By Carl Swanson
(This article originally appeared in the Journal of the American Harp Society Winter 2000 Issue, Vol. 127, No. 4.)

    It would be wonderful if printed music could tell us every nuance, every shift, every texture, that a composer intended for his piece. Or if all of the subtleties of a performance by a great artist, known for his interpretation of this or that composer, could be notated and passed down, right on the paper, from one generation to the next. How to make a particular crescendo or decrescendo. How much to linger over a note in a phrase.

    Unfortunately, musical notation is not that subtle, nor specific. And for this reason, we seek out the great artists of our generation, to learn that which cannot be written down, cannot be communicated by any means other than the oral tradition.

    Composers of the 18th century depended almost entirely on this tradition to communicate how their pieces were to be performed. Virtually nothing but the notes were written down, and even then, the rhythms were often notated differently from the way they were actually to be played. Louis Couperin, (uncle of the great Francois, and himself a terrific composer), says in his treatise ‘The Art of playing the harpsichord,’ “ It’s no wonder foreigners have such a difficult time playing our (French) music. It isn’t written the way it’s played.”

    It was only in the 19th century that composers began to concern themselves more extensively with interpretative details to help the performer. These dealt primarily with tempo and dynamics, and then only in a vague and haphazard way. By the early 20th century however, composers, and in particular the Impressionists, had become far more precise, indicating, as well as notation allows, how virtually every note of the piece is to be played. One can play Debussy’s Claire de Lune almost with a metronome, playing everything exactly as it says on the page, and the result will be the smoke and mirrors that we all know and love. But even for Debussy there were limits as to what could, and could not be notated. And because of this, the opening measures of his Dances are problematic.

    At first glance they appear to be nothing more than a series of chords, to be rolled in a harpistic way. But where is the melody? Is it the lowest note, or the highest? Does each arpeggiated chord start on the beat, and therefore continue after? Or does each chord anticipate the beat, so that the top note of each chord lands on the beat? And then, starting in measure 17- where, it appears, chords alternate with octaves-how does the player get all six beats to sound like an even melody line? More will be mentioned on these measures later on.

    Frequently when I hear the Dances performed, all rhythmic clarity is lost on the opening page because of the way the chords are arpeggiated. The harpist approaches this lovely sequence as if it had been written by Fauré or Pierné instead of Debussy, and the result is a cumbersome and blurry opening statement, devoid of a clear and rhythmic focus, and lumpy and lopsided as the chords alternate with octaves.

    There was an eye witness who coached with Debussy on the Danses, and who therefore had direct access to the great composer’s thoughts on this subject, and that was Pierre Jamet. In a wonderful video that was made about him when he was 97 years old, he alludes to this problem of the opening page. But first, the story of how this meeting of the minds came about.

    I think that all of us who studied with Pierre Jamet, were we to assemble together in a great room, could recite this story in unison, with very little variation from one person to the next. Jamet loved to tell it, and we all heard it, usually numerous times. What follows is a much more detailed account, taken from the Pierre Jamet memorial edition of the Bulletin de L’Association Internationale des Harpistes et Amis de la Harpe that was published after his death.

    In 1917 Jamet met Rose Féart, a singer at the Paris Opera and a well known interpreter of Debussy’s songs. “My boy, how would you like to meet Debussy?” she asked him. What a question! But it would take more than one intermediary to set up a meeting with the great composer. Féart passed her request to Jean Aubry, a journalist at the Figaro, and a friend and correspondent of Debussy.

    Debussy had recently written his sonata for Viola, Flute, and Harp, and had heard it performed only once, at a private concert at his publisher Durand. The performance featured chromatic harpist Susanne Dallies, Darius Milhaud on Viola, and Albert Manouvrier on flute. But Debussy really wanted to hear his piece played on pedal harp. Years later, Aubry reported that Debussy had said,”I’d like to hear my sonata played on something other than that horrible instrument. I want an Erard harp!”

    Jamet had been working on the piece with the same flutist as the private concert, and Sigismond Jarecki on viola. Debussy was told of this and was anxious to hear them, so a meeting was arranged. Rose Féart offered some sobering last minute advice. “Make sure you are well prepared,” she said. “If its not well prepared, he’ll simply get up and leave, without even saying good-bye!”

    On the appointed day, they went to Debussy’s home at 80 avenue du Bois-du-Boulogne(currently 60 avenue Hoche), to play the Sonata. Needless to say, they were nervous. Debussy was very hospitable, and they played. He said nothing, letting them play all the way through. When they had finished, he asked them to play it a second time. This time, he stopped them from time to time to coach them, to tell them what he wanted, to give them advice on the tempos he wanted. Debussy spent considerable time discussing the harp part at the beginning of the Final. He wanted the effect of a Tambourin* and experimented along with Jamet, with different ways of accomplishing this. The solution, which enchanted Debussy, was to play the notes using enharmonics. (Double the first note of each measure, playing E sharp/F natural, and then alternate right and left hand through the rest of the measure.) After this second run-through, Debussy, smiling, and in the presence of his wife, asked them if they would like to play the sonata at one of the concerts that he had organized to benefit the Red Cross war effort. Needless to say, they were delighted, and so they played at the first of Debussy’s concerts. It was at Laurent, on the Champs-Elysées on March 20, 1917. Also performing that night were Rose Féart and the pianist Walter Rummel.

    As the next concert approached, the singer fell sick, leaving a gaping hole in the program. Debussy contacted Jamet and asked if he could play the Danses. This request was made on a Friday, and Debussy asked Jamet to rehearse with him on the following Tuesday.

    It must be interjected here that the Danses had been composed only 13 years earlier, and were not in any way the ne plus ultra of the standard repertoire that they are today. After all, they had been written for Chromatic harp, and while a transcription by Renie did exist(Yes. What we play is a transcription!) they had not yet worked their way into the standard repertoire for pedal harp. Jamet had looked over the Danses prior to this, but had neither studied nor performed them. So not daring to rebuff the greatest living composer in France, he accepted, and learned the Danses in 5 days! He had a rehearsal with Debussy, again in the composer’s apartment, and because there was not enough time to assemble a string quartet, the rehearsal and performance were done with Debussy accompanying him on the piano.

    At the rehearsal, Debussy played the 7 bar introduction, and then Jamet came in. Almost immediately Debussy stopped him. “It’s too loud,” he said quietly. Jamet gulped. Debussy played the introduction a second time, and Jamet, playing much more softly this time, came in again, only to have Debussy stop him again. “It’s still too loud,” the great one said. Jamet always finished the story by saying,”Let me tell you. At that moment I knew what it meant to play softly!”

    But that wasn’t the end of it. Jamet had arpeggiated the opening chords, only to have Debussy stop him yet again. That wasn’t what he was looking for. Jamet tried again, this time playing the chords flat (unarpeggiated). That wasn’t it either. The two of them worked on it some more until Debussy got what he was looking for.

    In the afore mentioned video, Jamet, referring to that rehearsal, says, “Debussy didn’t want the chords rolled. But he didn’t want them flat either. It was something in between, but something closer to flat than rolled.” In the memorial Bulletin, he is quoted as saying,” Debussy wanted a certain emphasis of the melodic phrase by the left hand, and felt that the right hand-very rapidly arpeggiated-should be a reflection and sonorous prolongation of the left hand. “

    When I went to Paris to study with Jamet after finishing my undergraduate degree, I told him I would play whatever he wanted me to, and for the next year, using etudes, exercises, and pieces, we focused mainly on technique. Then one day, he said, “I think it’s time for you to work on the Danses.” I was thrilled. I remember him telling me that, while there were parts of the Danses that were difficult technically, the first page was the most difficult to interpret, and that each time I played it, I would have to work on it to create just the right texture. He was right.

    You might think that, simply knowing what the final result should be, one could just do it that way. But it’s trickier than that. The impulse to roll the chords is just too strong. So I have found that using a four step process to practice this first page helps overcome the urge to roll each chord, and to establish a texture for this passage consistent with what Debussy envisioned. Each time I have worked on the Danses, I have used the same process to practice the first page. I honestly can’t remember whether I invented this process, or if it was one Jamet suggested to me. All I know is, it works, and he liked the results.

    It is important to mention here that simply being able to hit all of the notes in the Danses in tempo does not necessarily mean you are ready to play this piece. This is advanced repertoire, and there are three points of technique that you must have mastered in order to be able to interpret any music at this level. The first is the ability to place a chord or individual note at the moment it is played, and not a microsecond earlier. The second is the ability to place in sequence. This means placing only one note at a time beyond the one you are actually playing, rather than placing blocks of 3 or 4 notes at a time. The third is the ability to play at least two dynamic levels in one hand simultaneously, in either the left or right hand. You should also be able to play multiple dynamic levels simultaneously when both hands are playing. If you have not mastered these techniques, go back to easier repertoire and work these things out before working seriously on the Danses.


    Whether you have already learned the Danses, or you are studying them for the first time, start by playing all the chords from bar 8 to the end of the page completely flat, with no arpeggiation. What’s important here is to be able to grab each chord easily, with no adjustment or correction, and to grab the strings at the moment they are played, and not a hair before. Make sure all the notes of each chord sound equally. Do this many times until it feels completely fluid and comfortable.


    Place all of the notes of each chord, but play only the lowest note of the left hand and the highest note of the right hand(i.e. the melody two octaves apart). Play these two notes together. Bring out the lower note slightly, and concentrate on shaping the phrases. Do this many times. Again, place, but don’t play, all of the middle notes.


    Play as you did in step two, with one change. Instead of playing the two outer notes together, play the top note slightly after, but as close as possible to the bottom note. This applies only to the chords of course, and not the octaves. The lower note should be played on the beat. This step is critically important to the final result, so do this many times, until it feels and sounds natural and very comfortable.


    Play all of the notes of each chord, but keep the same rhythmic distance between the lowest and the highest notes as you had in step three, packing all of the middle notes into the almost nonexistent space between those two outer notes. Start each chord on the beat, and bring out the lowest note slightly, which is the primary melody note. The effect will be that of a slight shimmer to each chord, without the feeling that it is arpeggiated. This is the effect that Debussy wanted. Of course, some of the chords can, and should be rolled longer: primarily the half note chords that end each phrase.

    Measures 17, 18, and 19 present a different problem. They are not, as they are so often heard, an alteration between chords and octaves. Look carefully at how Debussy notated these measures. He wrote out the melody in quarter note octaves, and then wrote two note, half note chords on every other beat. The octave/melody has to be played unarpegiated, and at a slightly higher dynamic level, then the half note chords. The chords are there to support, but not interfere with, the melody, and therefore must be played at a lower dynamic level. The same thing happens on the last two measures on the page.

    Note- A Tambourin is a folk instrument played in several European countries, including rural France. It consists of a simple recorder or flute, played with one hand, and a drum strapped to the player, so that one person can play both the recorder and the drum simultaneously.