Negative harmony’s mirror world

Take a moment to listen to Joss Stone, doing the Ray Charles tune I believe it to my soul with the able assistance of David Sanborn on sax. It’s a 12-bar blues. Check the harmony on bars 10 and 11 of each chorus, where Joss sings “I believe you’re trying to make a fool out of me”.

The first two lines are standard blues harmony. For the last line, a basic blues would go V (B7), IV (A7), I (E7), ending with a plagal or ‘amen’ cadence. Instead of that, we have bVI (C7), V (B7), I (E7) – a 2-5-1 with the II chord (F#7) tritone-substituted. This circle-of-fifths progression obeys the usual law of harmonic gravity, with the roots descending in fifths.

On subsequent choruses the return from C7 follows a different route: C7, G7, D7, A7, E7. It’s still circle-of-fifths motion, but in the opposite direction, roots ascending by fifths : bVI, bIII, bVII, IV, I. To me it feels as if the harmony is climbing stairs, pushing against gravity and getting there all the same. We are used to ending the blues with a plagal cadence. This is what a chain of four of them sounds like.

Earlier this month I attended a talk by saxophonist Mornington Lockett about negative harmony, an approach which puts musical pitch through the looking glass, so that high notes become low and low notes go high.

Melodic inversion, of course, is older than Bach. But here we are talking about turning harmony upside down. Major and minor triads are interchanged. Regular major cadences become minor plagal cadences. Turnarounds, rhythm changes and other circle-of-fifths progressions become chained plagal cadences – slightly different from those you just heard, but with the same sense of going against harmonic gravity.

Mornington’s blog, which I urge you to read, gives a no-nonsense explanation of negative harmony, including a full set of diatonic triads for negative C major. He rounded off his talk by playing a version of Stella by Starlight, reharmonised through the looking glass of negative harmony. Andrea Vicari on piano and Dorian Lockett on bass supplied the potentially weird chords.

Actually it did not sound so weird. One reason could be that a negative sequence has equivalent voice-leading to the original. And voice-leading does a lot to make a progression sound convincing.

Stella was a good choice for several reasons. Everyone knows it. It separates neatly into eight four-bar phrases, each with a recognisable rhythm. So a listener can follow the form, whatever weird things are done to the harmony. As for the harmony, it constantly moves between major and minor. It was interesting to hear what happens to Stella when major and minor are interchanged.

 If you find the analogy between colour and harmony persuasive, the effect is to turn a chord sequence into a kind of photographic negative of itself.

Major-equals-happy and minor-equals-sad are simplistic equations that real music often refuses to obey. Nevertheless, it’s hard to ignore the conventional associations of mode and mood. Listening to the reharmonised Stella, I felt that some kind of mood-reversal had been applied to the tune’s narrative arc. The effect on the last line was strange. The mild surprise, where (in the original) three minor 2-5s resolve unexpectedly to B-flat major, was still there. But a comforting surprise became a somewhat disturbing one.

There’s a load of technical stuff to say about negative harmony. In my next post I’ll point you towards some seriously geeky material, and explore the idea’s roots in the work of Ernst Levy and beyond.

But a musical technique is worthless unless it can help to create emotion and shape a narrative. So I thought I’d start by saying why I think negative harmony passes that first test. I believe it to my soul.