The idea of studying by creating is one that has been central to tonal Western art music for centuries. For as long as the tradition has existed, composers, performers, critics, and even casual enthusiasts have engaged in compositional exercises as a means of deepening understanding and appreciation. Although there are many valid ways to approach the study of this music, we feel that creating it oneself, following guidelines derived from observing compositional practice, is one of the most immersive and exciting paths.
Many of the exercises in this book will ask you to write or complete chord progressions of various lengths in four voices. This type of activity is particularly well suited to the study of tonal Western art music because it strikes a balance between the melodic and harmonic aspects of polyphonic music. When completing such an exercise, you must consider the musicality and independence of four individual voices or parts (hence “part-writing”)—one soprano, one alto, one tenor, and one bass (hence “SATB”). But you must also consider how the voices combine to form harmonic sonorities and how the music shifts from each sonority to the next.
This type of writing has a lot of moving parts, so to speak, and you must track all of them to be successful in completing these exercises. Keeping track of so many things can, naturally, be a bit overwhelming at first. So, our approach here is to introduce elements of the practice gradually. Many exercises will ask you to write just one or two chords, focusing on specific techniques or aspects of the part-writing process. In these cases, there is often just one or two solutions that adhere to constraints dictated by the tradition. Other exercises are more open-ended, giving you a greater degree of flexibility—and thereby offering more opportunity for creativity—in how you realize a given progression. For these lengthier exercises, the following guide to part-writing practice and proofreading will be invaluable and should be frequently consulted.
Historically, part-writing exercises are based in large part on the compositional practices of Johann Sebastian Bach, specifically in the 400 or so SATB chorale harmonizations he wrote in the 18th century. Countless other pieces by other composers share many of the characteristics found in these works and so, despite Bach’s unusually prominent reputation, we may view these exercises as representative of a much larger tradition. Within this body of work, however, one finds as much variety as commonality and so it would be impossible for a guide such as this to capture every aspect of compositional style in tonal Western art music.
We encourage you, then, to think of this guide not as a set of rules—or, worse, laws—that govern how all tonal music works or should work. It is better understood as a window into a particular mode of thinking about and creating one style of music. The guidelines described here are far more strict than one actually encounters in historical compositions. Bach himself, for example, occasionally wrote parallel fifths in his part-writing—the afterword of Fundamentals, Function, and Form describes one such case—and countless other composers took an even looser approach. The idea here is to narrow your focus on specific aspects of musical composition. Remember as you compose your own music or study the music of others that, more often than not, art does not confine itself to any rigid set of constraints.
General characteristics and guidelines
When completing a part-writing exercise, one must pay attention to many different things. The shape of each individual melody, the way the melodies interact, and the structure of the chords that result from the combination of voices all require careful planning. With so many things vying for your attention, part-writing exercises can feel a bit like juggling! The following sections of this guide break down the general characteristics and guidelines for each of these considerations.
The melody found in each individual part—soprano, alto, tenor, and bass—of a part-writing exercise should be singable. These exercises are, after all, rooted in vocal music. It is highly recommended that you sing each voice out loud, both while working on the exercise and especially when you have finished. If a melody is overly difficult to sing or even has just one or two small problem spots, you should consider revising.
Melodies should be mostly conjunct—that is, consisting primarily of stepwise motion. Held notes—pitches that are sustained through one or more chords—are very easy to sing and are therefore very desirable in this style. Intervals larger than a minor or major second are perfectly acceptable, of course, provided the melody does not become too jumpy. In general, it is considered good practice to balance out larger leaps—leaps greater than a third—with stepwise motion in the opposite direction.
Some intervals should be avoided entirely. Tritones, sevenths, and intervals larger than an octave are difficult to sing and tend to stand out, making the overall combination of voices sound unbalanced. Diminished and augmented intervals—such as the augmented second between scale degree [latex]\hat6[/latex] and the raised leading tone in a minor key—should also be avoided.
One should also be mindful of tendency tones. As the following example illustrates, certain scale degrees have a tendency to resolve by step in predictable directions:
Scale degrees [latex]\hat2[/latex], [latex]\hat4[/latex], and [latex]\hat6[/latex] tend to resolve down by step to [latex]\hat1[/latex], [latex]\hat3[/latex], and [latex]\hat5[/latex], respectively. This is not to suggest that these scale degrees must resolve in the prescribed fashion. Rather, you should be aware that resolving a tendency tone will have one effect and avoiding the expected resolution will have another.
There are two types of tendency tones which should be resolved in a predictable way. Scale degree [latex]\hat7[/latex], the leading tone, has an even stronger tendency to resolve up to [latex]\hat1[/latex]. When the leading tone appears in a dominant chord—V(7) or viio(7)—it almost always resolves by step up to the tonic. Occasionally it will leap down to scale degree [latex]\hat5[/latex] instead, but this should be restricted to cases where it appears in an inner voice (the alto or tenor). The other type of tendency tone that should be resolved consistently is not a scale degree but a chord member. Chordal sevenths, notes that form a dissonant seventh above a root (or a dissonant second if the chord is inverted), should always resolve down by step.
Rhythmically, part-writing exercises tend to be very clear and simple. For the most part, all voices follow the same rhythm and the duration of each note tends to be equal to—or longer than—the duration of a single beat in the meter.
As much as you must consider the musicality of each melodic line in a polyphonic exercise, you must also consider the counterpoint resulting from how each voice interacts with each other voice. An abundance of contrary motion between voices will help project a sense of independence between the four parts, but you should strive for a variety of motion types by including parallel, direct, and oblique interval progressions as well.
The use of standard interval progressions should be your primary concern here. The majority of interactions between voices in a part-writing exercise should be comprised of standard interval progressions. (See Chapter 12 and Appendix A of Fundamentals, Function, and Form for detailed descriptions of standard interval progressions.)
Of course, other interval progressions are found in this style of music as well and may include what we refer to as “resultant intervals” in Chapter 14 of Fundamentals, Function, and Form. Unlike the intervals used in standard interval progressions, resultant intervals may include dissonances. They arise when two voices separately form standard progressions with some third voice but not with each other, and therefore appear only in textures with three or more voices.
Some interval progressions, on the other hand, are strictly avoided in this style of music: parallel unisons, parallel fifths, and parallel octaves:
Pitches forming unisons, fifths, and octaves blend together so well it can become difficult to distinguish between the two voices. When two or more of these intervals are used with parallel motion, the effect is more like a single, strengthened melody than two unique voices. Standing out in the texture in this way, parallel unisons, fifths, and octaves undermine the independence of the voices and should therefore be considered impermissible.
Parallel motion from one fifth to another is considered acceptable in this style, however, when the second fifth is diminished:
The reverse progression—a diminished fifth leading to a perfect fifth—is typically avoided. (See Example 17-7 in Chapter 17 of Fundamentals, Function, and Form for further discussion.)
Note that guidelines concerning forbidden interval progressions may vary. Many instructors, for example, may advise you to avoid consecutive fifths by contrary motion since such progressions are effectively equivalent to parallel fifths with an octave displacement:
Some instructors advise their students to avoid direct motion to a perfect fifth or octave. This type of progression is sometimes referred to as “hidden fifths” or “hidden octaves” because in historical performance practice a singer might embellish their part by adding passing tones to fill in a melodic leap, thereby creating a parallel interval progression:
Generally speaking, direct motion to a perfect fifth or octave is acceptable when it involves either of the inner voices. When it involves both outer voices, it is best if the soprano moves by step. Students working with an instructor should seek clarification on which progressions to avoid.
Keeping track of the contrapuntal relationship between two voices is hard enough, but in an SATB setting there is a lot more happening. With four voices, there are six unique voice pairs: 1) soprano against alto, 2) soprano against tenor, 3) soprano against bass, 4) alto against tenor, 5) alto against bass, and 6) tenor against bass. When completing a part-writing exercise, it is important that you be mindful of each of the six pairs of voices.
At its core, a part-writing exercise is a four-part realization of a chord progression. The notes that appear among the four voices, in other words, are determined by a given series of Roman numerals or other chord symbols. To spell a chord, you must determine the letter name (e.g., F or Eb) of each chord member. Familiarity with different chord types, then, is a necessary prerequisite.
The key of a particular exercise is indicated by the key signature as well as in letter-based shorthand below. You can recognize the key name at the beginning by the colon (“:”) which appears after the letter or accidental. Uppercase letters represent major keys and lowercase letters represent minor keys. The shorthand “G:” stands for “G major:” while “f#:” stands for “F-sharp minor:” and so on. Note that some exercises modulate, changing keys one or more times before they reach the end. This is represented by an asymmetrical bracket with a new key name and Roman numerals written on a lower level:
In most cases, the Roman numeral represents the scale degree of the chord root in the key at hand. In G major, for example, a IV chord will have C as its root since C is scale degree [latex]\hat4[/latex] in that key. The quality of the chord is indicated by the case of the Roman numeral as well as any additional symbols appearing to the left or right. (See Chapter 13 of Fundamentals, Function, and Form for review with Roman numeral representations of triads and Chapter 18 for the same with seventh chords.)
Pay careful attention to the quality of each chord, particularly in minor keys. One of the most common errors students make in chord spelling is forgetting to raise scale degree [latex]\hat7[/latex] in minor-key dominant chords. In the following example, a V7 chord in G minor must have F# instead of F§ to match the case of the Roman numeral:
Chord position—the indication of which chord member appears in the lowest voice—is indicated by bass figures. These stacked Arabic numerals are written to the right of the Roman numeral. The tables below summarize the common figured bass representations of triads and seventh chords in various positions:
|position:||root position||first inversion||second inversion|
|common abbreviation:||(no bass figures)||6||6/4|
|position:||root position||first inversion||second inversion||third inversion|
|common abbreviation:||7||6/5||4/3||4/2 or 2|
It is often wise to start by writing out the bass line since that part is restricted by the provided progression.
A triad consists of a root, a third, and a fifth. In part-writing exercises, the various notes of each chord are performed by the soprano, alto, tenor, and bass. This raises a question that must be answered time and time again when realizing a chord progression: How should one distribute the three notes of a triad among the four voices of the exercise? The matter of deciding which chord member should appear in two (or more) voices is referred to as chord doubling.
The preferred doubling in a triad is generally determined by the position of the chord. Since the soprano, alto, and tenor tend to work together as a unit against the bass, it is desirable to have all three chord members represented in the upper voices. In other words, doubling whichever chord member appears in the bass is often a good idea. First-inversion chords, however, offer a bit more flexibility. When the third of a chord appears in the bass, any of the chord members may be doubled. The following table summarizes:
|root position||double the bass (the root)|
|first inversion||double any chord member|
|second inversion||double the bass (the fifth)|
There are several additional restrictions to keep in mind as well. The leading tone should never be doubled, regardless of where it appears in the chord. It has such a strong tendency to resolve to the tonic that a pair of leading tones suggests parallel octaves even if they resolve to different scale degrees. A V(7) chord, then, should never have a doubled third and a viio(7) should never have a doubled root. Chromatic chords like augmented sixth sonorities also have strong tendency tones which should not be doubled. (See Section III of Fundamentals, Function, and Form for guidance on voicing and resolving various types of chromatic harmony.)
Occasionally, a chord member will be tripled. In the resolution of V7, for example, it is common for the tonic resolution to have three roots, one third, and no fifth. (See Chapter 19 of Fundamentals, Function, and Form.) In other cases, a chord may have two roots, two thirds, and no fifth. This is sometimes found when a chord shifts positions (e.g., I moving to I6). The fifth is generally considered the least essential member and for this reason is sometimes omitted.
Seventh chords have four chord members: the root, third, fifth, and seventh. This works out nicely in a four-voice setting since each voice can be assigned a different chord member. To avoid voice-leading errors, a seventh chord may occasionally be written without its fifth. As with triads, the root is generally doubled in these cases. Note that since chordal sevenths have such a strong tendency to resolve down by step, they should never be doubled since, like a doubled leading tone, it would imply parallel octaves.
The way the individual notes in a chord are distributed among the voices and laid out on the staff is called “chord voicing.” The part-writing exercises in this book consist of four voices: soprano, alto, tenor, and bass. To save space, all four parts are traditionally written on a grand staff. The soprano and alto occupy the upper (treble) staff while the tenor and bass share the lower (bass) staff. Stem directions are used to indicate which notes belong to which voice:
Each of these voice types has a standard range of about one and a half octaves. Soprano notes range from C4 to A5, alto notes from F3 to D5, tenor notes from C3 to A4, and bass notes from F2 to E4:
These ranges are, of course, approximate. Different singers have different ranges. Some sopranos, for example, can sing up to C6 or even higher! Nonetheless, for part-writing exercises you should restrict each of your notes in the four parts to their respective standard ranges.
In addition to making sure that each voice stays within its respective range, you should also ensure that the four parts remain in the same vertical order: soprano above alto, alto above tenor, and tenor above bass. When chord voicings do not follow the typical ordering of voices—soprano above alto, alto above tenor, and tenor above bass—it is known as a “voice crossing.”
Voice crossings are found in this style from time to time, but only rarely and usually to avoid a more glaring problem. Note that voice crossings are generally easy to track between the soprano and alto or tenor and bass, since these voice pairs share a staff. They can be harder to notice between the alto and tenor, however, since these voices are written on separate staves:
You should also be mindful of the intervals found between the voices in each chord. When considering the spacing of a chord, it is helpful to think of the upper voices—the soprano, alto, and tenor—as forming a unit separate from the bass. Neighboring upper voices are rarely more than an octave apart. Again, careful attention is required to avoid spacing errors between the alto and tenor.
Unlike the spacing of the upper voices, the tenor and bass are often more than an octave apart since the bass note is determined by the position of the chord.
Basic part-writing strategy
This book includes several different types of part-writing exercises (see below), but they are all variations on the same basic activity. We recommend taking the following steps to convert an abstract series of chord symbols into a fully-realized four-voice progression.
Collect your materials: Part-writing exercises often involve a bit of trial and error. Having a few sheets of staff paper handy will be very helpful. Use a sharp pencil with a good eraser. Be prepared to identify and correct mistakes.
Spell each chord: Determine the letter names of each member of each chord in the progression. You may find it helpful to work with an extra sheet of staff paper, writing the chords in simple form first before voicing them in four parts. Write the key name and signature under a single staff as well as the provided Roman numerals. Write each chord in a simple, visually intuitive format. Do not worry about voice-leading yet. This sheet will simply serve as a reference later on, a catalog of notes to be included in each chord. Here is an example for the progression i–i6–iiø6/5–V–i in C minor:
Write the bass line: Since bass notes are pre-determined by the Roman numerals, write the bass part first. Keep within the standard range for a bass voice and follow the melodic considerations described above as closely as possible, but remember that bass lines are often more disjunct than the other parts. Here is one possible bass line for the progression shown above:
Add the upper voices: After writing out the bass line, you could complete each of the upper voices, one after another, but we recommend working on all three upper voices simultaneously. Work carefully, chord by chord, from beginning to end. Follow the guidelines listed above and check for errors as you go. It is much easier to fix problems when they appear than to go back and make corrections later on. Here is a full realization of the progression shown above:
Proofread: Checking for errors is an essential part of the process. Plan on setting aside plenty of time for proofreading, especially if you are new to this type of exercise. Use the proofreading checklists below to help ensure that your work is problem-free in every regard. Address any issues by rewriting the chord(s) in question and note that fixing a problem often requires rewriting everything from that point until the end—or at least rearranging the voices. Once you finish with any changes, proofread the whole progression once again.
Although the different types of part-writing exercises contained in this book are all variations on the same basic idea, each provides you with different preliminary information and each has its own idiosyncrasies. Below, you will find detailed directions and strategies for each of the three types: figured bass realization, Roman numeral realization, and melody harmonization.
Figured bass realization
In this type of exercise, you need only complete the soprano, alto, and tenor parts since a fully realized bassline is provided. Below the given bass line, you will see a series of Arabic numerals and accidentals. These symbols, combined with the corresponding bass note, provide all the information needed to determine the chords to be used at any given moment. The first step, then, is to convert the figured bass line to Roman numerals. Here is a quick guide (see Chapter 21 of Fundamentals, Function, and Form for a more detailed explanation):
- Arabic numerals under a bass note indicate the position of the chord according to the tables shown above. A bass note with 6/5, for example, is the third of a seventh chord in first inversion.
- A bass note with no accompanying Arabic numerals is the root of a triad.
- A numeral with an accidental (e.g., [latex]\hat6[/latex]# or b[latex]\hat3[/latex]) indicates that the voice forming that interval above the bass should be raised or lowered accordingly. (Note that in some cases raising or lowering a voice may cancel out part of the key signature. In these cases, the accidental in the figured bass may not match the corresponding accidental on the staff.)
- A numeral with a slash through it indicates that the voice forming that interval above the bass should be raised one semitone with an accidental.
- An accidental with no numeral attached alters the voice that forms a third above the bass.
The following example shows realizations of several figured bass notes as well as their corresponding Roman numerals:
To complete a figured bass exercise, first find and write out the corresponding Roman numerals for each chord. Then, follow the basic part-writing strategy described above, starting at Step 4.
Because the bass line is already set, figured-bass exercises are the most restrictive type of part-writing activity included in this book since one part is locked in and cannot be changed to avoid voice-leading problems. But, for the most part, this also makes figured-bass exercises the easiest type of part-writing. More restrictions make for fewer decisions and fewer decisions mean less risk of errors. The biggest difference between this type of exercise and those described below is the added task of converting the figured bass line into Roman numerals.
Roman numeral realization
In this type of exercise, you are provided with a key, a starting chord, and a series of Roman numerals. Here you will need to realize all four voice parts. To complete a Roman numeral realization, follow the basic part-writing strategy described above.
Compared to figured bass or Roman numeral realization activities, this type of exercise is much more open-ended. Here you are presented with just a soprano melody and you will need to determine how you want to harmonize it. The chords used to harmonize each note must be selected carefully, keeping in mind the common chord progressions found in this style of music.
Take the following melody, for example:
We recommend following these steps to harmonize it:
Consider the key: Write the scale degree of each melody in the key at hand. This will help you identify potential chords in the next step. This melody is in C major, so the G at the beginning is labeled [latex]\hat5[/latex]. The next note, B, is the leading and is labeled [latex]\hat7[/latex]. We continue in this manner until every note is labeled:
Consider the possibilities: Write out the Roman numerals for every chord that may be used to harmonize each note in the melody. Each soprano pitch can act as the root, third, fifth, or even seventh of a chord. Note that some possibilities will make more sense than others. Regardless, don’t worry about picking chords just yet. For now, simply catalog the full list of possibilities for each note. In this case the first note is a G which, in C major, may serve as the root of a V chord, the third of a iii chord, the fifth of a I chord, or the seventh of a vi7 chord:
Plan the beginning: Since most phrases begin with a tonic, you will likely want to start your harmonization on a I chord. Since this melody starts with scale degree [latex]\hat5[/latex], we will begin with a tonic triad:
Plan the ending: Determine the type of cadence you would like to have at the end of your phrase. Depending on the melody, this will most likely be a half or authentic cadence. Consider, too, the lead-up to the cadential dominant. Is it possible to include a cadential 6/4 or a pre-dominant chord in the cadence? In this case, the melody will allow for a pre-dominant ii chord as well as an authentic cadence with a cadential 6/4:
Plot a course through the middle: Consider the various types of progressions one commonly encounters in this style of music. Does this progression have a place for a passing 6/4 chord? How can you prolong the tonic at the beginning of the phrase? The diagram shown in Example 24-22 in Chapter 24 of Fundamentals, Function, and Form is a useful tool for plotting idiomatic chord progressions. (Hint: You can also use the progressions provided in other exercises as models.) In this case, we can prolong the initial tonic with a deceptive gesture on beats two and three of the first measure:
Write out the bass line: Once you have settled on a sequence of chords, write in the bass melody. Keep an eye out for any counterpoint errors such as parallel fifths or octaves. In some cases, forbidden parallels can be avoided by inverting one or both of the chords. In other cases, you may need to select a new chord altogether. Here is a bass line to go along with our melody:
Fill in the alto and tenor parts: Follow the voice-leading conventions described above as closely as possible while you add the inner voices. Here is our finished progression, now with inner voices:
Proofread: As always, check for errors as you go and when you are done. Make the necessary changes and proofread one more time.
As we have indicated in the preceding sections, proofreading is essential to good part-writing. Checking for voice-leading errors will immeasurably enhance your ability to write music in this style. Desirable results, however, do not come easily and you should expect this part of the part-writing process to be fairly time consuming. Although slow at first, you will eventually find the process becoming faster and faster as you become more and more familiar with common patterns and voice-leading formulae.
As with any music theory assignment, you should also attempt to perform your work in some way. Try singing through each line independently. Try playing through the progression on the piano. Music notation software is also useful for quickly converting notation to sound. Listening to the notes you write is, perhaps, the most efficient way to both check the musicality of your composition and internalize the underlying concepts in a meaningful way.
The following checklists will help you avoid common errors. Use them repeatedly as you write and revise your part-writing. When you finish an exercise, go back and systematically check each note, melody, interval progression (in all six pairs of voices), chord, and overall shape.
Melody (check each voice part individually):
- the melody in each part is singable and musical
- melodies are primarily conjunct (stepwise)
- larger melodic leaps are balanced with stepwise motion in the opposite direction
- melodies contain no tritones
- melodies contain no diminished or augmented intervals
- leading tones in dominant chords—V(7) or viio(7)—resolve up by step, especially in the bass or soprano
- chordal sevenths resolve down by step
Counterpoint (check all six voice pairs):
- counterpoint comprises mostly standard interval progressions
- voices that form resultant intervals form standard interval progressions with some other part
- counterpoint includes no parallel unisons
- counterpoint includes no parallel fifths
- counterpoint includes no parallel octaves
- counterpoint includes no d5–P5 progressions (P5–d5 is permissible)
- instances of direct motion to a perfect fifth between soprano and bass have soprano moving by step
- each notated pitch is a member of the chord at hand
- scale degree seven is raised to form a leading tone for dominant chords—V(7) or viio(7)—in minor keys
- chord member in the bass corresponds with the position—root position or inversion—of the chord at hand
- root position chords double the bass (the root)
- second inversion chords double the bass (the root)
- chords contain no doubled leading tones
- chords contain no doubled sevenths
- chords contain no strong (chromatic) tendency tones
- soprano notes are all within the C4 to A5 range
- alto notes are all within the F3 to D5 range
- tenor notes are all within the C3 to A4 range
- bass notes are all within the F2 to E4 range
- voice crossings are avoided (hint: take special care in checking for voice crossings between the alto and tenor)
- neighboring upper-voice pairs (soprano and alto or alto and tenor) are never more than an octave apart
A time-saving tip
Nothing is more frustrating in completing a part-writing exercise than to think you have finished only to find a glaring error. Re-writing the offending chord often requires further revision as the new voicing creates new problems with each chord that follows. In many cases, however, you do not need to abandon all your hard work.
Take the following progression, for example:
As you can see, there are parallel octaves between the soprano and bass at the end of the first measure. To solve this problem, we could revoice the V chord by putting the C in the soprano, F in the alto, and A in the tenor. However, revoicing only the V chord leads to a new problem since it will create parallel fifths between the soprano and alto at the beginning of the second measure:
Since the voice-leading in the rest of the passage was fine in the first place, we can keep the individual melodies in mm. 2-3 intact. We simply rearrange them among the voices. The following example takes the soprano melody in mm. 2-3 and puts it in the alto part. It also takes the alto and tenor melodies and puts them in the tenor and soprano parts, respectively:
The passage is now error free and we did not have to rewrite mm. 2-3 from scratch. With a bit of rearranging—and some scrap paper and a good eraser—we were able to save ourselves a lot of extra work. Keep this strategy in mind whenever you need to revise your part-writing.
A note to instructors
There are many different ways to evaluate student part-writing exercises and the specifics of assessment are a matter of personal preference. The following mark-up symbols may be used to highlight errors in student part-writing exercises:
|a curved, vertical line may be used to highlight a spacing error (e.g., alto and tenor are more than an octave apart)|
|a pair of straight lines may be used to highlight objectionable parallels (e.g., parallel fifths or parallel octaves)|
|a straight arrow may be used to highlight some voice-leading issue within a single voice (e.g., a melodic augmented second) and may be paired with another arrow to highlight voice-leading issues between two voices (e.g., hidden fifths)|
|the abbreviation “sp.” above a chord may be used to indicate the presence of a spelling error (e.g., a problematic doubling or omission of a chord member)|
a circled note may be used to highlight a more specific chord spelling issue (e.g., a failure to raise the leading tone in a minor-key dominant)
The following example shows these symbols in use in a sample exercise: