Let me just deal with question 2 for now.
Yes, it's fair to say that the classical E
for n photons are n times those for one, because all those values are exactly zero
only acquire non-zero expectation values for states which do not have a precise number of photons. For more typical states, with indefinite photon numbers and non-zero expectation of the fields, the field magnitude goes as the square-root
of the average photon number. This makes sense because the energy goes as the square of the field and also as the photon number.
On question 1, it's easier for me to think in terms of classical electric and magnetic susceptibilities of a solid, i.e. a large collection of atoms. These are frequency-dependent complex numbers, with the imaginary part of the complex number showing that the response of the material is not quite instantaneous. For a wave pulse with a narrow frequency width, and hence a broad width in time, the oscillations from the material slightly lag those of the incoming wave. For pulses with a broader frequency range, this lag is different for the different frequency components, so the pulse shape is distorted.
(published on 06/09/12)