Divas and Scholars

by Philip Gossett

  DONIZETTI (?) TRANSPOSES DONIZETTI:

  LUCIA DI LAMMERMOOR

  Just as Bellini’s most famous single passage in his most enduringly successful opera was written in a different key from the one we are used to hearing, so too were three entire pieces in Donizetti’s masterpiece, Lucia di Lammermoor, including both solo numbers sung by Lucia herself and her duet with her brother, Enrico.37 The pieces involved are (1) the entire scena and cavatina (N. 2) for Lucia (“Regnava nel silenzio”), originally written with the scena beginning in E major and the cabaletta concluding in A major, thus, a semitone above the tonalities of the first and subsequent Ricordi editions (D major/G major); (2) the duetto for Lucia and Enrico that opens the second act (N. 4), which is in A major in Donizetti’s autograph manuscript, but has normally been printed a full tone lower (G major); (3) the entire scena for Lucia (N. 8), from after the close of the introductory section for Raimondo and the chorus (unchanged in E major). The “mad scene” was written in F major, but has usually been printed a full tone lower (E major).38 There is evidence that Donizetti himself was responsible for at least some of these transpositions, but there is also evidence that, given the opportunity, he was prepared to return to his original keys.39

  Thus far, no one has advanced an explanation for these transpositions. When were they introduced and why? Did Donizetti agree to them to accommodate the vocal needs of his original Lucia, Fanny Tacchinardi Persiani? Rather than advancing this hypothesis, those who have studied the career of the prima donna have preferred to invoke the original keys as a way of understanding the character of her voice.40 Yet from the autograph manuscript of Lucia di Lammermoor it is clear that at least the transposition of “Regnava nel silenzio” was the work of Donizetti. Not that the scena and cavatina itself bears any sign of a transposition: Donizetti wrote it in E major/A major without hesitation, and at no point did he make annotations suggesting it should be sung a semitone lower. E major is a suitable key for a scene in a park near a fountain, since this tonality has a long associative history with hunting horns and woodlands: Rossini’s use of it in La donna del lago and Guillaume Tell must have been very much in Donizetti’s ear. And the elaborate harp introduction in E major would have called to mind a similar passage for harp, announcing the first appearance of another doomed heroine, Desdemona in Rossini’s Otello. Furthermore, the passage from the G major of the opening scene to E major provides a good but not excessive contrast (the tonic triads share the pitch g). In all these respects, the transposed D major/G major, with its conclusion in the same key as Enrico’s preceding solo, is poorer. Yet, poorer or better, either E major or D major can follow fluently Enrico’s concluding G major.

  Technical problems arise immediately after the Lucia cavatina, in the recitative for Alisa that precedes Edgardo’s entrance. The first Ricordi edition continues the transposition down a semitone for the first part of this recitative (through Lucia’s “E me nel pianto abbandoni così!”), but when the recitative picks up again with Edgardo’s “Pria di lasciarti” to introduce the duet for the lovers (“Sulla tomba che rinserra”), it abruptly returns to the original tonality. The chord linking the two parts is actually printed once in the lower key, then repeated in the higher key, an absurdity Donizetti would never have sanctioned. But the modern Ricordi edition is no better. After Lucia’s cavatina closes in the transposed G major, it returns immediately to the original key for the recitative, so that Alisa begins with the pitch c. This provides no link whatsoever between the two passages, although it was a perfect link when the the cabaletta (“Quando rapito in estasi”) concluded in its original A major (c is the third degree of the A-major chord). That Donizetti was aware of the problem is apparent from his autograph, since at the beginning of the recitative he revised the opening of Alisa’s vocal line to link the music more smoothly to Lucia’s transposed cavatina (example 10.6).41 Another hand added the phrase “nel tono” (that is, in the original key). While we do not know when Donizetti sanctioned the transposition of “Regnava nel silenzio,” that he participated in no production of Lucia between the 1835 premiere and the French revision of 1839 certainly suggests that the change was introduced for the first performances or soon afterwards.

  EXAMPLE 10.6. GAETANO DONIZETTI, LUCIA DI LAMMERMOOR, THE CONCLUSION OF THE SCENA E CAVATINA LUCIA (N. 2), TRANSPOSED DOWN A HALF TONE, WITH THREE VERSIONS OF THE OPENING OF THE ENSUING RECITATIVE.

  In fact, Tacchinardi Persiani soon abandoned altogether “Regnava nel silenzio,” with its unusually constructed narrative cantabile, a ghost story about an earlier unhappy love affair between an Ashton and a Ravenswood. The very next time she sang the role of Lucia, in a revival of the opera that opened the carnival season of 1837 (26 December 1836) at the Teatro Apollo in Venice, she substituted a much more conventional aria, “Perché non ho del vento,” from an earlier opera written for her by Donizetti, Rosmonda d’Inghilterra, which she had performed in February 183442 Tacchinardi Persiani continued making this same substitution in subsequent years, and other singers followed suit, using both the Rosmonda piece and others. Ultimately Donizetti himself, when preparing his version of Lucia di Lammermoor in French for the Théâtre de la Renaissance, decided to follow the lead of his prima donna and to substitute the Rosmonda aria, as “Que n’avons nous des ailes,” for “Regnava nel silenzio.” Thus, it is the Rosmonda cavatina that introduces the heroine in Lucie de Lammermoor.

  In Lucia di Lammermoor, however, there is no reason whatsoever not to sing “Regnava nel silenzio” in its original key. Nothing in either its recitative or its cantabile strains the voice, nor does the melody of the cabaletta. Only in the concluding cadential runs at the end of the cabaletta theme and in an exposed d in the final cadences of the composition is the singer’s upper register highlighted. The sustained singing Lucia must accomplish in that register in the subsequent duet with Edgardo, “Verranno a te sull’aure,” is considerably more arduous. Furthermore, the single most audacious moment in the melodic line of the cabaletta theme in “Regnava nel silenzio” has been misrepresented in printed editions of the opera. Not only has the theme sometimes been printed incorrectly, but Donizetti’s own alternative to avoid the very highest note has been suppressed. Example 10.7 gives the original version of the passage from Donizetti’s autograph manuscript (which also appears in recent editions), the version of the first Ricordi edition (transposed back up from G major to A major), and Donizetti’s alternative in the autograph. The equivalent of the high d of the original version was erroneously rendered as the equivalent of c in the first Ricordi printed edition (with its transposed key), but the error was corrected in subsequent editions. More important, in his autograph manuscript Donizetti suggested an alternative for the concluding measure of the theme. The triplets at the beginning of the measure and fermata in the second half provide a perfectly appropriate way to avoid the registral extreme without transposing the entire composition.43 Finally, given that the piece concludes either in A major (Donizetti’s original) or G major (the transposition), Lucia can easily interpolate a concluding high note—if she chooses—in either key.44

  EXAMPLE 10.7. GAETANO DONIZETTI, LUCIA DI LAMMERMOOR, SEVERAL VERSIONS OF THE FINAL MEASURES IN THE CABALETTA THEME OF THE SCENA E CAVATINA LUCIA (N. 2).

  The transposition of Lucia’s mad scene presents greater problems. Donizetti wrote the piece in F major, but it is traditionally sung a full tone lower, in E major, and that is the key in which it is printed in the first Ricordi edition and all subsequent ones.45 As originally written (in F major), there is really nothing that cannot be sung effectively by a good soprano. Even Donizetti’s notated passages in which a solo flute and Lucia either play off against one another or interact in close harmony (thirds and sixths) are perfectly idiomatic for the voice.46 As we saw in chapter 9, the interpolated cadenza for Lucia and flute at the end of the cantabile, substituting Donizetti’s short notated cadenza, is based upon a late nineteenth-century interpolation for Nellie Melba, and has nothin
g to do with Donizetti. The composer and his original prima donna might have expected a singer to handle this “cadenza opportunity” with a certain liberty, but it is unlikely that the endless coloratura display in the very highest reaches of the voice would have made sense to a contemporary of Donizetti’s, so different is it from anything else in the opera.

  Why, then, was the mad scene from Lucia di Lammermoor transposed down? In a private communication, Will Crutchfield suggests that for many sopranos the lower key is more gracious to their voice, because those short high notes in the main theme (“Spargi d’amaro pianto” and “mentre lassù nel cielo”) are easier for them to sing with color and resonance. The same is true, he feels, about the repeated, syncopated notes at the end of the theme. He goes on to argue, though, that in the wedding scene Lucia needs to dominate the ensemble, and her part is written in a register equivalent to that of the E major lower transposition of the mad scene, not the F-major original key. If you want a singer who will handle the wedding scene well, in short, you need to perform the mad scene in E major, not in the original F major.

  There may well be some truth to all of this, and it certainly would help explain why the transposition has such a long and enduring pedigree and why it took root even before singers habitually interpolated concluding high notes. But I cannot help suspecting that some singers, at least, who would be perfectly capable of handling the scene in F major, have their own reasons for accepting the transposition. They do it in order to jump around in the very highest register during the interpolated cadenza and then to conclude both the cantabile and, later, the cabaletta on a high e. Even the best of coloratura sopranos (and the role of Lucia was not written for a coloratura soprano) is hesitant to undertake the cadenza with the flute in the higher key or to risk a high f at the end of a long and difficult scene. To flub the cadenza or to crack on a concluding interpolated high note would cast into oblivion everything that preceded; to omit the cadenza or not to sing the interpolated final high note would bring down the fury of the loggionisti on even the finest singer. Much safer to transpose the entire piece down, even if you can sing it in the original key. That Donizetti’s careful citation of earlier themes in their original keys is thereby compromised matters little. That the lovely modulation from the end of the previous chorus (“Oh! qual funesto”) to Lucia’s entrance is thereby sacrificed matters not at all. We come to the opera house to hear concluding high notes, don’t we? Audiences should understand the cynical price they are sometimes paying for those notes.

  Althought the first Ricordi edition of 1837 printed the mad scene in E major, when Donizetti prepared Lucie de Lammermoor in French in 1839, he had the piece printed in its original key, F major. Similarly, the duet for Lucia and Enrico that opens the second act was returned from its lower transposition (G major) to its original A major.47 This evidence need not be taken as definitive (when performing Italian opera the word “definitive” is almost always suspect), but it is certainly worth considering.48 The important thing to remember is that no one is obliged to sing this music in the lower keys. Donizetti would never have sacrificed a singer’s performance by insisting on a key that was awkward for a particular voice, nor should we. But in thinking about transpositions, we should be aware that the composer’s own choices were rarely casual.

  TRANSPOSING BELLINI: LA SONNAMBULA

  AND GIOVANNI BATTISTA RUBINI

  The problem of transpositions in the operas of Bellini becomes intense whenever La sonnambula is mentioned, and in 1999–2000 the Metropolitan Opera of New York risked walking into this hornet’s nest by preparing a new production of the opera featuring Cecilia Bartoli in the role of the sleepwalker, Amina. Although Ms. Bartoli and the Met changed their plans, the problem remains. It is not merely a matter of trying to accommodate a role written for a slightly higher soprano voice than Bartoli’s to her mezzo-soprano instrument (which, in any case, has gradually been ascending into a higher tessitura). No, the core of this opera has been significantly compromised by transpositions introduced into the score, and the music published in all modern editions of La sonnambula is not the music Bellini wrote.

  Only a single source for La sonnambula can be associated directly with Bellini: his original manuscript, as prepared for the singers Giuditta Pasta and Giambattista Rubini when they performed the opera at its premiere at the Teatro Carcano of Milan on 6 March 1831.49 There has been much talk of a “Malibran” version of La sonnambula, as if Bellini himself arranged his opera for the great mezzo-soprano, but no evidence has been produced in support of this hypothesis. In fact, Maria Malibran first sang Amina in Naples in March 1833, at the Teatro del Fondo, before she ever met Bellini.50 When Bellini did finally come to know her in May 1833 in London, she was singing Amina in the Drury Lane Theater (and later, in June, at Covent Garden) in a version “adapted” for the English stage by Henry Bishop. Such adaptations into English were all the rage in London, one of the most famous being the 1830 Rossinian adaptation by Michael Rophino Lacy, also for Covent Garden, Cinderella, or the Fairy-Queen and the Glass Slipper. That was how La Cenerentola was generally known in England and the United States during the nineteenth century, restoring the very slipper Rossini had removed!51

  The story of Bellini’s first encounter with Malibran has been told repeatedly on the basis of a supposed letter from Bellini to his friend Francesco Florimo, the director of the Library at the Naples Conservatory. After extravagantly praising Malibran’s singing in La sonnambula, despite the English translation in which she sang (“the language of birds, or more specifically parrots”), the letter relates that the public recognized the composer sitting in the audience. Everyone sought to congratulate him:

  The first to approach me was Malibran, who, having thrown her arms about my neck, declaimed for me in the most exalted transport of joy these four notes of mine: Ah! m’abbraccia! adding nothing else... My emotion was at its heights: I thought I was in Paradise; I could not speak a word, and remained dumb; I do not remember anything else... The enthusiastic and repeated applause of an English public, which becomes frantic when it gets excited, called us both to the stage; we presented ourselves while holding hands: you can imagine the rest... What I can tell you is that I do not know whether in my whole life I will have a more intense emotion. From that moment I became an intimate of Malibran: she expressed to me all the admiration she felt for my music, and I all I held for her immense talent; and I promised to write for her an opera on a subject that suits her. It is a thought that already electrifies me, my dear Florimo.52

  Florimo, unfortunately, is known to have invented or rewritten many a letter from Bellini, seeking in this way to “correct” the historical record in a way more favorable to his friend, and this scene is probably Florimo’s invention.53 Still, it seems likely that the composer did indeed present himself to Malibran for the first time after hearing her sing the role of Amina in the Bishop adaptation. That he admired Malibran is certain; but at the moment of their first meeting, Amina was already in her repertory, both in Italian and in the English adaptation. There is no evidence that Bellini sought to intervene in her interpretation of La sonnambula, but he did hope to collaborate with her in a future project. The Neapolitan version of I puritani was to be that project, his homage to Malibran. As we have seen, however, that version never reached the stage during the nineteenth century.

  A trace of the English Sonnambula exists in the form of a reduction for piano and voice of several numbers from the opera, issued by Boosey in London later in 1833. They bear the explanatory phrase: “as sung by Mme Malibran at the Theatres Royal Drury Lane and Covent Garden.” The music is often modified in this publication: “Ah! non credea mirarti” (the cantabile of her final aria) is in F minor, transposed down by a major third from Bellini’s original key, A minor, while the final cabaletta, “Ah! non giunge uman pensiero,” is in F major, a fourth below Bellini’s original tonality.54 If it is legitimate to speak of a “Malibran” version of La sonnambula, reflecting the
vocal characteristics of a mezzo-soprano, it must be attributed to Bishop not to Bellini, although the singer could have made similar changes when she sang the role a few months earlier in Naples.

  The problem of tranpositions in La sonnambula, however, arises not because of Amina but because of the tenor role of Elvino. Rubini had what was considered an unusual voice, even in his own day.55 Without entering into the vexing question as to how he was able to sing in the extraordinarily high tessitura that characterized his art (obviously he did not maintain what is usually referred to as a “chest voice” into the upper part of his range), it is clear that composers, to favor his natural or acquired gifts, placed music written or revised for Rubini in a tessitura that the singer found grateful. For La sonnambula, this meant that three pieces were written in keys that seemed even in the early 1830s to be stratospheric. Indeed, one of the most amusing comments about Elvino’s range is found in a midcentury edition published by Novello, Ewer and Co. in England, where the following note appears: “The principal pieces in which Elvino is concerned are here transposed in accordance with Ricordi’s modern edition of this Opera. The voice for which the part was originally composed would seem to be happily extinct; even as it stands now the part is in some places beyond the ordinary tenor range.”56 By this point, of course, the “ordinary” tenor range was measured by Verdi’s practice. The three pieces are (1) Elvino’s cavatina “Prendi: l’anel ti dono,” whose cantabile and cabaletta were both written in B major; (2) the duettino for Amina and Elvino “Son geloso del zefiro errante,” in G major; (3) Elvino’s aria “Tutto è sciolto,” which begins in B minor, modulating to its relative major, D major, and then closes with a cabaletta, “Ah! perché non posso odiarti,” in D major. It is not merely a question of all the high, and often sustained, cs and ds. What is truly difficult is the overall tessitura of the vocal line. In its original form, for example, the cabaletta “Ah! perché non posso odiarti” moves at once to a high f and then sits between that note and a still higher b. From the publication of the very first edition by Casa Ricordi, during the summer of 1831, a few months after the Milanese premiere of La sonnambula, all three pieces were transposed downward: the cavatina and the duettino by a full tone (to A major and F major, respectively), the aria by a semitone (to B minor and its relative major, D major).