Not all performance conditions are ideal, to be sure, and Rossini himself had to compromise on many occasions. As we saw in chapter 3, the absence of an appropriate cellist at the small Teatro Re in Milan seems to have compelled the composer to rewrite “Per lui che adoro” from L’Italiana in Algeri with a solo flute in place of a solo cello. But what happened in Italy when Guillaume Tell returned from the French capital to Rossini’s native land? In Paris, his 1829 opera was composed for and performed by a string section of 12 first violins (in addition to the violon principal, who also directed the orchestra), 12 second violins, 8 violas, 10 violoncellos, and 8 double basses.64 In the overture, furthermore, Rossini scored the initial Andante for five solo cellos, with an additional five cellos playing two accompanying parts, three cellos on the first, two on the second. Since not a single Italian theater during the 1830s or 1840s had ten resident cellos, the overture to Guillaume Tell was certainly not heard on the peninsula as its composer conceived it.65 For a large opera house such as the Teatro alla Scala of Milan, there would have been enough cellos to cover the five solo parts, and it would have been relatively satisfactory to have the two accompanying parts played by some combination of cellos and double basses. Few theaters throughout Emilia-Romagna in this period, however, employed as many as five cellos, so that further compromises would have been necessary.66 How significant might this have been? It all depends on your point of view. I wonder whether any viewers shared my discomfort with a scene in Franco Zeffirelli’s 1988 film Il giovane Toscanini (The Young Toscanini).67 As Toscanini and a traveling orchestra make their way across the Atlantic on their way to Brazil, a musician dies. When his body is committed to the waves, the other members of the orchestra gather on deck, and we witness a solemn moment of musical commemoration: the string players are all sawing away (including the violins), the flutes are fluting, and the trumpets trumpeting. From the sound track, however, come the strains of the opening of the William Tell overture, that is, a passage for cellos alone (give or take a double bass).
EXAMPLE 12.6. GIOACHINO ROSSINI, TANCREDI, FINALE PRIMO (N. 7), VIOLA PARTS, MM. 26–28.
The problem was not always the presence of an insufficient number of string players. Later in his life, as orchestras grew larger, Verdi feared that in a particular theater the size of the ensemble might overwhelm the singers. In a letter of 1885, his erstwhile student Emanuele Muzio explained to Tito Ricordi that the composer felt there were too many players in the orchestra of the Teatro Apollo in Rome for a projected revival of Don Carlo. In Muzio’s words:
Our maestro, who wrote Il trovatore and Ballo in maschera for the Apollo in Rome, also finds that the orchestra of that theater is too numerous, and that it must certainly damage the vocal part of the operas. He is not able to write and perhaps run the risk of a refusal if he asked for a reduction in the number, but you and Lamperti can state that the opinion of the master is such that he believes the following orchestra would be more than sufficient for the Apollo.68
Verdi’s suggestion was for a maximum of 12 first violins, 10 second violins, 8 violas, 8 cellos, and 7 double basses, an ensemble with five fewer string players than Rossini used at the premiere of Guillaume Tell.69
There were other factors. Early in the century it was typical in Italian opera houses to divide the string players so that half of them sat on the left side, as seen by the audience, and half on the right. Sometimes this meant dividing equally first and second violinists, sometimes placing the first violinists on one side of the audience and the second violinists on the other, while dividing violas, violoncellos, and double basses.70 The musical purpose was to produce a sound that would seem uniform throughout the theater. With his wide European experience, though, by the late 1850s Verdi had come to favor the typical northern European seating, in which most of the string families were kept together (but sometimes with the double basses arranged on both sides of the orchestra). Apparently he imposed such a seating at the Teatro San Carlo of Naples when he staged Simon Boccanegra during the fall of 1858, but a decade later the theater had returned to the older Italian system, as Verdi complained to Ricordi in a letter of 5 April 1869:
Just imagine [...] that in the orchestra there are still violas and violoncellos scattered one here, another there, etc., etc. This is a vestige of olden times, when violas and violoncellos were always in unison with the basses, and it has not yet been eliminated, despite the demands of modern works, where the violas and violoncellos play independent parts like all of the other instruments. What is even stranger is that ten years ago I had these instruments united together as a group for Simon Boccanegra. It seems that the innovation was found to be absurd.71
He repeated this complaint in a letter to Francesco Florimo of 23 June 1869:
How can you tolerate that the violas and violoncellos are still not placed together? How can there then be a downbow, dynamic shading, accent, etc.? Besides, the group of string instruments will lack fullness. This is a vestige of former times, when violas and violoncellos played in unison with the double basses. Wretched customs!72
Surely no one is suggesting that in “historically informed” performances we must divide the string families spatially for operas from Rossini through middle Verdi, but unite them for operas beginning with Simon Boccanegra. In any event, Verdi’s historical explanation is faulty: while there are passages where violas and cellos are col basso in Rossini’s operas, there are also many important passages where each family of string instruments contributes in a distinctive fashion to the orchestral sonority.
More influential still on orchestral balance is the physical location of the players. While northern European opera orchestras were regularly placed in “pits” (Wagner took the process further at Bayreuth, submerging many players below the stage), Italian orchestras throughout the nineteenth century were positioned on the same level as the audience. Indeed, the Teatro alla Scala did not introduce a lowered orchestra pit until the 1907–8 season, their aim being “to obtain a more integrated sound with the singers on stage.”73 But is such a sound—perfectly reasonable for Pelléas et Mélisande or Salomé—necessarily appropriate for an opera by Donizetti or an early work by Verdi? Might other values be more important, such as the possibility of more interaction between orchestral musicians and singers? Such interaction may have been particularly significant in performances led by the principal violinist. When an effort was made in 1984 to stage Ernani at the Teatro Comunale of Modena in a way approximating its first Venetian performance 140 years earlier, the orchestra pit was closed and the orchestra was seated at audience level, although a modern conductor (Roberto Abbado) took responsibility for the proceedings.74 Two of the three major performing venues at the Rossini Opera Festival, the Sala Pedrotti and the sports stadium turned Palafestival, have no orchestra pit. Only the more traditional Teatro Rossini has a standard pit. The impact on orchestral sound is considerable.
In short, there is no simple response to the question of how many strings should be used in a performance of an Italian opera written during the first two-thirds of the nineteenth century. It depends on the specific repertory, it depends on the instruments being employed, it depends on the physical space in which the modern performance is taking place. No musician should find this surprising.
ALTERNATIVES FOR MODERN PERFORMANCES:
TWO DONIZETTIAN EXAMPLES
If a critical edition takes as its task not simply to produce a single performing text of an opera, but instead to explore the sources and transmission history of nineteenth-century Italian operas, it can suggest interesting and sometimes compelling alternatives for modern performers. Here are examples from two of Donizetti’s most beloved operas: many others could be adduced.
Glass Harmonica or Flute?
Since at least 1941, when the autograph manuscript of Lucia di Lammermoor was issued in facsimile, it has been widely known that Donizetti originally wrote a part for glass harmonica in Lucia’s mad scene.75 But Donizetti crossed out this part a
nd substituted a flute before the first performance (Naples, Teatro San Carlo, 26 September 1835), and no reference to the glass harmonica is found in nineteenth-century sources. Some critics have not considered this earlier notation particularly significant.76 Bini and Commons, on the other hand, write:
The quivering, supernatural sound of this instrument would have been perfect, we believe, for communicating the sense of madness, but for some reason we cannot now know—perhaps because no player of the glass harmonica was available or perhaps because the sound was too weak to be heard in the Teatro San Carlo, the part of the “armonico” in the autograph was decisively canceled and rewritten for flute.77
There is indeed something otherworldly about the glass harmonica, as Heather Hadlock has shown, for the instrument was quite explicitly related in the minds of the public with women, with mesmerism, with madness:
The history of women, trance, and the armonica coalesces in the Mad Scene of Lucia di Lammermoor, where the instrument’s sound was originally intended to conjure up anxieties about young women’s vulnerability to nervous derangement, taboo eroticism, and alienation from healthy, normal society. This use of the instrument is no sign of charlatanism, however, but a genuine invocation of the uncanny.78
Mary Ann Smart, too, speaks of “the far more uncanny glass harmonica that Donizetti originally wanted.”79 Nor was this the first time Donizetti had employed the instrument. As Emilio Sala has pointed out, there is a glass harmonica in the composer’s Il castello di Kenilworth (Naples, Teatro San Carlo, 6 July 1829), associated with another “female victim unjustly persecuted.”80
Since 1970 there have been several successful attempts to restore the glass harmonica to the mad scene in Lucia, whether in the theater or on recordings.81 Yet over these efforts has hung the suspicion that perhaps it is wrong to employ that instrument when Donizetti himself replaced it with a flute. Study of the autograph manuscript makes it clear, however, that Donizetti’s part for glass harmonica is beautifully laid out in every detail, with extensive articulation and, twice, the characteristic indication “ondeggiante” (wavering or undulating), an indication that would make no sense for a flute and, indeed, is not copied into the flute staff. There are some lovely interactions between the glass harmonica and the flute, too, interactions that are lost when the harmonica part is rewritten for flute. The harmonica appears during the introductory recitative (to introduce “Il dolce suono”), returns extensively in the Larghetto (“Ardon gl’incensi”), where it accompanies Lucia’s notated fioritura, and then is found again in the cabaletta “Spargi d’amaro pianto.”
Recent research during preparation of the critical edition of Lucia di Lammermoor has finally clarified the historical situation.82 Donizetti apparently wrote the part for a specific performer, Domenico Pezzi, who rehearsed the music not only with the original Lucia, Fanny Tacchinardi, but with the full orchestra. Pezzi, however, had sued the theater concerning his services for a ballet entitled Amore e Psiche, first performed earlier that year at the Teatro San Carlo on 30 May 1835.83 Court papers preserved in the Neapolitan state archives make clear that there was bad blood between Pezzi and the theater, which at some point during the run of twenty-six performances of Amore e Psiche had substituted a flute for Pezzi’s glass harmonica. Surely Donizetti was told to avoid Pezzi. With no other experienced player available, the composer fell back on a flute.
Performers today should feel free to choose between the two alternatives, and the critical edition of the opera will make both versions available. Those who wish to preserve the traditional approach to the mad scene will continue to use a flute, while those who are intrigued by the mysterious aura of the glass harmonica will want to go back to Donizetti’s original inspiration. Scholars must clarify the history and make both versions available; performers must make choices for particular performances under particular circumstances.
Horn or Violoncello?
After five quick measures, in which the entire orchestra scurries from the tonic (D major) to an unexpected dominant of F major, Donizetti focuses his overture of Don Pasquale (first performed at the Théâtre Italien of Paris on 3 January 1843) on one of the most beautiful melodic inspirations of the opera, the serenata in the garden that Ernesto sings to Norina, “Com’è gentil,” which opens the final scene. In the third act of Don Pasquale the melody will be presented by the tenor alone, in A major, accompanied by two guitars and tambourines, but in the overture the tune is played in F major, a pastoral interlude before the beginning of the principal movement in D major, derived primarily from the melody of Norina’s cavatina “So anch’io la virtù magica.”
EXAMPLE 12.7. GAETANO DONIZETTI, DON PASQUALE, SINFONIA, ORIGINAL HORN SOLO, MM. 6–38.
Example 12.7 shows the melody in the overture as Donizetti originally conceived it, played throughout by a solo horn, tuned “in C” (that is, sounding an octave below the written pitches).84 From the appearance of the autograph manuscript, it would seem that the melody was doubled by a solo bassoon at A2; otherwise A1 and A2 were essentially unaccompanied. At B1 and B2 pizzicato strings provided a simple chordal accompaniment, while at A3 this accompaniment continued, with bassoons now sustaining notes of the chords.85
This is not the orchestration as it is generally known. Donizetti himself modified his autograph manuscript, so that A1 and A2 are played by a solo violoncello (with the solo bassoon continuing to join the melody at A2). While B1 and B2 remain a horn solo, A3 is played by a solo flute and violoncello, with the horn joining only in the second half of the phrase. In its original form as a horn solo, this is not a simple melody to play, largely because of its insistence on the written melodic f in the middle register, already in m. 7, a “stopped note” on the natural horn, obtainable, according to Berlioz, only by “closing the bell (the lower aperture of the horn) to a lesser or greater extent with the hand.” While some of these notes are of “good” or “very good” quality, according to Berlioz, this f “is muffled,” an unfortunate quality for a note to which the melody returns again and again.86
And yet, thanks to the work of Bini and Commons, we now know from contemporary reviews that the entire melody was originally played in Paris by the solo horn. In Le Corsaire of 9 January, for example, we read, “After a charming overture in Re whose adagio, played by the horn, announces well the subject of the story, there is a romance for Tamburini [Doctor Malatesta].”87 The reviewer for La France Musicale of 8 January says, “The clarinet and the horn begin the adagio, whose theme is the first phrase of a melody full of tenderness and originality that Mario [Ernesto] sings in the third act.” while the reviewer might have confused the ensemble of wind instruments, he is unlikely to have mistaken a solo violoncello for a horn. A third review, however, suggests that Donizetti had good reason to make a change: the horn may not have accomplished his task very well. Les Coulisses of 5 January, in a distinctly hostile review, refers to the passage as being played by “the bassoon, and very badly by the horn of Forestier.”
Donizetti apparently decided that a prominent passage for unaccompanied horn at the very beginning of his opera was simply too risky, and hence introduced the modification in his orchestration before leaving Paris on 7 January for Vienna.88 With the knowledge that this change was made after the premiere, a modern production might certainly consider experimenting with Donizetti’s original orchestration, especially with the availability of a horn with valves.
Examples of this kind allow us to glimpse the realities of the musical forces at the disposal of Italian composers during the first half of the nineteenth century. Though they can offer no sure guide to modern performers, they do reveal the extent to which the composers’ problems remain our problems. Only by learning to recognize the elements of these problems and to appreciate the tentative and evolving solutions that composers undertook can we deal responsibly with instrumental problems of our own time.
Before concluding this chapter, I want to address one final question: in performing nineteenth-cent
ury Italian opera, how should one accompany secco recitative in the modern opera house?
SECCO RECITATIVE
While there is a considerable literature about the accompaniment of secco recitative in the eighteenth century, relatively little has been written about the nineteenth. That is understandable, of course, since until recently only a few nineteenth-century Italian operas using secco recitative were regularly performed, and those were exclusively comedies (Rossini’s L’Italiana in Algeri, Il barbiere di Siviglia, and La Cenerentola, as well as Donizetti’s L’elisir d’amore). Nineteenth-century opera buffa was largely thought of as a retrograde art, given the overwhelming presence of the serious, even tragic operas of Rossini (after his move to Naples in 1815), Bellini, Donizetti, and Verdi, which were generally accompanied by the orchestra throughout. Thus, it was easy to assume that eighteenth-century traditions of accompaniment lived on, unchanged, wherever composers continued to employ secco recitative between closed musical numbers.
Divas and Scholars Page 58