We can ask ourselves what the subject ‘looks like’, at the single-particle level, of polarized light. We know that at the level of wave-mechanics, both plane-polarized and circularly-polarized light are easy to understand: Either way, the dipole-moments are at right angles to the direction of propagation, all the time, even if randomly so. But there also needs to be a particle / photon -based explanation for all the properties of light, in order to satisfy the demands of Quantum Mechanics.
And so a key question could be phrased as, ‘If we pass randomly-polarized light through a simple linear polarizer, which consists of a gel-block, and which absorbs EM vibrations along one disfavored axis, maybe because it has been made ohmic along that axis, why is the maximum intensity of plane-polarized light that comes out, in fact so close to 50% of the intensity, of the randomly-polarized beam that went in?’ Using wave-mechanics, the answer is easy to see, but using particle-physics, the answer is not so obvious.
And one reason fw the answer may not be obvious, is because we might be visualizing each photon, as being plane-polarized at an angle unique to itself. In that case, if the polarizer only transmits light, which is polarized to an extremely pure degree, the number of photons whose plane of polarization lines up with the favored angle perfectly, should be few-to-none. Each photon could then have an angle of polarization, which is not exactly lined up with the axis which the polarizer favors, and would thus be filtered out. And yet, the strength of the electric dipole-moment which comes out of the polarizer, along the disfavored axis, could be close to zero, while the total amount of light that comes out, could be close to 50% of how much light came in.
If each incident photon had been plane-polarized in one random direction, then surely fewer than 50% of them, would have been polarized, in one exact direction.
(Updated 04/10/2018 … )
(As of 04/08/2018 : )
And what I suspect the answer to this question entails, is that the polarization of photons, is always a superposition of states ! We may have heard, that a superposition of states can only take place, until a particle has been ‘witnessed’. But at what point in time has a photon been witnessed? This question is especially crucial to understand, because it contains the word ‘Time’. Being Luxons, photons travel at the speed of light, and according to Relativity, time stands still, according to the frame of reference, of the photon. The photon can only be witnessed, when it reaches the end of its trajectory and becomes absorbed !
This creates a bit of a clash, with how Humans may think. Humans will often think, that because they have become aware of something, it has been witnessed, according to Physics. But this is not true. Humans may or may not be aware of the states of photons, without those photons actually having been witnessed. And further, the photons could also be witnessed, without Humans becoming aware of their states.
To state this more-directly: It’s perfectly possible for a Human to set up an apparatus, in such a way that he or she knows, what the states of the photons are at one stage of the beam. But as long as those photons are not absorbed, they have not been witnessed, according to the Physics definition of the term. Their state is still superposed.
This could be similar to how a Quantum Computer can only work with particles – usually photons – that are in a superposed state. A Human, aided by a non-Quantum-Computer, could reproduce the computation which the Quantum Computer is being made to perform, and could therefore reproduce what the states of the QuBits are, midway through a computation. Nevertheless, the Quantum Computer can perform its calculations, because inside its main component, the photons have still not been witnessed.
They are witnessed when the Quantum Computer is made to output its final state, at which point the Quantum-Computation must end.
I suppose that I should mention, that other types of QM interactions exist, that are defined for particles, and that may witness the photon. For example, a LASER is a device based on Light Amplification through the Stimulated Emission of Radiation. What this means is that in the resonant cavity of a LASER, an existing photon interacts with a stimulated atom or molecule, so that a second photon is emitted by this atom or molecule, resulting in two photons where there was one. I would not rule out, that this event also witnesses the first photon.
The case of a LASER is particularly ambivalent, because it tends to produce coherent light. This means that both resulting photons will be exactly in-phase, with the incident photon.
regardless of whether we conceptualize circularly-polarized light to originate from two phase-shifted plane-polarized states, or whether we conceptualize plane-polarized light to originate from two circularly-polarized states that are in a specific phase-position, the LASER will cause either set of states to be duplicated.
(Edit 04/09/2018 : )
Actually, the idea that the LASER must witness the first photon, before emitting the second, actually suggests that its behavior must logically be a bit more-complex than a first glance would have it. Because, the witnessed photon may not be plane-polarized exactly along the axis that the beam is collectively, for which reason a Brewster Window, and the amplification of light which becomes successively more-polarized, may be necessary, to emit plane-polarized light.
Furthermore, conservation of (angular) momentum, as well as competition, may prevent gas-lasers from emitting circularly-polarized light. But, the ability of gas-lasers to emit plane-polarized light, hints at the possibility, that polarized light has plane-polarized photons as its basis.
(Edit 04/10/2018 : )
Casual inspection of the subject might suggest, that because, When circularly-polarized light gets reflected, its direction of polarization is also reversed, this should be the reason for which lasers fail to produce circularly-polarized beams. But in reality, when the light makes one round-trip, it actually gets reflected twice.
For that reason, hypothetically, a situation should be possible, in which light traveling in one direction is consistently right-handed, while light traveling in the other direction, is consistently left-handed. Yet, practical examples of lasers don’t emit circularly-polarized light.
The same concept of competition that was at work form plane-polarized light, which is, that there is a finite supply of excited atoms or molecules, and that if the amount of light plane-polarized ‘horizontally’ increases, fewer excited particles are available to amplify ‘vertically’-polarized light, should in principle take place between left-handed and right-handed photons as well, and cause either one species or the other to be emitted. But in practice, this does not happen.
Well, even if we conceptualize circularly-polarized light to consist fundamentally of horizontally-polarized light, which is also phase-shifted 90⁰ with vertically-polarized light, competition will either favor the horizontally- or the vertically-polarized component, and circularly-polarized light won’t emerge.