Relating melody to chords

In general, harmonic accompaniments are based on the principle that the pitch of a melody note belongs to one of the pitch classes in the chord that accompanies it. For instance, if the melody note is E, the accompanying chord will most likely contain an E as well. For brevity, we’ll refer to a chord’s constituent pitch classes as its “chord tones.”

When a melody note is a chord tone, it creates a “consonant” or “harmonious” relationship between melody and accompaniment. However, it often happens that a melody note is not a chord tone. For instance, if the melody moves scale-wise while the chord doesn’t change, there will be “passing notes” heard briefly between one chord tone and another (since chord tones are normally two scale degrees apart; see Chords). On the other hand, if the melody sustains a pitch while the chord changes, what was at first a chord tone may not be a chord tone of the new chord, creating what is called a “suspension” or “suspended dissonance.” We will use the word “dissonance” to describe any melody note that is not a chord tone of the chord that accompanies it.

The effect of a dissonance is usually one of discord or tension between melody and accompaniment, often creating a feeling that the melody should move to a chord tone before the chord changes. Dissonance doesn’t mean that the music sounds bad or unpleasant. Most harmonic music uses some combination of consonance and dissonance, and the contrast and balance between them, to create its effects.

If we want to capture how music does this in our notation, we need a way of showing what the chord tones are and when the melody note is and is not a chord tone. A starting point for this might be the way we notated a melody with drone. In effect, a drone could be considered as a chord that doesn’t change throughout an entire piece. By writing the melody and drones in the same space with different thicknesses of line, we were able to see where the melodic pitch does and does not coincide with the drones.

To do the same for changing chords, we could replace part of a pitch line with a thicker or colored line wherever the pitch in question is a chord tone. But remembering that some chords contain extra-scalar pitches, we may want to distinguish those chord tones which are scale degrees from those which are not. Also, since any combination of (for instance) scale degrees 2, 4, and 6 forms a ii chord, the chord tones may not all be actually sounding in every octave, so it might be misleading to notate them in the same way as audible sounds (e.g. with color). On the other hand, we do want to mark the chord tones in the octave(s) where the melody lies, whether they are sounding in that register or not, so that we can easily relate each melody note to the chord tones that are nearest to it in pitch.

To meet all these requirements, global notation resorts to a somewhat subtle solution (and one that is probably not worth trying to write by hand). Pitch lines are drawn dotted by default, and continuous only when they represent a chord tone. Where a chord tone is an extra-scalar pitch, it is drawn in as an additional continuous line, only for the duration of the chord. The root of a chord is shown as a thicker continuous line. (It might seem that this could be confused with the tonic, which will also appear as a thick continuous line when the tonic is a chord tone, e.g. in chord IV or vi, but in fact it will be unambiguous that if there is only one thick continuous line in each octave, the tonic is the root, while if there are two, the one that is not the tonic is the root.) Symbols for sounds of definite pitch are then drawn in grey or (probably better) a color, allowing the pitch lines and chord tones to show through.

Applying this to “When the Saints Go Marching In,” we extend the notated pitch range downwards to show the basic triad form of each chord, initially keeping the chords as simple as possible.

It’s clear that most of the horizontal red lines (especially the longer ones) have a continuous black line running through them, which shows that most of the melody notes are chord tones, in other words the melody is largely consonant with the accompaniment. However, occasionally the black line running through the red one is dotted, which indicates a dissonance. The first bar doesn’t count, as the harmonic accompaniment hasn’t started yet, but at the end of bars 3 and 5 we can see “passing notes” where the melody is moving scale-wise between chord tones. In the second system, at the moment where the IV chord begins, there is a brief “suspended dissonance”: the highest pitch, which was a chord tone of the I chord, is held over while the chord changes, becoming a dissonance which is then “resolved” by moving down to the nearest chord tone below it.

At the beginning of the second system, there is another dissonance which is not prepared by being first presented as a chord tone. This seems to create more tension; perhaps it is related to the feeling of longing expressed in the words at this point: “Oh Lord I want…”? Many gospel singers would intensify this feeling by lowering the pitch slightly to create an extra-scalar pitch—a so-called “blue note.” Any extra-scalar pitch that isn’t a chord tone will appear in the notation as a colored horizontal line without a black line in the middle (either dotted or continuous). We’ll include this at the beginning of our next example.

Here we show the second system of “When the Saints Go Marching In” with the more interesting chords that would probably be used in an actual performance. The melody line is shown in grey so you can compare it with the red for legibility. We also include a bass line to emphasize that the lowest note in a chord is not necessarily the root. When it isn’t, the chord is said to be in an “inversion.”

In global notation, inversions are specified by indicating which note of the chord is in the bass. Here again, the notes of a chord are numbered by counting upwards from the root, and the sharp and flat signs are used to specify whether the “major” or “minor” version of the interval is involved. The resulting indication is added to the end of the chord symbol after a slash, so that, for instance, iii/b3 refers to a iii chord with the minor third in the bass. (The octave transposition marking at the beginning of this example indicates that only the sounds represented by double black lines sound an octave lower; see Wide pitch range.)

Admittedly, the distinction between melody and chords is not always as clear as this form of notation might suggest: a melody note can always be interpreted as a part of the chord that accompanies it, and it’s often a matter of analytical judgment whether a particular melody note is better understood as a dissonance or an added chord tone. Indeed, the whole exercise of identifying chords with symbols might seem more like analysis than notation. But the answer to that is that any attempt to notate music involves making analytical decisions, whether conscious or intuitive, and what global notation aims to do is to make the basis of such decisions explicit so that what seems “normal” or “natural” in one tradition is not unthinkingly applied to all the others.

2 comments on Relating melody to chords

  1. This approach to notating chords does not seem very applicable in a broad range of contexts. Consider chords in a scale other than 12TET – western chord symbols would be near-useless. Not to mention, western notation often has multiple redundant ways to notate one chord.
    Instead of chord symbols, you could use something that references the scale the music is played in – for instance, a vertical stack of circles (corresponding to the scale degrees, with the reference pitch on the bottom), which are filled or empty depending on whether each scale degree is present in the chord.
    Admittedly, this cannot notate chords with extra-scalar pitches, but this begs the question: what is a chord? What is a scale? If a set of notes is played long enough to define a chord, shouldn’t those notes be important enough to be scale degrees?
    Notating chord inversions is redundant if there is already a notated bassline – although there may be other ways of notating a bassline.
    Distinguishing pitch lines by whether they are chord tones or not, while it can be useful for at-a-glance analysis, contributes to the proliferation of line types in global notation. The decision to include or exclude this feature is ultimately a decision between the burden on the writer, and the convenience of the reader (personally I would vote in favor of the writer).
    Regardless of whether you accept my suggestion, the chord system needs an overhaul.

    1. You’re right, the suggestions on this page are so far limited to Western-type triadic harmony. That seemed a good form of harmony to start with since the chords have recognized identities that can be specified independently of the individual tones that they are made up of, and the relationship between chords (in that sense) and melodies has been a very big deal in Western musical theory and practice. I think a similar approach could be applied, say, to African music that uses cycles of chords each formed of two pitch classes and that may be based on a quite different system of tuning such as seven equal divisions of the octave.
      Your idea of having stacks of circles to show which scale degrees are chord tones at a given time sounds useful – with practice one would probably learn to associate particular patterns of filled and open circles with particular chords – although I think it would still be helpful to give names to the chords (as Western music theory does with its roman numerals etc.) so that we can talk about them clearly.
      Decisions as to which pitch classes form part of a chord or scale will often involve judgment. I still find it useful to make these distinctions, but if you’d rather avoid them you can treat any pitch as forming part of a chord and a scale by definition. As always, the aim of global notation is to allow you to specify just the information that you want to.

Leave a Reply

Your email address will not be published. Required fields are marked *