The aim of this set of experiments (Ref. 5) is similar to that of the Wheeler Delayed-Choice Experiment in that if one assumes the existence of particles, then one gets into causality (and locality) troubles. Before explaining the experiments, we should reiterate that there is No Evidence for Particles; all the particle-like properties of matter can be shown to be properties of the wave function. So our view is that, as in the Wheeler delayed-choice experiment, one is using an unsupported idea—particles—to derive seemingly surprising results.

Experiment 1. The experiment uses two entangled photons, originally in a spin 0 state, that fly apart from each other. If one photon has polarization in the x-direction, then the other has polarization in the y-direction, no matter what orientation is chosen for the x and y axes. A series of four experiments is done. In the first experiment, photon 1 goes through a double slit and is then detected on a screen while photon 2 is simply detected. The results are not surprising; when the experiment is run many times, one gets the usual Double-Slit interference pattern on the photon 1 screen.

Experiment 2. The light from the two slits in photon 1’s path are put through special crystals that tinker with the polarization so that the polarization of the light that goes through one slit is completely different from the polarization of the light that goes through the other slit. In that case, the light from the two different slits cannot interfere, so one gets a single-slit interference pattern on the screen.

Experiment 3. The light from photon 2 (which doesn’t go through the double slit) is now polarized and detected. The polarization of the second photon (according to quantum mechanics) has an effect on the first photon; the crystals in front of the two slits no longer give different polarizations to the light from the two different slits. Thus the two beams interfere and one again gets a double-slit interference pattern spread out over the screen.

From the particle point of view, this seems remarkable. The experimental arrangement for photon 1 was not changed from experiment 2 and yet the results for photon 1 are different from those of experiment 2. A change in particle 2 affects the results for particle 1.

Experiment 4. Same experimental setup as in experiment 3 except that the polarizer and the detector for photon 2 are moved far away, so photon 1 is detected before photon 2. One still gets a double-slit interference pattern. (But if photon 2 were not detected, or if it were detected as in experiment 2, one would get a single-slit pattern.)

Again from a certain point of view, this is saying that the state of particle 2, determined only after particle 1 is detected, retroactively affects the trajectory of distant particle 1. That is, causality appears to work backwards in time.

Quantum mechanical correlation explanation. In all the experiments, there is a coincidence circuit that accepts only those results where both detector 1 and detector 2 register a photon (within a certain time frame). So in experiments 3 and 4, one is measuring the pattern made by many 1 photons given that photon 2 has a certain polarization. If the polarizer for photon 2 were rotated 90 degrees, one would still get a double-slit interference pattern, but if the two double-slit interference patterns for 1 ( from 0 degrees and 90 degrees) were superimposed, they would just give the single-slit pattern of experiment 2! Thus the polarizer in path 2, plus the coincidence counter, effectively selects for a certain set of 1 photons and these give the double-slit pattern even though the set of all 1 photons gives the single-slit pattern.

If one postulates particles, and if one requires that each particle be in a definite state at each instant, then experiment 3 seems to require action-at-a-distance between the two particles. And experiment 4 seems to require retroactive action-at-a-distance. But if one postulates no-particle quantum mechanics, the experiments simply verify the correlations predicted by quantum mechanics between the two entangled photon-like wave functions.