- Hi everyone,
I'm pleased to report that my paper on "Einstein's Boxes" has appeared
in the current (Feb.) issue of AmJPhys. I am quite happy with the final
version and I would invite anyone who is interested to check it out!
There is also a fascinating critical comment on my paper (in the same
issue) written by Abner Shimony, who was one of the referees for my
paper. I personally think he is rather confused about the whole business,
but I have no complaints since his comment will be good publicity for my
I would of course love to discuss the paper or Shimony's reply (or other
critical comments or thoughts that people may have) here...
- On 29 Apr 2005 16:07:53 -0700 "Israel Silverman"
> But in the case of space-like separated events,No, Aspect eliminated the possibility of subluminal
> the only possibility for causal influence is superluminal
> causal influence. Forbidding any such influence
> (by imposing Bell Locality) generates a
> prediction which is violated by real experiments.
> So these experiments involve superluminal causation.
> Further, besides the polarizer efficiency argument,
> can we really say that Aspect eliminated all possibility
> of subluminal causation?
causation between A and B during any given
coincidence interval. However, if the correlations
aren't produced by A and B affecting each other
(and most physicists believe that they aren't),
then spacelike separating them doesn't matter.
You've got crossed linear polarizers analyzing the
same light. So, you can separate them by a
billion light years, and as long as the light
from the emission events is undisturbed in
transit, then you'll get cos^2 theta correlations.
The roadblock to understanding why Bell
test results don't imply superluminal signalling
in nature is a flawed analysis of the comparison
between an incorrect (the usual lhv) formulation
of the probability of coincidental detection and
a correct (qm) one. You could certainly generate
an lhv that gives the correct cos^2 theta correlation
curve, but it would proceed along the same line of
reasoning that's used to generate the qm prediction.
The point is that if you don't change the basis for
calculating the probability of detection at B once
a detection is registered at A, then you're not doing
it right. This doesn't imply that the light incident
on polarizer B has changed in any way via superluminal
signalling or whatever. In fact, the correctness of the
cos^2 theta formula *depends* on there being *no
change* in the light incident on polarizer B following
a detection at A. It has to be the same (meaning
in phase with and polarized the same, via emission)
as the light that was incident on polarizer A (which
light, via transmission of some portion of it by
polarizer A, ultimately produced a photon via the
PMT at A).