I've been studying the designs of the GH antennae and wondering about something. Do the reflectors really actually "bounce back" the electromagnetic waves? Or do they just set up resonating currents within themselves that re-emit EM waves at the same frequencies thereby boosting the signal received by the driven elements? I recall reading something along those lines, but I don't remember where, so I'm wondering if I'm just remembering incorrectly.

If they do reflect, by how much would they attenuate any signal coming from a broadcast tower that happened to be behind the reflectors? In other words how directional do they make the antenna?

Do the reflectors really actually "bounce back" the electromagnetic waves? Or do they just set up resonating currents within themselves that re-emit EM waves at the same frequencies thereby boosting the signal received by the driven elements?
A bit of each. Depending on antenna design, reflector size and reflector distance one effect may be greater than the other. I suspect that standard yagis provide more of a reflection aspect while NARODs provide more of a resonance and coupling effect. Designs like the CM 4228, with short elements and a long reflector, may reflect more at UHF frequencies but resonate and couple more at VHF frequencies. Directors (elements in front of the driven element) also provide a focusing effect but I'm not sure how that works. Parabolic designs provide pure reflection gain.

If they do reflect, by how much would they attenuate any signal coming from a broadcast tower that happened to be behind the reflectors?
This is stated by the front to back (f/b) ratio. Figures from 10db to 20db are fairly common. Directors also increase the f/b ratio. Each 3db is approximately 1/2 the signal so 10db means the rear signal is about 1/10 the strength of the front and 20db about 1/100.

Also see Questions about antennas - how and why they work (http://www.digitalhome.ca/forum/showthread.php?t=140261)

I suspect that standard yagis provide more of a reflection aspect while NARODs provide more of a resonance and coupling effect.
NARODS dont reflect at all, thats part of the name Not A Reflector Or Director. NAROD reflectors reflect the NAROD though.

Do the reflectors really actually "bounce back" the electromagnetic waves? Or do they just set up resonating currents within themselves that re-emit EM waves at the same frequencies thereby boosting the signal received by the driven elements?
Youre basically saying the same thing in both cases.

Kind of like saying does a ball bounce back off of a wall or does it store up kinetic energy and redirect that energy backwards.

With rod reflectors, there is a specific optimum resonate size of the reflector for maximum gain. For mesh or solid surface reflectors, its a bit more complicated. For example, with a CM4221a mesh reflector, AND DISREGARDING VHF_HI GAIN, increasing the size to silly proportions doesnt reduce gain, but doesnt really increase it either over a certain point. So the excess is just a waste of material and space.

Bear with me guys as I figure this stuff out. The reason I asked was because "bouncing a signal back" would, it seems to me, with my limited knowledge of EM radiation and antenna theory, more definitely attenuate any signal originating from the opposite side of the reflectors. After all the signal is not making it past the reflectors. Whereas an absorb-and-re-emit scenario would have less of an attenuation effect on the 'dark side' of the reflectors, since the reflectors aren't really 'reflectors' they're actually 'emitters' - in all directions.

Signals coming from the rear must be treated as a SCATTERING problem....some of it is reflected, but there is also
a significant portion that goes straight thru....and some is captured and reemitted by the reflector elements....

PS: And if you start thinking of light as a collection of individual photons that "bounce" around the structure,
you will completely MISS the inherent WAVE nature of light...you know, the part that is RESONANT....