Debussy extrapolations
For my final artistic presentation, I chose to perform my recreated performances of the two fragmented Debussy recordings made by Ricardo Viñes with the “reconstructed” missing beginnings of each piece. These “extrapolations” were played by intersecting the live performance with the original recording. I divided each of the two pieces in three parts and performed them in the following manner:
- The “extrapolated” missing first third of Viñes’s recording (played by me)
- Viñes’s performance (played in the room on loudspeakers)
- My recreated performance (played by me)
The following text reflects on the “extrapolation” process and the choices I made. I have included annotated scores with similar markings as before, but this time to illustrate my performance, not Viñes’s.
Images, 1re série:
Hommage à Rameau
Live performance at Dokkhuset in Trondheim, Norway at November 18, 2021:
Annotated score (only the first third of the piece):
The reconstructing experience
With regards to the Hommage a Rameau, bars 1-30 are missing in the recording by Viñes. The process of making this “extrapolation” started as a sort of jigsaw puzzle where several of the pieces are missing. In order to complete this in an “authentic Viñes manner,” I firstly examined the score and applied my knowledge from the earlier recreating process, looking for sections that relate to bars 31 and onwards (the part which Viñes recorded). Secondly, I looked for relatable sections in the already recreated recordings, in this case especially in the pieces of Debussy. Thirdly, it was interesting to hear Marcelle Meyer play the beginning of the piece and to (carefully) use this as a sort of inspiration in the extrapolation process. As discussed earlier, her recording does have some striking similarities with Viñes's, but it is very important to be clear on what she does differently. As stated earlier, she does not have the general “forward moving” tempo of Viñes and tends to slow down at the end of several phrases, as opposed to the longer “sweeps” of Viñes.
From my experience of recreating recordings by Viñes, I know that he tends to be quite consistent in his choice of tempos when repeating sections of a score. With this in mind, bars 5-8 were already “solved” because we can hear Viñes play the repetition of this material towards the end of the piece. I chose to play these bars in a very similar manner at the beginning. The tempo in which Viñes plays this towards the end of the recording is also an important indicator of the main tempo at the start of the piece. Despite the Lent et grave tempo indication of Debussy, I chose to play the beginning in a “forward-leaning” manner – making a four-bar phrase, rather than a two-bar one - by slightly accelerating towards the end of the second bar and then rhetorically slowing down at the end of bar 4. I also very slightly “dislocated” the melody, playing the top right hand melody slightly delayed after the left. I made these "dislocations" wider at the first notes of bar 1 and 3, as well as at the high C sharp in bar 4. I also generally use a lot of sustain pedal, with subtle rapid changes combined with a non-legato approach, and even a brisk staccato at times. Also, based on my experience from the earlier recreations, I tried to favor an expressive “forward-moving” line, rather than holding back to emphasize the pianissimo indication in the original score. Like with bars 1-4, a similar practice was used also in bars 10-13, and with a slight arpeggio in bar 11 to clarify the G sharp melody note from the underlying E natural. In bars 14-16, I used downwards “arpeggiation” (played briskly) to more clearly separate the thirds from the sustained top notes. This is not only typical “Viñes behavior,” it is also a technical way of finger playing which allows for a greater use of the sustain pedal without “muddying” the sonority. In bars 17-23, I went for a similar approach – much sustain pedal, but with the individual voices clearly separated within the sound. In these bars, I felt that playing the middle voices last produced the most clarity in the polyphony. Thus, I first played the octave bass, then the octaves in the right-hand chords, and finally the middle voices (thirds for the most part). In addition to this, “arpeggiation” was used to separate the voices in the left-hand chords starting from bar 19. All in all, a very complex and technically demanding interplay between overall sonority (rich use of pedals) and polyphony (“dislocated attacks” and arpeggios), but perfectly in line with how Viñes achieves the same effect in bar 35 (which is heard in the original recording). The final bars of my “extrapolation” (24-30) are, like the opening bars, slightly “dislocated” and slightly rushed in the middle of the phrase.
Etude No. 10:
“Pour les sonorités opposées”
Live performance at Dokkhuset in Trondheim, Norway at November 18, 2021:
Annotated score (only the first third of the piece):
The reconstructing experience
I used the same strategy to reconstruct the missing first 30 bars of this etude by Debussy. As an “inspirational recording,” it was natural to listen to the recording by Adolph Hallis. As discussed earlier, I found that this recording has some remarkable similarities to Viñes’s performance. In this case, Hallis's general tempo is closer to Viñes's than Meyer's is to Viñes's in the Hommage à Rameau. However, although Hallis greatly rushes and slows down in the first bars, he arrives at and stays in a steadier tempo by bar 7 and onwards (Animando poco a poco). In addition to this, Hallis does not showcase a rich sonority through the heavy use of the sustain pedal in the way Viñes does later in the piece. Also, Hallis plays the first bars quite dryly, with the pedal often not catching and sustaining the low bass notes in the sonority. Although this seems closer to what Debussy has indicated in the score (there are rests after the bass G sharp quarter note in the second bar), Viñes instead favors a lusher overall sonority and sustains the pedal throughout these bars in similar sections later in the piece, such as from bar 44 and onwards. Because of this, I chose to have a long sustained pedal in the beginning, even intentionally letting the dissonance between the G sharp and the A linger in the sonority. I also dislocated the right-hand G sharp octaves from the G sharp of the left hand. Subtle “arpeggiation” was used in bars 5-6 to separate the individual voices in the polyphony. From bar 44 and onwards, the complex sonority we hear in the recording of Viñes is realized through heavy use of the sustain pedal, “arpeggiation,” and “dislocated attacks” (in this case the middle voice octaves, which are played last in a very brisk and accentuated manner). Having worked quite a lot to figure out this puzzle in the recreated recording, I believed this to be the solution for the “unrecorded” bars 7-14. As can be seen in the annotated score, bars 7-14 involve a combination of rich sustain pedal, “arpeggiation,” and the octave melody being delayed through “dislocation” in a brisk and accentuated manner so that it rings out in the sonority. These octaves are always delayed, with the exception of bar 13, where they intersect between the hands. In bars 15-30, I played with crossed hands so that the left hand (which was on top) “arpeggiated” the chords upwards and the right hand (which played in the lower register) rolled them downwards, thus ending on the low note, which in this case is also the melody note in the polyphony. In addition to these things, I added some rushing and slowing down to the general tempo, especially in bars 4, 6, 13, 14, 24 and 25, always with the aim of driving the piece in a forward-moving manner and generally phrasing in longer “sweeps.”
Final thoughts
For my final artistic presentation, I chose to play the first and third parts of these pieces live, intersected by Viñes’s original recording in the middle. I chose to tell the audience that these were based on incomplete recordings, but not which parts were missing. Hopefully they were not able to say which ones…
