OK, Nick's conjecture would generalize Stapp's theorem.
Nick posits that there is no way to control (a'z3|z1a') locally at a'. It is determined inside the nonlinear interferometer which would be a common cause for both outputs a' & b'. That is, there is no way to construct an effective encoding equivalent of the Heisenberg microscope in Kaiser's Fig. 9.1.
Could be Nick is right.
On Jul 9, 2011, at 5:53 PM, JACK SARFATTI wrote:
Let x be the position of a photon detector at a place where the two components of the b' beam overlap. The nonlocally controllable fringe pattern is simply
(x|Tra'{rhoa'b'}|x)
this is proportional to the count rate for a photon detector placed in the b' output path without any local oscillator or beam splitter. It depends upon
(a'z3|z1a')
that should be a controllable variable at a' - we may need the local oscillator there?
Nick could quote Avshalom Elitzur who thinks that entanglement signals would violate the Second Law of Thermodynamics.