(Pagodas)
(Here are my thoughts about playing this piece. Please add your thoughts in the comments at the bottom.)
Estampes is considered by many to be the point of departure for Debussy’s exploration of new sounds for the piano. Before 1903 Debussy’s piano music is very beautiful, but (as Ravel pointed out) contained little that is new. After 1903 we have a completely new approach to the piano, leading to the Images and Preludes. Pagodes starts this exploration by using strong influences from oriental music, particular the Javanese gamelan.
Debussy saw a Javanese gamelan ensemble perform at the Paris International Expositions of 1889 and 1900. Debussy was to refer to this experience for the rest of his life, clearly indicating its influence on his musical style. In a letter to the poet Pierre Louys in 1895 (reprinted in Debussy Letters), Debussy said “Remember the music of Java which contained every nuance, even the ones we no longer have names for. There tonic and dominant had become empty shadows of use only to stupid children.” Writing as a music critic (quoted in Roberts, pp. 156-157) Debussy said that the complex polyphony and counterpoint of Javanese music made “Palestrina seem like child’s play”, and speaking of European percussion in contrast to the Gamelan, Debussy said “you have to admit that ours is no more than the primitive noises of a traveling circus.”
Pagodes is clearly written in imitation of this gamelan experience, though it is solidly in a European form both structurally and harmonically. The piece is basically percussive and resonant, like the gamelan, and contains many pentatonic scale structures to emphasize the oriental flavor. These are admittedly tricks that in a lesser hand would have given us a rather trivial cliche-laden piece. But of course Debussy uses these tricks to create and explore a beautiful sound. In later pieces, such as the prelude Canope, Debussy took the Javanese influence in a far more mature, and less imitative, direction.
What Pagodes is not is a homage to Javanese music or oriental music in general. Beyond the musical tricks described above there is no sign of actual oriental musical style. Debussy never saw a real pagoda, and it is not clear that he knew that there are no pagodas in Java. But this does not matter. Pagodes is an integration of oriental music into Debussy’s style, based on a deep respect for Javanese gamelan music in particular.
Paul Roberts has a thorough and enlightening discussion of of both the background and interpretation of this piece in his book Images, which was my primary source for this page.
Emotional Content: To me, this piece is one of the less emotional of Debussy’s piano pieces, even though it is not just abstract music and shows real passion. It is mostly quiet, exploring resonant layers of sound in a somewhat meditative way. At the same time it has something of a classical European structure, with a good old-fashioned climax in the middle and near the end.
Shape and Flow: The structure of this piece can be seen in both terms of layers and sections. There are three major themes: the pentatonic melody that first appears in m. 3 and is repeated in variation throughout the piece, the melody that first appears in octaves in m. 11 and develops into the theme in the left hand at m. 37, and the melody that appears in m. 33. There are also subsidiary motifs: the middle melody in m. 7, which reappears in m. 84, and the chords starting in m. 15. These themes and motifs weave in and out in various layers over the resonant pedal point of the bass notes and chords. This structure is reminiscent of gamelan music, with complex melodies playing over deep, resonant gongs.
At the same time there is a sense of sections, with section A extending from m. 1 to m. 22, an interlude from m. 23 to 33, section B from m. 33 to 40, the climactic section C at m. 41. Then the piece roughly has the structure ABCBACD, with D being the ending section.
As Roberts points out (pp. 289-292), this piece “offers the pianist some essential lessons in touch and pedaling…Unless the pianist draws myriad colors and textures from the instrument, the piece will not hold the listener’s attention, but will lapse into repetitive prettiness.” While prettiness is good enough for some of us, this piece is an opportunity to discover more. The problem is: how to find these myriad colors in the piano. From this perspective we can understand why most of the Pagodes needs to be very quiet, allowing the subtlety of touch and resonance to bring out the many sounds the piano can make. In sections like the opening (and in many other places in Pagodes) the chords in the left hand excite overtones that are reinforced by the right-hand melodies. It is this interplay between overtones and melody that gives the piece its unique effect. But overtones are very quiet, so to hear them we must play quietly in these sections. The sustain pedal must be held for long periods to allow the overtones to ring. Pagodes gives us an unusually explicit lesson in how Debussy gets his unique sound.
Even though the pedal is used heavily throughout, a careful technique must be used to maintain transparency and clarity. The percussive nature of most gamelan instruments inspires us to us a light, percussive staccato touch even when the pedal is held down. As Roberts says, we need “a refined touch for creating a variety of simultaneous dynamic levels, and, in conjunction with the sustaining pedal, a rapid release of the key – in the manner of gong-playing – the moment the sound is created.” Roberts says this gong-playing technique is “in reality a private illusion” because the piano action releases the note as soon as the string is struck no matter how the key is released. I’m not so sure this is purely illusory: while it is true that how you release the key does not matter, thinking “gong-like” causes me to strike the key somewhat differently, with a higher velocity but still soft. This produces a clearer, more bell-like tone.
Disclaimer: I am just now studying Pagodes for the first time. It is difficult and I have not mastered it. So what’s below may change and should be taken with a grain of salt.
The piece is marked “fairly lively”. This tempo should be maintained throughout with the exception of the occasional ritard. There should be no rubato or sloppiness in the tempo.
Opening: This section is marked “2 ped” (one of the very few pedal markings in Debussy’s piano music), indicating that both the soft (una corda) and sustain pedals should be held down. Clearly the sustain pedal is to be used heavily throughout the opening, the questions is how to maintain clarity. Elder (p. 236) tells us that Gieseking taught that the sustain pedal should be held without change for mm. 1-6, changed for m. 7, held without change for mm. 7-10, changed at m. 11 then held again for mm. 11-14. Debussy is explicit that the two pedals should be held without change for mm. 11-14. This indicates to me that we have a little more freedom to change the pedal before m. 11, but not for mm. 11-14, and that the soft pedal should be held all the way to m. 16.
mm. 1-2: The timing in these measures should be absolutely strict and carefully counted. Debussy is unusually explicit in marking which hand plays which chord: the left hand plays the first, the right the second and the left hand the third chord (the first in the treble) to be played with the left hand. While this is obviously not technically necessary because of the sustain pedal, playing the third chord with the left hand will facilitate a more gong-like tone.
mm. 3-6: Marked “delicately with hardly any nuance”, these measures should be played very evenly and transparently, with the melody not rising much above the chords. If you change sustain pedal at the beginning of m. 5, be sure to hold the low F-sharp from the previous measure during the change, and similarly for later measures.
mm. 7-10: These measures are much like those before, with the addition of a middle melody. Again, the playing should be even. The idea is to create equal layers of sound. Note the rhythmic difference in the low F-sharp at the end of mm. 8 and 10, which I think subtly increases the dynamics without increasing volume.
mm. 11-14: The volume level is increased to piano, but we are explicitly told to keep the soft pedal down. It is tempting to increase the volume as these measures are played, but I don’t think this is what Debussy had in mind. In another unusual pedal marking we are told to keep the sustain pedal down without change throughout these measures.
mm. 15-22: For these measures I release the soft pedal. The primary challenge of these measures is to keep the D-sharp-C-sharp triplet figure as regular as possible in its various forms and registers while playing the melodies and bass over it. Note that there is a small crescendo marked in m. 18, but the following measure is again piano. Most recordings I’ve heard have a steady increase in volume through these measures, reading the mark in m. 18 as a crescendo to the dynamic level of m. 19. But I read this as a small crescendo in m. 18 with a return to piano at m. 19, which builds the real but small crescendo in m. 21. m. 19 is marked “a little animated”. I think the effect of this reading is quite nice, with a mini-crescendo at m. 18, backing off to a dynamical restart at piano in a more animated feel at m. 19 leading to the crescendo in m. 21, rather than one continuous build from m. 17 to m. 21.
mm. 23-30: Marked “Still lively”, these measures keep the momentum from before but return to pianissimo. I use the soft pedal throughout here. In mm. 27-30 we are again instructed to hold down the soft and sustain pedals without change through the end of m. 31, where I also release the soft pedal. These measures are very technically challenging to perform at speed and pianissimo while allowing enough transparency for the left-hand melody.
mm. 31-36: These measures contain an exquisite melody against a syncopated background. Roberts (p. 292) says that these measures need “the greatest care, with the pedal sustaining the soft lower gongs without in any way affecting the clarity of the melody. The effect must be of a pedaled left hand and an unpedaled right (a tall order…).” There is a nice rhythmic thing at the beginning of m. 35. If you change pedal inside m. 35, Roberts suggests holding the D-sharp in the left hand throughout the measure, giving a bell-like effect.
mm. 37-40: Here the right hand is the background layer and so should be somewhat quieter than the left hand, which carries the melody. The high C-sharp in m. 38 should be clear, sounding like a third instrument (bell) added to the ensemble, as opposed to a chord in the melody.
mm. 41-44: These measures are a straightforward climax and should be loud. Keep the timing strict in the syncopation re-introduced in m. 44.
mm. 45-52: In mm. 45-49 the effect should be of three instruments: the melody in the left hand, a chime in the middle, and a different chime playing the two notes at the top. All very rhythmic and pianissimo. mm. 50-52 are somewhat rhythmically involved and should be in strict time (’till the retenu, of course). Again the feeling should be an ensemble of instruments.
mm. 53-77: This is a straight reprise of the opening sections, leading directly into the reprise of the climax in m. 73. The only extra twist is the maintaining of the D-sharp-C-sharp triplet figure throughout the climax measures in mm. 73-77. Note that the dynamic climax is maintained throughout mm. 73-78. Elder (p. 237) quotes Gieseking as saying that we should catch the low D-sharp in m. 75 and hold the sustain pedal from there through m. 79, changing at the beginning of m. 80.
mm. 78-79: These measures start ff and diminish all the way to pp.
mm. 80-98: Here we have a non-stop, rhythmically complex glissando based on the opening melody, played against a recall of some of the themes in the piece. These measures are entirely pianissimo, I play them with the soft pedal. These measures are very difficult to perform at tempo and pianissimo. I find that mastering mm. 84-87 helps me get the rhythm of the glissando. Other than that I don’t know of any tricks that can help.
mm. 1-2: The timing in these measures should be absolutely strict and carefully counted. Debussy is unusually explicit in marking which hand plays which chord: the left hand plays the first, the right the second and the left hand the third chord (the first in the treble) to be played with the left hand. While this is obviously not technically necessary because of the sustain pedal, playing the third chord with the left hand will facilitate a more gong-like tone.
mm. 3-6: Marked “delicately with hardly any nuance”, these measures should be played very evenly and transparently, with the melody not rising much above the chords. If you change sustain pedal at the beginning of m. 5, be sure to hold the low F-sharp from the previous measure during the change, and similarly for later measures.
mm. 7-10: These measures are much like those before, with the addition of a middle melody. Again, the playing should be even. The idea is to create equal layers of sound. Note the rhythmic difference in the low F-sharp at the end of mm. 8 and 10, which I think subtly increases the dynamics without increasing volume.
mm. 11-14: The volume level is increased to piano, but we are explicitly told to keep the soft pedal down. It is tempting to increase the volume as these measures are played, but I don’t think this is what Debussy had in mind. In another unusual pedal marking we are told to keep the sustain pedal down without change throughout these measures.
mm. 15-22: For these measures I release the soft pedal. The primary challenge of these measures is to keep the D-sharp-C-sharp triplet figure as regular as possible in its various forms and registers while playing the melodies and bass over it. Note that there is a small crescendo marked in m. 18, but the following measure is again piano. Most recordings I’ve heard have a steady increase in volume through these measures, reading the mark in m. 18 as a crescendo to the dynamic level of m. 19. But I read this as a small crescendo in m. 18 with a return to piano at m. 19, which builds the real but small crescendo in m. 21. m. 19 is marked “a little animated”. I think the effect of this reading is quite nice, with a mini-crescendo at m. 18, backing off to a dynamical restart at piano in a more animated feel at m. 19 leading to the crescendo in m. 21, rather than one continuous build from m. 17 to m. 21.
mm. 23-30: Marked “Still lively”, these measures keep the momentum from before but return to pianissimo. I use the soft pedal throughout here. In mm. 27-30 we are again instructed to hold down the soft and sustain pedals without change through the end of m. 31, where I also release the soft pedal. These measures are very technically challenging to perform at speed and pianissimo while allowing enough transparency for the left-hand melody.
mm. 31-36: These measures contain an exquisite melody against a syncopated background. Roberts (p. 292) says that these measures need “the greatest care, with the pedal sustaining the soft lower gongs without in any way affecting the clarity of the melody. The effect must be of a pedaled left hand and an unpedaled right (a tall order…).” There is a nice rhythmic thing at the beginning of m. 35. If you change pedal inside m. 35, Roberts suggests holding the D-sharp in the left hand throughout the measure, giving a bell-like effect.
mm. 37-40: Here the right hand is the background layer and so should be somewhat quieter than the left hand, which carries the melody. The high C-sharp in m. 38 should be clear, sounding like a third instrument (bell) added to the ensemble, as opposed to a chord in the melody.
mm. 41-44: These measures are a straightforward climax and should be loud. Keep the timing strict in the syncopation re-introduced in m. 44.
mm. 45-52: In mm. 45-49 the effect should be of three instruments: the melody in the left hand, a chime in the middle, and a different chime playing the two notes at the top. All very rhythmic and pianissimo. mm. 50-52 are somewhat rhythmically involved and should be in strict time (’till the retenu, of course). Again the feeling should be an ensemble of instruments.
mm. 53-77: This is a straight reprise of the opening sections, leading directly into the reprise of the climax in m. 73. The only extra twist is the maintaining of the D-sharp-C-sharp triplet figure throughout the climax measures in mm. 73-77. Note that the dynamic climax is maintained throughout mm. 73-78. Elder (p. 237) quotes Gieseking as saying that we should catch the low D-sharp in m. 75 and hold the sustain pedal from there through m. 79, changing at the beginning of m. 80.
mm. 78-79: These measures start ff and diminish all the way to pp.
mm. 80-98: Here we have a non-stop, rhythmically complex glissando based on the opening melody, played against a recall of some of the themes in the piece. These measures are entirely pianissimo, I play them with the soft pedal. These measures are very difficult to perform at tempo and pianissimo. I find that mastering mm. 84-87 helps me get the rhythm of the glissando. Other than that I don’t know of any tricks that can help.
Playing Debussy's Piano Works 