Ravel Edition

Unlike a considerable number of Ravel’s other works, the editorial history of La Valse is extensively documented: from the initial drafts scribbled in pencil by the composer himself in 1919 to the very first printed orchestral parts in 1921, we have an extremely accurate chronological overview of exactly when each of the various manuscripts and Durand engravings came into existence. It is also rare for all the manuscripts mentioned by Ravel in his correspondence with his peers to be accessible at two research libraries: the Morgan Library in New York and the Bibliothèque nationale de France[1] in Paris.

The urgency of this long-awaited “choreographic poem”[2]  in the composer’s own words, intended since its beginnings for the theatre and not the concert hall (initially composed for Diaghilev’s “Ballets Russes” in June 1920 at the Théâtre national de l’Opéra de Paris), forced the composer, who had found refuge far away from the stir of Paris in a hamlet in Ardèche, to produce his manuscripts for Wien[3]/La Valse within a tight timeframe of just four months. A private performance of the new score under the watchful eye of Diaghilev had been scheduled for early spring, and Ravel, in letters addressed to his close friends and family, mentions the “grind”, his long days of “labour”, and waking up “well before sunrise”, sarcastically alluding to his “lazy lie-ins until 7.30 am”. Furthermore, the letters that Ravel wrote in Ardèche during the winter of 1919-1920, oscillating between enthusiastic outbursts and depressed admissions, reveal the following chronological events[4]:  

• On 20/12/1919, Ravel wrote to Lucien Garban: “I’m tearing myself away from the joy of dancing to reply to you (…)”
• On 22/12, to Roland-Manuel: “I’m going back to Wien. It’s going well. I’ve finally managed to make a start, and I’m in 4th gear.”
• On 06/01/1920, to Mrs Fernand Dreyfus: “(…) at the moment, I’m frantically waltzing; I started orchestrating on 31 December.”
• On 10/01, to Misia Sert: “We’ve been referring to it as Wien… Sorry, it will now be called La Valse (…) My chorographic poem will definitely be finished, and even orchestrated, by the end of this month, and I’ll then be able to present it to Diaghilev.”
• On 13/01, to Nicolas Obouhow: “(…) I’m in the process of orchestrating my Viennese Valse.”
• On 26/02 to Lucien Garban: “This morning I gave the manuscript for La Valse for piano for two hands to the postman. (…) As long as it doesn’t get lost! I don’t have another draft.”
• On 28/02, to Georgette Marnold: “I’m starting to tidy up the orchestral score.”
• On 06/04 to Lucien Garban: [he explains that he submitted 28 pages of orchestral score to Durand on 3 April].
• On 06/04 to Georgette Marnold: “28 pages of the orchestral score have already been posted. I’m on the 40th.”
• On 13/04 to Georgette Marnold: “Phew! I finished last night – 75 pages (…). I still need to proofread my score – I have that to do today (…).”
• On 14/04 to Lucien Garban: “I finished the copy today (…). I’m leaving Lapras on Thursday and I’ll bring the rest of the orchestra.”

A study of the various manuscripts still intact reveals several insightful dates on the last page, near the composer’s signature and “Lapras” marking: the score for solo piano mentions the period between December 1919 and February 2020, whilst the orchestral ink manuscript reads “December 1919 – March 1920”.

Oddly, neither of the two copies of the Reduction (or “transcription” according to the terminology used by Ravel himself) for two pianos, kept at the Morgan Library, give any indication as to when the composer started and finished. These two complementary versions – the “black and white” piano scores required for Ravel’s presentation of the work to Diaghilev for the Ballets Russes and for the choreographic process, alongside the colourful injection of the orchestral parts –  were therefore seemingly conceptualised at the same time. On the piano parts, we also find several instrumental part markings, clarifying various features or glissandi between two staves, often far too cramped to truly express the full richness of the musical text.  

Although we are lacking specific details surrounding Diaghilev’s rejection of the score, what we do know is that a piano performance was given in May 1920 and that the Russian impresario turned it down[5]. It’s easy to imagine the composer’s frustration, but also his editor’s relief at the prospect of a now somewhat less hectic schedule, with the work being engraved simultaneously in three different forms.

Through his correspondence during the course of 1920, we learn that Ravel went back to his “Reduction” for two pianos in June. He considered his first draft from the winter of 1919-20 a preparatory  sketch with several “dubious passages” and finished his final version in July with the addition of a dedication to Misa Sert, and the very first explanation of the Ballet’s “Viennese” synopsis. In September[6], we discover that the composer was planning a première of La Valse at the Opéra de Paris, whilst Lucien Garban was working at Durand on the orchestral parts[7]. However, these parts could only have been temporary and very certainly handwritten or in the form of a simple proof, as there is absolutely no trace of these parts used for the work’s very first performance at the Salle Gaveau with the Orchestre Lamoureux, under the baton of Camille Chevillard on 12 December 1920[8]. On 17 December, Ravel instructs Lucien Garban “not to engrave La Valse before [he proofreads] the score.” Furthermore, the Durand plate number book clearly indicates that the orchestral score[10] and the separate instrumental parts were engraved in March 1921 (legally deposited on 19 and 30 March of the same year)[11]. Unfortunately, we therefore do not know from which score Chevillard conducted the very first performance. It certainly was not from the complete ink manuscript that we have in our possession, which is entirely void of any blue or red pencil markings that would have been added in advance of a performance or even during the preparatory work conducted before the engraving. The most likely hypothesis is that there was another “draft” orchestral score, either handwritten by Ravel or a scribe, which was used by Chevillard on the conductor’s podium and by Garban for the engraving, and which in all likelihood also contained alterations and corrections added by Ravel following the première: an essential document that has since disappeared or was perhaps destroyed in order to avoid the publishers getting confused with multiple draft copies (and which was not kept by Ravel[12].).

The archives of the Orchestre Lamoureux provide modern-day proof of this performance with the full set of Durand parts from the first print run, complete with concert dates added by the musicians on the front cover[13]: an essential source of information in understanding the corrections,  alterations, string desking[14] and famous initial division of double basses, which was also adopted by Manuel Rosenthal.

Following a comparison of the musical text contained in the various different manuscripts available, we felt that it was important to analyse one specific source that, in our opinion, still had not been truly appreciated: the orchestral sketches, previously the possession of Roger Désormière[15], which provide a fascinating insight into the composer’s initial intentions. We know that Ravel was somewhat hesitant and had little desire to copy out his parts in a hurry, amplified by his proneness to forget entire bars or considerable orchestral details, without ever going back over them again[16]: aspects that we had the opportunity to appreciate once again through a reading of this highly advanced preparatory work, containing several instrumental variations that we subsequently chose to incorporate into our 2022 revised edition.

Another essential component of our work was a meticulous analysis of the notes, corrections and queries from two of Ravel’s contemporaries, whose scores we were thankfully able to access: Charles Münch[17], whose score beautifully introduces the “rhythmic agility” that the composer so desired[18], building up to the final apocalypse and its violent collapse, and Roger Désormière[19], whose score was regrettably never recorded. Furthermore, notes left by Pierre Boulez[20] on his pocket score consolidate his elders’ corrections, primarily regarding transposing instruments.

And so, it will be down to the musicians of 2022 to recreate the “glitz and glamour” Ravel strived to convey in their performances of this homage to the Viennese Waltz: the sprinkling glissandi of the strings and decorative portamenti he heard both from the auditorium and conductor’s podium. All this, as per the tastes and performance codes that defined the start of the 20th Century, mirroring the 78 RPM record of Philippe Gaubert conducting the Orchestre de la Société des Concerts du Conservatoire in 1927, performed from one of the work’s very first engravings[21].

François Dru - August 2022 (Reproducing this text in part or in full is prohibited without prior authorisation from Ravel Edition.) 

[1] The Lehman Collection at the Morgan Library holds three different ink manuscripts: the orchestral score (75 pages) (R252.V214 – 115541), the sketches of the score for two pianos (22 pages) (R252.V214 – 115544), and the finalised score for two pianos (25 pages – with the ballet synopsis and dedication – belonging to L. Garban) (R252.V214 – 115543). Elsewhere, the Mary Flager Cary Collection at the Morgan Library contains the manuscript for solo piano (R252.V214 – 115542), which can be freely accessed on the library website, and the Music Department at the BnF is home to the full orchestral pencil sketches, previously the possession of conductor Roger Désormière (MS-17140).  

[2] Labelled “symphonic” by the press when it was eventually performed.

[3] Initial title. In Ravel’s musical library, we find the following “Viennese” scores (for solo piano) owned by Johann Strauss II: Kuss-Walzer op. 400 BnF VMA-3145(2), Myrthenblüten RV 395 VMA-3145(16) and Frühlingstimmen VMA-3145(13). Whilst Ravel clearly stated that the Viennese Waltz was his main source of inspiration, we can certainly hear the direct influence of Rapsodie viennoise Op. 53 n°3 by Florent Schmitt and Grande Valse di bravura Op. 6 by Franz Liszt. One critic also mentioned the work’s similarity to Valse infernale de Robert le Diable by Giacomo Meyerbeer.

[4] C.f. René Chalupt, Ravel au miroir de ses lettres, Robert Laffont, Paris 1956. Maurice Ravel, Lettres, écrits et entretiens, Flammarion, Paris 1989.  Maurice Ravel, L’intégrale, Le Passeur, Paris 2018. Catalogue de la vente des Archives Lucien Garban, 8 April 1992; Drouot-Richelieu, Paris.

[5] The opening performances of the 1920 Ballet Russes season were neoclassical ballets Pulcinella by Stravinsky and Astuce féminine by Cimarosa/Respighi. La Valse only entered the ballet repertoire in November 1928, at the same time as when the composer’s Bolero was composed, upon the initiative of Ida Rubinstein and her productions at the Opéra de Paris.

[6] Letter dated 26 September 1920 to Ida Godebska.

[7] Letter dated 9 September 1920 to Lucien Garban. In this correspondence, Ravel alludes to an attached explanation regarding a “troublesome tremolo”: a document that unfortunately has never been found.

[8] On the same day and at the same time, Philippe Gaubert was conducting Mother Goose Suite with the Société des Concerts du Conservatoire.

[9] Reproduced by Nigel Simeone in Mother goose and other golden eggs: Durand Editions of Ravel as reflected in the firm printing records. IAML Publications UK, 1998 Vol. 35 n°, pages 58-79.

[10]  Published for its first print run of 200 copies.

[11] The version for solo piano was printed in November 1920 on the same date as the version for piano for four hands. The reduction for two pianos was engraved in December of the same year and had been performed in Vienna at the Kleiner Konzerthaussaal by the composer himself and Alfredo Casella on 23 October 1920 before the première of the orchestral version.

[12] According to the reference found in Ravel’s music library taken from his house in Montfort-l’Amaury, which became part of the BnF collection in 1975. Oddly, there is no surviving printed copy of La Valse previously belonging to Ravel. 

[13] One Violin 1 part (6th desk) mentions 12 December 1921 [sic] and 23 November 1924. A concert conducted by Paul Paray, which showcased, amongst other works, Symphony K. 550 by Mozart and Prélude à l’Après-midi du faune by Debussy. 

[14] In photos of the Orchestre Lamoureux at the Salle Gaveau, like the one used on the programmes for the Durand concerts in 1910, despite the narrowness of the stage (which still had its organ), we can clearly see 16 first violins and 8 double basses.

[15]  Although we do not know how this manuscript was obtained, with its entry into the BnF collection dated as 1972 (the majority of the Désormière collection was deposited in 1969 to the Paris Conservatoire library), we know that Ravel was particularly fond of the conductor and composer.

[16] “But he would lose interest in his work: as soon he finished a score, it was no longer his own. This disaffection explains the ease with which he would let his compositions be transcribed for flute, trumpet or small café orchestra,” in Hélène Jourdan-Morhange, Ravel et nous, Éditions du Milieu du Monde, Geneva, 1945. On 10 January 1923, Ravel wrote to Garban: “(…) upon studying – very little – my scores (La Valse and Mother Goose), I discovered a significant number of mistakes, which I’ve taken note of. I will undoubtedly find others, and I will send you all of them.” Sales catalogue from the Lucien Garban Archives Op. cit.

[17] Durand D.&F. 9885 large orchestral score from 1921 (Jean-Philippe Schweitzer private collection). Münch also had a pocket-size version – D.&F. 10 080 from 1925 -  from the “Rhené-Baton music library”.

[18]  According to Ravel’s glowing report addressed to Ernest Ansermet after hearing him conduct this work.

[19] Durand D.&F. 10 080 1947 pocket score (Fonds Roger Désormière – Musée de la Musique, Paris). Désormière also had a Durand D.&F. 9871 part for the reduction for solo piano from 1947.

[20] Durand D.&F. 10 080 pocket score from 1947 (BnF 4-VM FONDS 148 BLZ-416 (1 and 2).

[21] Columbia 12502 78 RPM record released on 7 May 1927. Released on CD by Alpha 801 in 2006.