I want to tread carefully in this post because not everyone agrees with what I am about to discuss. I have good friends who will disagree with it in fact. That is all good. It is not a personal thing; it is just about music and I have no one in mind as I am writing.
The topic is chord substitutions, a topic that generates a lot of questions here. Pianists are fascinated by the subject; they instinctively know that the harmony they are often asked to play is pretty boring.
I started learning about chord substitutions when my older brother started coming home from college and telling me what he was learning. Eventually, I went to college and learned the same basic method. And ever since then, I have heard many teach the same method.
The method that is widely taught is called common tone substitutions and it goes like this: find chords that share common tones with the original chord and those become your chord substitutions. In other words, you can substitute Amin and Emin for CMaj because those two chords share two common tones with the C chord.
Using this strategy is pretty simple. Basically, you go up and down a diatonic third from the too-common I, IV, and V chords. That leads you to these conclusions:
Substitute vi or iii for I
Substitute ii or vi for IV
Substitute iii or vii(dim) for V
There are at least two big weaknesses here. First, if this is really the only way to do chord substitutions, we are working with a pretty small toolkit. That method gives you a whopping six possible substitutions. I suppose that is better than nothing but is hardly worth getting excited about.
The second weakness in this method is it ignores some very important things about how music works. At best, it leads the musician to try to choose a chord that works with the melody note. But it does not help the musician at all in considering the ultra-important relationship between the chords before and after the chord substitution.
For an easy example of the problem, find a V chord in any hymn and use the common tone formula to try to change it to a iii chord. Almost all of the time, the iii chord will sound worse than the V chord. It is not hard to understand why: V chords almost always resolve to I and a iii chord does not move smoothly to I. That is not how a iii chord works; it has a different function.
Or try this example. Let’s take the first 4 bars of a song in the key of C that has this harmony (one chord per bar):
C C G C
Using common tone substitutions, you could expect to substitute Amin or Emin for either or both of those first two bars. Let’s say that you decide to get clever and do both: Amin for one bar and Emin for one bar. That gives you two possibilities:
Emin Amin G C
Amin Emin G C
Now, here is a question for you. Which one is better and why? There is a clear answer though the common tone substitution method will not give it to you. The first is superior because Emin naturally moves to Amin which in turn naturally leads to G. The second option is unnatural.
So, the common tone method is flawed for one simple reason: it ignores function (how chords interact with each other). That is not to say that chords derived with that method do not often work, but the reason they do is not because they share common tones. It is because of the function of the chord and what melody notes work are compatible with that chord.
When I teach chord substitutions, I usually start with common tones and explain the weaknesses. I then move on to another perspective I call functional equivalence. Functional equivalence means that you substitute chords that share the same function. For example, a ii7 and V7 chord have the same function so a ii7 can substitute for a V7.
While functional equivalence is a big improvement over common tones, it is not ultimately where I want students to end up. The long term goal is to stop thinking in terms of chord substitutions at all but rather to think in terms of chord progression substitutions. We end up completely moving away from the idea of substituting one chord for another. The new goal is to replace entire chunks of existing harmony with chord progressions that are more sophisticated and more functionally sound. That is the mindset that generates the harmony from my students that I have been sharing recently. (If you haven’t seen them, you can watch demonstrations here and here.)
As beautiful and complex as the harmony might sound, the good news is that there is a startlingly small amount of information that you really have to know. I am always amazed at how quickly I can go through all the rules. It takes some time to learn to apply it in real time of course, but any musician can absorb the basic concepts pretty quickly.