First, I will present the possible new interference technique, and

then explain why it might be very important in fundamental issues of

quantum mechanics and the question of superluminal correlation and

communication.

It is generally believed that no detector can distinguish between a

stream of random H-V linearly polarized photons (i.e., oriented zero

or 90 degrees) and a stream of random "diagonal" photons (45 or -45

dgrees.) A linear detector at any angle will simply show random 50%

hits, as will a C-pol detector (and presumably any more complicated

device, optimized for a given degree of circularity such as elliptical

light, using say some fraction wave plate combined with a linear

filter or calcite splitter.) We don’t have a "photon characterizer",

just yes/no assessments of some orthogonal trait such as perpendicular

linearity or chirality.

However, suppose that we first sent polarized photons through a

quarter-wave plate oriented H-V (optical axes). Then H-V photons will

come out the same orientation as before, but diagonal ones will be

converted into either a RH or LH circular photon (wave function). Then

we send the ray into a beam splitter, of perhaps classical Mach-Zender

type. Before recombination, one leg passes through a set of two

half-wave plates, one being H-V and the other with diagonal optical

axes. Their effect on linear photons will be to rotate them 90

degrees, but circular waves will be double flipped back to their

original state. The implication is that upon adjustment of path

lengths and recombination we can have interference with the originally

diagonal photons (same circular chirality having been maintained in

both legs) but not with the H-V ones since their split waves become

perpendicularly polarizated. Of course, a single hit on a detector may

not prove the point since a 100% "peak" is replaced by a 50% chance,

but we would soon know that the source was mixed diagonals rather than

mixed H-V with enough hits to show concentration in one peak or

detector versus no preference.

If this is possible it is important because the supposed limitation on

polarized photon detection is vital to most assessments of the EPR

paradox with entangled photons and especially to the true superluminal

communication issue (for example, see Nick Herbert’s *Quantum

Reality*, *Faster than Light: Superluminal Loopholes in Physics*,

etc.) The correlation of entangled pairs of emitted photons rules out

simple emission of definite, pre-determined pairs of linear photons

each with a specified angle of polarization. Somehow the measurement

of the polarization angle of one photon constrains the other one

("instantly") to have the same (or perpendicular) character at the

other distant detector. If the first detector changes from say H-V to

diagonal measurements, then the second detector must now be sure to

detect diagonal photons of the correct angle. The distinction is

masked by the random noise, but in Herbert’s and many other’s

interpretations of QM collapse, we could know that the first detector

had been tilted 45 degrees if we could detect the change to diagonal

photons at the other detector – which is presumed not possible. Yet if

the above interference set-up actually works, could we exploit it in

quantum entanglement experiments? What other implications might it

have? Is the interference argument as straightforward as it sounds?

Neil Bates