home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: sci.physics
- Path: sparky!uunet!psinntp!scylla!daryl
- From: daryl@oracorp.com (Daryl McCullough)
- Subject: Re: hidden variables
- Message-ID: <1993Jan6.174455.15423@oracorp.com>
- Organization: ORA Corporation
- Date: Wed, 6 Jan 1993 17:44:55 GMT
- Lines: 86
-
- rkoehler@author.gsfc.nasa.gov (Bob Koehler) writes:
-
- >Would someone be so kind as to summaries the argument that hidden
- >variable theories violate causility? It's been a long time since grad
- >school, and this has been scratching at the front of my mind recently,
- >while I know the answer lies burried deep in the back. I just haven't
- >had the time to wade through my books.
-
- The simplest inequality to derive was given by Wigner in the American
- Journal of Physics, Vol 38, page 1005, 1970. (There are also several
- derivations in the book _Speakable and Unspeakable in Quantum Mechanics_
- by J.S. Bell.) A simple derivation of a "no hidden variables" theorem
- goes as follows:
-
- Let A,B, and C be three events. Then the joint probabilities obey the
- following inequality:
-
- P(A and not-B) + P(B and C) >= P(A and C)
-
- To see this, simply note that
-
- P(A and C) = P(A and not-B and C) + P(A and B and C)
-
- the first term is less than or equal to P(A and not-B), and
- the second term is less than or equal to P(B and C).
-
- Now, let's apply this inequality in the case of measuring spins of a
- spin-1/2 particle such as an electron. Assume that there are hidden
- variables that determine, for each direction, whether the spin will be
- measured to be up or down in that direction.
-
- Let a, b, and c be three unit vectors in the x-y plane such that the
- angle between b and a is 60 degrees, and the angle between a and c is
- also 60 degrees (so that the angle between b and c is 120 degrees).
- Let A be the event "The particle has spin-up in the a direction", B be
- the event "The particle has spin-up in the b direction", and C be the
- event "The particle has spin-up in the c direction".
-
- Now, it is only possible to measure the spin of the particle in one
- direction at a time. However, in a twin-pair experiment, two spin-1/2
- particles are produced with exactly anti-correlated spins; whenever
- one particle is measured to have spin up in one direction, the other
- particle is measured to have spin down in that same direction.
-
- Therefore, it becomes possible to measure the values of the spin in
- two different directions, by measuring the spin of one particle in
- one direction, and measuring the spin of the other particle in the
- other direction.
-
- The predictions of quantum mechanics for such an experiment are:
-
- P(A and not-B) = 1/8
- P(B and C) = 1/8
- P(A and C) = 3/8
-
- This violates the derived inequality:
- P(A and not-B) + P(B and C) >= P(A and C)
-
- This implies that our assumption was wrong; it can't be the case that
- the result of a spin measurement simply tells us the value of a
- pre-existing property of the particle.
-
- The other possibility is that the value of spin measured in a
- particular direction depends not only on the hidden variables
- associated with the particle, but also on the choices of orientations
- of the measurement devices. A slightly more complicated derivation
- shows that the following hypothesis is incompatible with the predicted
- results of quantum mechanics:
-
- The value of the spin of a particle depends only on (1) hidden variables
- associated with the particle, plus (2) the choice of orientation of the
- measurement device that the particle passes through.
-
- The only way to get a hidden variables theory that is consistent with
- quantum mechanics and probability theory is if the value of the spin
- measured in a particular direction for a particle depends on the
- choices of directions of *both* measurement devices (the one for
- measuring the particle, and also the one for measuring its partner).
- In other words, the only way a hidden variables theory can work to
- explain the twin pair experiment is if the orientation of a distant
- measurement device instantaneously affects the measured properties of
- a particle here.
-
- Daryl McCullough
- ORA Corp.
- Ithaca, NY
-