Line by line and passage by passage the poem comes to the poet from sources he feels strongly but does not pretend to understand. At least at the beginning of the poem.Very soon however, whatever comes to him from the je-ne-sais-quoi had better start coming in response to what he has already written down or he is not going to have a poem. Maybe the first line comes from nowhere. And maybe the second comes from a second-theme nowhere. But whatever part of what follows continues to come to him from his nowhere, it has to come from those first lines as well. The poem, that is, is forever generating its own context. Like a piece of music, it exists as a self-entering, self-generating, self-complicating, self-resolving form.

—JOHN CIARDI, HOW DOES A POEM MEAN?

Having worked our way from idea, to phrase, to section, in the next several chapters we will apply what we have learned to a discussion of the listening principles at the heart of some of classical music’s most fundamental forms. In considering these forms, I am not interested in musical taxonomy—laying out and labeling each “species” of form with an accompanying schematic diagram—but rather in getting at the different kinds of musical logic these forms contain and the different kinds of listening they require. Learning to define a form is much less important than learning how to experience it. How does the experience of listening to a fugue differ from the experience of listening to a sonata or a minuet? What do composers want us to hear? To use Ciardi’s language, how do these different forms “mean”?

The concept of form is not only at the core of musical meaning, but it is also an intrinsic part of the very definition of composition. The dictionary defines compose as “To put together, put in proper order or form.” The word itself comes from joining com (with) and poser (to place):“to place with.” Exploring the different ways musical ideas are “placed with” other musical ideas so that they create “proper order or form” will be the focus of these chapters. To quickly establish some basic listening principles, I want to first look outside the world of classical music at one of the clearest and most ubiquitous forms in all of popular music: the thirty-two-bar song form.

It is an article of faith in almost every critical discussion of artistic form that form and content are inseparable. Form, the argument goes, is not simply an existing mold into which new content is poured, but rather something that is shaped in each case by the particular details of the work in question. Because the content of every piece is unique, all forms are, in a sense unique, and the listener’s job is to hear how each particular form grows out of its particular content. Though all of this is absolutely true, it slightly oversimplifies the situation. Every sonata of Haydn’s is unique, and it is almost impossible to find one that fits any textbook definition of sonata form, while the enormous variety of formal structures in the fugues of Bach has led critics to stop even referring to fugue as a form, calling it instead “a way of writing.” However, though it may be true, for example, that what is valuable about any particular haiku is the unique way its individual thought is expressed—if in three unrhymed lines of five, seven, and five syllables—it is actually the complex interplay between what is fixed and what is free that is interesting: the way each individual haiku finds its own unique shape within a seemingly rigid, external mold. And the process by which each haiku constructs its particular seventeen-syllable shape, word by word, is its form. Form is happening at every moment, as every word is chosen. Form is an action. Form is a verb.

Popular Music

Perhaps the closest musical parallel to the haiku’s rigid form and clearly defined interplay between what is fixed and what is free is the form at the heart of the Broadway musical: the thirty-two-bar song form. The overwhelming dominance of this form in the repertoire of the musical is astonishing. Though no one I know of has ever done any statistical computations, I would venture to guess that the thirty-two-bar song form provides the structure, with slight variations, of nearly 90 percent of all theater songs. It is as standardized as a haiku or a sonnet, and along with the twelve-bar blues, it is one of America’s most enduring musical forms. It is also an amazingly courageous and self-confident one. In the standard version, twenty-four of the thirty-two measures consist of the same music. That in and of itself is an extraordinary fact. Seventy-five percent of the music consists of a single eight-measure melody repeated three times. In a conventional Broadway song, beginnings literally are everything. Let’s walk through the form and see how it works.

“I Got Rhythm” Redux

Since the opening idea is the key to the entire song, the form invariably starts with a catchy idea, eight measures in length, like the opening of “I Got Rhythm,” which we looked at in chapter 2. Call the opening idea A.

061

062

063

The form then repeats the music of A with different words (“I got daisies, / In green pastures, / I got my man, / Who could ask for anything more?”), followed by a contrasting section of eight bars, call it B. (“Old Man Trouble, / I don’t mind him, / You won’t find him, / ’Round my door.”) The standard form finishes by repeating the opening A music a third time, yet again with new words. (“I got starlight, / I got sweet dreams, / I got my man, / Who could ask for anything more?”) In a slight variation on the form,“I Got Rhythm” adds an extra two measures to finish the song by wittily repeating the words “Who could ask for anything more?” The complete standard thirty-two-bar form is: A (eight measures), A (eight measures), B (eight measures), A (eight measures).

Let’s look closely at some key listening principles embedded in this clear, simple form. The poet John Ciardi said,

A poem may well be conceived as a machine for making choices. At any given point in the poem, the poet must select the next thing to do. He must choose a word, an idea, an image—all these together. He must choose from the total language one word and not another.

Though the image of a machine perhaps conveys a more mechanical, automatic, deterministic quality than is appropriate, the idea is directly applicable to musical form. A musical form is also a “machine for making choices” and a standardized, highly structured form like the thirty-two-bar song form is a “machine” that influences (without determining) every musical choice, from first note to last. As in a haiku, there is a subtle interplay between what is fixed and what is free. Though each theater song shapes its particular content to meet the form’s structural requirements in its own unique way, simply knowing that your opening idea must be exactly eight measures long—not four, six, or ten—acts like a filter to exclude an enormous number of musical possibilities. It tends to lead to songs (like haikus) with short, striking ideas rather than ideas that develop gradually over time. Unlike the openings we looked at in chapter 2, popular-song openings like the Gershwins’, do not ask the listener to wait. They instantly create character and atmosphere and unfold their musical stories within clear, self-contained, eight-measure units. Like haiku, the thirty-two-bar song form is not for novelists or essayists, but for aphorists: musical quick-sketch artists like the Gershwins.

Let’s use “I Got Rhythm” as a way of getting at the listening principles built into this form. We have already seen in chapter 2 how the syncopated rhythm that opens the song immediately hooks the listener and is then repeated just enough times to become a pattern, set up an expectation, and prepare us for the surprise punch line (“Who could ask for anything more?”). We also saw how the melody’s four simple pitches (C, D, F, and G) played forward, backward, and then forward again, make the punch line’s new melody surprising and witty as well. Since all form is ultimately about how phrases are “placed with” other phrases, it is always important to listen closely at cadence points to hear how phrases connect—how seams are made. Notice the elegant way that the ending of the tune leads back to the beginning with three jazzy “turnaround” beats at the end of measure 8. (The term turnaround refers to the way the chords “turn around” the end of the phrase so it can repeat itself.) Remember this ending when we come to the final phrase of the song.

After the A section, the Gershwins repeat the music with new words. Repeating something is always a key form-defining moment, and repetition is at the heart of nearly all large-scale form. As we will see in later chapters, the element that is repeated may be different in different forms (e.g., a phrase, a fugue subject, a bass line, a harmonic pattern, a theme group, a rondo tune, an exposition, etc.), but some repeated element is a core organizing principle in nearly all forms. Here, repeating A lets us know that A is a unit. Repetition allows the listener to retrospectively group the opening eight measures into a single idea.

“Home”-“Away”-“Home”

What happens next exhibits another fundamental principle of form: contrast. Since three of the four phrases in a thirty-two-bar song form repeat the same music, the only contrast in the entire form is its eight-measure B section. In the Gershwins’ B section, not only is the melody completely new (though it does keep the A section’s syncopated rhythm), but the harmony is as well. The chords are completely different from those in the A section. They’re dynamic, unstable, and restless. The fundamental purpose of this B section is summed up by its letter designation: it is “not A.” To be “not A” is a fundamental principle of form. Contrast on one level provides welcome, Stravinsky-like “variety” to balance the song’s overwhelming “unity,” but more important, B allows A to seem new and fresh when it returns. Leaving A—“home,” the song’s core tune—generates the desire to hear it again. This movement of “home-away-home” is basic to a wide range of forms. It is, for example, the essential principal behind all rondos. Though a rondo by Beethoven, Brahms, or Antonín Dvořák might have a “home” theme that is longer and more complex than that of“I Got Rhythm,” and might move “away” to elaborate B, C, and D sections with transitions and developments to provide contrast, on the most fundamental level,“I Got Rhythm” and the rondo share the same core aesthetic. Put simply, rondos, like thirty-two-bar Broadway songs, are “about” their opening tunes. When we are “home” listening to that tune, we are structurally content. When we move “away” and leave the tune, as interesting and exciting as the contrasting sections (called episodes in a rondo) might be, on the most fundamental level we are always waiting to return “home,” to the rondo tune. The longer we wait, the more intense the desire to return, and playing with listener impatience is the source of all of the witty passages of preparation in rondos (called retransitions) that delay this key moment. Leaving generates the desire to return in “I Got Rhythm” as it does in any classical or contemporary rondo, but it creates a problem for the Gershwins and every other Broadway song-writer when they try to end their songs.

Cadences Are Everything

As I mentioned earlier in the chapter, since all form is about how phrases are “placed with” other phrases, it is always important to listen closely at cadence points to hear how phrases connect—how seams are made. In the thirty-two-bar song form’s AABA structure, the ending of A must perform three very different functions. The first time through, it must lead seamlessly back to the beginning of the tune. The second time through, it has to lead to the B section, while the third time through, it has to conclude the entire song. Remembering our discussion of the different degrees of finality of various cadences, in a thirty-two-bar song form, the same cadence must be a comma, a semicolon, and a period. In order to solve this formal problem, composers have come up with a remarkable variety of solutions. Some composers, like Irving Berlin with “Cheek to Cheek,” repeat the melody absolutely identically all three times and simply alter the accompaniment underneath to create the punctuation the moment demands.

064

065

066

Some composers, like Richard Rodgers with “I Wish I Were in Love Again,” subtly alter the melody at each of the three endings, and the three different vocal cadences beautifully shape the structure of the entire song.

067

068

069

The Gershwins, like Berlin, use the same vocal ending to return to the opening as well as to lead to the B section, while simply altering the accompaniment underneath to make the two different seams. However, the brilliant solution for the ending is all the Gershwins’. In the last A section, they alter the melody for “Who could ask for anything more?” so that it ascends this time, and they put a questioning harmony underneath. (A deceptive cadence to delay the final cadence and make it more conclusive.) They then add two extra measures to the form (eight, eight, eight, ten) to emphatically resolve the whole piece by repeating the words “Who could ask for anything more?” with the original vocal ending. A surprising yet utterly perfect conclusion.

Ex-Post-Facto Forms

One of the main reasons I have begun our discussion of form with the thirty-two-bar song form is because it is so clear and well defined, and one of the reasons it is so clear and well-defined is because composers were consciously aware that they were using it. Like someone writing a haiku today, the Gershwins, Berlin, and Rodgers knew the structure of the form before they began composing. Most musical forms, however, were “defined” after the fact by theorists looking back and generalizing from earlier practices. As we will see in chapter 11, the three greatest composers of Classical-period sonatas—Haydn, Mozart, and Beethoven—had no idea they were writing in sonata form. The textbook descriptions we know today were created by nineteenth-century theorists looking back on these masterpieces and were intended to be guides to composers for the production of future sonatas. Though these textbook forms are often highly inappropriate as descriptions of Haydn’s, Mozart’s, and Beethoven’s sonatas, they are almost perfect guides to sonatas written after the Classical period, when composers began to follow these ex-post-facto descriptions. (Charles Rosen’s two books, The Classical Style and Sonata Forms, contain insightful, comprehensive discussions of nineteenth-century descriptions of sonata form.)

Detailed forms with specific, spelled-out structural requirements, like the thirty-two-bar song form, are rare in music history. Baroque composers did not consult a textbook for the structural plan of a chaconne, a passacaglia, or a fugue before they wrote one any more than Beethoven looked up the plan for a sonata. What they did was look at other composers’ music, and what they discovered there were not strict forms with fixed rules but storytelling patterns and procedures that arose naturally. These patterns were then imitated, varied, and developed from composer to composer and from piece to piece as the material demanded. When similar musical situations arise over and over again, it is not surprising that recurring plot solutions and variations on these solutions tend to evolve over time, and invariably theorists have come along to generalize from these similar solutions and turn them into abstract forms. Yet though two pieces of music might share similar or even identical external forms, it is the measure-by-measure way each piece traces that form that counts. Saying that two theater songs both begin with eight-bar A sections is a starting point for discussing their content but nothing more. Form here is not a noun—the diagram of a shape (AABA)—but a verb: the experience of tracing that shape. To make this crucial point in the clearest way possible, I would like to look briefly at an absolutely perfect, thirty-two-bar song form piece written approximately a hundred years before the term was even invented: Schumann’s famous “Träumerei” (“Dreaming”) from Kinderszenen.

070

071

“Träumerei”

Although there are enough wonderful details in this piece to fill an entire chapter, I want to focus solely on the ones that show the enormous range of musical possibilities offered by a single form. First let’s make sure the overall form is clear. In “Träumerei,” measures 1 through 8 are the opening, eight-bar A section of the piece. The double dots at the end of measure 8 are a sign to repeat the whole opening, and this repeat makes up the second eight-bar A section. Measures 9 through 16 are the contrasting, eight-bar B section, and measures 17 through 24 make up the final, eight-measure A section. “Träumerei” is an absolutely perfect thirty-two-bar song form: A (eight measures), A (eight measures), B (eight measures), A (eight measures).

The differences between Schumann’s and the Gershwins’ approaches to the form begin with the opening phrase and reflect two very different concepts of what constitutes a musical idea. The first four notes of “I Got Rhythm” immediately give us both the signature syncopated rhythm that dominates the entire song, and the four-note melodic idea that generates all but the punch line of the opening melody. Nothing about the Gershwins’ opening asks the listener to wait for even a second. In “Träumerei,” however, though the meaning of Schumann’s opening idea dominates his entire composition as well, it will take the whole piece for that meaning to unfold.

Since this idea is the key to the song, let’s look closely at its subtle construction. Rhythm is as central to this idea as it is to the Gershwins’ but in a completely different way. Schumann’s opening starts with three and a half beats of “settling in” music as the melody (the upbeat and downbeat are indicated by “And-1” on the music) waits for the chord underneath to fully come into play. The melody then rises in a graceful five-note arpeggio to an F (remember this F for later) and repeats the note with a beautiful held chord underneath to finish the idea. (Note the lovely quick grace notes in the left hand to emphasize the arrival.) If the Gershwins’ opening rhythm “has rhythm,” Schumann’s subtle rhythm “dreams.” A normal version of Schumann’s opening would shorten the “settling in” process by one beat so that the long second chord could arrive on a downbeat, like this.

072

But Schumann’s second melody note “dreams” for an extra beat, which causes the second chord to arrive poetically on a weak beat in the middle of measure 2. Once Schumann reaches his high F, the rest of the idea gradually winds its way back down to complete this phrase in measure 4 and begin a consequent phrase in measure 5.

Berlin’s, Rodgers’s, and the Gershwins’ opening eight-measure phrases were all constructed in classic musical-comedy, unidirectional fashion to lead to “punch lines” in their final measures: “Who could ask for anything more?” (Gershwins), “When we’re out together dancing cheek to cheek” (Berlin), and “I wish I were in love again” (Rodgers). Schumann’s opening has a completely different kind of construction. It is a classically balanced, four-measures-plus-four-measures, antecedent-consequent phrase, and the entire consequent phrase is a beautiful variant of the antecedent. Measure 5 begins by repeating the “settling in” gesture of measure 1 (And-1) and then starts to repeat the graceful five-note arpeggio. But this time, in a classic compared-to-what moment, the music exquisitely arches up higher, to an A instead of an F, while the striking “dissonant” chord underneath the A has the first accidental (note outside the key) in the piece. Because the arpeggio this second time has risen higher, the route back down is different, and the details repay careful listening.

Having finished the first eight-bar A section, like the Gershwins, Schumann repeats it, but the purpose of Schumann’s repeat is completely different from the Gershwins’. In a popular song, everything is dependent on the opening idea. As Richard Rodgers said, “If the song is successful, it’s the idea that you walk out whistling.” On the most basic level, in a theater song, all of the repetitions of the opening idea, not only within the song itself but in the overture, entr’acte, scene-change, and incidental music are designed to imprint it in every listener’s ear. Schumann, however, is not repeating the opening eight measures so that the listener can whistle them, but so that he can retrospectively group the measures into a single unit and remember that unit as something to be developed and varied as the piece moves forward.

“Away”-“Home”

We saw that the Gershwins’ eight-measure B section functions primarily as contrast, a move “away” from A so as to generate a desire to hear it again. The melody in the Gershwins’ B section is new, though it keeps the A section’s catchy signature rhythm, while the harmony is more restless. Schumann’s B section (beginning in measure 9) also provides contrast, and has more restless harmony, but that is only the beginning of its story. It starts with the “settling in” music and the graceful ascending arpeggio, as if to start a third repeat of A. We have already heard this arpeggio ascend to an F and to an A, and the repeat has cemented these two versions in our ear. The B section now introduces yet a third version, which this time ascends to an E-FLAT in measure 10. (As always, the sound is what’s important. Think of the letter names as labels.) Then, once we have gotten this third version in our ear, he transposes it in measure 13, using it to make the second half of the B section. At the same time, Schumann is not just quickly whisking through some restless chord changes like the Gershwins do in their B section but actually changing keys twice; first to G minor in measure 12, then to B-flat major in measure 13. And then, just on the verge of arriving in yet a third key (D minor in measure 16) at the last instant, with a ritardando (an indication to slow down) marked so the listener will not miss the moment, he shifts direction and magically arrives back “home” in F major to begin the final A section.

By now the listener has already had to process an enormous amount of sophisticated musical information in a short space of time, but Schumann has saved the best for last. Measure 17 begins the return of A, and the entire antecedent phrase is repeated exactly in measures 17 through 20. The whole climax of the piece, however, turns on an unbelievably subtle yet exquisite reharmonization. We come to our graceful ascending arpeggio for the final time. We have already heard four different versions. The melody now floats up to an A exactly as it did in measure 6, but Schumann completely changes the harmony underneath, discovering a miraculous, otherworldly chord somewhere in dreamland that transforms the color of this A. (Schumann writes a fermata to make sure we luxuriate in this incredible chord.) And then, as if that were not enough, he rewrites the final descent back home from this climax, with a last cadence that, like the Gershwins’, sounds utterly right, but is truly surprising.

My purpose in looking at these two very different treatments of the same rigidly defined form was the hope that it would serve as a kind of cautionary example before we begin to explore other forms in upcoming chapters. In the end, all listening is particular. Though a form like the thirty-two-bar song form may exist as an abstract scheme in a textbook, what counts is the way it is lived piece by piece. Schumann’s complex developmental approach to this thirty-two-bar form offers a completely different (not better or worse) kind of listening experience than “I Got Rhythm.” Yet both pieces are shaped, in very different ways, by their 8-8-8-8, AABA formal structure. In the end, what is important is the listener’s ability to enter into the world of each piece and hear how its contents shape its particular form. A form may be a “machine for making choices,” but the machine is wielded by human hands in unique ways. Measure by measure, phrase by phrase. Form is happening at every moment, as every note is chosen. Form is an action. Form is a verb.