Originally posted here. Listed as: rev. 3–13–02. Page date: 7/30/2003.
This list of rules for 4-Part Harmony was taken from the web page of a teacher in the music department at Walnut Hill School for the Arts. This page disappeared some years ago, so I am publishing these rules here for the help of others. I’ve also added a few notes on doubled thirds and parallel octaves from a friend of mine, who translated the notes from a Romanian Conservatory book on Harmony.
When I was learning how to harmonize hymns properly, some of these rules came in very handy. Of course I wouldn’t say you have to necessarily follow these rules. They come from classical training, and were devised to help with clarity of voices.
However, some forms of music (medieval, folk) do not follow all these rules and still produce music of great beauty. So, if you know what you are doing, feel free to carve out your own niche. But for beginners, learning these rules can help make your harmonization more smooth.

Simply following these rules does not necessarily make good voice leading. We will also discuss more subjective criteria, as well as exceptions to these rules.

To preserve clarity of harmony:

  1. The root and the third are needed to establish a harmony.
  2. The root of a triad is usually doubled. When this is not possible, the third is usually doubled. Notice the word “usually,” rather than, “always.”
  3. The distance between soprano and alto never exceeds an octave. The same is true for alto and tenor. (Note the presence of audible overtones in inverted pyramid chords.)

To preserve clarity of line:

  1. Voices do not cross.
  2. Voices do not overlap.
  3. Unstable tones are resolved by step. (Without this, we ask, “Where did that line go?”)
  4. In dominant harmony, the leading tone resolves to the tonic.
    1. In dominant harmony, the tritone resolves in contrary motion by step.
    2. In any seventh chord, the seventh resolves down by step.
    3. Any unstable tone may transfer from one voice to another before resolving.

To prevent one voice from sounding like two (i.e., compound melody):

  1. Dissonant leaps are avoided.
  2. Melodic augmented seconds are avoided.
  3. Leaps do not exceed an octave.

To prevent two voices from sounding like one:

  1. Consecutive perfect fifths or octaves between the same pair of voices are avoided.
    1. In parallel motion.
    2. In contrary motion.
    3. Doubling is different, a one voice texture; not part of this study.
  2. Perfect consonances are typically approached in contrary or oblique motion, but may be approached in similar motion under one of these conditions:
    1. If the harmony is not changing; or
    2. If the higher of the two voices moves by step; or
    3. If a perfect fifth is approached with the lower voice moving by step, and the higher voice leaping by a third.
  3. Unstable tones are not doubled. (Doubling an unstable tone necessitates either poor resolution, or parallel octaves.) This includes:
    1. The leading tone,
    2. The seventh of any chord, and
    3. Any part of a tritone.

Additional rules of thumb:

  1. The inner voices are normally monotonous.
  2. The upper voices—soprano, alto, and tenor—tend to move in opposition to the bass, in this order of preference: contrary, oblique, similar, parallel.
  3. Within each part, maintain a sense of line, with these preferences:
    1. Best: repeated (or tied) notes; it is usually possible to find a common tone between chords, and this will usually stay in the same voice.
    2. Good: stepwise motion.
    3. Dangerous: leaps.

Some Notes on Doubled Thirds and Parallel Octaves

  1. It’s compulsory to constantly double the thirds of the minor chords in the major keys, namely if we stay in C Major we double the thirds of Dm, Em, Am, and Bm. They are substituting the major chords, and as such they have not to sound as strong as the Tonic, Dominant, and Sub-dominant, but to affirm the 3 major functions of the key.

    If you want to modulate to a minor key then you don’t have to double the respective thirds. When their thirds are doubled, the minor chords of any key affirm the principal function of the chord they are substituting. Dm and Am chords substitute the Sub-dominant F chord. Em substitutes the tonic C major Chord, and Bm substitutes Dominant G major chord. When the third of one of these minor chords is in the bass then it’s affirming the best the major function for which it stands as a substitute. The best sonority it’s achieved when the other third is given to the tenor, or when the doubled thirds are given to the alto and tenor. You can also double the sixth of these minor chords, and the best sonority is achieved when the double sixth is given to the tenor and soprano.

  2. When you have the CM chord and Dm one after another you can also double the third of the Tonic.
  3. Em is an ambiguous chord in C Major, it can substitute both C Major and G Major, respectively the Tonic and the Dominant. When you double its third it substitutes CM chord. When you double its sixth it substitutes GM chord.
  4. Am can also substitute C Major or F major. It needs double thirds in both cases.
  5. You can double the thirds of the major chords if they are on the weak counts of the bar, for example on the second count in 2/4, etc. But beware of doubling it by giving to the alto, it doesn’t always sound the best.
  6. Parallel fifth and octaves are not weakening the strength of the functional harmony of a Key when done on the same chord. When done on purpose they sound special like in the minor Greek modes Dorian, Aeolian, Phrygian, etc., and they introduce a special antique atmosphere by weakening the harmony of the Key. Some consider them as bad musical practice, but they can be used very effectively when done on purpose, and their colorful effects can be mellowing and pretty nice. In such cases the soprano affirms the key in which the song is written.


Leave a Reply

Please answer this small math test (it helps prevent robots from spamming the comments): * Time limit is exhausted. Please reload CAPTCHA.