Re: Light slows down in glass?

[David Bowman <dbowman@tiger.gtc.georgetown.ky.us>]

Regarding Bill Beaty's comments:
> My own opinion is that there are various models for what light does, and
> if we insist that light is emitted and reabsorbed, then we really are
> saying that the particle model for light is correct and the wave model is
> false.

EM emission and absorption processes work well with a classical wave model
as well.  For instance, one does not need the quantum corpuscular nature of
the Maxwell field to understand the operation of transmitting and receiving
radio antennae.

>        After all, it is *photons* which are absorbed and re-emitted.

So what?  The single-photon emission and absorption processes of optical
phenomena are due to the weak field regime of the relevant emitted/absorbed
modes.  In such a weak field case the granularity of the quantum
discreteness of the exitations for that EM mode stand out and behavior of
the EM field is best and properly described in terms of discrete photons.

Consider any simple harmonic oscillator.  For interaction phenomena
(involving that oscillator) that occur in the realm of huge exitation
numbers, n, the discrete nature of the quantum oscillator can be usually be
ignored and a classical description suffices.  We don't need quantum
mechanics to understand the behavior of macroscopic spring-mass systems and
their couplings to external agencies which feed and bleed energy into and out
of the system.  OTOH, when the mean energy exchange with the oscillator is of
the order of the quantum energy level spacing and the oscillator spends most
of its time in in its ground state or a few weakly excited states then the
discrete quantum nature of those energy exchanges (of the size of the
energy level spacing) becomes crucial for understanding the interaction
process.

Now the EM field degrees of freedom can be considered a collection of
independent simple harmonic oscillators for each of the modes of the field.
The number of photons in a given mode is just the quantum exitation number
n for that mode. For strong-field interaction phenomena such as couplings
to a radio antenna where the interaction process involves modes with huge
n-values, then a classical wave picture for the quantized EM field is an
excellent approximation.  Such a wave-like description *does* handle
emission and absorption processes.  For weak-field interaction phenomena
such as couplings of an optical-frequency EM mode in the n=0 or n=1 state
to an atom (where the quantum number n of the EM mode increases by 1 unit
for photon emission and decreases by 1 unit for photon absorption) then the
quantum nature of the process is essential for proper understanding.  In
this case the frequency of the modes strongly coupled to the atom matches
the atom's energy level spacing according to [delta]E = h*[nu].  Whether
the couplings which emit and/or absorb EM radiation to and from the Maxwell
field are described by classical waves or quantum mechanically by photons,
is a qualitative consequence of the relative (mean) field intensity per
unit frequency ratio for the relevant modes involved.  There is nothing
necessarily and essentially particle-like about the emission/absorption
process per se in general.

> Light wavelength is far larger than the spacing between atoms, and I
> cannot see how "absorbed/emitted" applies to the EM fields or to EM waves. 

My (very old) hand-held transistorized AM radio absorbs EM radiation whose
wavelengths range from about 600 m to about 200 m and converts the signals
modulated on that absorbed EM wave energy to audible sounds - even though the
radio receiver (and its antenna) is about 10 cm in size.

> Instead, the transparent material acts as a non-vacuum medium, having an
> altered propagation velocity.  If we insist on "absorbed/emitted", then we
> fly in the face of wave/particle duality.

No we don't.

> Perhaps the real issue is a collision between classical and QM models? 

This is not a problem.

> For example, if I send an EM pulse down a pair of plastic-coated wires,
> the pulse doesn't travel at c, it moves slower depending on the dielectric
> constant of the plastic.  This effect is fairly independent of frequency.
> It is normally explained by ignoring photons.  This is obviously not
> QM physics.  Yet is certainly is not "wrong", and if it is wrong, then any
> reference to EM fields is also "wrong."

This is a case where the classical description works fine for the EM field
and its interactions with polarizable matter and with free charges.

I wonder if the purported incompatibility of emission/absorption
phenomena with a classical wavelike description for the EM field that you
(Bill) suppose is more likely an incompatibility between an emission/
absorption process that is *coherent* and one that is *incoherent* rather
than one which is particle-like and one which is wavelike.  Maybe you think
of incoherent processes as particle-like and coherent ones as wavelike?  It
*is* essential that a coherent process be described by some kind of a
wavelike description.  For some phenomena such as the slowed propagation of
EM radiation in a material medium involves *coherent* interactions between
the EM dynamical degrees of freedom and those of the bound and free charges
found in that medium.  In this case the field tickles the charges and the
charges radiate in a manner that coherently modifies the field.  Other
phenomena (such as black body radiation or florescence) involve processes
where the absorption and emission processes are incoherently related to each
other.

> A separate issue: the original message mentioned Feynman.  I'm curious. 
> Do most physicists believe Feynman's interpretation of QM, that the wave
> nature of EM is a consequence of the actions of vast numbers of virtual
> particles which take all possible paths and only sum to "real" particle
> events sometimes?

The mathematics of the path integral formulation of ordinary quantum
mechanics shows that is mathematically how DeBroglie waves behave in
determining the amplitude for a given process.

>                   I was under the impression that this was a minority
> viewpoint in physics, and most physicists still thought in terms of
> "instantaneous collapse of wavefunction."

This (wave function collapse) is for the nonunitary process of how a
measurement actually gets made, not for the unitary Schroedinger evolution
of the state of the system between measurements which determines the
probability amplitudes for all conceivable subsequent measurements -- whether
those amplitudes are calculated by Feynmann-type path integral summations
over virtual processes, or whether they are calculated by just solving
the Schroedinger equation and looking at the projected amplitudes of the
solution.  Both (Schroedinger's and Feynmann's) formalisms are equivalent
and give identical answers.  They are not necessarily equally easy to apply
in various special situations, however.

>                                            Feynman seems to say that
> photons are real and waves are not, while "Copenhagen" viewpoint says that
> probability waves are real, and they act to "cause" the appearance of
> photon particles.  And if light is "really" virtual particles, how can we
> explain polarizatino and other field phenomena?

Huh?  AFAIK everybody interprets real photons as free propagating quantum
single-particle exitations of the Maxwell (EM) field, and virtual photons
in terms of summed-over contributions in terms of photon propagators in a
perturbative expansion for the amplitude for some process involving
interactions between the Maxwell field and its sources.

David Bowman
dbowman@gtc.georgetown.ky.us