by David Schiff
E7 augmented ninth
B9 fifth
E major 9 third
D 9 sus fourth (sus)
A minor 7 seventh
F minor 7 flat 5 minor ninth
B 7 sharp 5 sharp fifth
C1 3 sharp eleventh
A7 thirteenth
If you sing the pitch G against each of the chords you will naturally find yourself adjusting the intonation in response to the varying degrees of tension between the root of the chord and the position of the G and you will hear how Mingus's melody is a sustained keening wail on E and G
HOMMAGE À RAVEL
Play “Satin Doll” as a slow waltz and it will suddenly sound like a valse sentimentale of Ravel. Though his name rarely appears in jazz theory books, you can hear the influence of Ravel clearly in the music of Billy Strayhorn (“Chelsea Bridge”) and Bill Evans (“Waltz for Debby”). Ravel prophetically employed a jazzlike harmony based on seventh chords in some of his earliest works, like the song “Sainte” and the “Pavane pour une infante défunte.” By 1903, with the orchestral song cycle Shéhérezade, Ravel had mapped out all the harmonies that Burt Bacharach and Stephen Sondheim would use seventy years later; Sondheim's 1986 musical Into the Woods sounds Ravelian from beginning to end. I even hear Ravel (via Miles Davis and Gil Evans) in Steve Reich's Music for 18 Musicians. You would have to go back to Corelli to find a harmonic style with a comparable impact.
Ravel reached his full harmonic maturity in Valses nobles et sentimentales, published in 1911. A century after that scandal these eight waltzes remain a living thesaurus of harmonic devices that stay just within the boundaries of functional harmony and yet employ a wide variety of dissonances. You can hear Ravel's relevance for jazz most clearly in the fifth waltz of the set, written in an exquisitely spiced E major. The eight-bar phrase beginning at measure 17 sounds like the bridge of a popular song, and with good reason. The harmonic progression is a ii-V-I in F# major that is then repeated, transposed up a step, in A This is exactly the same progression found in the bridge of “Satin Doll” and many other jazz standards.
Ravel's way of creating dissonances appears in a nutshell in bars 17 and 18. The music is written in a three-voice texture. If you just play the outer voices you'll see that they are not particularly dissonant in relation to each other. In the first bar only two of the six melody notes are dissonant, and both fall on unaccented parts of the beat. In bar 18 Ravel repeats the melody exactly but changes the bass to imply a harmony a perfect fifth lower. This device (often used by Debussy as well) turns consonant notes into dissonances, placing the dissonances on accented beats. Adding the inner voice you will see that it creates dissonances at nearly every point where the upper voice was consonant. This means that there is some kind of dissonance present on eleven out of the twelve eighth notes of these two bars. Yet the only egregiously illegitimate-sounding dissonance in both bars falls on the last eighth note where the right hand has a C## and E# against a D# in bar 17 and a C# in 18. In general Ravel is as scrupulous in resolving dissonances as were Bach and Chopin, though at times he was far more devious than his predecessors.
A different technique equally pertinent to jazz appears in the first eight bars. The melody decorates the pitch G# while the bass moves downward. The G# sounds dissonant against the A in the bass, more dissonant when the bass move down to F#, and more dissonant still when the bass moves down to a D. The bass seems to outline a D major triad that fits with the C and D in the inner voices but clashes with the G# and E# in the melody. At the end of the third bar, though, everything changes. The C and D slide up a half step, the bass moves down to B, and the harmony suddenly comes into focus as the dominant of E major. The deceptively dissonant G# turns out to have been the consonant element in a series of what theorists term appoggiatura harmonies, illusional chords created out of nonharmonic tones, dissonances substituting for consonances. In the next four bars the melody repeats but the harmony changes, again in a jazzlike way. For the second bar, which could have been harmonized with the dominant B7, Ravel used instead an altered E# dominant chord, in other words, the tritone substitute. No wonder jazz pianists pay attention.
Ravel's music converses with jazz because he thought like a jazz musician, from the bass up. We know this from the evidence of his sketches and also a few short examples of self-analysis.4 Though his harmonies shocked his contemporaries, Ravel arrived at them through a method he had learned at the conservatory. He habitually used the baroque shorthand of the figured bass, sketching the structural outer lines before adding inner voices. In other words, he started with a lead sheet. His music nevertheless sounds unbaroque because he often used types of parallel motion banned in older harmony but common in Debussy's music and because he did not always resolve a dissonance in an obvious way if he felt that it already implied its resolution. A dissonance could, in effect, substitute for a consonance.
Here are some highlights of harmonic devices in the waltzes:
I (G major). The opening measure flings harmonic provocations at the audience: a chord of six pitches (all the white keys except E) followed by an even more dissonant chord of five pitches, the apparently nonsensical piling of a d# minor 7 over a D. The next bar resolves these chords slightly with a consonant five-tone chord, which is followed, however, by the sharpest sound yet, an apparently unidentifiable combination of the pitches A#, C#, D#, and E. When you live with the progression for a little while, however, its shock value recedes to reveal a linear logic. The dissonant notes of each chord, once you figure out which they are, resolve by step, just as they would in Bach harmony, but the steps have been displaced up an octave so that, for instance, the C which seems out of place in the first chord, moves up a half step plus an octave to C# in the second chord and then comes to rest on D, back down a major seventh, at the beginning of bar 2. The opening phrase reaches its peak in three statements of a modern dominant, a stack of thirds from the root to the thirteenth, omitting only the third of the chord; jazz musician call this a “sus” chord. The harmony for the rest of this waltz mixes these newly spiced consonances with even more challenging combinations of notes that at first sound discordant. The most notorious progression occurs in the eight-bar passage leading back to the recapitulation. Here the bass line goes clear around the circle of fifths (and through all twelve tones of the chromatic scale). The bass powers high-voltage chords of clashing major and minor thirds, or major and minor sevenths. You can find just about every chord needed for jazz piano in this one phrase.
II (G minor). This slow waltz begins with a forecast of Billy Strayhorn's “Chelsea Bridge” and then goes into a gymnopédie-style homage to Erik Satie by using the Dorian mode instead of the usual minor. Although the main phrase sounds consonant compared to the din of the first waltz, it too predicts jazz practice by using a g minor 7 as the tonic and a d minor 11 as a combined tonic/subdominant. A closing phrase seems to wrap the Satie tribute in a warm embrace and pulls the waltz into the major mode. After a brief midsection (all the waltzes are in ternary ABA form), Ravel recasts the gymnopédie by immobilizing the bass line and compressing the major third (C-E) of the second harmony so that it sounds like a major second (C#-E). With this whole-tone harmony Debussy joins the party.
III (e minor-G major). The sound of the major second, introduced in the last third of the previous waltz, now defines the sonority of a whole movement. In his Jazz Harmony at the Piano, John Mehegan refers to Ravel-inspired voicings based on seconds rather than sevenths, and this waltz is one possible source. Its piquant quality is further enhanced by the use of Aeolian mode or natural minor, which gives a slightly bitonal feeling to the opening sixteen bars; the right hand sounds like it is in G major, but the left hand sits on the pitch E.
IV (A, vaguely). This waltz teeters on the edge of atonality by contrasting nonfunctional harmonic progression based on thirds with more traditional moves based on fifths. Its cadences in A, C, and E feel like arbitrary resting points, each one a plausible tonic. It
begins with what a jazz pianist might term a dominant thirteen with both a flat and augmented ninth, in other words, a blues dominant with both major and minor thirds.
V (E major). Already discussed, but note how Ravel here picks up the sliding triads from the closing phrase of Waltz II. These neighbor progressions set up a network of thematic connections but also anticipate the polytonality of Waltz VII.
VI (C major). We might call this “kitten on the keys,” if that title were not already in use. The wrong-note half-step melodic figure, played by the thumb, sounds accidental until we realize that it develops the sliding triad theme of the previous waltz. Again we hear second harmonies, but here Ravel develops a contrast of melodic minor seconds and harmonic major seconds into a blur that disguises a simple V-I progression in C major.
VII (A major). This is the most elaborate of the waltzes and the one that Ravel cited as most characteristic; he later incorporated parts of it into La Valse, the waltz to end all waltzes. The main tune appears as a thickened melody of three-note chords that eventually swells to become a melody of seventh chords. The ultimate harmonic shock comes in the middle section, which seems to superimpose a melody in E major on a bass in F. Ravel analyzed this passage as an example of “unresolved appoggiaturas,” which he illustrated, tellingly, using a figured bass and compared to the opening chord in Beethoven's Piano Sonata op. 31, no. 3.5 He may have been joking. Because the melody in the right hand sounds tonally close to the A major of the outer section, it feels consonant while the bass seems “wrong,” a half step too high. Not surprisingly, the surreal, dreamlike effect of this superimposition proved useful in Hollywood, but it also appears in several Richard Rodgers standards such as “Spring Is Here” and “The Sound of Music.”
VIII “Épilogue” (G major). A new melody, voiced in parallel triads, drifts slowly over a nearly static bass line that, halfway through, comes to rest on a low G. Between reorchestrated and reharmonized recurrences of the melody, reminiscences of all the waltzes flash by against a tolling B. The music comes to rest with the return of the mysterious passage from the end of Waltz II, cadencing, just as it would in jazz, not on a tonic chord but on a dominant ninth.
Ravel's harmonic discoveries did not end with Valses nobles et sentimentales. In later works he absorbed influences from Stravinsky, Schoenberg, Bartók, and Gershwin; but this was a kind of payback since he had already influenced all of those composers except for Schoenberg. Ravel's oeuvre provides models for harmonic idioms by composers from Bartók and Berg to Mingus and Monk that retain a functional bass line and a sense of progression while pushing the envelope of dissonance. I hear a particular connection between Ravel's sensibility and Billy Stray-horn's (though Strayhorn's distinctive harmonic style also resembles Alban Berg's at times). Strayhorn began “Lush Life,” composed in 1933 when he was eighteen, with an unusual harmonic progression, a major seventh on the tonic moving, out of the key, to a major seventh on the flat seventh degree, straight out of Ravel's Sonatine. “Chelsea Bridge,” with its parallel augmented eleventh chords, is perhaps the most obvious homage, but I also hear a tribute to Ravel in Strayhorn's own valse sentimentale, “Lotus Blossom,” Ellington's favorite composition by his alter ego.
GOING MODAL: DEBUSSY AND SHOSTAKOVICH
In 1959 Miles Davis's Mixolydian “All Blues” launched modal jazz, and just a few years later the Beatles gave popular music a modal makeover in “Eleanor Rigby” and “Norwegian Wood.” When music theorists speak about modes, they mean scale patterns that differ from the usual major and minor, although technically speaking those are modes also. The scale from D to D on the white keys of the piano is called Dorian; E to E, Phrygian; F to F, Lydian; G to G, Mixolydian; A to A, Aeolian. (The names of the modes derive from a misreading of Greek music theory, not from actual ancient Greek music.) If you build triads on the notes of each mode you will see that compared to the major scale they do not have major triads on the fourth and fifth scale degrees. They do not produce the usual IV-V-I progression needed to define the key. Harmonies based on modal scales either have a looser sense of harmonic progression or find different ways of producing tonal function. In “Lush Life” and in many Beatles songs the major triad a step below the tonic (the flat seven from the Mixolydian mode) substitutes for the usual dominant. In jazz the use of modal scales emerged from the rich harmonies of bebop, in which chords often contained five or six pitches. Thinking modally, jazz players treated the subdominant not as a 117 chord, which, by adding thirds, might contain (in C major) the pitches D, F, A, C, E, G, and B, but as the Dorian mode, which contains these same pitches. Modal harmony helped jazz musicians conceive their music more linearly, less chordally. Just as they had done in Impressionist music, the modes cleared the air.
In French music modal harmony came back to life with the three Gymnopédies of Erik Satie, composed in 1888. The first twenty bars of the first Gymnopédie employ only the pitches of the D major scale, but they don't sound like D major. The harmony rocks between two major seventh chords built on G and D; since there are no dominant-sounding chords we could be in either D major or G Lydian. Satie avoided major triads that would strengthen a sense of key; for a final cadence on D he used a minor chord as the dominant rather than the usual major triad or dominant seventh chord. In the third Gymnopédie the melodic line stays on the white keys throughout and the harmony falls on the white keys except for a few Bs, which support a general strategy of avoiding major triads. Even though the piece sounds harmonically simple, it uses only one major triad, an inverted one at that, and only one dominant chord, a dominant ninth on G, which does not resolve in tonal fashion to C. Dominant progressions made music march; Satie's harmonies allowed music to float. You could draw up a long list of his imitators, beginning with Debussy, who orchestrated the Gymnopédies, and Ravel, who imitated them in his Mother Goose as well as in the Valses nobles.
Debussy's most rigorously modal composition is the first movement, “Pour invoquer Pan, dieu du vent d’été,” of his Six épigraphes antiques. The music appeared as a piano duet in 1914, but it had been composed in 1900, at the height of Debussy's association with Satie, as (unpublished) background music for recitations of the faux Greek Chansons de Bilitis by Debussy's friend Pierre Louÿs. (Satie had established a precedent for using the “Greek” modes to illustrate Greek subjects in the Gymnopédies and Gnossiennes.) The movement uses only the notes of the G Dorian scale; its opening melody, presented unharmonized, is pentatonic, omitting the E and B of the mode. These excluded notes form the interval of the tritone, which would tend to focus the harmony on the tonality of F. The pentatonic subset of G Dorian, lacking that dissonant interval, is tonally ambiguous. It could be harmonized in C major, or F major, or any of the modes derived from those keys. Debussy, however, further limited the harmonic possibilities by allowing no chromatic alteration of the pitches of G Dorian. He nevertheless shifted the tonal center of gravity from one phrase to the next, so that at first we hear the piece in C Mixolydian, then B Lydian, then D Aeolian. Only at the end is there a cadence on G, and it is set up with a progression of major triads on B and C (III and IV) playing the roles of subdominant and dominant, respectively. Throughout the movement Debussy breaks rules of academic tonality just as he had begun to do back in his conservatory days, with parallel fifths and unresolved ninths, but the overall effect is of a rediscovered consonance, music minus anxiety.
There are many later examples of “white key” music in the twentieth century. Prokofiev in particular liked to give diatonic melodies his own personal stamp. In the first piece, a slow waltz, of his Visions fugitives, op. 22, written between 1915 and 1917, the melody sits entirely on the white keys except for a single D six bars before the end that is marked misterioso. Prokofiev harmonized the melody with a counterpoint of seventh chords that move in parallel but don't resolve or define a key. When the phrase repeats a new chromatic inner voice adds a little spice but no tonal direction. This time the phrase drifts downward to a b minor t
riad, which turns (by way of the logic of tritone substitution) into a kind of dominant. It allows the music to close on e minor, where, it turns out, it actually began.
The name of Keith Jarrett became associated with modal jazz in the 1970s especially thanks to the recording of his solo concert in Cologne. Not incidentally, Jarrett later recorded the complete Preludes and Fugues, op. 87 of Shostakovich, composed in 1950-51. It took the insights and devotion of a great jazz musician to unlock a work that was still a buried treasure for most classical pianists. This massive cycle is a summation of Shostakovich's musical universe and also a commentary, at times satiric, on many aspects of twentieth-century music, including Stravinsky's neoclassicism (Fugue no. 2 in A minor) and twelve-tone composition (Fugue no. 15 in D).
Jarrett's critics sometimes accused him of, in effect, rolling an orange on the white keys. The first fugue of the Shostakovich cycle can serve as a response to their accusations. It is a model of absolute white-key composition that shows how much musical coherence can be found in just seven pitches. Within that limitation Shostakovich followed Bach's practice closely. The fugue is in four voices. The subject is eight bars long and uses just six of the available seven pitches, so that it retains its exact intervals when transposed up a fifth to become the answer. The answer imitates the subject at the fifth, and there are two counter-subjects (counterpoints to the subject or answers that return rigorously). As in many Bach fugues there are also a couple of canons and pedal points, all part of the usual fugal machinery. Bach, however, would have constructed the harmonies of the fugue through a tour of related keys, going, say, from C major to G major, to a minor, to F major, and then back to C. In each of these harmonic regions he would use pitches from the chromatic scale not present in the C major scale. Shostakovich also visits different regions, but because he does not use any accidentals, they appear as modes. We hear statements of the subject in E Phrygian, B Locrian, A Aeolian, and D Dorian before we get back to C major, but Shostakovich saved the F Lydian for the end, where it seems to challenge the tonic key; the final cadence feels more suspended than resolved. Each of the fugue's modal regions has its own harmonic weight and flavor, so that the self-imposed harmonic constraints of the fugue feel like an expansion of resources. Only in the Lydian section, however, does Shostakovich allow modal logic, and the contrapuntal logic of a canon, to produce dissonant-sounding tritone clashes that threaten, in passing, the mood of calm objectivity.