The cycle's name appears in large letters as Abendröte. The first poem is without title and is simply marked 'Erster Theil' ('Part One'). On his mansucript of the March 1823 setting of this poem (D690, in the original key of A major) Schubert has copied Schlegel's heading exactly: 'Abendröthe—erster Theil'. This is one argument for the theory that the composer intended that this song, the last to be composed from the set, should stand at the head of a grand cycle. In any case, this is how a single Lied came to be known under the poet's collective title of Abendröte.

Thus ends the first part of Schlegel's cycle which has taken place at twilight where the mountains, the birds, the river, the rose, the butterfly, the sun and the winds have all described their feelings and spoken to us in their own words. Schlegel has also permitted human beings, a boy and a shepherd, to be part of the natural tapestry where, in words from the introductory poem (Schubert's Abendröte) 'the whole world becomes a single choir singing many a song with one voice'. And all these are speaking to the poet, the elect interpreter of these sounds, who has divined their meaning. Thus the poet's own poem (Der Dichter) serves as something of an epilogue to Part I. When in 1856 Ruskin dubbed the ascribing of human emotions and sympathies to nature 'the pathetic fallacy', he preferred 'the very plain and leafy fact of a primrose' to 'blossoms that breathe and twigs that pant with pleasure' (in the words of a parody of the time). Nevertheless, many other Romantic poets, including Wordsworth, Shelley and Tennyson, have employed the pathetic fallacy, like Schlegel, in order to emphasise the unity of creation.

The second part of the cycle takes place after night has fallen. There is another reflective poem beginning 'Als die Sonne nun versunken', again without title and not set by Schubert, and once again 'Der Dichter' supplies the epilogue.

It is possible to see that Schubert had found the poems by the beginning of 1819, and set three songs from Part II of Schlegel's cycle then. About a year later he moved back to Part I for certainly three (and probably six) songs from that set, as well as returning to Part II to set Die Sterne. At the same time, he set Der Schiffer D694, which is a poem by Schlegel outside the Abendröte cycle. And then three years later, in 1823, Schubert returned to set the very first poem in the cycle, Abendröte. It is just possible that he had decided at this late stage to try to make a cycle of the songs he had already composed, and this is why he at last set the lyric which gives the unifying tone to Schlegel's group of poems; this was after all the year of Die schöne Müllerin when he was taking stock of his achievements and thinking in terms of cycles (the ballad Viola written in that year also represents an attempt to make a larger structure from what might have been seen as a succession of smaller songs). But it seems more likely that the Schlegel settings represent 'a passive flirtation with some form of cyclic arrangement that yet refuses to emerge' (Kramer). Also interesting is the point made by Berke that the composer found the poems set in 1819 in the Musen-Almanach für das Jahr 1802, and that the 1820 settings were taken from one of the later collections of Schlegel's poems—the 1809 edition pictured above, or the Vienna edition of 1816; the songs were written before the appearance of the Gesamtausgabe. This is not quite the same as the composer's experience with Winterreise (which Schubert first encountered in the almanac Urania in a version with only twelve poems) in that, despite small differences between the two versions, Abendröte was printed complete in each publication (in the Musen-Almanach it is not yet divided into two parts), and the composer was aware of the scope of its cyclic nature from the start. He simply seems to have enjoyed setting these poems singly or in small batches, although Schlegel's conception is arguably best realised in cyclic terms when the texts are gathered together, and the various 'characters' in the set are to be found 'singing many a song with one voice'. This last phrase is Schlegel's fervent expression of Romantic pantheism, but to a musician it could also seem to be a pithy demand for a song cycle.

This 'one voice' is a problem for us in a way that it might not have been for Schubert. We know that Vogl sang Ellen's song Ave Maria to rapturous reception in 1825, so that it seems likely male singers of the time would have been unembarrassed to perform Das Mädchen or Die Rose in public. Today, however, we have more inhibitions in this direction, and Fischer-Dieskau performs neither song in his vast survey of Schubert. At the present time it seems more likely that the Schlegel songs could all be sung by a female singer - she could get away with Der Knabe as a Hosenrolle, and Der Wanderer poses no problem in a musical climate accustomed to the female Winterreise. But what type of female singer?—for here we must consider the question of voice types, and vocal Fach. Could Schubert really have envisaged the same voice singing Die Rose, poised so high in the stave, and Die Sterne where the tessitura is completely different—not to mention the low notes at the end of Abendröte? Whoever sings Die Sterne in its original key of E flat will find it difficult if not impossible to sing the next song in Schlegel's sequence, Die Gebüsche, in its original key of G. The 'failure of cycle' seems to me to be Schubert's failure to have a single singer in mind, or even a single singing type, when conceiving these songs in their original keys. A song like Die Berge for example seems suited to a bright young tenor, others to a light soprano, and yet others to a high baritone. Only by subjecting Schlegel's descriptive diversity, in the process of composition, to the practical considerations which would have enabled these poems to find vocal and cyclic unity in the person of a single singer, could Schubert have made Abendröte into a cycle as we now understand it. Admittedly, by transposing down the higher songs it is possible for a mezzo with 'gravitas' to encompass the cycle as a whole, both in terms of singability and textual suitability (more or less); but it still seems to me unlikely that Schubert ever thought about this as a possibility. The performer is faced here with a similar problem as in Berlioz's Les Nuits d'Eté, another so-called cycle in which it has to be admitted that different songs suit different sexes and voice types; every performance in which all songs are sung by the same singer is something of a musical compromise.

Nevertheless it seems to me that modern-day performers who appreciate Schlegel's poetry (and Schubert's music), and who sense an elusive yet demonstrable unity behind the poems and music, have the right to salvage something of a cycle from these songs. The question is how. One solution is to share the Abendröte set between two singers, or perhaps even more. Although the best-known shared cycles come after Schubert's time (Schumann's Myrthen and Hugo Wolf's Italienisches Liederbuch), Schubert was to publish sets of songs which called for at least two singers - the Gesänge aus Wilhelm Meister D877 for example, where both Mignon and the Harper participate, albeit unequally. In the Op 52 Lady of the Lake songs we have a larger cast: the famous Ellen's Gesänge are interspersed with songs for men's chorus, female chorus, tenor (Normanns Gesang) and baritone (Lied des gefangenen Jägers). Thus we have a type of operatic casting within a Lieder framework. It would be possible for Abendröte to be performed by a number of singers, a small operatic cast with a Lieder agenda, but here it would make best sense if each and every song were performed by a different singer keeping his own character, with the pianist as common denominator. The sense of unity might be enhanced by readings of the unset material whereby the character of the poet himself ('Der Dichter') could provide the thread which binds the work together. One could even envisage an adventurous college (or collage) performance of these songs given as a 'masque' in costume, a reversion to the Liederspiele of Wilhelm Müller's days in Berlin.

One practical difficulty, even when the songs are shared, is the unprepossessing and boring sequence of major tonalities of the songs (A - G - A - A - B - F - F) when grouped in Schlegel's order. A prime consideration for concert performance should be (as far as possible) a comfortable and interesting juxtaposition of tonalities, speeds, and moods. On this disc we have retained Schlegel's ordering as far as possible, but we have separated Die Vögel and Der Knabe because they are in the same key and mood. It may be argued that Schubert intended their enchainment, as did Schlegel, but in a set of predominately slow songs, putting the two merriest items together ensures a long sequence of slow songs elsewhere. Here the soprano sings Das Mädchen and Die Rose, as well as Der Fluss, the silvery colour of which (in the quintessentially Schubertian key of B major) seems particularly threatened by downward transposition. Every performer will rearrange the songs in a different way, and it seems unlikely that a definitive order will be found to suit every voice-type and every team of singers. It nevertheless seems desirable that the song Abendröte should begin the set, and that Die Gebüsche should end it. (In the same way it is impossible to improve on Wolf's decision to start his Italienisches Liederbuch, a work where the songs are often shuffled in performance, with the motto-like Auch kleine Dinge, and to end it with the celebratory Ich hab' in Penna.)

Under the fingers of the pianist this accompaniment feels different to any other Schubert song. This is chiefly because the left-hand trills associated with sunset (cf. the opening of Freiwilliges Versinken) are awkwardly placed (often in the fourth finger of the left hand) and ubiquitous; indeed this is one of only a handful of Schubert songs where embellishment becomes an important expressive feature in its own right (cf. also Hippolits Lied and Pause). The independence of each strand of the accompaniment (which is what makes those left-hand trills difficult) suggests the layout of a string quartet as well as the music of an earlier time. The song is cast as a slow dance in 6/8, and the ornamentation in both the vocal line (e.g. at 'scherzen munter') and the piano part recalls the grandeur of early eighteenth-century music. There is also in the left-hand line a strong suggestion of passacaglia or ground bass, an impression strengthened by the ever-recurring sound of the trills on the third quaver of each bar. As in many a set of variations in the old style, the internal note values quicken as the song progresses: quavers prevail in the accompaniment for the first twenty bars of the song, semiquavers for the next seventeen, and demisemiquavers for the last eighteen. The basic tempo, however, remains that of a stately procession, a measured moto perpetuo of the kind that this composer often writes when describing the stars in their courses and other ineffable workings of God and Nature. And of course the poem is also very grand in that it describes the regally repetitive rhythm and melody of life pulsating imperceptibly behind the activities of beast and man, something all too easy for the insensitive to miss. The use of quasi-contrapuntal texture seems to fit this concept: it is, after all, only the trained ear that can make sense of a layered texture of this kind, where each voice has its own part to play, at the same time as contributing to the glory of the whole. Only the true and poetic musician is able to pick out and hear inner melodies hidden to non-initiates, at the same time as enjoying the beauty of the combined strands: he can appreciate both the particular and the general at the same time.

The first appearance of the singer, like a solo instrument adding its sound to the texture of a Bach cantata, reinforces our feeling that in this song the human voice is one among many, although primus inter pares. 'Tages Arbeit ist vollendet' inspires a duet with the cellos, and the next line, about children at play, with the flutes. There is a distinct (and rather old-fashioned) feeling of a middle section in the relative minor at 'Milden Balsam hauchen leise in die Lüfte nun die Blumen'. These six bars, the first three of which are lightened in colour and made more delicate by the fact that both the pianist's hands are in the treble clef, return to the tonic key; 'Der die Seele zart berühret, wenn die Sinne selig trunken' repeats this process, relative minor back to tonic. The birds are then delicately placed in the treble clef, with a shift to the intimacy of the subdominant. All this while the trills in the piano have illustrated different things as the song progresses: at the beginning, the slow turning of the globe towards night as it shudders on its axis; at 'Und die Kinder scherzen' (where the trills transfer for the first time to the treble clef) the high sound of children's laughter; rustling breezes are evoked at 'Milden Balsam hauchen leise', and (particularly enchantingly) birdsong at 'Kleine Vögel'; the mountains of 'Berge, himmelan geschwungen' inspire a reappearance of the ominous bass trills which we have first heard in Auf der Donau (1817) where the poem contains the words 'alte Burgen ragen himmelan', a related image about castles soaring heavenward which prompts Schubert to move, in mirror fashion, to the depths of the keyboard. These rumblings introduce the song's final section beginning 'Und der grosse Silberstrom' where the rippling watery demisemiquavers of Liebesbotschaft from Schwanengesang are heard for the first time; these, in combination with more trills in the left hand, herald a true peroration, a drawing together of all the images, and a magisterial summing-up. When the singer reaches 'Und das All ein einzig Chor', with its matching drops of sevenths in the vocal line, it is as if he has fathomed the meaning of life, and is transfigured by the revelation. The trills here betoken shivers of delight, the vibrating music of the spheres, the glowing of the molten fires at the globe's centre. At 'aus einem Munde' the singer plunges into the depths as if to commune with the earth gods, speaking his last tones in the tessitura of an Erda or a Sosostris. If Schubert's music sometimes reminds us of Britten's in its textural clarity, these final pages awash with shimmering figuration and swirling in heady abandon bring to mind the Tippett of The Midsummer Marriage and The Heart's Assurance. The music's refusal to reach the tonic until the end of the penultimate bar adds to a state of rapturous suspense, miraculous even by Schubert's exalted standards. There is nothing quite like this greatly undervalued song anywhere else in the repertoire. What a magnificent beginning it might have been to an 1823 Schlegel cycle.

from notes by Graham Johnson © 1996