directional cables?

Grounding the shield on the cable at the "source" produces the least noise (hum). That is the reason for the "direction".

Don, I'd be interested in the science on which you base this pronouncement ;--))

Clio, the cable you describe is either:
(1) the old (pre-shotgun) coax that was used to connect all the early RCA enabled equipment; a 'hot' conductor in the center with a 'ground' conductor/shield around it, much like today's CATV cable, OR
(2) some current manufacturers (like Nordost, I think) are offering air-dielectric single-ended interconnects that use teflon tubing and no shield.

The reason for floating the shield is because if connected to ground at BOTH ends, it can conduct a current -- such current can be the music signal, but it can also include that created by nearby electromagnetic sources (power cords, transformers) or by airborne radio waves. If you disconnect the shield at one end, it can still drain interference to the ground end, but it can no longer conduct the 'minus' half of the music signal.

In "pre-shotgun" days, one of the biggest problems people had (with the old coaxial interconnect) was with their record players -- especially after the introduction of stereo, which meant having TWO parallel coaxial interconnects (with shields connected at both ends) creating a nice BIG loop antenna, enabling you to hear both your record AND a local radio broadcast at the same time!!

BTW, the term "shotgun" simply refers to the fact that instead of having a single center conductor (for the 'hot' signal) and a signal conducting shield surrounding it, the new "shotgun" cable had TWO signal conductors in the center (doubled barreled shotgun ;-) with a floating, non-conducting protective shield surrounding them.
.

One of the cables I was referring to was Mogami 2549. Two conductors tied to the hot pin and the shield tied to the RCA shell. Would one conductor to the hot, one to the ground, with the shield connected to the ground at one end be a better arrangement?

All cables are "directional" meaning that due to variations in the plug quality, soldering and/crimping etc. they will probably sound different when reversed. Usually this difference in inaudible or minute in scale. As I suggested even ones designed to run in one direction MIGHT sound better reversed. Also, they appear to require some time to readjust themselves when reversed, not being an expert on virtual photon exchange I would not attempt an explanation of this but it has been noted by various writers.

Jea48, you (and they) are talking about the crystalline structure of the conductor metal, and yes, it *can* sometimes have an asymmetrical geometry that lets electrons flow better in one direction than another (the basis of solid state devices by the way.) This is also the reason some cable manufacturers boast of "single crystal" wire which presumably has no impeding geometry.

However, I was referring to the issue of the signal itself having a direction, which it can't, because it's an alternating current, constantly reversing it's polarity.

Herman -- energy is only "transferred" when it does some work. Until then, it's only "potential energy" (as in 'voltage' potential.) It really doesn't matter whether the voltage potential is of constant polarity (direct current) or variable polarity (alternating current), there is no "work" being done in the transmission line itself (other than some heat generated if the conductors are too small for current to flow through them unimpeded.) In direct current, the electrons do indeed flow in one direction, and thus "the load" (where the "work" is done) becomes an impediment to the flow of electrons through the entire system -- and which is why DC can only be transmitted a short distance -- and which is why Edison lost to Westinghouse and AC ;--)) Nevertheless, none of these things affect (or are affected by) the way the metal crystals in the conductor material happen to line up (or not.) As with all things, there are exceptions, the most common being ultra-high frequencies. Normal high frequencies (like in audio) just take the easy way out, and travel on the surface of the conductor(s) if there's enough of it.
.