🚀 go-pugleaf

cescript@mtu.edu (Charles Scripter) writes:


>Here's a QED puzzler for you.  Suppose we examine the a free photon
>using Feynman diagrams:

[photon propagator with one-fermion-loop correction]

>Can these contributions be viewed as actual processes, or do we
>construe them be merely mathematical models?  

A fine thing to do a lot of thinking about.  I think of them as
terms in a perturbation expansion of what is really a continuous
process: the EM field is excited and it causes some collateral
excitation of the charged matter fields as well.  The Feynman
diagrams are a calculational device to get the outcome of the
continuous process to nth order.  

>This second order
>correction could be viewed as "photon decay", followed by an immediate
>recombination.  

I was, of course, using the usual definition of "particle 
decay" as a situation in which the decay products last "forever," but
as John Baez pointed out ominously a while ago, if we detect them, they
haven't lasted forever.  

Which implies another puzzler:  Are visible photons virtual?  What does 
this imply for causality-- since, after all, virtual photons don't have
to be *exactly* on their light cones?  I know that this doesn't
actually destroy causality, but I've never seen a very lucid explanation
of why not.  Any takers? 

>Although energy-momentum conservation is violated, we
>are allowed to violate this on quantum time scales (how short is "short"?).
>Energy and momentum are clearly conserved on macroscopic time scales.

No!  No!  Energy-momentum is *always* conserved!  It's the laws of
motion that are violated!  :)  (The two views are really equivalent,
and correspond, respectively, to old-fashioned and covariant perturbation
theory.)
 
-- 
Matt McIrvin, grad student, Dept. of Physics, Harvard University
mcirvin@husc.harvard.edu      mumble mumble mumble mumble mumble