Divas and Scholars

by Philip Gossett

  Where cuts are needed in operas written during the first three decades of the nineteenth century (and occasionally in later operas)—whether to reduce performance time, to assist a challenged singer, or to meet other particular conditions—cutting entire numbers is almost always less damaging than making internal cuts in those numbers. The numbers most likely to be omitted are usually bound to earlier dramaturgical paradigms. Many operatic texts from this period embody conventional modes carried over from the eighteenth century, the convenienze deplored by early nineteenth-century critics. Operatic dramaturgy in the first half of the nineteenth century focused ever more attention on primary characters and their interactions. Omitting compositions sung by secondary characters or by principal characters in subsidiary circumstances nudges a work forward along an idealized historical continuum, toward a time when pieces of this kind played an ever smaller role.

  We can watch the historical process at work by examining operatic librettos written and set to music early in the century, then revised for later composers. Donizetti’s Don Pasquale of 1842, for example, is derived from a libretto by Angelo Anelli, Ser Marcantonio, first set by Stefano Pavesi for the Teatro alla Scala in 1810.35 (Anelli was also the author of the text for Rossini’s L’Italiana in Algeri.) Don Pasquale contains an overture and thirteen musical numbers. Dramaturgically, it presents four characters: the elder Don Pasquale, who wishes to marry; his nephew, Ernesto, who fears losing his inheritance; his nephew’s beloved, Norina; and Malatesta, doctor to Don Pasquale and friend to Ernesto, with whose help Norina is disguised as Sofronia, the pretended “sister” of Malatesta, and is presented to Don Pasquale as a perfect wife. After a feigned marriage ceremony, Norina makes Don Pasquale yearn for his days of bachelorhood. In Anelli’s libretto for Pavesi, there are three additional characters. Ser Marcantonio (who becomes Don Pasquale) has not only a nephew (Medoro) but also a niece (Dorina). The niece is in love with the doctor figure (Tobia), and Bettina (the Norina character), loved by Medoro, is actually Tobia’s sister. There are also two servants with solo parts (Lisetta and Pasquino).

  All these characters must be given arias and duets. If one servant laughs at Ser Marcantonio’s foibles, the other must laugh; if Medoro complains of his lot, Dorina must have her chance. As a result, Ser Marcantonio contains nineteen numbers plus an overture, six more than Don Pasquale. The only numbers in Don Pasquale that could be considered expendable on dramaturgical grounds are two short choruses of servants (although, to my mind, the quality of the second, in particular, makes its omission a grave error).36 A revival of Ser Marcantonio in which some subsidiary banter is abbreviated would not have a negative impact on the most characteristic elements of Pavesi’s music or the dramaturgy of his opera.37

  It is not just that the history of Italian dramaturgy points away from the proliferation of secondary characters: composers themselves often treated the pieces for these characters as expendable. The strongest sign is that Rossini frequently allowed the arie di sorbetto of secondary characters or minor arias by principal characters to be composed by associates, often the very musicians who prepared the secco recitative. In the second act of L’Italiana in Algeri, two arias (for Lindoro and Haly) are by unknown musicians, the author of the Haly aria also being responsible for all the secco recitative.38 Occasionally history provides us with a name: Luca Agolini for La Cenerentola and Giovanni Pacini for Matilde di Shabran (as we have seen in previous chapters), as well as the young Michele Carafa, who in 1818 supplied Faraone’s first-act aria, “A rispettarmi apprenda,” for Mosè in Egitto.39

  In many cases, these pieces are omissible on musical and dramatic grounds: they are not by Rossini; they play no part in the essential fabric of the musical drama; and some (but by no means all) are artistically mediocre. Although Rossini occasionally replaced pieces set to music by associates with new music of his own composition (as he did with Faraone’s aria from Mosè in Egitto and all of Pacini’s arias for Matilde di Shabran), the dramaturgical issue remains unchanged. Indeed, on internal grounds alone, one could make similar judgments about a number of compositions prepared originally by Rossini: the arias for Isaura and Roggiero in Tancredi, the aria for Amaltea in the 1818 version of Mosè in Egitto (which Rossini borrowed from one of his earliest operas, Ciro in Babilonia, then omitted when he revived Mosè, still for the Teatro San Carlo of Naples, in 1819).

  But opera has always been more than a fabric of essential threads. Quite apart from the inherent musical quality of some of these pieces, there are pressing reasons why they cannot all be omitted. The charming Roggiero aria in Tancredi, “Torni alfin ridente, e bella,” for example, is dramatically expendable and tries the patience of a modern audience after a long evening.40 But several practical factors make its omission problematic. First, the singer playing Roggiero needs to participate in major ensembles: omitting the aria considerably reduces the pool of singers prepared to take the part. Equally important, the Roggiero aria separates a major duet for Tancredi and his beloved Amenaide from Tancredi’s elaborate gran scena. The Roggiero aria both allows a set change to be prepared (the concluding scene of the opera is a spectacular vista of Sicily with Mount Etna in the background) and offers the mezzo-soprano time to rest. No modern singer who has assumed the role of Tancredi, including Marilyn Horne, has ever countenanced cutting the Roggiero aria.41

  On the other hand, it is hard on aesthetic grounds to justify keeping the Albazar aria (written by Rossini’s collaborator) in Il Turco in Italia.42 The music is poor and the situation dramaturgically feeble. The aria, however, has two practical virtues. First, as in the case of the Aria Roggiero in Tancredi, it allows a theater to hire a better singer for the role, which many singers will refuse if they know the aria is to be cut. Second, it provides Don Narciso time to change his costume between his big aria (N. 11)—in which he announces that he will put in an appearance at the masked ball—and his actual arrival, in Turkish garb. With goodwill, such practical problems can be overcome. For the revival of Turco at Cremona and Milan in the spring of 1997, the first problem was solved by arranging for the tenor taking the part of Albazar to be hired for a few performances in the more important role of Don Narciso. The second problem wasn’t resolved until the dress rehearsal, when an army of personnel from the costume department lay in wait for Paul Austin Kelly, the Narciso, after his aria and poured him into his new costume.

  If we can avoid simplistic responses to the question of cutting individual numbers in modern performances of Italian operas, we can approach each situation with the necessary flexibility and sensitivity to the multiple and conflicting issues involved. The historical position of different works suggests different approaches. For Rossini, there are good reasons to countenance the omission of entire numbers; by the time of Verdi, where each number tends to have a precise role in the dramaturgical structure and emotional life of the opera, the practice makes less sense. The operas of Bellini and Donizetti sit between these poles. When we consider instead the matter of cuts internal to a musical number, however, the circumstances seem to reverse themselves, and the criteria we apply to operas written after about 1830 need to be very different from those applied to earlier compositions.

  MAKING CUTS: PORTIONS OF A MUSICAL NUMBER

  Performers of Italian opera have been making internal cuts in musical numbers since the operas were written, and it seems unlikely they will stop in the foreseeable future. Some of their reasons are frankly venal, a category that could be defined as “vanity cuts.” We have already cited an example in the closing measures of Assur’s “mad scene” in Semiramide, where Rossini wrote a four-measure, repeated cadential phrase in the orchestra, quoted from earlier in the composition, alternating two measures on the tonic and two measures on the dominant. By removing the measures on the dominant, the singer can sit on an interpolated high f while the orchestra chugs helplessly away on the tonic triad (see examples 6.3 and 6.4). The cut obscures Rossini’s citation of a theme from earlier in the composition and
produces a dull string of measures on the tonic that Rossini would never have written, all in the service of an interpolated and prolonged high note that singers during the first half of the nineteenth century would rarely have added.

  But most cuts are not “vanity cuts.” They stem from a sincere belief that omitting certain passages improves an opera. There are social arguments (in the modern theater, audiences listen to works differently from the way they did in the nineteenth century); aesthetic arguments (a section is musically weak); practical arguments (by the end of a duet, singers lack the strength to repeat a cabaletta theme yet again). The results, however, tend to have the same effect, and it is similar to the effect of cutting entire numbers in Rossini or Donizetti: they push the work, sometimes gently, sometimes roughly, further along the stylistic and historical continuum. Cuts that deform Rossinian symmetries make the operas sound more like Donizetti; cuts that tighten Donizettian dramaturgy make the operas sound more like early Verdi; cuts that eliminate cabalettas or cabaletta repeats in early and middle Verdi make the operas sound more like later Verdi.

  It is instructive, once again, to read Serafin, whose pronouncements regarding cuts in the Verdi operas remain those most widely invoked by self-described traditionalists. The same explanations recur time and again. Serafin recommends cuts that “do not change the equilibrium of the formal design but serve to avoid pedestrian, absolutely useless repetitions”; but he neither defines what he means by equilibrium nor provides us with a method to differentiate “pedestrian, useless repetitions” from ingenious, useful ones. He recommends trimming cadenzas, which he calls “efflorescences provoked by exhibitionistic demands of singers,” but provides no evidence that Verdi wrote such cadenzas with anything but complete conviction, however much the composer may have modified elements of his style in later years. He recommends eliminating entirely or shortening concluding cabalettas of arias, which are “of little musical worth” and are “inopportune, damaging the necessary, rapid, and natural development of the action.” Serafin has rarely met a cabaletta that is either musically worthy or opportune.43 In other words, if Verdi had only known better in 1850, he would have been writing Otello.

  This will not do. The operas forming the so-called trilogy of Verdi’s middle period (Rigoletto, Il trovatore, and La traviata) were written between 1850 and 1853, just before the composer turned forty and before his second appointment with the Paris Opéra—for the composition of Les vêpres siciliennes—left profound traces on his style.44 Like all works of art, those three operas show some traits that to contemporaries would have seemed conservative and others that would have seemed progressive. If Verdi sought to aim his compositional style in 1850 toward a lack of repetition, the absence of vocal virtuosity, and a rapid development of the action, he failed miserably.

  That does not mean Serafin’s recommendations for cuts can be disregarded. Verdi himself, less than four years after the premiere of Il trovatore, revised the opera for a Parisian performance, where he cut the cabaletta of Leonora’s aria at the beginning of the fourth act, “Tu vedrai che amore in terra.” 45 The omission is suggestive, even if it reflects a concession to practical requirements and to Parisian musical taste. Yet the disappearance of the cabaletta leaves the other sections of the aria—its cantabile, “D’amor sull’ali rosee,” and its tempo di mezzo, the famous “Miserere”—free-floating in a formally ambiguous ether within an opera largely tied to formal conventions of the period.46 To these contrasting considerations, one must add an aesthetic judgment about the musical quality of the cabaletta, a dramaturgical judgment about its resonance within the story, and a practical awareness that all but the studiest singers may come to grief over this difficult moment. None of the alternatives—omitting the cabaletta entirely, reducing its length, or performing the score as Verdi conceived it—provides an ideal solution to the historical and aesthetic problem, but we are obliged to pose the dilemma in all its ambiguity.

  The Operas of Rossini

  The cabaletta in Verdi’s operas, as with so many formal gestures Verdi was to abandon or profoundly transform during his later career, prolongs a practice established firmly in the operas of Rossini and carried on (with modifications) in the works of Bellini and Donizetti during the 1830s and early 1840s. A standard aria cabaletta consists of a theme, a short transition, and what is written as an exact repetition of the theme, followed by final cadences, often in phrases of decreasing length, each of which is immediately repeated. Similar forms held sway in ensembles, whether duets, trios, or principal finales (where they were referred to as strettas). Rossini did not invent the cabaletta. Indeed, efforts to assess priority in the use of cabalettas, crescendos, and other elements of the Rossinian formal arsenal, are unlikely to produce important results. What we can say with certainty is that at the start of his theatrical career, in 1810, Rossini used a variety of related forms to conclude a piece, as did his contemporaries. By the time he wrote Semiramide in 1823, the tradition of bringing practically every musical number in an opera to a close with a regular cabaletta was fully established. Rossini’s operas so dominated the repertory that it was with reference to them that younger composers developed their style.

  In Rossini, the cabaletta was tied to a crucial element of performance practice that remained operative through the 1830s but had been greatly reduced by 1850. The repetition of the theme and of cadential fragments provided opportunities for singers to ornament the melodic line, intensifying and personalizing its dramatic meaning and musical content. As we shall see, there is ample evidence that this practice was fully embraced by Rossini. Indeed, in a group of manuscripts Rossini himself showed singers the kind of interventions he considered appropriate. Manuscripts and printed books documenting the practice and teaching of important singers from the 1830s, such as Laure Cinti-Damoreau or Adelaide Kemble, demonstrate that ornamentation remained fundamental for the operas of Bellini and Donizetti.47 While there may occasionally be practical motives for abbreviating cabalettas and cadential repetitions in Rossini operas, doing so contravenes artistic principles that he himself developed and that are at the heart of his style. Intelligent performers must find ways to make Rossini’s leisurely formal schemes effective in the modern theater, just as they are doing for the operas of Handel.48 Indeed, singers, stage directors, and conductors who find this task unappetizing should not perform Rossini’s operas: denaturing them has persistently proved a sure recipe for failure.

  EXAMPLE 8.3. GIOACHINO ROSSINI, LE SIÈGE DE CORINTHE, INTRODUCTION (N. 1), OPENING CHORUS (MM. 23–26) AND CLOSING CHORUS (MM. 228–231).

  While clarity, a balance among parallel phrases, and formal coherence were fundamental stylistic qualities for Rossini, I am not suggesting that every cut within a musical number is equally disturbing. Let me suggest an operative principle: internal cuts in a Rossini number should not create a musical form different from what the composer uses elsewhere in the same work or in works from the same stylistic period.49 This principle would exclude many cuts that disfigured the revival of The Siege of Corinth at La Scala in 1969. The most notorious occurred in the opera’s introduction, a closed composition beginning and ending in E major.50 In a slow opening chorus, the Greeks, besieged by the Turks, seek counsel from their leader. After a trio in which the Greeks debate how to respond to the Turks, the introduction concludes with another choral movement, now quick, whose theme is a melodic transformation of the one previously heard (example 8.3). In this final chorus the Greeks swear to resist the forces of Mahomet or die fighting. In the Metropolitan revival, the movement was cut, leaving the shape and sense of Rossini’s music in utter shambles. It is the kind of cut no one in the nineteenth century would ever have made.51

  What kinds of internal cuts might be possible in Rossini? We have already discussed those the Metropolitan Opera made in 1989 in the choruses of Semiramide, reducing Rossini’s more leisurely

  A (orchestra) A (chorus) B (choral contrast) A (repeat of chorus)

  cadent
ial phrases

  to the more concentrated

  A (orchestra) A (chorus) cadential phrases,

  a form the composer frequently used in his mature operas. On structural grounds there can be no quarrel with such a reduction; whether it is advisable to do so in any particular case depends on other factors.

  We know that the composer himself was prepared for practical reasons to emend his scores in ways he could never have approved from a musical and dramaturgical perspective. Preparing the premiere of Guillaume Tell, for example, Rossini found that there was insufficient time during the orchestral introduction to the chorus that opens the second act to accomplish the complex action demanded by the mise-en-scène on the enormous stage at the Opéra: mimes carrying the carcasses of deer and wild boar, the chorus and extras (some of them on horseback) executing complex stage maneuvers before preparing to sing, and the rest of the court taking up positions.52 Rossini decided during rehearsals to repeat the orchestral introduction, a formal practice unique in his operatic production.53 The resulting form begins: