) implies C?. I agree that only A implies C and (A and -B) imply -C.
    I then applied truth table for rigorous analysis, (I used ">" to mean: implies and "^" to mean "and".
    When A implies C and (A and -B) implies -C, then this does not mean that A is false.
    Hereby I attach my analysis using a truth table which does not lead to a tautology in the last column because of the false result in the first row. So EPR should not prove that QM is inconsistent.
     

    Attached Files:

  21. Aug 13, 2016 #20

    Nugatory

    User Avatar

    Staff: Mentor

    I'm using the rule that if X implies Y, then (X and Q) implies Y for all Q. (If you're not sure where this came from, or if you are confused because you've seen what appears to be a counterexample, start a new thread in one of the math forums).
    Of course it doesn't, and I never said that it did. What I did say is that from the two propositions "A implies C" and "A and -B imply -C" we can conclude that (A and -B) is false. That being false is equivalent to (-A or B) being true: one or both of A and -B must be false.
    Of course not. That's not the claim being made.
    What EPR does show is that the proposition "QM is consistent and quantities not measured have definite values" is inconsistent. Use A to represent "QM is consistent" and B to represent "quantities not measured do not have definite values" and this proposition can be written as (A and -B). That's the proposition that leads to a contradiction so is known to be false. What we didn't know, until Bell provided an experimental method of testing the two possibilities, was whether (A and -B) was false because A was false or -B was false.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted



Similar Discussions: Conceptual catch in EPR
  1. On The EPR Paradox (Replies: 68)

  2. EPR doubts (Replies: 17)

  3. EPR paradox (Replies: 36)

  4. EPR paradox (Replies: 6)

  5. Entanglement and EPR (Replies: 4)

Loading...