Against the primacy of the third.
Western harmonic theory rests upon an assumption so familiar that it is seldom questioned: that the major third is one of the primary consonances from which harmonic language is built.
I am no longer convinced that this is true.
The harmonic series undoubtedly contains a major third, but it arrives only after the octave and the dominant fifth have already established themselves. More importantly, the major third found in nature is not the major third found on a modern keyboard. Twelve-tone equal temperament requires every interval to be compromised, and among the common consonances it is the major third that suffers one of the greatest distortions. We accept this because it allows us to modulate freely, not because it is acoustically ideal.
Traditional harmonic theory nevertheless places the third at the centre of its architecture. Chords are constructed by stacking thirds. Extensions continue the same process. Seventh, ninth, eleventh and thirteenth chords all arise from the same underlying assumption that harmony is fundamentally tertian.
My own experience as an accompanist has gradually led me elsewhere.
When a chord sounds, we do not hear isolated notes. We hear interacting spectra. Every note carries its own overtone series, reinforcing some frequencies while competing with others. The ear does not experience every component equally. Some tones possess greater perceptual weight than others.
That distinction matters.
If the third is treated not as the primary structural interval but as one colour among many, a different harmonic language begins to emerge. Instead of building sonorities by successive thirds, it becomes possible to think in terms of successive fifths. The resulting structures remain harmonically rich, but their centre of gravity changes.
Consider a simple C major triad.
Conventionally it is understood as C-E-G.
My ear increasingly hears a more stable framework in C-D-G, where the dominant chain exerts greater perceptual force than the tempered third. This is not an argument for abandoning thirds altogether. Thirds remain indispensable. Rather, it is an argument for demoting them from their privileged position within the hierarchy of consonance.
The practical consequences become clearer as chords become larger. Conventional extended harmony often produces dense interactions between competing overtone series, particularly when closely voiced. By contrast, sonorities organised around fifth relationships tend to remain open, stable, and remarkably transparent beneath a singer. The third continues to contribute colour, but it no longer determines the identity of the harmony.
For an accompanist, this changes almost everything.
The task is no longer simply to realise harmonic function. It is to distribute harmonic weight. Certain notes establish stability. Others imply direction. Others provide colour without demanding attention. A successful accompaniment balances these competing roles so that the listener's ear remains fixed on the narrative rather than the mechanism supporting it.
This is not an attempt to overturn Western harmony, nor to replace centuries of musical practice with a new system. Tertian harmony has proved itself extraordinarily fertile, and without it much of the world's greatest music would be unimaginable.
It is, however, an invitation to ask a different question.
Instead of asking what chord a collection of notes represents, we might ask which notes the ear experiences as structurally significant, and why.
That shift, from harmonic function towards harmonic weight, has shaped almost every accompaniment I have written in recent years. Whether it proves to be a useful theoretical model for others remains to be seen. As a practical approach to accompanying traditional song, however, it has transformed the way I hear.