The wordless manuscript of D1A contains a small section of this setting – also without an underlay of words. This flurry of activity from circa 1810 suggests that as soon as Schubert, aged thirteen, had abandoned Gabriele von Baumberg’s Lebenstraum after setting 59 lines of it, he determined to try again. The poem, with its symbolic significance for the composer’s future as a servant of the muses, was too important to be abandoned without a struggle. This means that Hagars Klage, hitherto acknowledged as Schubert’s first song, has been displaced not by one, but by two, songs (although both of these are incomplete fragments) which were possibly written somewhat earlier than anything else that has survived from Schubert’s hand. If the Deutsch catalogue were to be completely re-ordered (something which is now hardly practical), D1A and D39 would be strong candidates for D1 and D2. At the very least they are contemporary with the music for piano duet that we have always taken to be Schubert’s earliest surviving works.

Quite apart from the problem of textual underlay, the scholarly reconstruction of this work was made much more difficult by the fact the manuscript was divided between three different sources: the first part is in Paris, the continuation in Louisville (Kentucky), and another fragment in Vienna. This no doubt prevented the piece becoming better known at an earlier time. The correct identification of the text was also delayed. The work was only partly reassembled (without full text and identification of poet, and lacking another few bars unearthed in Vienna) in the Neue Schubert Ausgabe in 1969. And only since Hoorickx’s reconstructions of both these songs has it been possible for performers to compare these two settings of the Baumberg poem.

It is fascinating to imagine the tone of the composer’s youthful self-criticism in this, the first instance of what was to be a lifelong penchant for returning to poems that had eluded him. Schubert was obviously hopelessly over-ambitious in wishing to set the Baumberg poem in its entirety. But it seems that when he began this second attempt, he still intended to do so. If the work was to be kept to a manageable length he would have to cover more ground with less discursive illustration. In the first version – already a long piece – only a quarter of the poem had been set. He now aimed for a tighter construction which kept the story (such as it was) on the move. And there is also a reflection here of the war that was being waged by the spirits of the past and present for the possession of the young man’s musical soul. In D1A he had given his wholehearted allegiance to Gluck, but it was obvious that he should now embrace the lucidity and elegance of something less dramatic and more modern. The very look of the music on the page of D39 suggests the marshalling of another set of musical influences, now including the German ballad as well as opera: in the opening, a hint of watered-down Mozart (the overture to Joseph Weigl’s opera Die Schweizerfamilie begins in a similar anodyne manner in C major); in the recitatives, the newly discovered world of the ballads of Johann Rudolf Zumsteeg.

1. The introduction to the song is reduced to a mere ten bars (as opposed to the 33 bars of D1A). These comprise a modest – and not very memorable – right-hand melody (decorated by turns or mordents) and a flowing pattern of left-hand semiquavers suggesting the galant style with Alberti bass. The sense of mystery and wonder at being in the grove of the gods of D1A is sacrificed to an elegant opening section where the ‘Tempelhalle’ is made of the coolest marble, pleasing in its proportions rather than greatly imposing. The cascading waters of ‘Umrauscht vom nahen Wasserfalle’ are suggested by the tripping melismas of the vocal line: the sense of ongoing unity in the vocal line is not interrupted by pianistic illustration. This seems a pity, for this was one of D1A’s impressive moments. ‘Im sanften Abendschein’ is similarly suave, but the murmuring accompaniment beneath ‘Kein Lüftchen wehte’ seems to illustrate the very thing that the text denies – here there is a sense of air-borne rustlings. (The atmosphere of a later ballad, Ossian’s Cronnan, comes to mind.) On the other hand, for ‘die Sonn’ im Scheiden vergüldete’ the Adagio recitative (which takes over the end of the phrase) effectively mirrors the sun’s descent over the horizon. This is marred, however, by the insertion of a dramatic pianistic flurry of three bars which senselessly separates the verb (‘vergüldete’) from its object (‘die matten Trauerweiden’). This is a rare example of Schubert’s disrupting grammatical sense. He had become carried away by the task of painting the glint of sunlight on the weeping willows.

2. The act of quietly musing (‘still sinnend sass ich lange da’) inspires a recitative which includes the most outlandish modulation of the piece. The rise of a semitone in tonality (C major to D flat minor) has a sudden gear-change at ‘lange’ that might suggest the elapse of years spent in thought, rather than minutes, or perhaps a radical shift of consciousness. Perhaps it was Schubert’s intention to aim for the depiction of this sort of dramatically altered state, but in this piece it is difficult to tell the difference between inspired (and deliberate) eccentricity, and youthful awkwardness. The solemn insistence on the repetition of ‘auf meine Rechte’ within the relatively unimportant phrase ‘Das Haupt gestüzt auf meine Rechte’ seems unjustified. The contrast between future and past (‘ich dachte Zukunft und Vergangenheit’) is more effectively depicted in D1A; here there is more bluster than sense. On the other hand, the repetitions of ‘dem Tron der Götter nah’ emphasise the poet’s awe, and astonishment, at being in the presence of the throne of the gods. The whole of this ‘scene’ in F minor is impressively majestic despite (or because of) its repetitions of the text, as if the spirit of Gluck has briefly returned to give this section, with its pulsing quaver accompaniment, his blessing. A new Andante section beginning in the distant key of B minor, and soon settling into D major, introduces the idea of hymning the singers of olden and newer times. The musical rise and fall of the phrases for the two appearances of ‘Der Sänger alt’ und neuer Zeit’ have a mellifluous and singer-friendly quality about them, with a hint of graceful coloratura. We find writing of this kind (also in D major) in the Goethe ballad Der Sänger of 1815. All of this is gently lyrical and civilised, and very much in the galant style. Perhaps this harmonic continuity (a pedal grounded on D for 35 bars) is Schubert’s way of trying to emphasise that singers, old and new, are united in one great spiritual aim, in one greater tonality. This is probably reading too much into a section in which the sense of drama, a necessary element in keeping works like this glued together, falters and dies. The recitative of the rest of this verse is full of good ideas: the very simple setting of ‘Tod’ – a single crotchet introduced by two keyboard flourishes in contrasting dynamics – isolates death in just the right way for it to be ‘unbemerkt, und längst vergessen’ (unnoticed and long-forgotten); the settings of ‘Riesenschatten der Vergänglichkeit’ (the words are heard twice, as if echo and shadow come from the same bag of musical imagery) are suitably ominous.

3. The new tempo marking is here Adagio in 3/4, the key is C minor, and this is one of the most effective passages in this setting. This depicts the entry of the dark spirit and its attempt to offer the poet the poppy-drink of oblivion. In complete contrast to D1A, Schubert sets these words in such a way as to suggest the rumblings of the underworld: here we glimpse the same mind that was destined to create Fahrt zum Hades and Gruppe aus dem Tartarus – slowly pulsating right-hand quavers are pushed forward by a shuddering motif in the left hand, a device which the composer was to retain all his life, right up to those ominous left-hand trills in the first movement of the B flat piano Sonata, D960. The mood of all this shows how at ease Schubert already was in the playing-fields of the new romanticism where gothic horror danced hand-in-hand with classical mythology. The lolloping triplets under ‘entstiegen einem morschen Heldenmahle’ are splendidly grotesque and prophetic of Der Geistertanz. Now the sinister purpose of the spirit’s appearance is revealed, and the poet panics at the prospect. The sleeping draught is pressed on the narrator, first with undue haste and menace: a Presto vivace recitative accompanied by scurrying piano scales, and then with persuasive charm – a mini aria in 2/4, marked ‘Allegro’ and ‘dolce’ with an unearthly, and sudden, shift from G minor to A flat major. (There is further proof that the ‘Gesang in c’ shares the same text as D39 in that Schubert has already discovered this key-relationship for these words in the equivalent section of D1A). This hypnotic little aria promising relief from pain seems briefly to suspend time, although the horrified diminished seventh on ‘Lethe’ shows that the composer, already well-versed in the classics, knows how to provide a suitably frightened response to the idea of Hades. The temptation to yield to this dangerous wooing is broken by the sound of the terrible word ‘Vergessenheit’ which is separated from the rest of the sentence by a piano interlude, and resounds, unaccompanied, in the poet’s ears. After this, the sudden harsh diminished chord shudders with shock at the possibility of any true artist willingly accepting oblivion. The poet fights back as if to save her life, or at least her art.

4, 5. The rest of the song does not rise to this challenge. Indeed one can immediately sense the waning of the young Schubert’s interest in the project. The recitative beginning ‘Betroffen, wollt’ ich die Erscheinung fragen’ is more simple than the setting of the same words for D1A, but also much less effective. There seems to be insufficient resentment and disdain in the refusal of the sleeping-draught. The recitative marked ‘Vivace’ (beginning with oscillating semiquavers to introduce ‘Doch schon was sie entflohn’) is the insert passage of fifteen bars which was the last piece of the manuscript jigsaw to be put in place. There is nothing very exceptional about this music either, perfectly effective though it is. The dancing entrance of the ‘kleiner Genius’ of Verse 5 is already familiar to us from D1A. Schubert lifted it from the earlier setting but, after copying out ten bars, put a strong line through the passage. His dissatisfaction with the whole enterprise seems reflected (as is the case in a famous crossing-out in the Winterreise manuscript) by the graphical vigour of the deletion. It seems unlikely that the composer continued with this setting on manuscript sheets now lost to us – although this is always possible.

The dating of early Schubert’s songs, particularly fragments such as these, is always fraught with difficulties. Given that that D1A and D39 are probably the earliest of all Schubert’s works, it is difficult to be absolutely certain which came first. The presence of a fragment for D39 on the autograph for D1A suggests that we have got the order correct on this disc. But for a while, Reinhard Van Hoorickx believed that D39 pre-dated D1A, and it is easy to see why; I would be tempted to think the same based on the quality of the music. To my mind, D1A is more successful, and in some ways more interesting, than D39. The lesson that might be learned is that Schubert gave full rein to his responses to words and images in D1A where he was schooled by Gluck’s example, but not inhibited by it. In D39 he attempted something different, more controlled and faster-moving: he aimed for concision and classical lucidity, but he achieved this at the expense of the expressiveness which came naturally to him. He has not yet mastered what Zumsteeg had to teach him. Did anyone give him any guidance or make any comments about the earlier setting? We shall never know.

The fact is that both settings have their problems. And in both cases the composer got to a certain point and felt unable to continue. The stage was now set for a new type of approach to song composition. Schubert decided to take Zumsteeg’s ballads and model his own settings on them, bar for bar, and section by section. This was his self-improving scheme as he composed his next song, Hagars Klage, and his decision to work like this was no doubt influenced by his failure to complete the two Lebenstraum songs. Such humility in one so young is touching, but it is wonderful to think that Schubert should have shown us so much of his intrinsic genius at such an early age, and without having to refer to another composer’s work on his writing table.

There is a biographical note (with picture) on Gabriele von Baumberg accompanying Volume 15 of this series. The reader is also referred to the magnificently detailed chapter on Baumberg in Schubert’s Poets by Susan Youens (Cambridge, 1996).

from notes by Graham Johnson © 1999