Maybe I should get some ruby red slippers. :) 

I'm sure there's some great advice in there somewhere, S. Alf. :)  

Lynard Skynard did pretty well with D/C/G.

ZZ Top did pretty well with A/D/E.

From my perspective, part of the problem is that you're not asking the correct question. Mere chords being played one after the other completely ignores the linear aspect of harmony.  Voice leading is EVERYTHING.  There's a world of difference between someone who knows "all the chords" and someone who understands how all the lines in harmony move.  

So in my mind the problem to be solved is to have a unifying system that:

1. Obeys linear voice leading principles.

2. Is flexible to allow complete freedom of root motions, including those that are circle of 5ths-based (tonal), other cyclical root motions or completely free root motions that have no tonal centers.

3. Observes "vertical" tonality, i.e., it doesn't break the overtone series.  In plain English, b9 intervals are bad unless you're using a 7b9 chord and you REALLY want to avoid a b3 and b9 in the same vertical structure.

There IS an answer; but to get you started in a new direction, it's probably easiest to keep your melody diatonic (conforming to a single scale) and then trying different roots underneath.  Here is Schoenberg's system of distant tonal relations, all of which relate to the center "C" (courtesy of Sound on Sound).




Here's an explanation that will help you unpack it:

Thanks, Jonas.  That's cool.  Sheesh...I can't even ask the correct question!    


Thanks for this.  I'll look into it soon.  :)  

@jsaras:

I'm thinking she would have seen that - especially if those three chords are/were the only ones used.

I'm not disagreeing, but to expound on your voice-leading comment:

This is something pianists harp on - for physical reasons, but for composition/theory reasons, too.

For example, that I-IV-V you're stuck on, I hope you've been playing them in inversions because doing so allows for the most adequate voice-leading and the progressions sound smooth and connected.

I-IV-V all in root position. in C would move you "disturbingly" across the piano or keyboard. Not that staying in root position is a bad thing, but you will see and hear what happens when this is done (and your ears don't lie lol)

So, instead of root position only:

C (in root position) F (second inversion) G (second inversion) C (octave above, first inversion)

CEG-CFA-DGB-EGC (Bottom voice: C-C-D-E; Middle Voice: E-F-G-G; Upper Voice: G-A-B-C)

Or

C (Second inversion) F (First inversion) G (First inversion) C (octave above, root position)

GCE-ACF-BDG-CEG (Bottom Voice: G-A-B-C; Middle Voice: C-C-D-E; Upper Voice: E-F-G-G)

Note, not every voice has to move, but that is the reasoning behind using inversions. You want each voice to move as little as possible even if they don't move at all. The greatest distance acceptable is a perfect fourth. Anything greater is a no-no except for the bass voice. (Sevenths are easiest to see this, but when triads are used, doubling has to occur).

So, yeah, I will echo jsaras again by saying that harmony needs to move melodically. After all, it for the most part accompanies the melody, but the top voice is considered the melody note - especially if/when you use inversions:

C-C7-F-C / F-C-G7-C / C7-F-D9-G7 / A7b9-Dm-D9-G7 / C-C7b9-FMaj9-Em7-A7 / Dm7b5-Em7-A7-G9sus-G7b9-C

C-G-A-G / F-E-D-C / G-F-E-D / G-F-E-D / C-G-A-G / F-E-D-C

TTLS/ABC/Blacksheep.

When playing, pay attention to your fingers (they're the "singers" so-to speak) or you could sing each line yourself. A popular saying is, "If you can sing it, it's okay.")

So, if there's a high note (the chord is in root position, but the octave is fairly high), drop one of those notes an octave (preferably the fifth though any note can be moved). Just like a singer would when singing live.

The other thing to consider is how the chords are played. Whether or not, they are close or open, voice-leading still takes precedence; however, it's easier to deal with voice-leading when the chords are in close position because you will end up doing some serious error checking to make sure everything (interval wise) is legal.

Jazz tends to use open chords frequently; however, putting them under the voice-leading rule classically, there will be a ton of no-nos. That isn't to say that open chords don't exist in classical music or that even close chords don't exist in jazz, but again, he voice-leading guidelines start to rear their ugly head. So, it's best to try to really follow these.

@ Janet:

I don't think you asked the wrong question, just one you didn't think to ask.

@Rus: 

Your observations about voice leading are basically true enough, but I'll add my $0.02.

Strictly speaking (or even "freely" speaking) the inversion in tonal harmony is determined by the bass note and not the configuration of intervals in the treble.

I would add the following to your comment about the top note of the harmony as being the melody; every line of the harmony is melodic and it shouldn't necessarily be driven by the top note.  In fact, the motion of the root is the most important factor to consider when building the rest of the vertical intervals.

There are several different kinds of voice leading.  Strict voice leading has no parallel intervals, so the examples that you gave where there were two first inversion chords in succession, or two second inversion chords in succession, do not fit that category.  It IS perfectly acceptable, but it is technically parallel voice leading.  There's also bracket voice leading and voice leading that I'll call modal parallelism (moving intervals up and down through a scale) and even more types of voice leading 
Open harmony has been common practice for centuries, and it does not impact the horizontal/melodic aspects in any way.  However, using open harmony can create unintended vertical dissonances.  For example, a C major 7 chord in drop-2 open position, with a C in the top voice and the "B" dropped down an octave, creates a b9 interval between the B and the C.  The "fix" (if the context demands it) is either to move the "B" down to an "A" or to move the "C" to a "D".  

So far, our discussion has been stuck in the limited prison of tonal harmony.  There is a way to observe all these voice leading principles in contexts that do not have a tonal center and the mathematical key to unlock it is found within the overtone series itself.  The overtone series also provides the key to truly writing one line at a time.   It's so simple that it's kinda embarrassing.

I agree with you on that, but perhaps that's why it made out to be "complicated," because no one likes to be embarrassed.

What you're talking about here is solving the dissonance problem or solving suspensions properly (not chords although the same solving technique applies).

Csus2/4 = D and F move to E
C7sus4 = F moves to E

Of course, M7s. m2/9s are dissonant without question. m7s (b7 = Dominant 7ths), despite being consonant, don't stay put very long because that b7 (within the chord) says move up a fourth while the interval itself becomes the new chord's third (C7-F: Bb-A) - regardless of the inversion applied.

In essence, were talking about resolutions - that is the "fix", you speak of.  Within a progression, the chord after is a resolution of the previous one.

C-Am-Dm-G7-C

C can stand on it's own, but the three chords after imply tension, therefore, needing resolution. However, in isolation, we get three separate tension/resolve actions

Am-Dm (Resolution I)
Dm-G7 (Resolution II)
G7-C (Resolution III)

This doesn't change despite prolongation: The Secondary Dominant Song:

C-A7-Dm-B7-Em-C7-F-D7-G-E7-Am-Bdim-C

While C is fine on its own (yet, it can serve as a resolution if a V7 (G7) comes before it)

A7-Dm (Res. I)
B7-Em (Res. II)
C7-F (Res. III)
D7-G (Res. IV)
E7-Am (Res. V)
Bdim-C (Res. VI)

The Secondary Dominants and the diminished seventh degree (an incomplete V7) are the chords creating the tension and the chords after resolve each. (And most often, inversions are played).

I totally get what you're saying and I was amazed having seen the charts (although I have seen that "Close/Distant Key" one before. If you totally meant something else, my apologies, but I think that is the reason for the voice-leading guidelines - regardless of how the voices within chords are arranged.

Hi Rus,

We're not on the exact "same page" as your progressions are still basically derived from the "gravity" of the circle of fourths/tonal root motion (albeit with added secondary dominants).  I'll give concrete examples of leaving that "orbit" and moving away from tonality a bit later.

Before we do that, let's start with what you call chord resolutions or what I call substitute tones, which have both resolutions as well as reverse resolutions.

Strictly speaking there is no such thing as a Csus2/4 chord as you are using two different substitute tones (D and F) to substitute for a single tone (E).  If you drop the C from the treble structure, the D is a "2 for 1" substitute tone and the F is a "4 for 3" substitute tone. This also maintains the integrity of using three-note structures throughout the passage. The motions of substitute tones can also be reversed. At its essence, this is creating nice contrapuntal motion within what would otherwise be static chords.
  
The "fix" that I originally described is what I call a "compensating tone", i.e., I am getting rid of the dissonant b9 interval (it's not a good sound on a C major 7 chord in open position).  b9 intervals are easier to create than one may realize and things just sound better when that vertical interval is corrected by moving one of the tones involved by step, which also creates the additional benefit of a little contrapuntal motion.

I've taken your original progression in the key of C major and I put it through several permutations which ultimately move out of anything having to do with any tonal center, yet it all sounds perfectly logical. I'll leave it to you to do your own analysis.  Now just imagine what progressions are possible if the "original" triadic progression isn't confined to a tonal center, but still voice-lead correctly. 


MP3 audio: http://bit.ly/QoszU7

Btw, the sus2/4 meant sus2 and/or sus4. I didn't make that clear. Sorry.

I totally see what you mean. Coincidentally, though if it's just seen, but not heard either by the physical or mental ear ...

I totally get not being bound to a key or scale so stringently; however, one's ear trumps logic if you think about it. Yet, logic guides your ear or helps one grasp where the ear gravitates and why it does so. All the permutations you did work very well logically though aurally they may not while the reverse is true, too. I still wouldn't advise throwing something in there without thinking it through first - and that's with your ear (thinking with the noggin comes later)

I did misinterpret what you meant and I apologize for that

Hi Rus,

No apology needed as no offense was taken.  

My examples above are "over-baked".  It was more a demonstration of technique than an actual composition.  That said, the bottom progression with the moving lines removed is extremely usable.

Context is king.