Divas and Scholars

by Philip Gossett

  During the 2002–3 season, I had the opportunity to work with two fascinating productions in Scandinavia, in which major new “versions” of operas by Verdi and Rossini were successfully performed. While Gothenburg’s Gustavo III was the “hypothetical reconstruction” of an opera that Verdi had “completed” (the composer’s word) in 1858, and Helsinki’s Il viaggio a Reims involved considerable rewriting of the libretto by Dario Fo, each production demonstrated that the cooperative efforts of performers and scholars need not stop at the establishing of composer-sanctioned scores and their theatrical realization. In the best of senses, the theater is a site for play: so let’s play.

  RESTORING VERDI’S GUSTAVO III

  Sweden’s National Opera

  The new opera house is one of the most characteristic buildings on the waterfront in Gothenburg, Sweden’s major southwestern port and its second largest city. Founded in the seventeenth century to give the country a military foothold near the North Sea, where battles with adjacent Denmark and Norway were common, the town soon became a significant trading center. It attracted merchants from England and the Netherlands, who helped establish its East India Company and provided some of the economic strength needed to develop a major industrial and commercial area. As the town grew, a series of semicircular canals and ring roads provided a frame through which urban life could sprawl. The Gothenburg Opera, which opened in October 1994, is perched at the very center of this design. While many of the older buildings within the innermost canal are constructed of the yellow bricks that dominate the appearance of old Gothenburg, the side of the opera house facing the city is largely concrete. Its wondrous seaside facade, however, opening into the lobby where patrons enter the theater and congregate during intermissions, is made of glass, and its shape calls to mind the great ships that used to ply the port. The city’s shipbuilding days are largely over, although some large freighters and even passenger ships continue to use the port, but maritime museums, floating restaurants and hotels, and architectural landmarks (a parking garage in the shape of a tanker, an office building that resembles a smokestack constructed from Lego building blocks) nostalgically invoke past glories.

  Opera had been produced in Gothenburg with regularity since the nineteenth century, often by visiting companies, but the art form became an ever more central part of cultural life during the twentieth. Performances were held in the Grand Theater, built during the 1850s, but its small seating capacity and inadequate stage prompted efforts to provide Gothenburg with a modern opera house. After almost fifty years of discussion, in which some of Sweden’s most prominent performers, including Birgit Nilsson and Jussi Björling, were involved, the new theater was finally erected. During 2002 the city continued beautification efforts around its imposing presence, constructing a tunnel for the city’s principal roadway so that the space in front of the theater could become a park. Culture continues to matter in Sweden, and in Gothenburg the opera house is arguably the most important cultural site.

  Ever since the Gothenburg Opera opened its doors, there had been talk of producing Verdi’s Un ballo in maschera. The work had been tentatively scheduled for 1998, but the theater learned that the rival Royal Opera in Stockholm also intended to offer Ballo that season. Rather than mount a competing production, the theater asked the stage director already hired, Anthony Pilavachi, to substitute La traviata. As it turned out, Stockholm changed its plans, so that neither house produced Verdi’s Ballo that year. Bear in mind that Un ballo in maschera is not just any opera in Sweden. Despite a native tradition of operatic composition, Ballo is as close as the Swedes come to having a “national opera,” since everyone knows that Verdi’s work was based on a historical event: the assassination of that most European and cultivated of Swedish monarchs, Gustaf III. Indeed, one of the treasures in the fabled collection of musical manuscripts assembled by Rudolf Nydahl, housed today in Stockholm, is the autograph manuscript of the 1833 French opera by Daniel François-Esprit Auber, to a text by Eugène Scribe, Gustave III;3 this libretto, in turn, was the direct source for the libretto of Verdi’s Un ballo in maschera.

  There are star-crossed works of art, born under circumstances that challenge their very integrity. Donizetti’s Maria Stuarda and Poliuto, ravaged by Neapolitan censorship, could be rescued simply by locating and editing the original, complete manuscripts (that of Maria Stuarda, in fact, is in the same Nydahl collection in Stockholm): and so Maria once again refers to Elisabetta as “vile bastard” and Saint Polyeuctus goes to his martyrdom on the operatic stage. Undoing the censor’s last-minute effort to improve the moral tone of Rigoletto by pretending that the Duke has not shown up at Sparafucile’s inn looking for the assassin’s “sister” required no more than looking with care at Verdi’s original manuscript and realizing that his text had been altered by another hand. Even after Verdi decided that Stiffelio, with its explosive mixture of sex, hypocrisy, and religion, would never be permitted to remain in the repertory without extensive revisions, he carefully preserved almost all the pages he had removed from his manuscript, allowing scholars some 150 years later to reconstruct the work in its written form precisely as he had envisioned it.

  But Un ballo in maschera has quite a different history, and some of that history has been known for a long time.4 During the autumn of 1857 Verdi and his librettist, Antonio Somma, began to prepare an opera on the theme of Gustavo III for performance during the carnival season of 1858 at the Teatro San Carlo of Naples. That the assassination of a Swedish king in 1792 during a masked ball might raise censorial hackles did not surprise the composer, and by the end of November Verdi and his librettist had agreed to shift the action away from the Swedish royal court at Stockholm to a smaller ducal court in the Pomeranian province of Stettin, with the title of the opera becoming Una vendetta in dominò. For Verdi this change was not unlike what he had accepted during the composition of Rigoletto, moving his story from the Renaissance French court of François I to a ducal court in Mantua.5 What neither Verdi nor the Neapolitans expected was that on 14 January 1858, just as the composer was ending his sea voyage from Genoa to Naples, Felice Orsini and three accomplices would attempt to assassinate Napoleon III and his empress in Paris. The reverberations from the Italian anarchists’ bombs, which killed eight people and wounded another one hundred and fifty, were felt all over Europe. Thus, although Verdi was prepared to begin rehearsals for Una vendetta in dominò immediately, he soon understood that even this revised story had become impossible. The Neapolitans suggested a libretto set in medieval Florence, Adelia degli Adimari, the masked ball becoming an unmasked banquet. Verdi was outraged. A legal battle with the theater ensued, and as part of his evidence the composer had a copyist write the two librettos (Vendetta and Adelia) in parallel columns. Then, in a series of pungent footnotes, Verdi explained why he would not be party to this substitution, concluding, “A composer who respects both his own art and himself could not nor would not dishonor himself by accepting as the subject for his music, written to quite a different text, these absurdities, which manhandle the most obvious dramatic principles and insult the conscience of an artist.”6 Ultimately Verdi and the theater came to an agreement: the contract was dissolved and the composer agreed to return during the autumn of 1858 to mount an opera that had not yet been produced in Naples, Simon Boccanegra.

  But what was to become of Una vendetta in dominò? By coincidence, a play on the subject of the life and death of the Swedish monarch was staged in Rome during that same winter, apparently without significant censorship. Upon learning of it, Verdi sought to interest the Roman impresario Vincenzo Jacovacci in mounting the opera he had originally conceived. Toward the beginning of March 1858 he therefore sent Jacovacci a copy of the libretto for what he claimed was Gustavo III. In fact, it was Una vendetta in dominò, but with the setting returned from Stettin to Stockholm, thus, from a ducal court to the royal one. The enthusiastic Jacovacci forwarded it immediately to the Papal censors, only to find them as inflexible as their Neapolitan
brethren: their suggested substitute libretto, Il Conte di Gothemburg, heavily modified both the structure of the drama and the individual verses.7 Verdi was again outraged, but this time he did not suspend negotiations with the theater. Gradually a compromise was reached: if Verdi wanted to keep the essential characters and plot, he would be compelled to move the story out of Europe; if he insisted on keeping it in Europe, he would have to introduce most of the modifications demanded by the censors.8 Ultimately he and Somma shifted the setting to colonial Boston, and the composer acknowledged receipt of the new libretto, Un ballo in maschera, on 11 September 1858. During the remainder of the year, first in Sant’Agata, then in Naples, Verdi elaborated the new version, although not without constant harrassment from the censors, who continued to demand modifications in phrases they found immoral or politically daring. Having now decided, come what may, that he was going to stage his new opera in Rome, the composer accepted changes he might otherwise have resisted. And so, complaining all the while, Verdi arrived in Rome in early January 1859 with a completely orchestrated opera, whose first performance took place on 17 February.

  Un ballo in maschera has not ceased to be a problem ever since. There are, of course, those who claim that there really is no problem: if Verdi had wanted to change the opera, they say, he had ample opportunity to do so once Italy became united in 1860 and regional censorship disappeared.9 That argument, of course, fails to recognize that Rome, where Ballo was first performed, did not become part of a united Italy until 1870, by which time Verdi had no motivation whatsoever to alter an opera that was more than a decade old and continued to circulate successfully. It is easy, on the other hand, to understand why the Swedes would want to return Verdi’s opera to its original historical setting: the court of Gustaf III, one of the most important of Sweden’s monarchs, was often compared with that of the French king Louis XIV.10 The prospect of experiencing on the operatic stage the magnificence of Gustaf’s court and the unhappy history of his demise, especially when accompanied by Verdi’s music, seemed irresistible. The best-known adaptation was made by a Swedish poet, Erik Lindegren, who rewrote the text in Swedish, introducing characters and situations he claimed to be truer to the historical record, and adopting actual words attributed to the King.11 Lindegren did not simply “adapt” Somma’s verses: he rewrote the text, introducing references that did not figure in the original libretto. After being shot by Count Holberg (the “Renato” figure), for example, Gustaf addresses him with the classical and Shakespearean “Ej du, min Brutus!” This version, performed at the Royal Opera of Stockholm in 1958, makes no secret of Gustaf’s purported homosexuality, and the King seems every bit as interested in his page Oscar (a woman in the role of a young man) as in Amelia.

  But dissatisfaction with the Boston setting of Ballo has hardly been restricted to Sweden. Edward Dent had already made a version set in Sweden for Covent Garden in 1952, in which he restored the names from the Scribe libretto. At London’s Sadler’s Wells, which produced the work in 1965, a new English version was provided by Leonard Hancock, with a Swedish setting; this text was published in 1989, side by side with the Italian text as A Masked Ball, Un ballo in maschera, an absurd mixture.12 Many American productions simply transport the story from Boston to Stockholm, with as few interventions as possible, modifying explicit references to “America” and to the “New World.” The Metropolitan Opera’s production from the 1980s and 1990s, revived again in 2005, had the King of Sweden continue to send Amelia and Ankastrom across “l’immenso Oceàn.”13 For anyone who knows American history, of course, colonial Boston was hardly associated with masked balls, and conspirators named Sam and Tom evoke an unmistakably comic tone. Verdi knew that Gustavo and particularly Oscar were characters “cut in the French mode” and could exist only in a European court that had “breathed the atmosphere of the court of Louis XIV.”14 And his use of a can-can alla Offenbach (“Dunque, signori, aspettovi,” So, comrades, I’ll await you) to celebrate a visit in disguise to the den of Ulrica hardly captures the spirit of a culture known for the Salem witch trials. Boston, in short, far from being an example of “local color,” has been an embarassment to Ballo; most contemporary stagings—whether in Milan, Paris, or Chicago—have done their best to ignore or override it.15

  How should the change in venue be accomplished? That is the question. Do we really want the dying ruler to sing “Addio, diletta Amelia” (Farewell, beloved Amelia) instead of the “Addio, diletto America” of Ballo, thereby undoing all his efforts to avoid compromising the name of his beloved?16 And who wants to hear Renato referred to as “Ankastrom” (the Italianized form of the name of the actual assassin, Johann Jacob Anckarström)? Although both names have three syllables, their accents fall differently (on the second syllable for Renato, on the first for Ankastrom). Do we assume that a European Ulrica continues to be of the “immondo sangue dei negri” (even in those countries where political correctness does not automatically require “foul blood of the negroes” to be changed)? Is “immondo sangue dei gitani,” substituting one oppressed minority for another (“gypsies” for “negroes”) and awkwardly adding a syllable to the text, really any better? And to what “natio tuo cielo” (your native sky) does the ruler intend to send Amelia and her husband if they are not English colonists in Boston: surely not “in Inghilterra!” (to England!). These problems have attached themselves to Ballo for generations, like so many barnacles: we can pretend that it doesn’t matter, but it does. A Swedish setting for this opera is as relevant as a French one for La traviata. To place Ballo in Boston is like setting La traviata in Munich: one can do it, of course; indeed, one can make an appropriate postmodern statement in the process. But we shouldn’t delude ourselves into thinking that the physical setting and the opera Verdi wrote have much to do with one another.

  The early history of Ballo has always been told (just as I have told it) on the basis of its libretto alone. Yet during work on the critical edition of the opera, which opened the Verdi Festival in Parma in February 2001, celebrating the hundredth anniversary of the composer’s death, it became clear both to Ilaria Narici, editor of that edition, and to me that there was much more to be said. When Verdi wrote to his Neapolitan friends that he had “completed” Una vendetta in dominò as his opera for Naples, he meant it. Tracing the musical history of Un ballo in maschera and its earlier versions allows us to think about performing it in new ways.

  Una Vendetta in Dominò (Gustavo III)

  By October 1857 Verdi had proposed for the following carnival season at the Teatro San Carlo of Naples an opera to be entitled Gustavo III, based on the libretto Scribe had prepared for Auber at the Paris Opéra in 1833. Verdi himself wrote a prose outline, which he sent to Somma for versification some time before 24 October.17 Somma sent his original verses scene by scene to the composer, who commented on them and requested revisions (CVS 50–77, pp. 185–260). As Verdi’s musical sketches preserved at the Villa Verdi at Sant’Agata reveal, the composer seems not to have written down any music before the arrival of Somma’s textual revisions.18

  Throughout his career Verdi made clear that sketching was the most important creative task in preparing an opera. As we have seen, the heart of the sketching process consists of a “continuity draft,” in which Verdi traces in detail the shape and melodic content of the score. In some cases, such as La traviata, these drafts are primarily at the level of the individual piece or section. In other cases, such as Rigoletto and Ballo, drafts start at the beginning of an opera and essentially continue until the end. For Ballo there is a continuity draft for all of act 1, followed immediately by the scena before Amelia’s aria in act 2. Then there is a gap: no full continuity draft exists for the aria or for the following duet and trio, although there are isolated sketches. The continuity draft resumes with the finale of act 2 and follows without hesitation through all of act 3.

  Verdi spent much of November and early December 1857 sketching act 1 of Gustavo III in its original Swedish setting. In his
continuity draft, all references are to Sweden, to King Gustavo, and to Carlo, Duke of Ankastrom. In the first measures of the introduction, the chorus invokes the “King,” whom every Swedish heart adores; when Ankastrom enters, he addresses Gustavo as “Sire”; at the end of his solo he sings, “With you lost, what would remain of Sweden?” When the people stream in at the end of the Ulrica scene, they sing “Viva Gustavo” and “Glory to our King.” Verdi would not have continued in this way once he and Somma had agreed about the Pomeranian setting as Una vendetta in dominò. Thus, the continuity draft of act 1 and the ensuing scena must have been completed by early December.

  We can establish even more precise dates. Since Verdi did not like the poetry Somma originally sent for Ulrica’s invocation, the librettist forwarded new text on 17 November (CVS 63, p. 236), with which Verdi expressed satisfaction on the 26th (CVS 67, p. 242). Only the revised text is in the composer’s sketch: hence, he could not have drafted music for the scene until he had the new words. At that very moment, in fact, he was completing the introduction, for in a second letter on 26 November Verdi added this request: “In the stretta of the introduction it would [...] be wonderful if you could find me a dactyl [a sdrucciolo in Italian metrics] instead of “al tocco,” and allow me to say “alle tre” (CVS 69, p. 244). Meanwhile, though, as we saw in chapter 11, Verdi drafted the introduction using Somma’s words, except for this final passage, which he wrote without text. Thus, between mid-November and approximately 26 November, Verdi sketched the introduction of Gustavo III; from 26 November to early December, he sketched the scene in Ulrica’s den and the scena that opens act 2. Then he abruptly broke off work on the continuity draft.