Ah, ok. Now, suppose that atoms did not absorb light in quantized amounts (had continuous energy bands) then would the light hitting these atoms be percieved as having particle-like properties?

Yes, if the light's frequency is high enough to produce Compton scattering. In that effect, light interacts with electrons bumping them off the atom and computations shows that everything can be explained as if light were made of particles. The key-point however is still in the interaction between light and electrons (free or bound in atoms or inside a metal etc.)

What do you mean by 'quadrupole oscillations?'

Let's make the example with electrostatics, which is more simple. You have many point charges, of both polarities, in a specific region of space. If you want to compute the electric field generated by this group of charges away from it, you can add these terms:

1) the total charge of the group (algebric sum of all the charges)

2) the total dipole moment; it's the fact that a sub-total + charge can be slightly displaced from a sub-total - charge (example: only two charges, one + and one -)

3) the total quadrupole moment; it's a more complex unbalancing of charges

...

etc.

A system of charges can have total charge = 0 but total dipole moment ≠ 0; or charge = dipole moment = 0 but quadrupole moment ≠ 0, and so on.

When we compute the electromagnetic radiation generated by a system of moving charges, we see that we can sum the dipole moment variation, then the quadrupole moment variation...and so on.

The vantage of this procedure is that every term we add to the sum is less than the preceding one, so we can stop our sum up to the desired precision of the result.

General Relativity is much more complex than electrodynamics; it turns out that gravitational radiation (gravitational waves) cannot be produced by dipole moment variations (oscillations) only, as it is in electrodynamics, but they need at least quadrupole oscillations. Don't ask me why, however.