Alternate Paths on a Blues

Alternate Paths on a Blues

Today I’ll use “negative dominant” progressions to solo on a jazz blues. These ideas are from Steve Coleman – and I’m not the only one to have tried to interpret them. I had to cut them down a lot, so I recommend you read his stuff, with the warning that it is hard! After I do my best to explain the idea, I’ll show how these movements are present in typical jazz harmony, then play through entire alternate chord progressions built off them.

To understand a “negative dominant” progression, we should consider a traditional dominant to tonic cadence.

Trad 2

The tritone B F (actually tritone plus an octave in this voicing) resolves to C E, a major 3rd (plus an octave). The chord moves down a 5th (or up a 4th) – G7 to C. These resolutions are the basis of mainstream jazz harmony… but not the whole story.

Although this cadence happens all the time, spelling the notes of a plain V I progression makes a very corny melody. Jazz musicians have long avoided that sound in favour of altered and substitute chords.

Steve Coleman has characterised the harmonic/melodic techniques used by Charlie Parker to avoid the V I sound as “alternate paths” or “invisible paths”. He brilliantly uses symmetry to explain how they are the “dark side” of a normal V I. (He is also brilliant at coining names for these things, evidently.)

Symmetry emerges from mirror images.


What would be a mirror image of a V I? I can “reflect” it by inverting it, for instance around the axis note D. (For the nerds, this is because the C major scale is symmetrical around the note D.) So, every note in the original progression is replaced by one equally distant to middle D…

Inv Loran Fixed.png
The process of reflecting each chord of the original progression. G7 turns into D-6, C turns into A-.

… but on the opposite side of the D axis note. Below D if it was originally above, and above D if it was originally below. E.g., the B in the G7 ends up as the high F in the D-6.

Inv Full Progs.png
Comparing the original progression with its reflection in bar 2.

What is the new progression? D-6 to A-. It has a tritone resolving to a major 3rd (B F to C E, again separated by an additional octave) but the chord moves up a fifth, not down (D-6 going to A-, not G7 going to C). And both chords are minor, not major. Steve Coleman calls this a “negative dominant” resolution to A minor.

This by itself might explain how a lot of II V licks seem to be more tonally weighted on the II- than the V7. A melody over a II V I could imply a II-6 to VI- movement (negative dominant resolution) rather than a V7 to I movement. Although the chord tones are almost the same, the mental model and the tonal gravity would be different.

Such melodic shapes can be shifted to other positions and still retain their cadential power. This is due to the phenomenon of borrowed chords, or modal interchange in jazz speak. So, instead of the negative dominant that fits in the C major scale (D-6), we could use the one that fits C minor, i.e. F-6. It still resolves down a fourth (because it’s a negative dominant), but with the distinctively pretty sound of landing on a tonic major: F-6 Cmaj. Coleman notes that this IV-6 sound is often used over a V7 chord, creating a “dominant 7th complex” notated V11b9 (G Ab B C D F). The darkening substitution of D-6 by F-6 can be re-applied to the F-6, changing F-6 to Ab-6. The resulting bVI-6 sound is also used on dominant chords forming an altered V7#5b9 chord.

So, without going any further into symmetry, we have three melodic-tonal centres that can be used as dominant chords to target a tonic chord: II-6, IV-6 and bVI-6, targeting I. Crucially, these negative dominants are present as upper structures in most functional  jazz progressions.

Often, one of these negative dominant chords will be found as an upper structure of a jazz chord, followed by one of the darker versions (e.g. II-6 followed by IV-6) as an upper structure of the next chord. So:

D- F- is present in the following functional chord progressions:
B-7b5 E7alt
B-7b5 Bb7
D-7 G7b9 (probably in the key of C)
D-7 Ealt (probably in A minor)
Fmaj7 Bb7 (probably in F major)

I’m being flexible with chord spellings – to make the point clearest I could say F6 Bb9, because clearly D- and F- are the exact upper structures of those chords. But I’m using Fmaj7 Bb7 as a shorthand for two chord types, not exact voicings. Same deal with the G7b9, it technically should be the 11b9 mentioned above.

There’s another darkening movement, which is shifting up a tritone:

D- Ab- is present in:
D-7 G7alt
D-7 Db7

As well as these darkening movements, there are the actual negative dominant resolutions to a target chord.

D- A- is in:
B-7b5 E7b9 A-
E7b9 A-
G7 Cmaj
D-7 G7 Cmaj
G7 F#-7b5
C#7alt F#-7b5

F- Cmaj is in:
D-7b5 G7b9 Cmaj
G7b9 Cmaj
F-6 Cmaj (back door, same with the next two)
Bb7 Cmaj
F-7 Bb7 Cmaj
Bb7 A-
F-7 Bb7 A-

Ab- Cmaj is in:
G7alt Cmaj
Db7 Cmaj
Ab- A- (not seen as a written chord progression but I’ll be using it later)

Okay, let’s stop with the wall of chord symbols. The take-away is: a small set of negative dominant progressions (and their associated voice-leading and cliches) can be re-used on a huge variety of jazz changes. Today I’ll use the two basic types of movement: darkening and resolving – to navigate inside and outside the harmony on a jazz blues.

2 2 Stave

My alternate pathways in the first video, with the second staff showing example bebop harmony compatible with the alternate pathways.

My alternate pathways here are inspired by the original melody of Blues For Alice (transposed to C). Then I take a somewhat strange turn in bar 9. I work from II- VI7 II-7 V7, a common decoration of a II V progression, e.g. as implied by the melody of Billie’s Bounce…

BB Lick

… but I use a C#- to target the second D-, and then straightforward negative dominants to target the E-7 of the turnaround.


Here there are two main ideas: bar 1 has an unexpected B- (equivalent to Bb7alt) targeting Eb-6 in bar 2, which I interpret as a C blues scale shape (because it has the notes Eb, Gb, Bb and C). This is another way to use minor shapes – as blues colours, primarily I-6 and bIII-6 against a I or IV chord. But here the Eb- (bIII-) is also functional, implying a D7alt sound going to G-.

Then I use what could be standard bebop changes to reach the bar 9: interpretable as, say, Cmaj7 F7 E-7 A7b9 (bars 7-8). But I keep up this rate of movement to arrive at a tonic chord (A- which could be Cmaj) in bar 10 rather than bar 11 as expected. I create a cyclical feel by repeating the exact pathway for the next 3 bars. Every pair of chords involves a shift up a minor 3rd, but it’s not a strict pattern because the G- D- resolution breaks it.

The first pathway in video no. 3.


The second pathway in video no. 3.

If you find it hard to hear how this relates to a blues, here is the same solo with bass notes added (and abominable sound quality!).

I had to slow down even more to get some juice out of these progressions. (I play the first one once and the second twice.) The first uses those blues colours again. The second uses unexpected resolutions of a minor chord to major chord a fourth below (so, the F#- is an Amaj, and the C#- is an Emaj), with that major chord changing to a minor chord. It also strictly uses only the darkening and dominant cadential movements, lending it quite a lot of momentum.

There are so many more possibilities, of course. For example, diverging from the subdominant chords in the blues, i.e. the IV in bar 5 and the II in bar 9. In my examples I stick to the original subdominants. Obviously I’m only barely scratching the surface!

I had fun coming up with and playing through these progressions. To conclude, I think the relationship of these pathways to conventional jazz harmony is crucial. I’m thinking both ways as I play. Also, obviously, the pathways are only a technique. These sequences have a rather severe sound due to the unrelenting drive of the cadences and the minor colouration – that mightn’t always be what you want.

Hope you enjoyed it! Sorry for the late post. Please comment with any related ideas, thoughts, questions or criticisms!

Thanks to Loran Witteveen for correcting my examples!


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s