Divas and Scholars

by Philip Gossett

  It has been my purpose in this chapter to introduce some of the issues performers should be facing when they make cuts in Italian operas. Rarely are cuts neutral; rarely do all the elements suggest unmistakably that one procedure rather than another is correct. Cuts acceptable in Verdi may not be acceptable in Donizetti; cuts we can easily countenance in Bellini may be unthinkable in Rossini. Some cuts are so destructive as to be beyond reason. A performance of Rossini’s Guillaume Tell at the Paris Opéra in the spring of 2003 was the worst example of irrational and destructive cutting that I have ever had the misfortune to witness in the theater. Tell is a long opera, to be sure, and for a regular theater during a normal season it probably requires trimming. But why would one choose to perform Guillaume Tell by sacrificing some of its most characteristic music (“Enfans de la nature” in the first act, the opening chorus of the second act, the gathering of the cantons in the second-act finale), while preserving long passages of music (such as the dramaturgically leaden trio for Mathilde, Jemmy, and Hedwige in the final act) that the composer himself chose to cut?

  There may be practical reasons, as we have seen, for making decisions we know to be far from ideal. If a valued singer, beloved by the public, offers a conductor the choice between a vanity cut or no performance, the conductor may have no real choice. Such confrontations, however, are rare. Usually performers do have choices, and the care with which they exercise those choices affects the quality of their performances. Making cuts in an opera to suit the particular needs of a modern production is legitimate as long as those cuts are introduced with sensitivity to the individual nature of the work, to the stylistic characteristics of its composer, and to the relative position of each composition within the historical development of nineteenth-century Italian opera. What we must not do is abandon our own responsibilities to make such decisions in the name of standardized procedures of uncertain historical provenance and questionable aesthetic standing.

  9

  ORNAMENTING ROSSINI

  COME SCRITTO: THE WRITTEN AND THE UNWRITTEN

  In 1976, Alberto Zedda and I were installed at Zedda’s home in Milan, peacefully arguing about slurs and accents in Rossini’s La gazza ladra, the first volume to be published in the new critical edition, when we were interrupted by an urgent phone call from the Teatro alla Scala. Jean-Pierre Ponnelle was raging, Frederica Von Stade was in tears, Thomas Schippers was frantic, and the planned revival of Il barbiere di Siviglia threatened to dissolve before their eyes. A slow taxi ride through the appalling traffic of central Milan brought us to the theater. Von Stade had arrived for rehearsal with vocal ornamentation (variations and cadenzas) she planned to incorporate into her performance of Rosina’s well-known cavatina “Una voce poco fa.” Ponnelle, who had originally staged the production with Claudio Abbado on the podium in 1969, was adamant that Rossini’s vocal line should be sung precisely as written, just as in the performances he prepared with Abbado.1 They were using the critical edition of Barbiere and seemed to believe that interpolations by singers were therefore taboo. (The irony of a brilliant stage director’s insisting to musicians on the need for textual purity was not lost on any of us.) With bruised egos on every side, Maestro Schippers found himself in the role of mediator.

  In this case, thanks to Rossini, we were able to provide explicit guidance to the performers. The library of the Milan Conservatory, named “Giuseppe Verdi” after a country boy who had been refused admission in 1832, contains a manuscript in Rossini’s hand of ornaments for this aria. Prepared on 12 July 1852, thirty-six years after the composition of Il barbiere di Siviglia, during a visit the composer made to the baths at Montecatini, the manuscript was dedicated to Matilde Juva.2 She was a dilettante musician, sister of Emilia Branca, who was the wife of the most important Italian librettist of the first half of the nineteenth century, Felice Romani. Rossini’s ornaments for Juva, we explained, were representative of how the composer expected singers to shape his melodic lines in performance. Although giving ground slowly, Ponnelle relented. Schippers breathed a sigh of relief, Von Stade sang her ornaments, and disaster was averted.

  Only in a few cases do we possess this kind of unequivocal authorial support for modifying in performance the written notes present in an autograph manuscript (or its public face, a printed edition). Most modifications introduced into modern performances result from blind acceptance of recent practices, particularly for works widely represented in twentieth-century recordings: this is the kind of interpolation that Abbado sought to discourage. Other modifications, however, involve extrapolation from earlier historical models. Newspaper accounts often give valuable testimony concerning contemporary practice, while many specific examples of ornamentation are linked to individual composers or singers, derive from pedagogical treatises on singing, or reflect annotations in contemporary manuscripts or printed editions. Performers unaccustomed to the world of Italian opera, and wary of unexamined tradition, may or may not find extrapolation from such models convincing.

  The Italian pianist Maurizio Pollini, for example, is more associated in the public mind with Bach, Beethoven, Schoenberg, and Stockhausen than with Rossini. Nonetheless, infected by the virus that prompts many instrumentalists to take up the baton, he conducted La donna del lago in two sets of performances from the critical edition at the 1981 and 1983 Rossini Opera Festivals. Although Pollini approached the score from a viewpoint far different from traditional attitudes toward this repertory, he brought to it musical intelligence and dramatic insight that helped even those who had worked with the repertory for years to develop a new understanding of the composer’s art: the subtlety of orchestration, the richness of motivic development, the precision of harmonic pattern. That other elements of Pollini’s interpretation (especially his approach to vocal style) were less successful simply reinforced a general principle: you can never get it all right. In this case, though, when the tempo Pollini set for Malcom’s second-act cabaletta was impossibly fast, the tears flowing down the cheeks of a desperately frustrated mezzo-soprano, Martine Dupuy, night after night, were all too wet. When one tried to reason with him, Pollini would sit down at the keyboard and play the contested passage, saying, “You see, that’s how it should go.” He was right, but he was wrong. Musical works are complex organisms, and each performance can aspire only to reveal certain facets of their richness. Knowledge of a composition in the concert hall or in the opera house is a product of multiple performances; the more convincing a particular interpretation may be in certain directions, the less so it is likely to be in others.

  In the Pesaro Donna del lago, Pollini took a purist’s view toward the vocal lines, insisting that singers perform them as written. Even when cadential figures in slow movements sprouted elaborate written fioritura, he maintained a steady beat, eschewing accommodations to breath or grace. (The effect was still apparent, although somewhat tempered, when the 1983 reprise, with Katia Ricciarelli and Lucia Valentini Terrani in the roles of Elena and Malcom, was issued as a recording.)3 Rossini scholars cited historical evidence to plead the cause of ornamentation, of rhythmic freedom, of appoggiaturas. Only in one case was Pollini moved: for the final rondò of the opera, Elena’s “Tanti affetti in tal momento,” we could show him autograph manuscripts of three different sets of ornaments prepared by Rossini later in his life for particular singers.4 After staring at them, perplexed, for a few moments, Pollini allowed that he would permit the singer to use some of those ornaments if she wished, but no others.

  Scholars working as consultants in the opera house often find themselves in an untenable position: they are presumed to be rigorous upholders of abstract truth, when in fact they understand full well the difference between what is written and what is unwritten. A printed edition, no matter how critical, is a point of departure for a performance, not a blueprint to be followed with architectural precision. There is no unequivocal way to agree upon the distinction in decibels between forte and mezzoforte, the difference between a horizontal accent
and a vertical one, or the interpretation of melodic lines covered by slurs and those without them. But there are also conductors, ranking among the staunchest supporters of the new editions, who employ them with a rigor in some respects ahistorical. Scholars responsible for preparing those editions do not wish to quarrel publicly with distinguished proponents of their use. So we face the slings and arrows of outrageous journalists with resignation, while seeking patiently to clarify the relationship between historical knowledge and contemporary performance.

  Musical performance in the western tradition is a collaboration between composers, performers, and listeners: composers provide written instructions of varying degrees of specificity, performers respond to those instructions according to changing attitudes about the function of a written score, listeners develop expectations concerning the music they will hear that cannot lightly be resisted. The balance among these collaborators has shifted over the course of the past four hundred years, and will continue to do so, even for the same works.

  During the first three-quarters of the eighteenth century, composers preparing concertos, chamber music, operas, and cantatas provided only limited instructions about articulation and dynamics. They expected soloists (vocal and instrumental) to ornament lyrical lines, encouraged flexibility as to which instruments might be employed, and left much responsibility for introducing harmonic support to musicians who “realized” in performance the bass line annotated with signs indicating the harmony (the “figured bass”). Later, as composers offered ever more extensive written instructions, the necessity for performers to intervene in matters of pitch, dynamics, or instrumentation was more circumscribed. The expressive designs of the music, on the other hand, encouraged other kinds of interpretive freedom, deformations of rhythm, intensifications of dynamic levels, the inventive use of pedal effects on the keyboard.

  Opera, and Italian opera in particular, while participating in this new expressivity, remained bound longer than other nineteenth-century genres to eighteenth-century traditions concerning the relationship between composer and performer. Apart from some church styles, only in Italian opera did the use of a figured bass remain prevalent well into the nineteenth century. Apart from the improvisational exploits of solo instrumental virtuosos, only in Italian opera was ornamentation integral to the performance of newly composed notated works. Nowhere else in nineteenth-century music did performers have as much freedom to choose instrumentation as in the realization of the stage band, of which composers normally provided only a sketch.

  Even performers of the orchestral repertory or of German opera made modifications to suit local conditions: Wagner’s autobiography and the memoirs of Berlioz demonstrate that these composer-performers did what was necessary to produce the best possible realizations of their music, often under trying circumstances. The new critical edition of Tristan und Isolde provides extensive details concerning changes introduced by Wagner to accommodate his singers.5 Nonetheless, long after most European composers assumed that performers would try to respect their notation, composers of Italian opera continued to treat singers as collaborators. Indeed, to be certain that the first measures of the recitative preceding the trio in the last act of Rigoletto would be performed without modifications that singers might otherwise have considered self-evident, Verdi wrote in his autograph score: “This recitative must be declaimed without the usual appoggiaturas.”6

  When considering the relationship between musicians and the written text of an opera, we must bear in mind both conditions prevailing when an opera was written and those characterizing performance today. We cannot be indifferent to the intervening history, but neither must we give it undue weight, especially when that history is rooted in social and cultural practices or aesthetic preferences no longer operative. After all, the diversity of styles (both chronological and geographical) coexisting in American and European opera houses today was unknown to earlier periods. This has had a profound effect upon the way in which thoughtful modern musicians approach their task.

  In the late nineteenth century, for example, the few early nineteenth-century Italian operas remaining in the repertory were routinely reorchestrated in order to render their sound more similar to operas being composed at that time: Rossini’s characteristic use of the piccolo as a solo instrument was ruthlessly suppressed in favor of a flute; to operas with lighter scorings were added trombone and tuba parts, as well as elaborate percussion.7 In a world defining the coloratura soprano by the vocal exploits of Delibes’ Lakmé or Offenbach’s mechanical doll Olympia in The Tales of Hoffman, cadenzas and ornaments introduced by such singers into earlier Italian operas tended to be stylistically akin and structurally positioned in ways reflecting this new ideal of vocal virtuosity rather than the practices of a previous era. The evidence of early recordings reflects late nineteenth-century practice, and must be evaluated accordingly.8

  Yet we cannot pretend that knowledge of performance practice as it existed in the first half of the nineteenth century will suffice to answer the questions that invariably arise when we now perform operas in this repertory. As always, our knowledge of historical circumstances must be joined to aesthetic values and to the practical conditions of modern performance, the three-dimensional grid introduced in the discussion of multiple versions in chapter 7. This chapter examines the kinds of modifications that singers of nineteenth-century Italian opera might appropriately introduce into notated vocal lines, including ornamentation (whether improvised or learned) and puntature (changes made to accommodate singers unable, for whatever reason, to perform the text as written).

  VOCAL ORNAMENTATION: CONTEMPORARY EVIDENCE

  Rossini loved inserting notes to himself (and to posterity) in his autograph manuscripts. Some are pointed commentaries on the music he has written, transformations of the traditional “Laus Deo” (Praise God) of eighteenth-century composers. At the end of the sprightly overture to La scala di seta, for example, Rossini signaled his approval of the piece by commenting “Accidenti!” (probably best translated as “Awesome!”); at the conclusion of the overture to Il signor Bruschino, during which the second violins are instructed to beat their bows rhythmically against the metal shades of their candleholders (inauthentically replaced today by mere music stands), Rossini acknowledged his preoccupation with the reception of this unusual effect by writing, “Dio ti salvi l’anima” (God save your soul).9

  Sometimes his comments are pointed barbs, vulgar and funny, aimed at his own musical practice or at linguistic traits of the poetry he is setting. In the closing cadences from a second-act trio in Otello, for example, he wrote five parallel triads, strictly forbidden according to the rules of voice-leading, annotating the passage: “Queste cinque quinte sono per li signori Coglioni” (These five fifths are for——).10 When I tried to quote that phrase in a Metropolitan Opera intermission feature, Gerry Souvaine came storming out of her producer’s box: this is family radio! In Ermione, even the composer was unable to take seriously the florid verses with which Orestes greets Hermione after many years apart from her:

  Ah mio Nume adorato! ormai la sorte

  Quel piacer mi concede

  Che sospirai ben mille volte, e mille:

  Vagheggio alfin le amate tue pupille!

  Ah my beloved Goddess! finally fate

  Has granted me that pleasure I sighed for

  A thousand times, and a thousand times more:

  I gaze at last into the beloved pupils of your eyes.

  In his autograph score, the twenty-six-year-old Rossini put an asterisk next to the word “vagheggio” and at the bottom of the page annotated the word as “vaccheggio.” Suffice it to say that the word vacca (cow) is a vulgar Italian term for prostitute.11

  In two places, Rossini’s commentary refers to vocal ornamentation, demonstrating unequivocally that what he wrote was not what he expected to be sung. Felice Romani’s libretto of Il Turco in Italia is best known for the presence of a quasi-Pirandellian Poet who seeks to manipulat
e the action, but who instead is manipulated by it. (In fact, Romani derived much of his libretto, including the character of the Poet, from an earlier libretto by Caterino Mazzolà, best known today for reshaping Metastasio’s La clemenza di Tito for Mozart).12 Just before the stretta of the first-act finale (the boisterous ensemble that brings down the curtain), Fiorilla and Zaida, rivals for the affection of the opera’s Turk, Selim, engage first in verbal sparring, then in physical combat. While most characters try to separate them, the Poet delights in the spectacle:

  Seguitate... via... bravissime!

  Qua... là... bene; in questo modo

  Azzuffatevi, stringetevi

  Sgraffi... morsi... me la godo..

  Che final! che finalone!

  Oh! che chiasso avrà da far.

  Go ahead... on with it... wonderful!

  Here... there... good; that’s the way,

  Go to it, get closer,

  Scratch... bite... I love it...

  What a finale! what a great finale!

  Oh! what an effect it will have

  In his autograph, Rossini annotated the final verse, which leads directly into the stretta, with the following phrase: “The composer leaves it to the art of Sig.r Vasoli [the singer who first played the role of the Poet] to fill out properly this...” at which point he drew an enormous fermata (example 9.1). Vasoli, in short, was expected to interpolate a cadenza.13

  There is a similar case in La scala di seta. The romantic lead, Dorvil, was written for Raffaelle Monelli, the most capable tenor at the Venetian Teatro San Moisè, for which—as we have seen—Rossini prepared five one-act farse between 1810 and 1813. The young composer must have had a spirited relationship with Monelli. Over a notated cadenza at the end of the cantabile of Dorvil’s aria, “Vedrò qual sommo incanto,” the composer added: “Dolce per le cinque piaghe di Cristo” (Sweetly, by the five wounds of Christ). Just before the beginning of this cantabile, he suggested that the tenor interpolate a cadenza into the concluding phrase of his recitative by writing “a piacere del Sig.r Monelli” (at the pleasure of Signor Monelli). Actually he employed yet another vulgar term, adding an “a” in a conspicuous box between “Mon-” and “-elli”: “mona” is a vulgar term for vagina in Venetian dialect.14