Visiting Composer David Matthews on Mahler 10

©2012 Clive Barda

2017 Visiting Composer David Matthews and his brother Colin played a key role in the preparation of the final versions of Deryck Cooke’s Performing Version of Mahler’s Tenth Symphony. Below is an essay from David Matthews about the Tenth with some insights about his own work on the piece.

Gustav Mahler: Symphony No. 10 – a performing version of the draft, prepared by Deryck Cooke in collaboration with Berthold Goldschmidt, Colin Matthews and David Matthews


When Mahler died in May 1911, Das Lied von der Erde and the Ninth Symphony had not yet been performed, and his Tenth Symphony was left unfinished. The last of his works to reach the public during his lifetime was the Eighth Symphony, which Mahler conducted in Munich in September 1910 – his greatest personal triumph as a composer. Few of those in the audience who heard the heaven-storming sounds of the ‘Symphony of a Thousand’ could have guessed the new paths that Mahler’s music had taken since that work. In 1907, the year he completed the Eighth, Mahler discovered that he had a serious heart disease, and that he possibly had only a few years to live. He wrote to his friend Bruno Walter: ‘at one fell stroke, I have lost everything of clearness and reassurance that I had ever won for myself … now, at the end of my life, I must learn anew how to walk and stand’.

Mahler was, above all, a Romantic composer whose music expresses in an extraordinarily direct way his most intimate feelings about his own life. So it is not surprising that Das Lied and the Ninth Symphony reflect this new attitude. In both works the conflict between affirmation and negation, life and death – a familiar feature in earlier Mahler symphonies – is intensified to an unprecedented degree. In both works the conflict ends with a serene but resigned acceptance of death as an end. The religious faith that Mahler had proclaimed in several of his early symphonies, the humanist optimism of the Fifth and Seventh, or the all-embracing spiritual vision of the Eighth – all were now inconceivable. Against their loss Mahler could only set his continuing joy in the beauty of the natural world, and the tenuous peace of a quasi-Buddhist Nirvana.

Yet the infinitely lingering coda of the Ninth Symphony’s Adagio finale was not Mahler’s last word. In the summer of 1910 he sketched a Tenth Symphony at his holiday retreat at Toblach in the Dolomites, in which the conflict with death was to take a different and more vital form. One crucial reason for this change is again bound up with his personal life, for it was while he was composing the Symphony that Mahler discovered that his wife Alma was having an affair with the architect Walter Gropius (whom she was to marry some years after Mahler’s death). Gropius asked Alma to leave her husband, in a letter that he inadvertently addressed to Mahler. Mahler was devastated: he had perhaps never had a realistic understanding of his young wife, whom he all too easily turned into a Romantic symbol – as for instance Goethe’s ‘Ewig-Weibliche’ (‘Eternal Feminine’) in the Eighth Symphony. His helplessness convinced Alma that she should stay with him. Meanwhile, all his turbulent emotions went directly into the music he was writing, and out of them he drew the strength to reach, once again, a positive resolution: the finale ends with a passionate reaffirmation of his love for Alma. It should be noted, however, that a feeling of renewed vitality is already strongly present in the second movement, which was almost certainly drafted before the marital crisis occurred. Like Berlioz’s Lélio (though with added Mahlerian irony), the Tenth Symphony might be subtitled ‘The Return to Life’.

It was Mahler’s practice to sketch a symphony in his summer holidays and to revise and score it during the winter. But he had no opportunity to complete the Tenth Symphony, as during the winter of 1910-11 he was preoccupied with extensive revisions to the Ninth Symphony, and then came his final illness. He left behind the draft of a five-movement work in a complex state of incompleteness, which may be summarised under four headings:


(a) For the first movement, Adagio, there is a draft full score, from which, had he lived, Mahler would have made a fair copy. The existing score can be played more or less as it stands, though no doubt Mahler would have refined it.

(b) The second movement, Scherzo, exists in a full-score sketch. For about half its length this is fairly complete, but in the latter part the texture is often extremely sparse. This is the most problematic movement, chiefly because it is not entirely convincing in its shape. It is likely that Mahler would have made revisions – perhaps quite drastic ones – to the structure, just as he did with the Scherzo of the Ninth Symphony when it had reached its draft full-score stage (this was one of the things he did in the winter of 1910-11).

(c) The brief third movement: the first 28 bars are written out in full score, and from then onwards there is a short score with enough details of instrumentation to make realisation fairly simple, though for the repeat of the first section Mahler simply wrote ‘da capo’.

(d) For the fourth and fifth movements there is only a short score with a few hints of instrumentation. Both movements, however, are convincing as structures, particularly the finale.

For many years it was assumed that it was impossible to play the whole symphony. At first the sketches were shrouded in mystery, amid rumours that Mahler had asked his wife to destroy them. But in 1924 Alma arranged for a handsome facsimile of the manuscript to be published, and at the same time she asked Ernst Krenek, then her son-in-law, to prepare the first and third movements for performance. There was some hostile criticism of these moves at the time, but the first movement thereafter quite often found its way into concerts. After the Second World War, several musicologists began working independently on realisations of the other movements, including, from 1959, Deryck Cooke.

Cooke first became interested in the symphony while he was writing a booklet for the BBC’s celebrations of the Mahler centenary. Having examined the sketches, he decided to prepare for performance as much of the symphony as he could, for an illustrated broadcast talk. He soon found he could realise much more of the score than he had initially thought possible. The Tenth differs from the majority of unfinished works in that, although the texture is very thin in places, sometimes only a single line, there are no gaps in the music, so no free composition is necessary. What Cooke did was to fill out the texture where this was required, particularly in the latter half of the second movement, where extra counterpoint derived from Mahler’s thematic material had often to be added; to orchestrate the third movement partially (Cooke saw no alternative to following Mahler’s ‘da capo’ literally, using the same orchestration, though he was well aware that there is no precedent for such a literal repeat in other Mahler symphonies) and the fourth and fifth movements entirely; and to tidy up Mahler’s full score of the first movement. The result was that in the eventual programme, broadcast in December 1960, he was able to present virtually the whole symphony, except for some gaps in the second and fourth movements. Berthold Goldschmidt, who conducted this preliminary performance, had given Cooke much expert assistance with the orchestration (though a composer himself, Cooke had little experience of writing orchestral music) and he helped Cooke to revise the whole work after Cooke had decided he could without too much difficulty fill in the gaps in the score. This full-length version was first performed at a Henry Wood Promenade Concert in August 1964, with Goldschmidt again conducting. The first American performance followed in 1965, conducted by Eugene Ormandy, who went on to make the first recording.

It may be appropriate here to record may own small part in the history of this realisation. My brother Colin and I had heard the 1960 broadcast as teenage Mahler fanatics. We got to know Deryck Cooke in 1963, after Colin had written to him pointing out some errors of transcription in his score and making some suggestions for changes in the orchestration. (The BBC Music Library had, with remarkable generosity, lent us the Cooke score after we had gone to see it there.) The three of us became friends and from 1964 until the publication of the score in 1976, just before Deryck died, we collaborated on a comprehensive new revision of the orchestration. We would meet three or four times a year to go over ten or so pages of the score very thoroughly. Colin and I would put forward our suggestions for revision, Deryck would explain his, and eventually we would reach an agreement. Deryck’s concern for detail meant that he never tired of giving careful consideration to even the minutest suggestions for potential improvements.

The problem with the 1964 version was that, effective as it was, the last two movements did not sound altogether like Mahler. Mahler’s orchestration is among the subtlest of any composer, and Mahler himself sometimes laboured for years on small modifications to the scoring of his symphonies. How could we achieve an authentic Mahlerian sound? The first, most obvious thing to do was to increase the size of the orchestra from the original triple woodwind and three trombones, which Deryck had used for the simple reason of economy, to one of quadruple wind, plus an extra clarinet, and quadruple brass. It could easily be deduced from the manuscript that this was the size of orchestra Mahler had intended to use. The larger orchestra immediately enabled us to produce a more authentic sound. We could, for instance, exploit much more those powerful unisons of the upper wind that are so characteristic of Mahler. A specific example of one of our revisions came from the inclusion of a fourth trombone. In the first movement of Das Lied von der Erde Mahler had originally used three trombones and a tuba, but in his orchestral draft score he crossed out the tuba part, and in fact the tuba now plays in only one place in the whole of Das Lied, in the rowdy middle section of ‘Von der Schönheit’. We decided to do something similar in the fourth movement of the Tenth, which is close in the style to the first movement of Das Lied. So we kept the tuba back until the coda, where it makes a dramatic entrance with a descending chromatic scale, ending on the low A, from which it begins its rising-scale solo at the opening of the finale. It’s like a character making his appearance on stage, a dramatic touch that we thought Mahler would have appreciated.

As is often pointed out, Mahler used a large orchestra in his symphonies not so much for sheer volume of sound as for clarifying orchestral texture, and in rescoring the last two movements we always worked towards the principle of clarification. One passage we looked at with special care was the Allegro moderato section of the Finale. Originally Deryck had had quite a basic scoring, with the upper line almost continuously on the violins. We all agreed that it lacked the necessary nervous tension. So we muted the strings, which immediately made the sound more spectral and sinister, changed many of the string chords from arco (played with the bow) to pizzicato (plucked), and added stopped (muted) horn notes, giving much attention to dynamics, especially to sforzando markings. We made the upper line a dialogue between strings and wind. The result is, I think, one of the passages in the symphony that most closely approaches the elusive Mahlerian ideal.

It was only when the symphony was heard in its entirety that it became possible to judge just how far Mahler had progressed since the Ninth. And the quality of what was revealed was such as effectively to silence those purist critics who argued that an unfinished, imperfect work should not be played at all. For, as Deryck Cooke wrote:

Mahler’s actual music, even in its unperfected and unelaborated state, has such significance, strength and beauty, that it dwarfs into insignificance the momentary uncertainties about notation and the occasional subsidiary pastiche-composing, and even survives being largely presented in conjectural orchestration, so long as Mahler’s characteristic widely-spaced texture is faithfully preserved. After all, the thematic line throughout, and something like 90 per cent of the counterpoint and harmony, are pure Mahler, and vintage Mahler at that.


Nevertheless, it is important for the listener to bear in mind exactly what it is he or she is listening to. Cooke called his realisation ‘a performing version of the draft for the Tenth Symphony’, and he was always concerned to stress that it was in no sense a completion, since only Mahler himself could have completed the work. We were aware, while we were revising the score, of the paradox that what we were trying to make as perfect as we could was something that was intrinsically imperfect. And we were also aware that the closer we got to what we believed was a Mahlerian sound, the more likely it was that the audiences would forget that what they were listening to was only the realisation of a draft and accept it simply as Mahler’s Tenth Symphony, which it could never quite be. Yet we all believed the risk was worth taking.




The symphony is cast in two parts. The first two movements are both in the key of F sharp, and could almost stand independently as a two-movement work. The first movement is a 25-minute slow movement, an unprecedented way to begin a symphony; but it follows on directly from the last movement of the Ninth. The dark, searching Andante theme for violas alone which opens the work alternates with a consoling Adagio for strings and trombones, related not only to Mahler’s Ninth but also to Bruckner’s, and to a common background in the prelude to Act 3 of Wagner’s Parsifal. The whole movement, indeed, is a Parsifal-like quest, whose ending is a precarious and temporary haven. For midway through the movement there occurs a great outburst in A flat minor, leading to a terrifying nine-note dissonant chord, which casts its shadow over the rest of the movement, and is not exorcised until the Finale.

The F sharp major Scherzo, however, is full of confidence: a busy neo-Classical movement with constantly changing time-signatures. It has a relaxed Ländler trio. Mahler had not written a scherzo so free from malice since the Fifth Symphony; and its coda, with its exultantly whooping horns, directly recalls the scherzo of the Fifth. So ends Part 1 of the symphony.

If, as seems likely, the first two movements were drafted before the crisis in Mahler’s marriage, which occurred at the end of July, then the little B flat minor third movement, which Mahler entitled ‘Purgatorio oder Inferno’ (the last word has been crossed out), was his immediate reaction to the shock of the discovery. In several places in the manuscript Mahler wrote anguished exclamations over the music, including at one point ‘Erbarmen!’ (‘Have mercy!’ – the wounded Amfortas’s cry in Parsifal) and, at another, ‘Tod! Verk!’. This latter allusion is to the scene from Act 2 of Die Walküre where Brünnhilde appears before Siegmund to tell him he must die, known as the Todesverkündigung (‘Annunciation of Death’). The Todesverkündigung begins in F sharp minor, then moves into A flat minor. One of its brass motifs is strikingly similar to the A flat minor brass chorale in the F sharp minor first movement’s central outburst, which is clearly also an Annunciation of Death, and the tonal parallels are surely too close to be coincidental. Possibly this passage was inserted into the first movement after the July crisis: the existing sketches suggest that it was an afterthought.

The structural function of the third movement is to act as a prelude to the last two movements, and to supply them with most of their thematic material. The fourth movement, a lamenting fast waltz in E minor, is closely related to the ‘Trinklied’ (Drinking Song) from Das Lied von der Erde. A wild inscription on the title-page begins ‘Der Teufel tanzt es mit mir’ (‘The Devil dances it with me’). At its start the music is forceful and energetic, but it gradually exhausts itself under the weight of its climaxes, and at the end the dance runs down to a standstill, finishing with the dead sounds of percussion (Mahler’s instrumental indications here) and, last of all, a loud stroke on a muffled bass drum. Alma Mahler has described how Mahler heard this sound at a fireman’s funeral in New York, and how he watched from his hotel window, his face streaming with tears. The experience deeply affected his death-obsessed mind. ‘Du allein weisst was es bedeutet’ (‘You alone know what it means’), he wrote for Alma on the last page of the score.

The Finale brings redemption. It opens in D minor, in utter darkness. Fragments of theme grope upwards from the depths of the orchestra, to be beaten down relentlessly by repeated strokes of the muffled drum. Then in sudden calm a solo flute (another of Mahler’s own indications of instrumentation) plays a long, arching melody of serene and strange beauty, a metamorphosis of elements from the third and fourth movements. Violins take it up with a glorious modulation to B major and move slowly to a climax, whose apex is violently cut off by the drum stroke. The darkness of the introduction returns, heralding a central Allegro moderato – nervously agitated music derived from the third movement. Its middle section develops the flute theme, with a noble continuation for high trumpet (here too Mahler has indicated the specific instrument), after which the crisis comes quickly with the return of the dissonant chord from the first movement. It is followed by the opening theme of the symphony, played by the horns. A gentle modulation to B flat brings back the music that the flute had initiated: all tension subsides, and the climax of a long tranquil ascent is a radiant restatement of the flute theme in F sharp major on all the violins. The last pages are full of tenderness, and over the final passage Mahler wrote Alma’s name, as a last gesture of love. It is no coincidence that these final pages are reminiscent of the Adagietto from the Fifth Symphony, Mahler’s first music inspired by Alma. It was by recreating her image in its original purity that Mahler was able to come to terms with the trauma of her infidelity, and to end his life’s work with a victory of love over death.


©David Matthews