Until all contemporary sources are examined, both printed editions and manuscripts, the history of these transpositions will not be fully known. In the case of the cavatina and the duettino, however, printed editions have usually presented the pieces at this same transposition, a full tone lower than the autograph reading.57 The manuscript tradition, however, which more often reflects actual practice in Bellini’s day, is quite diverse. Of the two complete manuscripts of La sonnambula in the library at the “Santa Cecilia” Conservatory in Rome, for example, one (G. Mss. 714/15) transposes the cavatina even lower (down a minor third, to G major), while giving the duettino in its original key (G major). The other manuscript (G. Mss. 716/17) gives both pieces in their original keys.
Ricordi’s early downward transposition by a semitone of Elvino’s aria “Tutto è sciolto,” on the other hand, seemed insufficient from early in the history of the opera. In the Bishop version (later reproduced in several French editions), the key was lowered everywhere by a major third: the aria therefore began in G minor and closed in B major. But in Ricordi’s second edition of the opera, published in 1858, the cantabile is brought only to A minor (down a tone), while the cabaletta is printed in B major (a major third below the original). Thus, where Bellini provided a fluent transition between the cantabile and the tempo di mezzo (with the D-major harmony closing the cantabile serving as the dominant of the G-major harmony that opens the tempo di mezzo), in the Ricordi version regularly used in performance today (which is identical to the edition of 1858), the chords are more remote. Here, too, the readings of manuscript sources are instructive. In the two Roman manuscripts described above, one (G. Mss. 716/17) follows the original keys, while in the other (G. Mss. 714/15) the entire aria is transposed down even further, a full fourth below the original (beginning in F minor and concluding in A major). In this way the entire aria is transposed in a logical and coherent manner. As in the case of Norma, there is no proof that any of these transpositions were Bellini’s work, despite a statement in what may be an inauthentic letter to Florimo.58
That transposition presents an acute problem in La sonnambula results not simply from the tessitura of Elvino. After all, if tenors today cannot handle the original keys of the opera in an acceptable manner, practical considerations far outweigh possible objections to transposing this music.59 While the transposition of solo numbers would not normally affect the remainder of the score, however, in this case they have a profound effect upon the part of Amina. In Elvino’s “cavatina” (the title is Bellini’s), in which he presents a ring to his beloved and sings of his love for her, Amina is present everywhere, a pertichino who reacts with joy to his words and becomes practically a protagonist. Indeed, in printed editions of La sonnambula the cavatina has traditionally (although incorrectly) been labeled a “duet.”60 Even in Elvino’s second-act aria, which lacks a title in Bellini’s autograph manuscript but is correctly called an “aria” in printed editions, Amina has an important part, especially in the cantabile but also in the tempo di mezzo, and her part in the printed editions is unnaturally low. Thus, in all three pieces in which Amina sings together with Elvino, the tessitura of her part is much lower in the edition used today than in Bellini’s original, whereas all the music she sings alone (her cavatina and concluding aria) remains in Bellini’s original keys. In short, as the role is printed in modern editions, Amina is a mezzo-soprano when she sings with Elvino, a soprano when she sings alone. No wonder singers have such a difficult time wrapping their vocal cords around the part.61
Only once in La sonnambula is there direct evidence that Bellini himself changed a key after having completed his score: the conclusion of the first-act finale was originally written in G major, with a transition through a common tone (“g”), played by the horns, between the Adagio (in E major) and in the music leading to the stretta, “Non è questa, ingrato core.” A copyist then rewrote the entire closing section a semitone higher, in A major (Bellini himself seems to have indicated a cut in the cadential phrases in this copy). The copyist’s manuscript was then bound into Bellini’s autograph manuscript of La sonnambula along with the composer’s original, G-major version. In printed editions of La sonnambula, on the other hand, the stretta has always been printed in the higher key, A major, not in Bellini’s G major. Unlike the situation in Norma, the new transition here is clever: the g of the horns in the first measure of the transition is replaced by an e (the tonic of the previous key). This note then becomes the dominant of the ensuing music, now transposed up a semitone, to A major; example 10.8). In the same peculiar letter Bellini is alleged to have sent Florimo in 1834, he supposedly added, “I also lowered the largo and stretta of the finale by a semitone, since Rubini’s voice now makes its effect a semitone lower than the tessitura he required six years ago.” It is hard to believe that Bellini expressed any of these garbled ideas. First of all, Bellini wrote the part of Elvino four years earlier, not six. Secondly, in the Largo of the first-act finale, Rubini’s part in the original version never goes higher than an insistent high g and a, a piece of cake for Rubini and for any tenor specializing in bel canto. As for the stretta, in its higher version (in A major) the voice goes up to b, but no higher. And, finally, if Rubini’s voice had darkened in the way this letter suggests, how could Bellini have written for Rubini one year later (in 1835) the role of Arturo in I puritani with its legendary f in the final ensemble? Whatever the truth may be, if the stretta of the first-act finale of La sonnambula is brought down a semitone from the key in which it has been known in every printed edition of the opera (A major), the resulting tonality, G major, is simply the key in which Bellini originally prepared the conclusion of the first act of his opera.
EXAMPLE 10.8. VINCENZO BELLINI, LA SONNAMBULA, FINALE PRIMO (N. 7), THE TRANSITION BETWEEN THE END OF THE LARGO (“D’UN PENSIERO E D’UN ACCENTO”) AND THE TRANSPOSED VERSION OF THE TEMPO DI MEZZO (“NON PIù NOZZE”) AND STRETTA.
BELLINI ADJUSTS BELLINI: LA STRANIERA
AND GIAMBATTISTA RUBINI
Transpositions in Norma and La sonnambula occurred largely in the transmission of those operas, or were practical decisions taken by Bellini or—frequently—his publisher soon after or even before the premieres. For other operas, however, sources in Bellini’s own hand demonstrate precisely how he went about modifying a previously written part for a singer whose voice was more comfortable in a different range. It is no surprise that he undertook this work for singers he particularly admired, such as Rubini and Malibran. Since the Malibran version of I puritani was discussed at length in chapter 7, let us focus on revisions that Bellini made in La straniera to adapt its principal tenor role, Arturo, for his beloved Rubini.
When Bellini wrote La straniera for the Teatro alla Scala of Milan, where it was first performed on 14 February 1829, he very much wished to have Rubini as his tenor hero, Arturo. Rubini had already created the principal tenor role of Gernando for the first performances of Bellini’s Bianca e Gernando at the Teatro San Carlo of Naples, beginning on 30 May 1826. More important, Rubini’s extraordinary portrayal of the title role in the composer’s Il pirata, at La Scala on 27 October 1827, had contributed impressively to the success of that opera. But Rubini was under contract to Naples for the carnival season of 1828–29, and even though Domenico Barbaja was serving as impresario for both theaters at the time, he was unwilling to free Rubini from those contractual obligations. Thus, Bellini had to settle for a different tenor, the young Domenico Reina. The composer wrote to him about his new opera in September 1828, and the response he received was so encouraging that he hastened to inform Florimo about it:
I wrote to the tenor Reina in Lucca, and he answered graciously, telling me that his voice is virile, always in tune, with a chest register extending from b under the staff to a above the staff and with falsetto notes up to high e, that he has agility and sings evenly, and that his style will adapt itself to my music, which he is prepared to study like a dog and perform just as I wish him to, in my own fashion
, which he understands is the correct one.62
As it turned out, Bellini was indeed pleased with Reina’s willingness to work closely with him in developing the role, and Reina’s performance was well received by the Milanese public. Indeed, the vocal style of La straniera, much more declamatory than the still relatively florid Il pirata, may have been developed in part because Bellini could not count on having a singer who would be able to duplicate Rubini’s style and technique.63
Nonetheless, when La Scala, a year later, decided to include both Il pirata and La straniera in their carnival season of 1829–30, the one significant change in the cast of La straniera was the replacement of Reina by Rubini, who sang in both operas. For that occasion, Bellini rewrote the principal numbers of Arturo in order to adapt them to the unusual vocal qualities of Rubini, significantly raising the tessitura of the role. Despite Bellini’s efforts, Rubini did not please the Milanese in the role of Arturo. The composer himself could not be in Milan to prepare the performances, since he was in Venice composing I Capuleti e i Montecchi, but he told his uncle, Vincenzo Ferlito, on 19 January 1830:
They write me from Milan that on the 13th of this month La straniera was produced, that it gave the same pleasure as last year, and that Lalande and Tamburini truly distinguished themselves; but that Rubini in this opera is cold, and all wanted Reina: some claimed the revised part does not make the same effect as the original: I believe that; but it cannot be so bad as to make them desire a mediocre tenor instead of Rubini. Rather Rubini, by failing to animate a part such as this one, full of soul and fire, must have made everything languid. But enough, in general it has given pleasure, and the rest is unimportant since I did everything I could to accommodate the part for him.64
Although it was the original version that remained in circulation, and not the Rubini modifications, those modifications were undertaken with gusto by Bellini, and they provide a perfectly plausible alternative even today for the right singer.
Bellini entered his changes for Rubini in a copyist’s manuscript of the opera, which still exists in the library of the Conservatory in Milan.65 Not all the changes are in Bellini’s hand: another hand made a few preliminary modifications, which Bellini reviewed and revised extensively. Many of the changes are in recitative. They largely maintain the rhythm and chords of the original, but significantly raise the tessitura, as shown in example 10.9, drawn from the recitative that precedes the first-act trio.66 In the quartet of the second act, Bellini apparently instructed the copyist not to reproduce the part of Arturo in the section beginning “Che far vuoi tu?” Then the composer himself wrote a higher part for Rubini into the score, always remaining close to the original rhythm and always adopting the same harmonies.
EXAMPLE 10.9. VINCENZO BELLINI, LA STRANIERA, RECITATIVE PRECEDING THE TERZETTO (N. 5), AS REVISED FOR RUBINI.
The most extreme changes occur in the duet for Arturo (tenor) and Valdeburgo (baritone) in the second act. Not only did Bellini make small changes in the vocal line, but on four occasions he extracted pages from the original manuscript and provided entirely new ones, with revised vocal lines and orchestral parts. Adjacent measures in the copy are crossed out to ensure a smooth transition from the copyist’s manuscript to the new autograph pages, and back again. In some cases he transposed individual passages. In the opening section of the duet, for example, the Allegro giusto, “Sì... sulla salma del fratello” (Yes, on the corpse of her brother), Valdeburgo first sings a declamatory period in the tonic, B major; Arturo responds in kind, but then the orchestra launches into a lyrical phrase in the same key, which Arturo immediately repeats with an anguished phrase describing his sorrow, “è il dolor d’un cor piagato” (it is the sorrow of a wounded heart). In the revision for Rubini, Bellini (after inserting a cadenza for his tenor) moved the lyrical phrase a minor third higher, to D major. Example 10.10 shows how he inserted the new key (the oboe melody is doubled at the higher octave by piccolo and at the lower octave by clarinet).67 The composer was to use a similar subterfuge to get back to the original key. Later in the duet Bellini transposed another lyrical phrase (“Ah! non sai d’un core ardente il delirio tormentoso” [Ah! you don’t know the tormented delirium of a burning heart]) up a major second, where he felt it would sit best for Rubini’s voice. Despite all these internal manipulations, however, the basic shape and overall tonality of the duet remain untouched.
Bellini’s work for Rubini in La straniera was hardly profound. But it did not have a profound aim: he sought to allow Rubini’s vocal strengths to emerge effectively by giving greater emphasis to his higher tessitura. With some modifications of the vocal line and some transpositions of entire phrases, the composer succeeded splendidly.68 Furthermore, throughout the duet with Valdeburgo, Bellini also modified the baritone part, bearing in mind that the two voices needed to work together effectively. Merely raising the tessitura of the tenor was insufficient. That is a lesson that should be pondered carefully when transpositions are introduced into pieces involving more than a single singer.
If a singer today were to introduce similar manipulations to suit his or her voice, just imagine the reaction of scholars, critics, and loggionisti. For the last production in the Verdi Festival of 2001 in Parma, a great Lady Macbeth, Tiziana Fabbricini, was having difficulty singing Verdi’s final, soft, devastatingly sadistic d at the end of the sleepwalking scene. When I dared to suggest that she could simply take the note down an octave, I was assured that the boos would be so loud and prolonged that the woman’s career would effectively be ruined. At least I hadn’t suggested that she transpose the scene down from the original D major to B major. Bellini (not to mention Verdi) would have been astonished at the fuss and bother. Fabbricini was replaced at the last moment by her understudy, an unexciting singer with a high d that was mediocre but apparently acceptable to the Parma loggionisti.
EXAMPLE 10.10. VINCENZO BELLINI, LA STRANIERA, DUETTO ARTURO AND VALDEBURGO (N. 8), LYRICAL PASSAGE IN ONE KEY FOR REINA, IN ANOTHER FOR RUBINI.
The puzzles we face in deciding the keys in which to perform Bellini’s music (not to mention Donizetti’s) are not simple to resolve, nor will the new critical edition of Bellini’s complete works, launched in 2003 with the publication of I Capuleti e i Montecchi, be able to avoid them, whether in the case of La straniera, La sonnambula, Norma, I puritani, or many other operas. Unless a way is found to accommodate to responsible modern practice the music written for Rubini, the critical editions could languish unused on library shelves. Whether in the context of a critical edition, however, or simply to favor the wishes of a beloved singer, the problem of transposition will remain with us as long as Italian opera is performed. And in every case performers will need to consider the following issues. What is the best way to introduce a transposed section and, afterwards, to return to the original tonality so that the modification appears relatively seamless? Does transposing a particular passage for one singer have implications for the vocal ranges of other singers in the opera, and, if so, what can be done about it? What is the impact of a transposition on the instrumental parts, and how can anomalies in instrumental register be smoothly accommodated? Performing Italian opera well requires its practitioners to seek elegant solutions to intractable problems.
11
WORDS AND MUSIC: TEXTS AND TRANSLATIONS
Now I know what translation means, and I have sympathy for all the awful
translations that exist, because it is impossible to make a good one.
Giuseppe Verdi to Tito Ricordi, 6 July 1855
DO WORDS MATTER? GUGLIELMO TELL
AT THE TEATRO ALLA SCALA
The Teatro alla Scala of Milan opened its 1988–89 season with Gioachino Rossini’s last opera, a work based on the exploits of the famous Swiss hero William Tell, and written for the Opéra in Paris, where it was first performed in August 1829. Riccardo Muti, then music director of La Scala and conductor for those performances, is well known for his continuing determination to employ musical editions a
s close as possible to the most authentic sources and to develop his interpretation from their readings. In this case he chose to work from Elizabeth Bartlet’s newly prepared critical edition of Guillaume Tell, but Muti decided to perform not Guillaume Tell, using the French language in which Rossini composed the score in 1829, but Guglielmo Tell. This Italian translation, with which the composer had nothing whatsoever to do, was underlaid to Rossini’s score during the 1830s in ways that grotesquely modified his music.
Divas and Scholars Page 49