In article , "Richard Crowley" wrote:
If that's so, that's what I'm not getting. If ground means there's
a rod stuck in the earth, how can that be generating the other
"half" of the AC signal?

Maybe you are confusing yourself by thinking that conductors
*generate* signals? They DO NOT generate signals, they only
*conduct* them (hence, the name ;-) . Whatever generated the
signal uses the conductors to get the voltage from the source
to the the destination.

I know that conductors don't generate signals, they're just the
transport...but...

A circuit is a loop of wire that goes from the source, through
all the destinations, and back to the source. Your (b) and (c)
are the circuit [in a simplistic view]. It makes no difference
(for the purposes of your discussion) whether ONE of the points
in this circuit is grounded. Clearly, of TWO or more points
on this circuit are grounded, you have a dead short and no
signal.

What does it really mean to be grounded? I guess I'm confused there. In all
the books I've read, grounded means it's connected to a metal rod stuck inside
the earth's crust. That makes me think that the circuit is:

audio_generating_source - conductor wire A - destination device - conductor
wire B - earth.

When I think it needs to be:

audio_generating_source - conductor wire A - destination device - conductor
wire B - audio_generating_source

Obviously you can't stick another metal rod in the earth & connect it back to
the audio_generating_source.

How would the other device know how to separate the real
signal (b or c) from ground (a)?

The signal is the difference between your (b) and (c) without
reference to any other point in the known universe. If you now
"ground" either (b) or (c), it makes no change to the differential
voltage between (b) and (c). The receiving device looks at
the difference between (b) and (c) and does not care if one
of them is connected to ground or not.

OK...are you saying that when (a) is +12V, if (b) was not connected to ground
it should be -12V. But if it is connected to ground, it may be -14V? That
means that ground was adding 2 volts? Or that the whole audio signal is offset
2 Volts above the X axis...like a DC correction?

* Conductors DO NOT generate signals.
Understand that.

* Audio signals don't have "positive halves" and
"negative halves".
Understand that, I guess I'm just not using the right words. There is no (+)
or (-) in AC since they change roles frequently.

* It makes no difference to the circuit whether one
point is grounded or whether the entire circuit is
"floating" ("differential", "balanced").

OK, they why ground? Maybe I just don't get the whole point of grounding if it
makes no difference on the audio signal once they are differenced against
each other.