Corey Mwamba


Descriptive harmony style guide


Harmony has moved on: musicians do not necessarily write music that can be analysed functionally any more, but the descriptive language has not moved with the music.

The prevailing system of notating chords using letters and numbers has existed for approximately 300 years. There have been few changes to it in this time; what has proliferated more is the number of analyses and explanations as the language of Western modern music has become more complex and variegated. These explanations tend to bring in slivers of knowledge about mediaeval modes with synthetic scales and chords abstracted from early 20th Century music, classical functional harmony and a de facto convention in notation.

But this is not a synthesis; more a vague shifting of an orthodox position, which is tied closely to assigning transpositions of scales as "modes"; degree alterations of these "modes"; and triads built from these.

In music most closely related to jazz this has had consequences on accuracy (how correct something is) and precision (how reproducible or consistent something is).

  1. Chord symbols are now less descriptive and more functional – if a chord is rigidly married with a scale then the flexibility of description is diminished; moreover, as the chord has an associated scale, melodic notes may be added to the chord that actually break down the intended harmony.
  2. The capacity to describe has diminished – if a chord cannot be married rigidly to a scale or "mode" it is less likely to be described, or more likely to be described inaccurately.
  3. The symbolism is arbitrary and not truly representative – take "alt chords" for instance. Visiting any music theory forum will present a question on this: and each thread will run as there really isn't a consensus on what an altered chord is, other than the fact that it has been altered in some way. Thus the prevailing system is not precise, although it is intended to be.

The tools to purely describe a chord, with no reference to scales (or at most, with reference only to the chromatic scale), do exist. But the robustness of current Western descriptive harmony notation should not be ignored or thrown away: perhaps some reassessment is in order though. This is an attempt to do so, looking at different and personal approaches to specific problems in popular/jazz notation. In each case I have to satisfy these questions:

And in these questions it is worth exploring those terms accuracy and precision.

Accuracy in this case does not mean "the measure of detail", but "the measure of truth to the target". This means that if your artistic aim was to leave some openness about a harmony for a musician, allowing for various colourations, it may be better to write a more basic symbol. If this gets you your desired effect (or close to it) the symbol is accurate. This hinges on knowing and understanding things that are less to do with mechanics (which, let's face it, is what we're dealing with here) and are more to do with people and the depth of your work.

If however you need a specific sonority and you are unable to describe it fluently then the symbol possibly needs work, unless it is really a very difficult thing to describe. But if the notes are there, it can be described.

How efficient you are at realising the harmonies as symbols for others to read has a direct effect on the accuracy. If the notes are coming in the right order but the sound isn't working the way it did before then the precision might need some work – sometimes it's a matter of dynamics, but it can be the voicing of the harmony; again something the composer can influence directly.

I feel the style guide I present here gives me the flexibility of accurate description with a precision in emotional feeling. Some of it will be familiar – but some may not be. It's all intended to be helpful but not prescriptive.