CHAPTER
15

Harmony and Counterpoint

In This Chapter

• Understanding the differences—and similarities—between harmony and counterpoint
• Creating pleasing background harmonies
• Using different chord voicings
• Composing interesting two-part counterpoint
• Learning effective voice-leading technique

A song doesn’t have to be anything more than a melody and chords. Think of a folk singer and her guitar, or a solo violinist accompanied by piano. Melody and chords are all you need. However, when you turn on the radio you don’t hear a lot of solo folk singers. What you typically hear is a full arrangement, complete with keyboards and bass and drums, background vocals, and other types of instrumental backing.

Of course, these background vocals and instruments are doing nothing more than playing the notes in the song’s chord progression. But they also help to fill out the sound and make the piece of music more interesting.

To fill out your songs, you need to add harmony parts. These harmony parts can be either vocal or instrumental, and there can be any number of them. What they do is simple: they follow separate lines within the underlying chord progression, thus buttressing the harmonic structure of the song.

If you want to get really fancy, your backing parts can represent new and contrasting melodies when played against your original melody. When you create this type of complex harmony, it’s called counterpoint; it’s widely used in many forms of classical music.

Two Ways to Enhance a Melody

Two notes sounded together make a harmonic interval; three or more notes sounded together make a chord; and two or more melodies sounded together make counterpoint. Intervals and chords are used to construct harmony; counterpoint exists as separate melodic lines.

Another way to think of it is that harmony is a vertical (up and down) combination of notes, whereas counterpoint operates horizontally (side to side).

Counterpoint is horizontal.

In reality, harmony and counterpoint are related concepts; both involve “vertical” combinations of notes and both involve a “horizontal” movement of individual voices or instruments. Still, harmony is more about singing or playing parts of a chord; counterpoint is more about creating a second (or third or fourth) melody line—albeit one that adheres to the underlying harmonic structure.

A melody with backing harmonies is the difference between a solo vocalist and a full chorus. It’s the difference between a folk singer with an acoustic guitar and a pop singer with a group of backup vocalists. It’s the difference between a jazz trio and a big band. In other words, harmony makes music bigger.

You create harmony parts by using the notes in the underlying chord progression. If all you do is assign specific notes of a chord to specific instruments or voices, you’ve created harmony. It doesn’t have to be any harder than that. Harmony parts, whether vocal or instrumental, are typically less rhythmically complex than the main melody. It’s not uncommon to find harmony parts consisting of whole notes or half notes while the melody maintains a more complex rhythm.

Harmony parts can also mirror the rhythm of the melody; in these instances, the harmony resembles classical counterpoint—which you’ll learn about later in this chapter. Harmony can also be used to punctuate the melody, fill in breaks in the melody, and function as a kind of call-and-response mechanism. (Think of the classic pop tune “Midnight Train to Georgia”; Gladys Knight is the call and the Pips are the response—woo woo!)

In terms of harmonic complexity, you can have everything from a single accompanying voice to choruses and string sections with two and three and more voices. The more voices you have, the more challenging it is to create distinct harmony parts without doubling or duplicating other parts. Of course, writing a single harmony part is also challenging, but in a different way; that single part has to include just the right notes, suggesting the underlying chord without distracting from the main melody.

For the purposes of learning basic theory, we’re going to concentrate on simple two- and three-part backing harmony, without a lot of rhythmic or melodic complexity. Once you master this type of rudimental harmony, you can expand to include more complex types of vocal and instrumental backing.

Voicing and Inversions

The order of the notes of a chord (top to bottom) is referred to as the chord voicing. (It’s also called the chord inversion, as you learned in Chapter 9.) Voicing is very important when you’re writing harmony parts, because you have to employ different voicings to avoid parallel motion between parts.

Let’s consider the harmony you play as piano accompaniment. If you recall, we used the following chord progression as an example back in Chapter 13:

A common chord progression that needs harmonized.

As learned back in Chapter 13, a simple block-chord piano accompaniment (right hand only) to this chord progression looks something like this:

Simple block-chord harmonies.

The triads you play as accompaniment represent three-voice harmony; you just happen to play all three voices with one hand. All the chords are in the root (noninverted) form of 1-3-5—which represents a problem. When you change from chord to chord, all the harmony notes move in parallel to each other. In terms of voice leading, this is often frowned upon.

It’s also boring.

You can prove this by isolating the top note of this chord accompaniment. Now sing the succession of top notes as if you were singing backup vocals:

Isolating the top voice in the I-vi-IV-V chord progression.

Like I said, pretty boring—and not particularly melodic, either. A better approach is to vary the voicings of the chords so the harmony parts don’t have to move in parallel. For example, you might keep the C chord in its normal root position but change the Am to the first inversion (C-E-A), the F to the second inversion (C-F-A), and the G to the first inversion (B-D-G), like this:

Inverting the chords to vary the internal voicings.

Not only does this make the chord progression easier to play (all the notes are closer together on the keyboard) it also makes any individual part easier to sing. Take the top note part again; instead of moving G-E-C-D as it did originally, it now moves G-A-A-G, like this:

The new top voice harmony part, thanks to revoicing the chords.

Now the voice is fairly consistent; it doesn’t jump all over the place like it did before. And if you check out the other voices, you find that they’re also a lot more singable. (The middle voice moves E-E-F-D, and the bottom voice moves C-C-C-B.)

When you write out each of these parts separately, you use three different staves, like this:

Three-voice harmony—on three different staves.

These principles apply when you’re playing piano accompaniment; they also apply when you’re writing vocal or instrumental harmonies. If you vary the voicings, you open up a lot of possibilities as far as which voice goes where.

Swapping a few notes between the top two harmony parts.

See what we did here? We swapped the second and fourth notes between the parts, so that the first part now goes G-E-A-D, and the second part goes E-A-F-G. You’re still representing all the notes in the chord, but you’re making each individual line more melodically interesting.

A good tip when you’re creating either vocal or instrumental harmony is to physically sing each part yourself. If the part is boring or hard to sing, consider different inversions or swapping notes between parts. The best harmony parts sound great on their own!

Voice Leading

Voice leading is what you get when you follow one harmony part from start to finish; the different intervals between the notes follow a set of conventions and act to create a pseudo-melody out of the harmony line. You have to make sure that one note properly leads to the next to avoid having the harmony line sound like a bunch of totally unconnected tones.

When you’re writing harmony, there are three key voice-leading conventions to keep in mind. When you follow these conventions, the creation of harmony parts becomes somewhat easy:

• In general, voices should move smoothly. That means moving each voice the shortest distance possible, and retaining common tones between successive chords in the same voice. How big of a leap is too much? Try to avoid leaps larger than a fourth, except in the lowest (bass) voice, which can accept larger leaps that move along with the chord progression.
• When at all possible, you should avoid moving all the voices in the same direction. If some of the voices move up over the course of a line, at least one voice should move down.
• Avoid moving voices in parallel “perfect” intervals—fourths, fifths, or octaves. Voices can move in parallel thirds and sixths.

Here’s an example of bad voice leading. Notice how the two parts move in parallel, and how each voice has huge leaps from note to note?

Bad voice leading …

The second example shows how to fix the voice-leading problems merely by swapping a few notes from voice to voice or by picking other notes within the chord. It isn’t that hard when you realize that any note can be assigned to any voice; all you have to do is pick the right notes to create the best-sounding musical lines.

… corrected.

Making a Point—with Counterpoint

Counterpoint is the art of combining two or more simultaneous musical lines. Unlike the art of harmony, where the harmony parts are subsidiary to the main melody, both melodic parts in counterpoint have equal weight.

The two voices in traditional counterpoint must not only sound melodic when played separately; they also must fit together harmonically to suggest the chord structure of the song. That means if you play the underlying chords against the counterpoint, the melodic lines mustn’t sound dissonant; it also means you should be able to deduce the underlying chord structure from the melodic lines alone.

Classic two-part counterpoint, courtesy of Bach—notice the independence of the two melodic lines.

It’s also important that both lines be singable. That means either melody should be able to stand on its own as a main melody against the underlying chord progression. A melody in counterpoint should not be just a combination of notes to fit the chords; it has to be melodic, it has to have its own internal musical logic, and it has to fit with the other melody.

When composing melodies for counterpoint, call upon the skills you learned back in Chapter 8. Make sure each melody has a shape and a destination, as well as a coherent form. Also make sure each melody makes sense within the song’s general chord structure, it fits within a comfortable vocal or instrumental range (generally no more than an octave range), and the intervals within the melody are small and singable.

When you put your two melodies together, you should make sure they both fit within the notes of the underlying chord, and stay within the confines of the underlying scale. The two melodies should also not duplicate each other; if you have two identical melodies, you’re writing in unison, not in counterpoint. That means each melody should have its own motion—which should complement, but not interfere with, the motion of the other melody. The notes within each melody also should not interfere or clash with the notes in the other melody; avoid dissonant vertical intervals.

It’s okay for the two melodies to have their own rhythmic patterns. In fact, in classical music it’s expected. (If your two melodies have identical rhythmic structure you have what is called a 1:1 rhythmic ratio, which in some ways is easier to work with than more rhythmically complex forms.)

Creating Your First Counterpoint

The best way to learn about counterpoint is to dive into the deep end—and start composing!

For our first counterpoint, we’ll start by defining some parameters. These are not necessarily hard-and-fast rules for the contrapuntal form, but they do make it easier for beginners to create a working counterpoint.

Here are the parameters:

• Use a 1:1 rhythmic ratio; both melodies should use the same rhythmic patterns.
• Only consonant harmonic intervals are allowed; no dissonant harmonies.
• The lower part must begin and end on the tonic of the key.
• The upper part must begin on either the first or the fifth of the key, and end on the tonic—either in unison with the bass part or an octave above.
• Between the melodies, rely heavily on intervals of thirds and sixths. Avoid octaves and unisons, except for the final note.
• Move each voice the shortest distance possible.

Now let’s get started.

Some composers like to write both melodic parts simultaneously; this enables them to employ various advanced techniques, such as sharing a contrapuntal melody between the two parts. However, for beginners it’s a lot easier to write the first part first and then complete the puzzle by adding the counterpoint melody. Which is exactly what we’ll do.

We’ll start from the bottom up and create our first melody for the lower voice. (This is the staff labeled “2,” in bass clef.) Our melody is in 3/4 time, in the key of F Major. Because counterpoint works best when both parts have a lot of rhythmic and melodic movement, the melody is fairly lively, with a lot of eighth notes; it also starts and ends on the tonic note, F.

Now things get interesting. We need to come up with a second, higher, melody that doesn’t clash with the lower melody—in fact, it should complement the first melody. You can’t create this second melody in a vacuum; counterpoint is more than just putting two unrelated melodies together. The two melodies have to fit together logically, and have to work together to suggest the underlying chords.

The lower voice—the first of two parts in counterpoint.

So let’s start at the start. The lower part started on an F, which is the tonic of the underlying key. According to our parameters, the upper melody can start either on the tonic or the fifth. We don’t want to get too fancy with our first counterpoint, so we’ll start the upper melody (the treble clef labeled “1”) on the F two octaves higher.

In the first measure, the lower melody has a slight upward progression. To distinguish the upper melody from the lower part, we’ll give this second melody a slight downward progression. Remember, we want the intervals between the two voices to be pleasant-sounding, which means emphasizing the thirds and sixths, and avoiding dissonant intervals. We also want to avoid parallel movement, so we’ll leave the second note of the upper melody on F, which sounds good against the D on the bottom. (It’s a nice minor third.)

For the third note, it’s time to introduce some motion—in this case, a downward motion, to contrast with the upward motion in the lower melody. We’ll move down from F to C, which is a nice singable fourth. It also forms a pleasing minor sixth against the E in the lower melody.

The fourth and final note in the first measure also needs to show downward movement, so we’ll continue down from C to A, an easily singable minor third. The A also creates a pleasing major third against the F in the lower melody.

Now, if you look closely, you’ll see that not only did we introduce contrary motion in the upper melody (down, in contrast to the upward-moving lower melody), but we also used the upper melody to suggest the underlying chord (F Major). The downward melody is actually an arpeggiated F Major chord, in its first inversion (F to C to A). What we have so far is shown here:

Starting to add counterpoint to the first voice.

As you can see, creating counterpoint requires a lot of logical thinking—it’s a lot like thinking through the moves in a game of chess. The conventions are there and it’s all very logical, but you have to work hard to make it all fit together as it should.

We’ll continue adding to the upper voice line, following the conventions note by note. When we’re finished, we have two complete melodies, in counterpoint to each other, like this:

Your first counterpoint!

Your first counterpoint is a good example of simple 1:1 rhythmic counterpoint. Of course, classical counterpoint requires that the rhythms differ between the two voices, so let’s continue to work with this example to create some rhythmic variations.

How do you introduce different rhythms into the top melody? In this example, you can do it by essentially scooting all the notes to the left one beat. That is, you put the eighth-note figure on beat one instead of beat two of each measure, and extend the measure-ending quarter note into a half note.

Here’s what this looks like:

Your first counterpoint—but with different rhythms in the top melody.

Note that this approach created eighth notes in the top part when the bottom part is using quarter notes, and vice versa. This provides a simple rhythmic variation without going whole-hog on complex interrelated rhythms.

What to Avoid

Whether you’re writing harmony or counterpoint, there are certain combinations of notes that you want to avoid. In most cases, you’ll avoid these combinations because they don’t sound right; still, it’s good to know the conventions in advance so you can keep from making rookie mistakes.

Here’s what to avoid:

Extended parallel movement. Parallel movement is when two or more voices move in the same direction by the exact same interval. Parallel movement is frowned upon in some types of music—but perfectly acceptable in others. For example, in classical music you want to avoid parallel movement in perfect fourths or fifths and octaves; in popular music, however, parallel fourths and fifths are permitted and even somewhat common—even if they do sound a little boring if carried on for too long. In any case, if you try hard, you can generally come up with something that sounds more interesting than extended parallel voices. Go ahead and use a little parallel movement if you have to, but then vary things to keep from getting boring.

Big leaps. Whether you’re writing harmony or counterpoint, you should follow one of the key conventions used when creating melodies—keep the intervals between notes as small as possible. Avoid big leaps between notes; they not only sound disconnected, they’re also hard to play and sing. As with standard voice-leading, any interval larger than a fourth or fifth is probably too big a leap.

Dissonant intervals—unless they resolve. In this case we’re talking about the intervals between different voices—and this is as much a practical consideration as anything else. For example, if you’re harmonizing an F Major seventh chord, don’t put the seventh and the root (E and F) together on close voices. Instead, put one of the notes in a lower voice, so there’s an octave or so of space between the two notes. That said, it’s okay to introduce this kind of dissonance, if you then resolve it. That means you can include dissonance in passing tones, or even at the end of phrases, but only if one of the voices then resolves to a more pleasing interval. For example, in the F Major seventh chord, the voice with the E might resolve to F.

Exercises

Exercise 15-1

Find and correct the voice-leading errors in the following piece of harmony.

Exercise 15-3

Create a second melody part in counterpoint to the following melody.

Exercise 15-4

Based on the following chord progression, create a two-voice counterpoint.