Book Read Free

Divas and Scholars

Page 7

by Philip Gossett


  A composer was normally hired by a specific theater to write an opera for a specific season to be performed by a specific company of singers. The theater might turn to one of the acknowledged masters—Rossini, then Donizetti or Bellini, finally Verdi, but also significant were Mayr, Mercadante, Pacini, or Petrella. Most theaters depended on an impresario who controlled artistic and administrative decisions, though he in turn was beholden to regulations enforced by local royalty, political leaders, and censors. Impresarios bore significant financial responsibility for their companies: many suffered tremendous losses, a few grew rich. When capital was accumulated, it did not derive from operatic receipts alone. For a number of years the lobby of the San Carlo in Naples provided theater patrons with opportunities for gambling. The fortune of the great impresario Domenico Barbaja was largely made in that lobby. To keep Rossini in Naples between 1815 and 1822, effectively as musical director of the Neapolitan theaters, Barbaja ultimately offered the composer a stake in the gambling business. In Paris, Rossini talked his way into funding directly from the governmental purse; indeed, he made completion of Guillaume Tell dependent on his receiving a guaranteed pension from the state.

  In the typical Italian theater of the time, the more socially prominent patrons sat in several rows of private boxes built around a horseshoe-shaped auditorium. When not under direct governmental control, theaters were built or maintained by local families, each of which owned one or more boxes. These families constituted a society that hired an impresario to administer the theater, while attempting to keep some control over their investment, both financial and artistic. Such a society differed only marginally from the board of trustees of a modern American opera house. Still other theaters were run for limited periods by individuals who produced operatic seasons in cities without a regular opera company or in competition with more established theaters. An extraordinary venture of this kind occurred in 1830–31, when dilettantes took over the Teatro Carcano of Milan, commissioning Donizetti to write Anna Bolena, and Bellini La sonnambula.

  Fundamental to most Italian opera houses during the first half of the nineteenth century was the operatic “season.” Companies did not provide continuous entertainment throughout the year, but functioned instead during periods often tied to the church calendar. The most important was carnival, which began on the night of St. Stephens (26 December) and continued through the beginning of Lent or even until the start of Passion week, depending on local custom. Where operatic performances continued into Lent, works with a religious orientation were sought. A tradition of Lenten operas in Naples provided the context for Rossini’s Mosè in Egitto, first performed on 5 March 1818, or Donizetti’s opera about the biblical Noah, Il diluvio universale, on 28 February 1830, both prepared for the San Carlo.8 Seasons in different cities were more variable. At La Scala, opera was presented in the autumn, but the theater usually closed its doors for Advent. In Rome, both the Argentina and the Valle would produce opera during carnival, but in the spring or autumn one house would offer operatic entertainment, while the other might feature a company presenting prose drama.

  Singers were normally hired for an entire season, not for a single opera. If three different works were to be performed, the same singers would participate in them all. The artistic ramifications of this commercial organization are palpable. It is no surprise, of course, that composers prepared new works with the particular characteristics of individual singers in mind. Before contracting to write an opera, a mature Donizetti or Verdi would demand to know which singers were in the theater’s company for the season; if the company had not yet been determined, the composer’s contract was often contingent on his subsequent approval of the singers engaged.9 Popular operas written for certain singers, however, would be revived in a later season or a different city with new artists, whose talents might be more appropriate to another of the operas to be performed. Impresarios tried to program a group of works suitable to the available personnel, of course, but they were not always successful. Many of the revisions and adaptations common to Italian operas during this period result from this structural organization of the theaters. The situation was particularly devastating for operas written for an unusual combination of voices. That San Carlo regularly engaged two principal tenors in its company during Rossini’s tenure (Andrea Nozzari and Giovanni David) imposed characteristics on the composer’s Neapolitan opere serie that other opera houses could rarely meet.10 The tormented contemporary history of these works directly reflects the circumstances in which they were composed.

  Into this world stepped the composer, whose contract with a theater required that he would prepare an opera to be performed during a given season, sometimes with a specific date (if it was to open the season), sometimes simply as the second or third opera of the season. The contract might have been signed as long as a year in advance, or it might have been concocted at the last moment. Rossini was commissioned to write L’Italiana in Algeri less than a month before its premiere, when a desperate impresario at the San Benedetto of Venice found himself without the opera that had previously been commissioned from Carlo Coccia. We do not know why Coccia relinquished his commission, but—for obvious reasons—composers who did not fulfill their contractual obligations were rare. On the other hand, Rossini’s contract with La Fenice for Semiramide, which had its premiere on 3 February 1823, was signed six months earlier, in mid-August of 1822.11 But hiring a composer was not enough: his art depended on the words he would set to music. A librettist was needed.

  THE LIBRETTIST AND LE CONVENIENZE TEATRALI

  During the eighteenth century, composers typically set the same basic libretto again and again, often texts by Pietro Metastasio, the Caesarian court poet in Vienna from 1729 until his death in 1782, whose mellifluous texts continued to be used into the nineteenth century. By early in that century, however, new librettos were generally written for most Italian operas. They were often adaptations of spoken drama from the French or German theater. But librettists also found inspiration in Spanish drama or in English Romantic literature, particularly the poems and stories of Scott and the epics and verse dramas of Byron. Fashions changed considerably during the first half of the nineteenth century. Early in the century source plays might predate the derived opera by some forty or even fifty years. Rossini’s Tancredi (1813) is based on a play by Voltaire from 1760, while his Ermione (1818) goes back even further, to Racine’s seventeenth-century drama Andromaque. By the 1830s, Italian composers and librettists were turning to the latest creations of the emerging Romantic theater, and Donizetti’s Lucrezia Borgia (1833) was written the same year as the verse drama by Victor Hugo on which it was based.12

  While it had become less typical for composers to set anew texts previously used by other composers, this common eighteenth-century practice did continue into the nineteenth. Looking for a libretto that would force him to try his compositional skills on the more grisly elements of Romantic melodrama, Donizetti came across the Gabriella di Vergy of Andrea Leone Tottola (written in 1816 for the composer Michele Carafa) and composed his own setting of the text (“for my own pleasure”) in 1826.13 The haste required to get L’Italiana in Algeri on stage made it essential to choose a preexisting libretto, and Rossini employed one written by Angelo Anelli for Luigi Mosca five years earlier. (It is possible that Anelli himself made important changes in the text that brought the libretto within the orbit of Rossinian opera buffa.) Verdi’s Un giorno di regno of 1840 was based on a libretto that Felice Romani had written in 1818 for the Bohemian composer Adalbert Gyrowetz (Il finto Stanislao). For many Italian librettists of the time, French operatic texts were a rich vein to be mined: Antonio Somma’s Un ballo in maschera (1859) for Verdi, derived from Scribe’s 1833 libretto for Auber (Gustave III), is only the most famous example.

  The choice of a librettist generally fell to the management of the theater, and librettists often developed close ties with local theaters: Romani with La Scala, Ferretti with Roman theaters, Gaetano Rossi
and Francesco Maria Piave with La Fenice, and Salvadore Cammarano with San Carlo. They were responsible not only for preparing the text of an opera, but also for providing the modest staging required by contemporary theatrical practice. Most composers accepted the librettist suggested by the theater commissioning the work, but more prominent composers could impose their own choices. Bellini preferred Romani, and the two collaborated on operas first performed in Venice and Parma, but these collaborations (Zaira of 1829,I Capuleti e i Montecchi of 1830, and Beatrice di Tenda of 1833) were as a whole less satisfactory than the work they did together in Milan, where Romani knew better the ambience of the theater and the taste of the public. From Lucia di Lammermoor (1835) until his 1838 departure from Italy for Paris, Donizetti preferred Cammarano for his serious operas, even when first performed away from Naples: Belisario (Venice, 1836), Pia de’ Tolomei (Venice, 1837), Maria di Rudenz (Venice, 1837). In insisting on Cammarano, in fact, Donizetti specifically contradicted the wishes of the Venetian theaters.

  In theory, the choice of subject lay in the hands of the theater and its librettist, and composers were expected to set to music whatever text was put into their hands: there is no evidence, for example, that Rossini intervened in the librettos of his early operas or that Verdi modified the libretto of Antonio Solera’s Nabucco (1842). Increasingly, however, composers participated in these decisions. There is good reason to believe that Rossini played a significant role in choosing the subjects for many of his Neapolitan operas, while Romani certainly consulted Rossini about the subject of their 1819 collaboration for La Scala, Bianca e Falliero.14 Rossi wrote much of the libretto to Semiramide in the villa outside Bologna belonging to Rossini’s wife, the soprano Isabella Colbran, as Rossini then and there began composing the score. Donizetti’s correspondence from the 1830s demonstrates his active participation in choosing the subjects of his operas, and in a number of cases (particularly with comic operas) the composer was his own librettist. During his early career, on the other hand, Donizetti had significantly less choice about his texts. Verdi’s struggles with theaters, censors, and librettists are well known, yet almost from the beginning the composer was actively involved in the search for subjects and usually suggested them himself. In Verdi’s Copialettere—copybooks in which he drafted his correspondence and included other business dealings—there is a page (probably written in 1849) with a list of possible subjects for operatic treatment.15

  Verdi went even further: he frequently provided his librettists with a prose outline of a subject, often called a selva, indicating precisely how he wanted subjects to be organized in operatic terms, with the succession of musical numbers laid out and a great deal of the text specified.16 One of the most fascinating documents of this kind to survive is his selva for the second act of La forza del destino.17 In it he laid out the structure of the entire act, specifying the kind of music he envisaged for each section (whether recitative or a lyrical number—aria, ballata, duet) and often making specific recommendations to the poet as to the verse forms he should use. At the end of the duet for Leonora and the Padre Guardiano, for example, he indicated “lirici di metro piuttosto breve” (lyric verses in a relatively brief poetic meter—i.e., a small number of syllables in each line).18 At the end of the act, where the monks take their farewell from Leonora, who intends henceforth to live as a hermit, Verdi wrote the following words in his selva:

  La Vergine degli Angeli vi copra del suo manto, e l’angelo del Signore vegli alla vostra difesa.

  [May the Virgin of the Angels cover you with her cloak, and may the Angel of the Lord keep watch in your defense.]

  While the words seem to be written as four lines of poetry, they are not metrically consistent and are not rhymed, but the organization of the words on the page suggested that Verdi wanted a quatrain of seven-syllable verse (settenario) here. That is precisely what Piave gave him:

  La Vergine degli Angeli Vi copra del suo manto, E voi protegga vigile Di Dio l’Angelo santo.

  [May the Virgin of the Angels cover you with her cloak, and may the blessed Angel of God vigilantly protect you.]

  A document of this kind makes it clear why Verdi is said to have played such a significant role in the drafting of the librettos for some of his operas.19

  While it was primarily the responsibility of the librettist to derive from the subject a drama per musica, organized into a series of musical numbers, with the poetry written in standard verse forms, during the first half of the nineteenth century many conventions—some fixed, some changing—governed the way the librettist was expected to proceed. Similar eighteenth-century conventions were satirized in Benedetto Marcello’s famous treatise Il teatro alla moda. Donizetti’s wicked Le convenienze ed inconvenienze teatrali (in either its 1827 or its 1831 version) is only one of a string of operatic satires on the subject.20 Some of these conventions reflected the fact that opera companies were assembled for a specific theatrical season. They consisted of a certain number of soloists of varied artistic stature (with a prima donna clearly differentiated from a seconda donna) and vocal register (soprano, contralto, tenor, bass). Each principal singer was expected to participate appropriately in the work, with an adequate number of solo pieces, duets, and ensembles. And these pieces needed to be spread out over the course of the opera, so that there would be ample time for a singer to rest. During the first two decades of the century a secondary singer might also demand his or her moment in the sun, a so-called aria di sorbetto (sherbet aria), whose name suggests the level of audience attention to the piece. But this particular convention, left over from the aria-dominated style of eighteenth-century opera, gradually disappeared as librettos derived from Romantic melodrama became more streamlined and individual musical numbers grew more expansive.

  Some conventions had to do with matters of “priority.” For a certain time, at least, a prima donna presumed she would conclude the opera with a major solo number (often referred to as a rondò, even though its structure by this time had little to do with the classical rondo).21 Librettists, supported willingly or unwillingly by the composer, arranged the action accordingly, knowing that the success of an opera could depend on a contented prima donna. On some occasions the convention functions well: Rossini’s La donna del lago, with its happy ending, and Donizetti’s Anna Bolena, with its tragic one, are excellent examples. At other times the convention may seem to border on the absurd: the original version (1833) of Donizetti’s Lucrezia Borgia, in which Lucrezia sings a spectacular rondò about how much she loved the son she has just poisoned, “Era desso il figlio mio.” For a revival of the opera several years later (Milan, 1840), Donizetti replaced his original ending with a more dramatically pertinent conclusion. Sometimes a composer would take a principled stand. The soprano Sofia Löwe expected that Ernani would conclude with a stand-up-and-sock-it-to-them rondò for her alone, but Verdi, who early in his work on the opera had written to Piave, “For the love of God do not end with a rondò but write a trio: and this trio must be the best piece in the opera,” would not be budged.22 More than one prima donna has protested vigorously the structure of Lucia di Lammermoor, which produces the worst indignity imaginable for a soprano: she must die before the tenor. Unfortunately, the story would permit no other dénouement, and our heroines have had either to settle for a penultimate mad scene or to reverse the scenes and destroy all semblance of dramaturgical structure.

  But more was involved in organizing a libretto than assuaging the sensibilities of singers concerned primarily with the size and scope of their own part, rather than with the quality of the opera as a whole. Ernani, first performed at the Teatro La Fenice of Venice at the end of the carnival season of 1844, serves as a particularly good example for the problems librettists and composers faced in achieving a workable libretto, in this case one based on a revolutionary work of French Romantic drama, Victor Hugo’s 1829 Hernani. Not only was Ernani Verdi’s first collaboration with Piave, who was later at his side for operas such as Rigoletto and La travia
ta, but it was one of the librettist’s first efforts. By 1843, Verdi had had considerable experience in the theater, both as an assiduous operagoer during his formative years in Milan and as a composer (he had written four operas for La Scala, the last two of which, Nabucco and I lombardi alla prima crociata, had been very well received). Thus, he felt obliged to point out serious problems in the libretto Piave proposed, while showering praise on Piave’s poetry (which the composer actually considered secondary to matters such as dramatic action, timing, and structure).

  When Piave began to bristle under Verdi’s criticism, the composer wrote directly to Guglielmo Brenna, one of the administratiors of La Fenice, explaining his motivations. His letter includes a clear, explicit description of what needed to be taken into account when planning a libretto:

  For my part, I would prefer never to trouble a poet to change a verse for me; and I wrote music to three librettos of Solera [the poet for Oberto (1839), Nabucco (1842), and I lombardi (1843)], and comparing the originals, which I have kept, with the printed librettos, only a very few lines will be found to have been altered, and these because Solera himself wanted it.23

 

‹ Prev