🚀 go-pugleaf

In article <1992Oct2.191819.29150@newshost.lanl.gov> jlg@cochiti.lanl.gov (Jim Giles) writes:
)In article <1992Oct1.224027.8421@lmpsbbs.comm.mot.com>, bhv@areaplg2.corp.mot.com (Bronis Vidugiris) writes:
)|> [...]
)|> Anyway, if you set up an experiment with a pair of photons in the 'singlet'
)|> state (i.e. opposite spins) and measure the spin of  both photons along
)|> an arbitrary axis with an angle from the vertical of theta1 (for detector1)
)|> and theta2 (for detector2)
)|> 
)|> detector 1     <======     source           =======>    detector 2
)|> 
)|> 
)|> you get the following result from QM
)|> 
)|> p(theta1,theta2) = p1(delta) = cos^2(theta2-theta1)
)
)This is actually:
)
)   p(theta1,theta2) = p1(delta) = cos^2(theta2-theta1)/2
)                                                      ^^
)
)Otherwise, just by lining up the two detectors, *all* the photons
)must start being detected!  And we know that each observer actually
)only sees *half* the photons on average.

Nope.  Two photons are generated per event.  Cascade photons are but one
example of a process that generates two photons with opposite spins
heading in opposite directions - *pairs* of photons in the 'singlet state' -
which is the idea behind this experiment.

I *was* assuming 100% detector efficiencies - if you want the really rigorous
analysis which includes detector efficiences and suchlike,  you'll have to
track downthe paper by Clauser et al that I referenced (the one that has
escaped my filling system).

)Note also that the above predicts a probability *distribution*.  It
)does not predict specific events.  There is no reason to expect that
)the photons are each - individually - reifications of the function
)at all.

Yeah, we've been going around on that one for a while.  So that's why
I started talking about things that can be measured.  Probability
distributions can be measured.  You just have to repeat the experiment
several times.

)|> Anyway, there *is* a way to distinguish between the two condtions I
)|> was describing.  Local Realistic theories predict that you will have a
)|> constraint on p1(delta).  A simple 'speed-of-light-collapse' theory predicts
)|> that p1(delta) *will* change its value when the spins are measured in a
)|> 'timelike' manner (i.e within a lightcone) vs a 'spacelike' manner.
)
)This is like being able to distinguish between "leprechauns 
)have red  hair" and "leprechauns have green hair" without first 
)demonstrating that leprechauns exist at all!

p1(delta) is measureable, so it's not (IMO) a leperchaun or like a
leperchaun.