Substitute chords – a simple jazz technique

When I was learning to play jazz piano, I was doing it pretty much by myself. Having learned rudimentary harmony as a trumpet player, I began to explore the piano keyboard, and in so doing discovered for myself many of the tricks and techniques that I could hear on my records of Art Tatum, Oscar Peterson, Duke Ellington and so on.

One thing I discovered completely revolutionised the way I thought about voicing chords in progressions. I found this out back in 1976 or so, but the way of thinking it led to still helps me now, not just in playing jazz but in many other branches of musical style. I called it “chord substitution“; I’m not sure if this was a term I’d heard others use, or one I dreamt up myself.

I’ve never seen this principle explained in quite the way I that I see it, so I thought I’d muster my meagre talents and put together a kind of tutorial.

Introduction

First, let us consider one of the basic building blocks of Western harmony – the II-V-I progression. It’s a snippet of the “cycle of fourths” (or, as the classical players call it, the “circle of fifths“), that never-ending sequence of resolving 7ths:

(If this isn’t familiar to you, I’d advise you to look further into this before continuing here.)

The vast majority of songs and other pieces of Western music have, at the end of their phrases, three of those in sequence, with the final chord’s dominant 7th omitted to finalise the progression. In other words, something like (in the key of C):

In jazz, and other music with blues elements, that actually sounds a bit uncool. The F# not of the D7 chord doesn’t fit the key, and makes it sound brash and, well, too damned happy by half. So, for the more mellow and sophisticated sounds of post-1920s jazz, that F# is lowered to F and the progressions becomes:

And those three chords – in one key or another – crop up again and again. Pretty much any jazz standard you hear or play will have those that progression at last a couple of times, usually (but not only) at the ends of phrases.

Consider the first eight of the standard All The Things You Are, by Jerome Kern:

The colours show three different sequences of fourths, separated by key changes to keep things sane. OK, so I’m cheating with this tune – there is no other I know of that’s both well-known and such a clear example of the Cycle of Fourths.

Here, for illustration only, is a rough chart showing all the II-V-I progressions there are, connected together by colour in the Cycle of Fourths. The colours in this diagram and the previous are not meant to co0incide.

The colours indicate II-V-I chord sequences, going from the outside in. Click to enlarge

The defining notes

Now consider these three intervals:

You could, of course, voice those three chords in other ways – you don’t have to use those particular notes.  But what that shows are the defining notes of the progression; the important ones that make the progression work the way it does. No matter what other notes you add to those basic intervals, it’s still these six notes that tell our ears where we’re going. Omit any of them, and the progression won’t work properly; as it is, it actually sounds complete with just the six. Boring, but complete.

Sure, we can add more notes, and then it will sound more interesting, but let’s not do that just yet. Instead, let’s look at those notes in terms if which part of the chord each one is.

Let’s look at the red part. The G7 is voiced by its major 3rd (B) and its dominant 7th (F). It’s crucial. Just playing those two notes makes the listener want to hear the resolution that comes here in the third chord.

We’re now going to play a trick on those two notes.

Substitution

It so happens that B and F are also the major 3rd and dominant 7th of another chord, the chord of Db7:

We’ve not changed a note (yet!). All we’ve done is to swap the roles of those two notes. Instead of the B being the major third, it’s become the dominant 7th, and the F has switched from dominant 7th to major 3rd. The only chord that fits that description is Db7. Let’s fit that back into musical notation:

Again, the notes are the same. This sequence is identical to the example above, except we just think of the middle chord in a different way. If we now expand on it, and sound the other notes of the chords, we may feel we’ve not really improved matters much. Try it and see – here are the “before” and “after” chords, with the other chordal notes:

Before (II-V-I):

After (II-IIb-I):

Now, I don’t know about you, but I don’t find either of those particularly exciting. Don’t despair, however, because the hard work has been done, and now it’s time to reap the benefits.

Combining normal and substitute chords

The simple fact is that simple chords sound simple, and our Dm7-G7-C and Dm7-Db7-C are, let’s face it, simple chords. Let’s get inventive and combine them:

It’s starting to sounds a little more sophisticated now, and this is only the beginning. Of course, we can add more notes – 6ths, #9ths, 13ths, 11ths, and who-knows-what, but now we have two sets of each additional note to play with, one for each of our substitute chords. For example, you can have the 6th of G7 (E), or the 6th of Db7 (Bb). In this way, we’re simply adding to our palette of possibilities, remembering that our 7th chord is ambiguous. Yes, it may say G7 on the sheet music, but you and I know that a Db7 is also possible, and that the notes of both can be mingled in all kinds of ways.

And finally …

For a grand finale, let’s make some good sounds. We’re now no longer in C, we’re now in F (I never liked the key of C much anyway), and we’re faced with sheet music that looks like this:

Surely the mind of a Country & Western banjo player at work? Let’s get creative, never forgetting the principle of chord substitution:

What our banjoist had as C7 now looks like nothing so much as Gb7(b10). Or is it C7(b9/13)? Or is it, quite simply, a combination of Gb7 and C7? I know what I think.