Polarized Light Weirdness
The same weirdness problem arises when we pass light through polarized devices as in the figure at the left. The devices are calcite crystals in which the light is split into two parts, a horizontal (H) and a vertical (V) channel. If we try to send individual photons through, they go through only one channel or the other, never through both, and those that come out of the H channel are always horizontally polarized, those that come out of the V channel are always vertically polarized as we might expect.
It is possible to orient photons to other angles at the input. One such arrangement is to adjust them polarized so that they are tilted 45 degrees right or left as illustrated in the same figure. If we orient the input to 45 degrees, tilted right, we get half of the photons coming out the H channel and half out of the V channel, one at a time, but these are always horizontal and vertical polarized, no longer polarized at 45 degrees right.
Now comes the weird part. See the figure at the left. If we put a second calcite crystal in line with the first one, but reversed so that the H channel output of the first goes into the H channel of the second and the V channel output of the first goes into the V channel of the second, we expect the output to consist of one photon at a time (and it is), but since the first crystal only outputs horizontal or vertical polarized photons we expect only horizontal or vertical polarized photons out of the second crystal.
Quantum Weirdness at work.
However, if we test the polarization of the output, we find that the photons coming out are oriented to 45 degrees right, exactly like the input. Individual photons go in at 45 degrees right at the input, are still individual photons but horizontal or vertical oriented in the middle, but come out oriented 45 degrees right again at the output! Somehow the two channels combine as if the individual photons go through both channels at the same time, despite rigorous testing that detects only one at a time. Quantum Weirdness at work.
The polarization problem, like the double slit problem, is often called a quantum measurement problem. An often-quoted theory is that the photon does go both ways, but any attempt to detect/measure one of the paths disturbs the photon such that the measurement results in a change in the path of the photon.
My theory reafferms the idea that it does go both ways, but in a manner you would not expect. We will get to that later. Next I want to mention Quantum Weirdness in Glass