An email asked for help on the subject of writing countermelodies. This subject is part of the wider field of counterpoint and is covered extensively in the book. My reply is reprinted here for the benefit of anyone else having trouble.
A countermelody may be subservient to the principal theme or it may be of equal importance. It will also tend to avoid a similarity of vertical stresses and accents. This helps prevent the combination of the two (or more) melodies from developing a homophonic (vertically conceived, harmonic) structure. On this same subject, too many chord changes in close succession will create a homophonic rather than polyphonic style. To quote an obvious example, if we placed a chord change on every note of the melody, there’d be little space for counterpoint at all.
The number of changes is governed partly by the tempo of the music. Fast tempos are usually characterized by harmonic inactivity (see old standard songs such as Limehouse Blues, Lover etc.).
The rhythm and phrasing of the ‘original’ can be used to create effective countermelodies. Switching bars around in their vertical correspondence (diagonally, for lack of a better word) can achieve results that are inherently homogeneous because everything has a common origin. Similarly, rhythms within a bar (or within a phrase) can be divided into cells, which can also be moved around to form effective contrasts in adjacent voices.
Tip: write in the rhythm of the proposed countermelody below the staff where the part will go. Just indicate the ‘pure’ rhythm, the actual intonations will follow. Check for excessive similarity in vertical cross-sections, keeping the tyranny of the bar lines under control with tied-over (or accented over) phrasing to enhance the horizontal, linear style.
Everything we do will probably be altered and refined, especially when the harmonies exert their influence.
Or we can try making the rhythm of the countermelody the same as the original but entering a bar or half bar later. We might have to tweak all the parts to make this work.
Another technique is to use activity where there is a sustained note in the melody voice and vice versa. The countermelody can split into voiced section writing (to use jazzspeak) for this purpose (under sustained or slow moving notes in the melody). The example of The fully versatile ensemble in the book is a fair example.
The countermelody can use the same rhythm as the principal but with doubled or halved note values (augmentation and diminution). Doubled values can be amazingly effective in producing lead voices in background harmonies that truly ‘belong’ to the style of the piece, especially where the harmonies allow duplication of the intonations of the melody. But, in any case, the resulting melody/harmony correlations may produce higher extensions and altered chords, with surprising results.
A simple countermelody of equal note durations may already be present (or potentially present) in the existing harmonies. It will usually move downwards but it may be important to the form and structure of your arrangement to have an ascending, rather than descending line. This can always be created, if you know how. The movement may not always be in semitones, although semitones predominate in the most effective cases when moving downwards. The basic movement can be decorated with rhythmic fills or grace notes.
Fugue writers often use retrograde (backwards), inverted or retrograde/inverted (upside down and backwards) versions of the principal for the counterpoint. They justify this by pointing out that the music hasn’t changed, just the ‘point of view’.
Pop and Jazz writers may rebel against what they perceive to be excessive cerebralism but it really can be a good way to start. Thinking about it, the selection of rhythmic patterns is the main problem since intonations are governed by the harmonies, plus available unessential notes and chromatics, and by the need to avoid undesirable parallels or other voice leading problems. Attend to these essentials and there is often very little else left to do.
Another tip: where there is a marked contrast of rhythm, accent and mobility between two (or more) lines, voice leading problems hardly ever occur because the ear can’t recognize individual vertical correspondences. (Look up Arabesque forms in counterpoint.)
The countermelody and the melody (or lead voice of the section playing it) should avoid getting tangled up with each other. Very brief encounters are OK and a countermelody can cross over the melody voice where a distinct advantage occurs, especially when there’s a marked difference in tone colour. The countermelody can also be in octaves. The lower octave is heard as an amplification of the top part, not as a note of harmony, so that clashes with harmony voices are not a problem except for long notes at slow tempos (especially in the low register). We often need the robustness of octaves to fight loud ensembles.
A mixture of parallel, oblique and contrary motion is good, too.
It’s generally better to avoid beginning the theme and countermelody simultaneously. Use a lead-in (anacrusis in straight-speak) or a delayed entrance.
The rhythm of the countermelody will often feature rhythmic retardation both within the bar and tied over the bar lines. Where the melody is intended to dominate, a countermelody that ‘lags behind’ rhythmically will help achieve this aim.
There’s another useful, quasi contrapuntal technique which can be most effective and useful without modification. It’s not suited to the ‘mainstream’ development of music but may be used in bridge passages and codas, for example. A feeling of immobility is created, like treading water.
(1) Take one (or two…) bars of a passage in three or four part harmony. (2) Write out all the voices, beginning from the top down, continuously on one staff. (3) Repeat the process in the second staff, this time beginning with the second voice of the original harmony, and carry on until all bars are used up (the top voice now comes last because these are circular permutations). (4) Continue the process from the third voice (etc….) of the original harmony until the sequence is exhausted. Because the original harmony worked, this new scheme, which has converted simultaneity into continuity, will also work. Call and answer figures are created automatically, being handed about from one voice to another.
There are reservations. Higher extensions may be forced below their acceptable acoustic limits. Where the melodic lines are rhythmically active, the ‘offending’ notes will be too short in duration to cause problems, and can create intriguing textures, but we can rearrange the octave placement of parts as required.
We will not usually employ this technique for too long because of its lack of formal content but if we do (the book never bosses you around), we simply continue with the permutations. Or we can change key or mode, or key and mode, using any of the 36 seven note scales.
Where bluesy, interwoven sections combine to produce a jazzwise version of three part counterpoint, the methods discussed above may produce results that are too rigid (uncool in jazzspeak). With practice and ingenuity, we can produce movement where initially it seemed impossible. Parts demonstrating a marked difference of rhythm, accent and stress can coexist even when brief semitone ‘kisses’ occur. Each voice develops its own ‘logic’ and stands up in its own right. There will be a mixture of notes on the beat and notes tied over. The unique (to jazz and blues) tendency for a major/minor conflict to occur between the melody and harmonies will also be present.
Personally, and I know this might upset a few people, I prefer section by section writing in jazz counterpoint: trumpets with trumpets, trombones with trombones, saxes with saxes, because the instruments share a common form of articulation and similar methods of producing legato phrasing etc. It’s easier for instruments of like-kind to capture the subtle nuances of interpretation. This kind of ‘vocalization’ has been an integral part of North American music since it all started.