Cecil,

I think you are conflating models with nature, and trying to champion
one correct model over another correct model! It's confusing to
onlookers and boring.

There is NO inconsistency between saying "there's only one
electromagnetic field in a transmission line" and "a circulator
seperates the forward wave from the reflected wave" if you've suitably
defined what all those terms mean and you do the correct math.

The electromagnetic field as a function of space and time in the
coaxial transmission line is a three-dimensional time dependent field.
There's a description wherein one single vector valued function
E(r,phi,z, t) describes the electric field and another describes the
magnetic field, and of course, you can get one from the other, so in
some sense, all you need to describe what's going on is E(r,phi,z,t).

Now, in the coaxial TEM mode the radial and azimuthal dependence of the
fields becomes trivial, and you're just left with some function E'(z,t)
to describe the electric field, and one B'(z,t) for the magnetic field
(once again, you can of course, get one from the other) It turns out
that mathematically you can represent this function as a superposition
of other functions, forward and reverse traveling waves. It's just a
DIFFERENT WAY OF WRITING IT DOWN.

A circulator *doesn't know math*. Its operation may have a simple
description in the language of forward and reverse waves, but it does
what it does no matter what model you use to describe it. If you get
different answers using a forward and reflected wave description than
some other description, then one or both of your descriptions are
wrong. The conversion of one mathematical description of the
electromagnetic field into a series of statements in English and the
argument based on those words never gets you anywhere on this topic.
Why not pick up a copy of Jackson's Electrodynamics and write down what
you're trying to say mathematically. If you're right, everyone will
have to be convinced.

73,
Dan