This work was commissioned by the Holland Festival for the Schönberg Ensemble. The concepts for generating musical light effects are radicalised, and rhythmic patterns are used in a more expansive way. At the outset, the composition develops from a six-pitch aggregate which permutes constantly within each of the orchestra groups (winds, brass, percussion and strings), given new colour accentuation through voice exchange and change of position. This permutation procedure is also used individually in every orchestra group. Asynchronicity predominates simultaneously among the groups; piano, celesta, harp and vibraphone (as instruments with “built-in decrescendo”) are active between the other orchestra groups, pulsing through large portions of the composition with rapid pitch sequences. The pitch inventory broadens in the course of the work (which lasts ca. 30 minutes); at the same time, the continuous voice mixing is joined by rhythmical mixing with relatively rigidly motionless chords among the orchestra groups. Feldmann used this compositional technique in a way analogous to the method of producing Turkish carpets in the 19th century, in which the patterns were hooked together with innumerable tiny loops, resulting in irregularities due to the nature of the materials used. No sonic shape seems untransformed. They are newly colourised – even later, when discrete groups of bars are repeated ever more frequently. These repetitions are varied mostly by marginal differentiations, not simply “do it again.” The music attains its inexorable self-absorption through a mixture of orchestration and minuscule rhythmical shifts when forming chordal sounds. There are no leaps, no margins (like many of Rothko’s pictures) – there is only the limit of the composition’s beginning and end. Feldmann died on 3 September 1987 in Buffalo New York, only a few months after finishing For Samuel Beckett. Martin Hufner

Webern wrote his last instrumental work in 1940. It was premiered in Winterthur in 1943, conducted by Hermann Scherchen, with Webern in attendance; it was the last public performance of one of his works which he was able to attend. Once again, they are variations, a form which especially suited Webern’s use of dodecaphony and which facilitated his amalgamation of several types of form, due to their comparative lack of restriction. Once again, his style took another decisive turn: his orchestration is more “orchestral.” For instance, doublings at the unison of a melody line occur again and again; although the mixing covers up the instruments’ distinctive sound, it does contribute much to the work’s aural transparency. The importance of the harmony has grown considerably; broad chordal expanses predominate long passages, such as mm. 21 – 55 of the first variation. In a letter to Willi Reich in 1941, Webern gave a terse analysis of the piece, “which, in its overall result, is a kind of overture:” “The theme of the variations … is conceived periodically, although it has an ‘introductory’ character. – Six variations follow … the first fully unfolds the overture’s main theme (andante form), so to speak; the second is the transition, the third a subordinate [or “second”] theme, and the fourth is a recapitulation of the main theme – it’s an andante form! – but in a developmental way; the fifth, reiterating the fashion of the introduction and the transition, leads to the coda, the sixth variation. Everything that happens in the piece is based on the two ideas appearing in the first and second bars (contrabass and oboe), except reduced still more, since the second actor (the oboe) is already in retrograde: the second pitches are the retrograde of the first two, but in augmentation. Then the trombone again plays the first actor (contrabass), but in diminution and in retrograde of the motifs and intervals. So that is how my row is built.” Manfred Angerer

With his third opera Jen?fa Leoš Janá?ek succeeded in making his breakthrough as an operatic composer. Since the premiere of this moving story about the fate of the sexton and her stepdaughter Jen?fa at the Brno National Theatre in 1904, it has become one of the composer’s most frequently performed works. Janá?ek was the first to succeed in transforming everyday speech directly into music. His method of using speech-melodic motives is clearly distinctive. It is known that the composer preserved everyday conversations in the form of little musical sketches: ‘Jotting down genuine speech melody is, as it were, music’s life class,’ he said. Janá?ek’s third opera, commonly known in the non-Czech world as Jen?fa, was based on the play Její pastorky?a [Her Stepdaughter] by Gabriela Preissová (1862-1946). Preissová’s ‘tale of Moravian folk life’ was given at the National Theatre in Prague on 9 November 1890. The next year a new production was given in Brno (10 January 1891), and the play was published by the Prague firm Fr. Šimá?ek. Janá?ek had already had dealings with Preissová as the writer on whose short story his second opera, Po?átek románu [The Beginning of a Romance], 1891, had been based and seems to have approached Preissová for permission to set the play even before The Beginning of a Romance had been staged in Brno, in 1894. Preissová recalled in a memoir fifty years after the event how ‘the highly talented, quick-tempered Moravian Leoš Janá?ek applied to me. He said that he had fallen in love with Jen?fa and already whole sentences of it rushed into his mind which he immediately dressed with his music. He did not need to put anything into verse, the words and sentences apparently spoke with their own music fully in accord with his. We came to a happy arrangement.’ In fact Preissová at first tried to put Janá?ek off: ‘I think that the material of P[astorky?a, i. e. Jen?fa] is not suitable for musical setting – but perhaps in time we’ll find something more suitable’ (Preissová to Janá?ek; 6 November 1893). Janá?ek, however, was single-minded in his choice and, in Preissová’s words, soon ‘began to dress the action of Jen?fa with his passionate endeavour. He studied the cries of young men at their folk dancing, he went off to the mill where he listened to and took down the noises of the turning and rumble of the mill wheel.’ With Janá?ek’s letters to Preissová lost, the earliest history of Janá?ek’s attrac­tion to the play is difficult to document. So is his actual composition of the work. Janá?ek usually dated his autographs, but in this case he destroyed the autograph. Asked to provide an account, Janá?ek wrote the following in a letter to Otakar Nebuška dated 22 February 1917: All that was possible to gather from the old manuscript about the beginnings of Jen?fa is as follows: 1. My copyist Josef Štross (in his time an excellent oboist from the Prague Conservatory during the directorship of D. Weber [Friedrich Dionys Weber] noted only when he finished Act 1 of the vocal score; I then rubbed it out. I don’t know why. 2. Act 2 of the vocal score was completed (by the copyist) on 8 July 1902. 3. Act 3 of the vocal score he completed with the words ‘End of opera’, 25 January 1903, 3.30 p.m. It should be noted that I compose first in full score and do the vocal score from that; thus work on the full score was finished earlier. Between Acts 1 and 2 there was a long break. At that time I was working with Fr[antišek] Bartoš on folksongs published by the Czech Academy. My maid remembers that in her second year with us I began compos­ing J. P. [Jen?fa]; i. e. in 1896. For me then composition was done only on the side: being choirmaster and organist, imperial and royal music teacher at the Teachers’ Training Institute, director of the Organ School, conductor of the Beseda Philharmonic concerts – to have at home a mortally ill daughter – and [day-to-day] life. In short it was hard to compose, and thus little was done. Therefore it’s also hard for me to remember. There might be a date hidden somewhere in my copy of the first full score; I don’t have it to hand. I don’t possess the [original] manuscript of the full score. A witness to the composition of the opera was Janá?ek’s maid, Marie Stejskalová (1873-1968), mentioned in his own account. In her memoirs she recalled the following: When I went to [work for] the Janá?eks, the master was beginning to write Jen?fa. [...] He seldom had time for it during the day, but he devoted all his free evenings to it. He rarely stayed out longer than he had to: at concerts, in the theatre, in the Readers’ Club, in the Old Brno Beseda – he never hung around anywhere, and while others went to sleep when they got home, he sat down to work. In the morning I brought a lamp filled to the brim with paraffin into his study, the next day I took it away empty. The mistress would look at it: ‘He’s been writing through the whole night again.’ Today I find it strange that the whole of Jen?fa was written by the light of a paraffin-lamp. Sometimes it seemed to me that the master was battling with Jen?fa, as if he went into the study not to compose but to fight. He got up from supper, stood and thought a moment and, really more to himself, sighed: ‘Lord God and the Virgin Mary, help me!’ [...] In those happy days when we were still all together, the master would often talk about Jen?fa: what he was currently working on, how he thought it might continue, whether the work was going well for him or not. He said this with such fire that he convinced us all what a great work it would be. We were quiet as mice whenever he played through on the piano what he had written, often we crept up on tiptoe to the door of his room, and all three of us would listen. Little Olga wouldn’t fool around, she wouldn’t even laugh aloud when her father was working. These accounts can be supplemented and to some extent corrected by dates that Janá?ek inserted in his copy of the play (end of Act 1: 18 March 1894; end of Act 2: 17 January 1895; end of Act 3: 11 February 1895), and by brief references to the work in Janá?ek’s correspondence. An important date is 31 December 1894, when, according to the last page of his copy of the play, Janá?ek completed the prelude, originally intended as the overture to Jen?fa. A comment in the play at the beginning of Act 2 is especially interesting: ‘instrumentation begun 16 February 1895’. Although Janá?ek wrote his first two operas in piano-vocal score and then orchestrated them, we have his word (in the above letter to Nebuška) that he wrote Jen?fa straight into full score – his normal practice in subsequent operas. However, it is clear that at least some of the score was worked out in a very rough two-stave version, as for instance he did with sketches of his later, unfinished opera, Paní mincmistrová [The Mintmaster’s Wife], 1906-1907. A single page has survived with decipherable fragments of Act 1 Scene 2. This suggests that Janá?ek made sketches in this fashion, obviously finishing them on the dates noted in the play at the end of each act. If this is so, then this stage would have been completed by 11 February 1895 (taking in a full version of the overture on the way), and that a few days later, on 16 February, Janá?ek began detailed work in full score. This appears to be in conflict with Janá?ek’s statement: ‘My maid remembers that in her second year with us I began composing J. P.; i. e. in 1896.’ But Janá?ek got Marie Stejskalová’s starting date wrong: she entered the Janá?eks’ service on 27 August 1894, so that her second year would have been 1895. From Janá?ek’s letter to Nebuška above we know that ‘between Acts 1 and 2 there was a long break’. What is not known, however, is when the break began and thus when Act 1 was completed, at least in its initial state. In addition to his many commitments as a teacher, several other major pre­occupations took up all of Janá?ek’s spare time from 1896 onwards. Janá?ek himself mentioned his work with Bartoš on the collection of folksongs, Národní písn? moravské v nov? nasbírané [Moravian folksongs newly collected], JW XIII/3. Proofs began arriving in September 1898, but for such a vast enter­prise (we are talking about a book of 1200 pages) it is clear that collecting and organizing must have been going on for a couple of years before that. Another demand on his time was Janá?ek’s proof corrections of his harmony manual O skladb? souzvuk?v a jejich spoj?v [On the composition of chords and their connections], JW XV/151, a process mostly completed by 21 March 1896. What with the beginnings of a large-scale creative project (the canta­ta Amarus) in the summer of 1896, it would seem that Jen?fa had been tem­porarily abandoned well before then. A psychological turning point may well have been the Brno première, on 16 January 1896, of Tchaikovsky’s opera The Queen of Spades. Although such elements were reduced in Janá?ek’s revisions in 1906-1907, Act 1 of Jen?fa still contains the semblance of a number opera, with identi­fiable arias, duets, a trio, choruses and even a ‘largo concertato’ – a slew of operatic conventions not that far from those in Cavalleria rusticana. Janá?ek was much taken with Mascagni’s opera when it was first given in Brno in 1891 and wrote a long and enthusiastic review of it in Moravské listy (JW XV/137). What particularly struck him was the ‘destructive passion’ in the encounters between Lola, Turiddu and Santuzza and it is perhaps the sim­ilar jealous triangle of Jen?fa, Laca and Števa that attracted him to Preissová’s play. An additional attraction were the folkloristic scenes it offered – it was easy enough for Janá?ek to adapt his earlier arrangement for chorus and orchestra of Zelené sem sefa [I have sown green], JW III/3, as the basis for the Recruits’ scene since this was territory familiar from the folkloristic adapta­tions in his previous opera, The Beginning of a Romance. So far Jen?fa presented no insuperable challenges to Janá?ek. The intense enclosed drama of Act 2 – no chorus and just four characters – was quite another matter, but Tchaikovsky’s opera offered new solutions. Again there is the evidence of a long and enthusiastic newspaper review by Janá?ek (JW XV/149); the fact that by this time he was no longer a regular critic makes his review especially significant. The key to his fascination is his description of the Herman-Countess scene: ‘Jerky, fragmentary, it lacks tightly linked big tunes. The orchestra simply throws up random piercing notes in all direc­tions. And yet the composer’s highly developed musical thought weaves all these tiny particles into such a magnificent whole, with such an overwhelm­ing effect, seldom achieved in all of musical literature.’ This new method, as Janá?ek saw it, of composing an operatic scene without ‘big tunes’, and instead with ‘tiny particles’ organized by the orchestra, pointed the way for his development as an operatic composer. In Act 1 of Jen?fa the musical dramaturgy is slow (much of the time the characters sing melodic paragraphs rather than sentences), and those unaware that Janá?ek was writing to a prose libretto might be surprised to learn this fact, given the carefully struc­tured nature of the vocal lines. This was achieved by extensive surgery on Preissová’s prose text, thereby creating a sort of quasi-verse to provide regu­larly structured lines of text to fit to the regularly structured music. When, in Act 2, musical organization was delegated to the orchestra, this type of rearrangement of Preissová’s text was no longer necessary. Such a radical rethinking of his operatic style did not come easily. One can build up a good case for Janá?ek’s stopping Jen?fa soon after hearing Tchai­kovsky’s opera in January 1896 and, a few months later, trying out the new techniques in a smaller work (but nevertheless for similar forces: orches­tra, soloists, chorus) – his cantata Amarus. The experiment seemed to have worked: Amarus is the first major completed work by Janá?ek that sounds as if it belongs to the mature composer, and is particularly striking in its new bal­ance of orchestral motives and vocal line. Perhaps it was because of this new style that he sent it to his friend and mentor Antonín Dvo?ák, the first piece that he had shown him for eight years. And Dvo?ák, once Janá?ek had prised a comment out of him, made an influential comment: ‘The piece is interest­ing particularly from the harmonic respect, only I would like more melody and then perhaps rather more correct declamation (Dvo?ák to Janá?ek, 21 May 1897). ‘Correct declamation’ and ‘melody’ soon fused together strikingly in Janá?ek’s creative armoury when, a few months afterwards on his Hukvaldy holiday in August-September 1897, he began to write out systematically what he later called ‘speech melodies’ in his notebook. These were fragments of overheard everyday speech, notated in ordinary musical notation (showing rhythm and approximate pitch), often with annotations about the time of day, circum­stances, emotions depicted and so on. Instead of adhering to the current models of ‘correct’ Czech declamation that he could have found in the operas of Dvo?ák and Smetana, Janá?ek established his own theory of declamation by doing field work – not so very different from the way that his collecting and studies of Moravian folksong had earlier transformed his compositional style. The fact that most of the speech melodies that he notated and stud­ied were from Moravia allowed him, at least until it became a disadvantage rather than an asset, to claim that his compositions of the time, including Jen?fa, when he returned to it, were ‘Moravian’. Speech melodies thereafter dominated Janá?ek’s discourse and were described and demonstrated in the many articles he wrote about them up to his death. As far as Jen?fa was con­cerned this new way of looking at vocal lines allowed him to renounce such operatic props as leitmotifs and later to assert that leitmotifs did not figure in the opera at all. The discovery of speech melodies and the new Tchaikovskian model of orchestra-led operatic composition took place in the early part of Janá?ek’s five-year sabbatical from writing Jen?fa. Act 1 had been written without the benefit of these insights. With Amarus and other smaller works, he began to explore how to incorporate them in his musical and operatic style and by the time he returned to the opera in 1901 he was a substantially differ­ent composer. The first indication that his thoughts were returning to the opera appears on the envelope of a letter sent to him in Hukvaldy by his daughter Olga on 30 December 1901, on which he jotted down a voice part to words from Act 2 Scene 3. This is corroborated by a brief reference to the work in a letter to Olga (17 April 1902): ‘I am working very hard so as to finish the second act before the [summer] holidays.’ But as Janá?ek got back to the opera, its composition was overshadowed by the greatest personal tragedy in his family life. On 22 March 1902 Janá?ek escorted his surviving child, Olga, to St Petersburg. His brother František had settled and married there and had invited his niece to spend half a year in St Petersburg to improve her Russian. From an early age Olga had had prob­lems with her health, which meant that she could not pursue a career as a schoolteacher, the traditional calling of generations of Janá?eks. There was also another reason for sending her off to Russia. Now getting on for twenty, she was strikingly handsome and lively, and had several beaux. Janá?ek con­sidered the most recent as particularly unsuitable and had forbidden further communication. When the young man retaliated by issuing threats of killing her, it seemed a good idea to get her out of town for a while. So Janá?ek left Olga in St Petersburg, returning to his busy life as a teacher at the Teachers’ Training Institute and at the Brno Organ School. In the little time left over for composition Janá?ek resumed work on Jen?fa. Within days of writing to Olga on 17 April 1902 Janá?ek had news from St Petersburg of his daughter’s illness from typhoid fever. Her recovery was charted in a series of telegrams and letters which her uncle František sent to Brno; finally Olga began writing herself. For a while she seemed to be making a reason­able, if slow, recovery, but after a relapse her mother set off immediately to St Petersburg to help nurse her. Zdenka Janá?ková departed on 11 June, leav­ing her husband to be looked after by Marie Stejskalová. In a letter to Olga and her mother in St Petersburg, dated only ‘Sunday 6 a.m.’ but probably written on 22 June 1902, Janá?ek reported: ‘So I got down to work – until I finished Act 2! At least when we meet the holidays will be more pleasant for us.’ Act 2 thus seems to have been written, apart from preliminary sketch­ing in 1894-1895, between late 1901 and the early summer of 1902, and was copied by Štross in piano-vocal score (apparently after completing the full score) by 8 July 1902. Soon, however, there was rather more disturbing news from St Petersburg, as is clear from one of Janá?ek’s letters to his daughter. It is undated but was probably written on Monday, 7 July, a few days before the end of the school term and before Janá?ek’s departure for Hukvaldy, where he usually spent the summer holidays with his family. Dear Olguška I’m crushed by the repeated sad news about the fever. Ask the doctor whether they wouldn’t allow you to travel in a sick state. Perhaps the different air would immediately stop the return of the fever. The jour­ney wouldn’t be bad – you could lie down. I’d come for you. Your grieving father It was against this background that Janá?ek composed the third and final act of Jen?fa. Instead of spending time with his convalescing daughter as he had hoped, he spent the first half of the holidays waiting for bulletins from St Petersburg. So, unlike his later method of writing in intensive bursts with long reflective pauses in between, Janá?ek simply went on with the next act, working more intensively on a single composition than he had ever done before. At the end of July Olga was at last deemed well enough to travel and with great difficulty completed the journey to Hukvaldy. In mid-September Janá?ek returned to Brno to teach, joined a little later by Olga and his wife Zdenka. The summer stay in Hukvaldy had not worked a miracle cure and as the winter settled in, Olga continued to get worse. By Christmas it was evident to Olga herself that she was not long for this world and she began to take leave of friends and family. Janá?ek himself slipped off to Hukvaldy straight after Christmas to get away from the morbid atmosphere, and prob­ably to work without such distractions on the last pages of his opera, com­pleted, if the testimony of the piano-vocal score is correct, on 25 January 1903. On Sunday, 22 February 1903, Olga received holy unction. That after­noon, as Mrs Janá?ková recounted, we all sat by her. My husband had then just finished Jen?fa. During the whole time that he was composing it, Olga took a huge interest in the work. And my husband also used to say later that his sick daugh­ter was his model for Jen?fa. Now Olga asked him: ‘Daddy, play me Jen?fa, I won’t live to hear it.’ Leoš sat down at the piano and played ... I couldn’t bear it and ran off again to the kitchen. Four days later Olga died. Janá?ek continued to tinker with his opera, but finally put a date at the end of his copy of the play: 18 March 1903, the third week after the terrible mortal struggle of my poor Olga. Completed. The work was dedicated to her and when she was buried he placed a sheet from his manuscript of the opera in the coffin. 20 years later Janá?ek wrote in his memoirs: I would bind Jen?fa simply with the black ribbon from the long illness, suffering and laments of my daughter Olga and my little boy Vladimir.