The Symphony No. 10 by Gustav Mahler was written in the summer of 1910, and was his final composition. At the time of Mahler's death the composition was substantially complete in the form of a continuous draft; but not being fully elaborated at every point, and mostly not orchestrated, it was not performable in that state. Only the first movement is regarded as reasonably complete and performable as Mahler intended. Perhaps as a reflection of the inner turmoil he was dealing with at the time (Mahler knew he had a failing heart and his wife had committed infidelity), the 10th Symphony is arguably his most musically dissonant work.
Mahler started his work on his Tenth Symphony in July 1910 in Toblach, and ended his efforts in September the same year. He never managed to complete the orchestral draft before his premature death at the age of fifty from a streptococcal infection of the blood.
Mahler's drafts and sketches for the Tenth Symphony comprise 72 pages of full score, 50 pages of continuous short score draft (2 pages of which are missing), and a further 44 pages of preliminary drafts, sketches, and inserts. In the form in which Mahler left it, the symphony consists of five movements:
- Andante – Adagio: 275 bars drafted in orchestral and short score
- Scherzo: 522 bars drafted in orchestral and short score
- Purgatorio. : 170 bars drafted in short score, the first 30 bars of which were also drafted in orchestral scoreAllegro moderato
- [Scherzo. Nicht zu schnell]: about 579 bars drafted in short score
- Finale. Langsam, schwer: 400 bars drafted in short score
The parts in short score were usually in four staves. The designations of some movements were altered as work progressed: for example the second movement was initially envisaged as a finale. The fourth movement was also relocated in multiple instances. Mahler then started on an orchestral draft of the symphony, which begins to bear some signs of haste after the halfway point of the first movement. He had got as far as orchestrating the first two movements and the opening 30 bars of the third movement when he had to put aside work on the Tenth to make final revisions to the Ninth Symphony.
The circumstances surrounding the composition of the Tenth were highly unusual. Mahler was at the height of his compositional powers, but his personal life was in complete disarray, most recently compounded by the revelation that his young wife Alma had had an affair with the architect Walter Gropius. Mahler sought counselling from Sigmund Freud, and on the verge of its successful première in Munich, dedicated the Eighth Symphony to Alma in a desperate attempt to repair the breach. The unsettled frame of Mahler's mind found expression in the despairing comments (many addressed to Alma) written on the manuscript of the Tenth, and must have influenced its composition: on the final page of the short score in the final movement, Mahler wrote, "für dich leben! für dich sterben!" (To live for you! To die for you!) and the exclamation "Almschi!" underneath the last soaring phrase.
The instrumentation of the symphony cannot be defined precisely, owing to the incompleteness of the orchestral draft. However, in the short score there are occasional indications of instrumentation, and some of the orchestration may be surmised from the three movements of the orchestral draft, from which the probable forces include: four flutes, one piccolo, four oboes, four clarinets in B flat and A, with one doubling E flat, three bassoons, two contrabassoons, four horns, four trumpets, four trombones, a tuba, two sets of timpani, tam tam, a large muffled military drum, harp and strings (first and second violins, violas, cellos and double basses). The surviving orchestration does not specify a cor anglais (English horn) or a bass clarinet, bass drum, cymbals, and triangle, although Mahler regularly used these instruments in his other symphonies.
Realisations of the work
After Mahler's death there was no immediate attempt to complete the symphony, or render it in a state where it could be performed, although figures such as Paul Stefan described the high quality of the work as drafted. Arnold Schoenberg famously expressed the opinion that no one could possibly write a Tenth Symphony without being close to the hereafter (see Curse of the ninth); and a mistaken report led Richard Specht to suggest Mahler wanted the manuscript burned after his death. Hence it was only in the 1920s that Alma Mahler-Werfel asked the composer Ernst Krenek to make a fair copy of Mahler's orchestral draft for a festival of performances of Mahler works, and at about the same time some of the manuscripts were published by the company of Paul Zsolnay in facsimile (1924). The facsimile made evident that the stress of Mahler's final year had not adversely affected the composition, and that the draft contained passages of great beauty. Much of the manuscript, however, was too difficult to read and seemingly too chaotic for the unbroken continuity of the music to be clearly apparent.
In 1924 Krenek made a fair copy of only the first (Adagio) and third (Purgatorio) movements, and might have made a fair copy of the second movement, but as Mahler's draft of the Scherzo was very much patchier this was evidently less feasible. Alban Berg was enlisted to proofread the work, but his suggested corrections were never incorporated, while at the same time some unauthorised changes were introduced, possibly by one of the conductors of the first two performances, Franz Schalk and Alexander von Zemlinsky. Krenek is supposed to have renounced the changes to his version, which was subsequently published. Performances of the Krenek-Schalk/Zemlinsky version have been moderately successful, but the third movement is not generally convincing when taken out of context between the second and fourth movements: it is possible that some of the conductors who have refused to perform the Tenth, most famously Bruno Walter and Leonard Bernstein, took exception to such a piecemeal representation.
In 1923, Alma had also sent a copy of the score to Willem Mengelberg in Amsterdam with the addition that two parts (obviously, the Adagio and Purgatorio) were 'absolutely performable'. Briefly after Schalk performed Krenek's score (with his own additions) on October 12, 1924, Alma sent what is believed to be Schalk's score to Mengelberg, who subsequently prepared his own edition with the aid of his secondant Cornelis Dopper. This version uses a larger orchestra and makes significant changes in dynamic markings and tempi. It was premiered on November 27, 1924, in the Amsterdam Concertgebouw, and subsequently played a number of times under Mengelberg's baton.
It was soon realised that a performing version of only two movements did not give listeners a clear idea of the entire symphony, let alone constitute a complete artistic statement, so in the 1940s the American Mahler enthusiast Jack Diether tried to encourage several notable composers to realise the work. Figures such as Shostakovich, Schoenberg, and Britten (all of whom had been considerably influenced by the works of Mahler) refused, and instead the task was taken up by musicologists: early attempts at realising the entire work were made in America by Clinton Carpenter (completed 1949, subsequently revised 1966), in Germany by Hans Wollschläger (1954–1962, withdrawn), and in England by Joe Wheeler (1953–1965) and Deryck Cooke.
Deryck Cooke's versions
The various realisations produced by Cooke have, since the mid-1960s, become the basis for most performances and recordings.
A first, still incomplete performing version by Cooke (1959–1960) stemmed from a performance and an associated lecture for radio broadcast on the BBC Third Programme, marking the centenary of Mahler's birth. This was aired on 19 December 1960, with the Philharmonia Orchestra conducted by Berthold Goldschmidt, who also assisted with the orchestration of Cooke's edition. At its first performance Cooke's realisation of the final movement proved to be a revelation to listeners, and Cooke resolved to complete the orchestration and elaboration of the Scherzo movements, which required much more compositional work than he had time for.
Alma Mahler, who had at one point taken the views of Bruno Walter to heart and demanded a veto on further performances of the Cooke performing version, actually changed her mind upon seeing Cooke's revised score and hearing the recording. She wrote Cooke a letter in English, postmarked New York, 8 May 1963, which Cooke includes in the preface pages to the score:
Dear Mr. Cooke,
Mr. Harold Byrns visited me here in New York. Today he read me your excellent articles on Mahler's Tenth Symphony and [showed me] your equally authoritative score. Afterwards I expressed my desire to finally listen to the London BBC tape. I was so moved by this performance that I immediately asked Mr. Byrns to play the work a second time. I then realised that the time had come when I must reconsider my previous decision not to permit the performance of this work. I have now decided once and for all to give you full permission to go ahead with performances in any part of the world. I enclose [a] copy of my letter of even date to [the] BBC.
- Alma Maria Mahler
Cooke's revised and completed version, conducted by Goldschmidt, was premièred at the Proms on 13 August 1964 and recorded soon after. After Alma's death in December 1964, her daughter Anna allowed Cooke access to the full set of manuscript sketches, many of which had not been published four decades earlier. In the light of these, Cooke made a revised performing version in association with the British composers Colin and David Matthews between 1966 and 1972, and thereafter his final version before his death in 1976. The release of these pages also prompted the International Gustav Mahler Society in Vienna to issue another, more complete collection of Mahler's manuscripts in facsimile (Ricke, 1967). This revised edition of Cooke's first complete score was published in 1976, shortly before Cooke's death. A further revision, with mostly minor changes made by the three surviving collaborators, appeared in print in 1989.
Summary of the Cooke versions
Cooke's performing editions of the Tenth Symphony may be summarised as follows:
- Cooke "0" – (1960, unpublished): BBC performance, complete realisations of 1st, 3rd and 5th movements plus partial realisations of the 2nd and 4th movements; presented as part of a lecture-demonstration
- Cooke I – first complete performing version (1960–1964; unpublished)
- Premiered on 13 August 1964 by Berthold Goldschmidt; basis for the recordings by Eugene Ormandy (1965/66) and Jean Martinon (1966)
- Cooke II – second performing version (1966–1972; appeared in print in 1976)
- premiered on 15 October 1972 by Wyn Morris; basis for all recordings from 1972 to 1992
- Cooke III – a slightly revised form of the 1976 score (printed in 1989)
- various reading errors corrected, some minor changes made in orchestration, and the whole enhanced by considerations concerning performance; the editorial input coming from David and Colin Matthews and Berthold Goldschmidt. (Those pages which have been altered are marked with an asterisk.)
Other complete versions
Clinton Carpenter (1921–2005) started working on his edition long before Cooke, and called his score a "completion" rather than a "performing version". Although he finished his version in 1949 (revising the work in 1966), it had to wait until 1983 for a performance. Carpenter did not merely review Mahler's symphonic output to guide him in his effort, but went so far as to include actual quotations from every Mahler symphony in his edition. The view has been expressed that much of this process of recomposition gives the impression that Carpenter has effectively written his own symphony using Mahler's as a basis.
The completion by Joseph Wheeler dates from 1953 to 1965, and like Cooke he also refined his ideas several times, so the final version of 1965 was actually the fourth iteration; the American composer Remo Mazzetti Jr. considers Wheeler's fourth version to be the closest to Mahler's late orchestral style. Wheeler's interventions are at the opposite end of the spectrum to Carpenter's, and he is less interventionist even than Cooke: he only makes additions to the score where performance is otherwise impossible. The effect is sparer than other completions, although Wheeler does increase the brass part to a greater degree than Cooke.
In recent years several further realisations of the symphony have been attempted: Remo Mazzetti initially made his 1989 version from dissatisfaction with the existing Cooke, Carpenter, and Wheeler editions, though the spur of preparing a performance of Wheeler's version in 1997 led him to recant his earlier view. Of his own revised version he remarked, "I really believe I got things right this time". Two more completions have been produced since, by the conductor Rudolf Barshai (2000), and a joint effort by Nicola Samale and Giuseppe Mazzucca (2001). All have been performed and recorded. The version by Samale and Mazzucca was commercially released in 2008 on Octavia Records, through Exton from Japan, with Martin Sieghart conducting the Arnhem Philharmonic Orchestra.
Another new version, by the Israeli-American conductor Yoel Gamzou, was premiered in Berlin in September 2010. Its author conducted the International Mahler Orchestra.
A chamber orchestra 'recreation' by the Italian composer Michelle Castelleti premiered in November 2012 in Canterbury, UK, by the Canterbury Chamber Orchestra under Castelleti's direction.
A piano transcription of the first movement (in the pre-Cooke 1950s UMP edition) was made by the British composer Ronald Stevenson; to this the English pianist Christopher White added solo transcriptions of the other movements in 2010. This composite version (whose last four movements do not follow Cooke's edition at all points) has been recorded with White as soloist.
Recomposition of the First Movement
A project to recompose and recontextualise the first movement using samples and electronic effects was completed by Matthew Herbert and released by Deutsche Grammophon in 2010.