Etude No. 10: 

“Pour les sonorités opposées” 

Claude Debussy

(fragmented recording)

Original recording by Ricardo Viñes:

My recreated performance:

Cross-cut video:

Annotated score:

This is the first of two fragmented recordings by Ricardo Viñes of an unknown date. As I wrote in the distilled Viñes biography chapter, my little “treasure hunt” ended with the email correspondence I had with Jonathan Summers at the British Library Sound Archive. I had a small hope that the missing first one third of these two Debussy recordings would magically re-appear if I searched thoroughly, but they seem to be lost forever. I therefore decided that I would use all my experience and knowledge from the other recreations to “reconstruct” the missing first parts of both recordings. To be able to do this as truthfully as possible, I started by recreating both pieces in their fragmented state (in the case of this etude, bars 31-75). This text concerns the experiences from the recreated recording of the fragmented original. 

The recreating experience

At this point in my research I had already experienced how Viñes is using “arpeggiation” and “dislocation” to emphasize the more important voices of the polyphony in the pieces I worked on. What I had found to be even more interesting was that this practice was not limited to the upper register of the piano (in historic recordings we often hear the melody in the right-hand being “dislocated” so that it is played after the accompaniment in the left). This Debussy etude turned out to be the most striking example of this practice out of all the pieces I recreated. It was also the most challenging with regards to recreating the sonority Viñes produces in the original recording. As with all the recreated pieces, I had to experiment with different technical approaches before I could get the right sound in my own playing. All pianists have different bodies and minds, but having experimented with several technical approaches, I eventually found that there were not many alternatives, and that I always seemed to end up with one specific approach which made my playing sound similar to Viñes’s. 

In the annotated score, I have marked Viñes’s use of rushing/slowing with thick red lines. Omitted notes in his performance are marked by red crossings. There are also a couple of red circled markings towards the end of the piece which are explained in this text. My markings which are meant to illustrate the different practices of “arpeggiation” and “dislocation” not marked in the original score, are in this piece the most substantial part of my annotated score. These different practices can be divided into three categories:

1. “Dislocated” chords - (“un-rolled”), marked as red circles with arrows showing the “dislocation.”
2. Expressive arpeggios - used to separate the melody from the chord notes, or to create expressiveness in the melodic line.
3. Directional arpeggios with "dislocated" melody - marked as arpeggio lines with arrows to demonstrate if they are played upwards or downwards, and with blue circles and arrows showing the “dislocated” melody notes.

The original recording starts at bar 31. The typical “forward moving” tempo we often hear in the other pieces recorded by Viñes is clearly evident already by the end of bar 32. We also hear a very subtle “dislocation” between the hands. Viñes slows down the tempo a bit in bar 33, probably to emphasize the decrescendo from the original score. As marked in the annotated score, we hear a subtle “dislocation” between the chords and the bass line. In bar 36, Viñes clearly rolls the chord and delays the E natural, which is the melody note in this case. In the Animato e appassionato section starting on bar 38, it becomes really interesting. First of all, Viñes omits the second low G sharp in the first three bars, but he does play them in 41 and 42. Taking into consideration that Viñes is usually quite consistent in such deviations from the scores, there could be some logic in omitting them and thus creating a link to bar 44 and onwards, but then it seems strange that he chooses to play them in bars 41 and 42. It could also be edition related, or perhaps in this case unintentional playing from Viñes’s part. What is more interesting is how he uses a variety of arpeggios in this section. In bars 38-39, he rolls both chords upwards. When the hands are crossed in bar 40 and onwards, it becomes more complicated. He first rolls the left-hand chord (F sharp, A sharp, E sharp) in an upwards manner. Secondly, he rolls the right-hand chord (A sharp, B sharp, E sharp) in a downwards manner, so that the last note of the chord, which in this case is also the main melody note of A sharp, is automatically delayed (marked in blue in the annotated score). The result is that the middle melody voice is “dislocated” from the chords. Even though this practice is complex and basically involves two different varieties of arpeggios as well as “dislocation” of the melody, the simple logic to it is that the main melody (which is in this case in the middle) is delayed and thus emphasized in the sonority. This is perfectly in line with what I have experienced in Viñes’s playing earlier and I believe that it’s particularly interesting to observe how he applies this practice as a “solution” to the objective of this etude (“pour les sonorités opposes”).  It did require a lot of experimentation and deep listening before I realized the order in which these arpeggios are played. In bar 44 and onwards, he slightly changes the “arpeggiation” practice and brings in a new element: he now firstly rolls the notes of the chords that are not part of the melody first (F sharp, A sharp + F sharp, A sharp, E sharp) very briskly, then he plays the octave melody of two D sharps in the left hand very accentuated and extremely fast (staccato). These bars are perhaps the most complex I have worked on with regards to the sonority. Viñes uses the sustain pedal constantly and I struggled for some time with this. It was at first impossible for me to project a clear melody line in the sonority with such a constant use of the sustain pedal, without “muddying” the sound. And, on top of this, there should be a very distinct melody line projected through the rich sonority. I found the very subtle dislocation of the melody to be the solution. If the octave melody is played after the chords in a very rapid and accentuated manner, it will ring out on top in the sonority without “muddying” the sound. This means that I can leave the sustain pedal down without making clear shifts. The molto sostenuto marked in the original score is thus realized by playing the octave melody accentuated and staccato with a rich use of the sustain pedal. In bars 47-50, the practice is continued, but now the octave melody is in the right hand. Experiencing how Viñes’s use of articulated staccatos and rich sustain pedal work together, it is interesting to repeat the quote from his student Francis Poulenc: 

The art of pedaling, this essential ingredient in modern music; no one could teach it better than Viñes since he managed to play clearly in a wash of pedaling, which seems paradoxical. And what science he demonstrated in staccato![1]

From bar 51, Viñes plays the same way as in 40, but rushing so that he almost misses the last G sharp (here he does not seem to intentionally omit any of them). In the Calmato section that follows, we hear a very subtle use of "dislocation" between the two hands playing the chords on the first beat. In bar 60, the third beat is accentuated with a rapid arpeggio, delaying the F sharp in the middle register (very subtly). In the proceeding bar, I believe that he uses "finger pedal" and keeps the sonority of the low B natural in the bass by holding it down through the pedal shift. In bars 63-65, we hear yet another variety of “arpeggiation” and “dislocation.” I would argue that Debussy’s way of notating this in the original score was a suggestive description for how to play the arpeggios and bring out the melody in the aforementioned bars 44-50. In bars 63-65, rather than marking the chords and the melody on the same beat, Debussy wrote out the “arpeggiation” of the chords as appoggiatura notes (downwards in the right hand and upwards in the left), followed by the “dislocated” and thus delayed melody. Viñes, though, further complicates things by accentuating the first high note of the right-hand appoggiatura. This results in a sort of anticipation of the melody, one octave higher than the middle voice. This accentuated note is played in-between the two left-hand appoggiaturas. He also "dislocates" the eighth-note F natural in the left hand and the D sharp in the right. This practice is continued through the following two bars. This heavy use of “dislocation” also results in a triplet-feel in Viñes’s performance; in other words, this is also a rhythmical deviation from the original score. In the remaining bars of the piece we hear Viñes alternating between a very subtle and a more evident use of “arpeggiation” to "dislocate" the chords from the bass line. In bar 69, he plays an A sharp instead of the written A natural. In bar 74, he makes a “guitar-like” embellishment of the originally written arpeggio. In my experience, he starts by playing a downwards arpeggio of B – G sharp – C sharp, followed by an upwards arpeggio from the low C sharp to G sharp and C sharp on top (the octave above the low C sharp is added by Viñes and not in the original score). The arpeggio continues in the right hand as written in the score with E, G sharp, B, C sharp and E. The result of all this is a prolonged, embellished sort of arpeggio.

Comparison to other recordings

To investigate how Ricardo Viñes's interpretation of this piece stands in relation to other recordings, I have chosen four to compare it with. Because of the fragmented nature of Viñes’s recording, I have only compared the bars 31-75.

Maurizio Pollini’s recording is, compared to Viñes’s, very strict in rhythm.[2] His tempo is steady and slower than Viñes. He allows himself very little rubato or nuances in the timing other than what is marked in the original score. There are a few bars that have a feeling of subtle “dislocation,” but far from the coherent use of it as in Viñes’s case. Pollini does not add any arpeggios that are not marked in the score. 

What’s interesting about Mitsuko Uchida’s recording, is that she does apply some “dislocation” in her playing.[3]  We can see an example of this in bars 44-46. It is also very clear that she uses “dislocation” to bring out the tricky middle voices of the piece. She does also add a few arpeggios not marked in the score, somewhat in the manner of Viñes, although far from as consistently, as they only appear on some occasions. Examples of this can be heard in bars 38, 41-43 and 51-52. Uchida’s tempo and timing is however quite far from Viñes’s. Her general tempo is much slower and does not have the slightly rushed “forward moving” direction of Viñes.

In Walter Gieseking’s recording, we hear by far the “strictest” recording with regards to the general tempo in this comparison.[4] Besides adding little or no rubato, he also makes little contrast between the tempo directions marked in the score. There is no evident use of “dislocation” and he does not add arpeggios that are unmarked in the original score. 

Because these etudes are far from the most recorded works composed by Debussy, it was difficult to find comparative recordings from “Viñes’s time”. I did however eventually find the recording of South African pianist Adolph Hallis (1896-1987). In 1938, Decca released a recording of the complete set of 12 Etudes by Debussy, and Adolph Hallis consequently became the first pianist to record the entire cycle.[5]  This recording is to my knowledge not issued on CD, but it can be found on YouTube.[6] This recording clearly has some similarities to Viñes’s performance. The general tempo of Hallis is more so than the other compared recordings “forward moving.” Although not as extreme as with Viñes, Hallis uses both rushing and slowing in his range of rubato. Hallis also uses “dislocation” to separate the voices throughout the piece. In bars 38-43, he uses “arpeggiation” in a similar manner of Viñes. In 44-46, he continues to use “arpeggiation” and “dislocation” to separate the melody, but unlike Viñes he seems to play the octaves before the arpeggios, not after. In bar 47 and onwards, he appears to do the opposite and very slightly plays the octaves after the chords, more in line with Viñes’s practice. Perhaps the most striking similarity between Viñes and Hallis is Hallis’s tendency of playing the melody of bars 63-65 more in a “triplet-feel” as opposed to the notated straight eighth notes.

Final thoughts  

To summarize, I would argue that there is also in this case an evident romantic performance practice in the recording of Viñes. This particularly applies to his tempo modifications, which include unmarked rubato and the general “forward moving” tempo. In addition to this, there is the complex use of “arpeggiation” and “dislocation.” If there is such a thing as a definable “Viñes sonority,” bars 44-50 should serve as a great example. Even in such an old recording, the incredible rich sonority Viñes achieves through his special approach of touch and pedaling is in my view truly fascinating. 

Previous                                                                                                                                                                              Next