youngpokie
Senior Member
I think he goes there because it makes a lot of sense on a certain level.For Riemann the subdominant is the SUB-dominant: the dominant from below, from the inverted series... The inverted series is deducted from the inversion of the tonic triad. But why does he go here?
We usually consider the tonic as the first node in some kind of linear representation of functions. It is certainly the mainstream view in many countries and it is very convenient. It can be easily represented in a straight line, analogous to a scale. Something like this illustration (just a random image from Google) - a linear model I to VII:
But the other way of describing the same thing is by putting the tonic in the center and the mediants, subdominant and dominant as radiating outwards (or up and down) from the center:
This latter modeled is Riemann-derived and used (and enriched extensively) by the Russians, from what I understand. You can conceptualize this model as a kind of oscillation between S and D, crossing the T, with clearly defined and measurable tension qualities, and that eventually settles on the tonic. It can be compared to the oscillation of a string or a metronome. More recently, it's been described as something not unlike Fibonacci spirals to show how it can incorporate microtonality, temperaments and other types of sectioning and division.
In my opinion, this German-Russian model (in contrast to the familiar linear model) is far more visual, deep and elegant. It better explains the leading tones, chromatic alterations, the very concept of cadence and all of its types, plus chord substitutions. The Russians used it to explain the merger of the major and minor modes into a single mode, a far more elegant concept than "chord borrowing" the linear model uses even today.
"Negative harmony" is a very unfortunate and misleading term, but it's central idea - the mirror inversion up or down, is very reasonable. If you build a C major chord and then repeat the same exact intervals in the opposite direction from the tonic, you will indeed get an F minor chord. All of the major chords will become minor chords, the diminished chords will remain diminished. All authentic tensions will become plagal. All this is part of "negative harmony" but is more simply and elegantly explained by the Russians in the model shown above, and used by composers since well before 1895.
What Levy introduces that's totally new is the idea of an axis - a theoretical tone that's equal distance between the tonic and the dominant. This specific tone (and not the tonic any longer) is the point from which the inversions are created up or down, retaining all of the principles shown above. The entire subdominant group becomes displaced, so to speak. And:
- we get a whole new group of (mirrored) chord substitutions, while all of the leading tones and tensions of the given key are preserved
- the dissonances from layering of these polar chords are very striking (Bartok concerto for orchestra is a great example)
- we can reharmonize existing material differently and we can also deliberately modulate between the two groups in the same period or sentence and get even more color
I'm not a big fan of negative harmony but I do think it's an interesting inversion technique for enriching harmonic color and specifically progressions and chord layering. It sounds a little jazzy, often unusual and sometimes fresh. That's all it is, imho.