I am trying to understand the harmony of this intro of a song but I can't seem to make sense of it. The song is in C and the intro is four bars:

| C | Ebmaj7 Dm7 | C | Ab Gm7

And then the verse starts on a C chord.

I think that it sounds nice and that the chord progression makes sense.

But why? The Ebmaj7, Ab and Gm7 are all non diatonic but both sets of two bars seem to resolve to a C chord.

Can I interpret the Ab-Gm7-C as actually resolving to a Cm with the Gm as the dominant, but then with a Picardy Third?

That still leaves me clueless as to what is happening with the Ebmaj7 and Dm. It feels similar but the Dm isn't in the key of Cm.

Any help would be appreciated!

  • 2
    The intro to what? – Tim Dec 27 '19 at 11:11
  • "And then the verse starts on a C chord." ...but is that verse in a clear key? – Michael Curtis Dec 27 '19 at 15:19
  • 1
    Yes, the verse is clearly in C. Not a non diatonic note in sight. – Bob Dec 27 '19 at 15:21

It's modulating ... or at least changing modes between C major and C minor, a commonly used thing. Changing between parallel major and minor keys. You can also see this as a movement of tonic by +/- 3 semitones, if you think Eb becomes the tonic in the three-flats parts. The chords could even be seen as diatonic, just in different keys. There are different (justified) opinions on whether a change between parallel major and minor keys should be called a "modulation", because the home note stays the same in such a change.

The modulation sounds interesting, because it makes you re-think and re-imagine what's happening. "It's a cat!" "No, it's a dog!" "Now it's a bird!?" "It's Superman!"

The chords do not explicate all notes of the scale, so they leave room for imagination and soloing. And even if you play and explicate all the notes you're imagining, the memory of the key from the past few seconds is still in your mind and you're in a state of re-thinking. The new harmonic context needs some time to settle in, and the feelings during the settling-in time is part of why modulations sound fascinating.

variation 1

In the picture above, I've written out key signatures denoting scales you might be imagining. See video at the bottom for an audio example. That's not the only way to see it though, you might imagine e.g. C mixolydian harmony (1 flat) over the C major, or Eb lydian on the Eb (2 flats). On the Dm you could imagine a lot of things.

Michael Curtis suggests in a comment that declaring these changes as separate keys is a bit questionable, because the time spent in each "key" is so short. And of course, one wouldn't write key signature changes for these, because in 4/4 you would have two key signature changes in a single bar. (Which is why I wrote the illustrations in 2/4) But in my opinion, even during those brief moments you have to have a means for reasoning and handling the situation. The only way is to try to keep up with the change and act in each moment as if that was "home", while thinking about the next temporary home already. Papa was a rolling stone, wherever he laid his hat was his home.

To understand what's happening in the harmony, try these variations:

variation 2

In the second picture I just wrote Cm in place of the other chords. Does it still carry some of the characteristics of the original?

variation 3

In the third variation, I replaced the C majors with C minors. Now it's kind of doing the same thing as the original, but in a slightly more predictable or boring way. The settling-in and re-thinking of the modulations is missing! The only hint of that is the Dm7, which has an A note (when in natural Cm it should be a flat Ab). If you replace the Dm chord with Dm7-5 i.e. D half-diminished i.e. Fm6/D, then even that little bit of ambiguity is removed.

variation 4

In the fourth variation, I've written out the missing G dominant chord that leads to the next C-based chord.

If we make the Gm7 a "proper" G7 dominant seventh and replace the Dm7 with a D half-diminished seventh, we get the ultimate version with all ambiguity removed:

variation 5

Here's an audio example, how I think the key or mode changes work in this chord progression:

| improve this answer | |
  • Interesting! I do think it works quite well when playing just C and Cm. Thanks for your answer and the audio example! – Bob Dec 27 '19 at 18:27
  • 1
    @Bob No problem, it was interesting to try and explain this. I can play this stuff without even thinking, but having to explain it to someone else is a different story ... Having to explain clarifies my own thinking as well. I think this was like the fifth question this week where the essential thing that made a chord progression unclear for the OP was switching between parallel major and minor keys. They are all just tricks, and even one trick can throw you off, if you aren't familiar with the it. Let alone multiple simultaneous tricks. – piiperi Reinstate Monica Dec 27 '19 at 18:33

| C | Ebmaj7 Dm7 | C | Ab Gm7 ...And then the verse starts on a C chord.

It isn't clear if the verse is in C major or just starts on a C major chord. Knowing what happens in the verse might help with understanding the intro.

@Tim already pointed out that Ab Gm7 could be viewed as a tritone substitution. But where is the tritone in the Ab chord. A major chord rooted a half step above the next chord is not necessarily a tritone substitute. It could be some kind of Phrygian coloring.

@Piiperi brings up modulation, but how is it modulating when a key isn't even established in the first place?

You might talk about borrowed chords or some kind of modal coloring. But I think the first significant thing to say regarding function is: it isn't fulfilling any functional harmonic roles. In fact it seems to eschew harmonic function.

It provides non-functional, modally colorful contrast to the C chord.

It seems worth pointing out three details:

  • the Dm7 and Ab sort of work in opposition - presenting A natural and A flat - making modal borrowings murky. If it weren't for those A's in those two chords, we could identify a single mode from which to "borrow."
  • there is a kind a sequential harmony potentially at play Ebmaj7 Dm7 and Ab Gm7 are essentially the same pattern sequenced by a descending fifth. Depending on the chord voicings and the rhythm, that sequence may be noticeable and add some structure to what otherwise seems a bit amorphous and non-functional.
  • the progression from C to the two sequenced patterns involves a chromatic mediant change in both cases: C Eb and C Ab. (Chromatic mediant changes like those always catch my eye.) It's sort of like chromatic mediant stepping stones to two modal regions: first to Dorian-ish and then to Aeolian-ish. I wrote "-ish" because I'm not sure this really works as typical chord borrowing or an outright change of mode.
| improve this answer | |
  • 1
    +1 for 'it isn't fulfilling any functional harmonic roles' Is there actually anything wrong with 'it's there 'cos it sounds good'? – Tim Dec 27 '19 at 16:30
  • Who said anything about whether it sounds "good?" Are you mistaking "non-functional" for "sounds bad?" That isn't what non-functional means. – Michael Curtis Dec 27 '19 at 16:53
  • I said it - in my comment! Non-functional to me means it has no particular reason that can be explained in theory, like V>I etc. – Tim Dec 27 '19 at 17:42
  • Of course it's modulating. :) If you can't sense what's happening, try playing it at a slower speed maybe? ;) It's utilizing the mechanisms of functional harmony. The Ab is screaming "I am a subdominant"! And when the Gm comes, it is just stating the obvious: it's some kind of a dominant, the supposed tonic is Cm, though everyone knows it's probably going to give a C major instead, because it's doing the same switch at a steady rhythm and the change from C minor and C major is the "invention" here. Wax on, wax off, wax on, wax off, C major, C minor, C major, C minor, ... – piiperi Reinstate Monica Dec 27 '19 at 18:07
  • 1
    My thinking goes, even during the brief moment spent in each key ... you have to be able to operate and have a way to look at things. Where is "home note" if you're a wandering kind of person? "Papa was a rolling stone, wherever he laid his hat was his home". I guess harmonic progressions like this are designed to exploit the feelings of settling down. – piiperi Reinstate Monica Dec 27 '19 at 18:46

TTS! Tritone substitution! A♭ > Gm at least. There are many many changes in music which incorporate V>I. Here, V>I where I is G (or Gm) is D7>G (Gm). Looking at the make-up of D7, its 3rd is F♯, 7th (♭7) is C. Looking at A♭, its 7th is G♭ (enharmonic to F♯) and its 3 is C. Swapped over, maybe more convincing if it was labelled A♭7.

The E♭ to Dm could have been the same, except it's not E♭7, but E♭maj7. Containing three notes from G minor. So, I'm saying it's a chromatic passing chord.

Don't think Tierce de Picardie is the reason, as pieces are in a minor key before finishing on parallel major, and also Gm7 isn't a good dominant chord.

| improve this answer | |
  • So you would say that the Ab/D7 idea is only related to the Gm, and not to C? – Bob Dec 27 '19 at 10:27
  • 1
    Tts is only related to the target chord and the one before it. – Tim Dec 27 '19 at 11:09
  • I can't see tritone substitution as a good explanation here, because Ab is an in-key chord at that point, and it feels like a subdominant, not a secondary dominant. Exchanging it for a D7 or F#dim7 or something feels like taking the progression to a different genre. – piiperi Reinstate Monica Dec 27 '19 at 18:14

I don't think we need modulations here. Or to force a 'functional' explanation where every chord acts as the dominant of the next. Just an acceptance that chromatic chords are normal, common and permitted.

The A♭, B♭, C sequence (of which A♭, Gm, C is a variant) is complicated to explain functionally (though I've seen attempts). It can be labelled easily enough as 'planing'. But that's a description, not a reason.

Functional analysis often has problems with ♭VII. Yes, it has two notes in common with V7, but only one important one! Perhaps we should just let it be 'the chord next door'?

| improve this answer | |
  • But what do the chords do? They ”are chromatic”, is that what they do? Is there any difference between such ”chromatic” chords, just pick any of them, it doesn’t matter which one, any one does the same thing as any other? Random noise? I don’t think so. There is logic and it’s not random. – piiperi Reinstate Monica Dec 28 '19 at 3:46

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.