Okay, if you'll indulge me, I'd like to try to spell out what little I think I know about black body radiation and absorption. If you might correct me wherever I stray from truth I'd be much obliged :)
Since 0 degrees Kelvin is unattainable, the coldest black body we can sensibly talk about would be some fraction of a degree above absolute zero. If we're talking about a chunk of matter; i.e., a cluster of particles, then it would be a cluster of neutrons. The fact that these neutrons are not at absolute zero doesn't just imply logically that they're vibrating ever so slightly, or "oscillating" I should say, it literally means it, semantically. But these oscillations don't make waves because neutrons don't cast EM fields into space. If the neutrons' vibrations were to speed up, they reach a speed at which some of them (but not all of them) would collapse into protons and electrons. At this rate of oscillation (still very close to absolute zero) the oscillations produce waves, because at this rate of oscillation EM fields are produced, because, again, at this rate of oscillation some neutrons become protons and elections and protons and electrons cast EM fields into space. And so it seems as if there can never be any such thing as a perfectly still EM field, because the necessary conditions for the field's creation and perpetuated existence are the very oscillations which cause the waves in it. In order for a dropped pebble to makes ripples on the surface of a pond, the pond must already be there. But we're talking about a pond that owes its existence to the oscillations of the pebble; a pond that can never be still. In contrast, a gravitational field is a pond that is about as still as the EM pond is turbulent. We have yet to even detect a wave in the gravity pond, but it's not forbidden by the laws of physics in the same way that a perfectly still EM pond is forbidden - that's where the dichotomous symmetry breaks down - it's at least theoretically possible to generate a gravitational wave.
Back to oscillating particles. Only two things can cause a particle to oscillate: collisions with other particles and the absorption of a photon; i.e., the absorption of an EM wave. So particles emit EM waves and they absorb them. Oscillations cause wave-emission, but wave-emission causes oscillations to decelerate, while wave-absorption causes oscillations to accelerate. This is where it breaks down for me. If we were talking about atoms and molecules and not just protons, neutrons and electrons individually, then I'd say that, because electrons can only take certain orbits around atomic nuclei, token-atoms and any atom-type, as well as token-molecules of any molecule-type, can only absorb certain wavelengths of EM waves. That is to say, if there's no electron in a given orbit around a nucleus, then that atom will not absorb EM waves whose frequency corresponds to that empty orbit. Electrons do the absorbing, not the orbits, but it just so happens that electrons are always in certain discrete orbits. So we have orbits, and for each orbit, there's a frequency of EM-waves that resonates with it, and causes the electron at that orbit to take the next higher orbit. And the taking of this higher orbit causes the atom or molecule as a whole to oscillate faster?
So energy doesn't come in discrete units. Atoms and molecules can only emit and/or absorb energy in discrete units, but this is a feature of atoms and molecules, not an intrinsic feature of the energy that they emit and absorb?
I've given myself a headache. Maybe I've given you one, too. If so, I'm sorry