Divas and Scholars

by Philip Gossett

  With these historical issues in mind, we can offer some guidelines for making cuts in recitative (whether secco or accompanied) in modern performances of Italian opera:

  1. Unless there is a physical change of scene or a complete break in the action between two closed musical numbers, enough recitative should be preserved to highlight the formal boundaries of these numbers. Failure to do so threatens to distort the intelligibility of an opera and create the impression of a concert in costume. When passages of recitative between musical numbers exist primarily to bridge scenes, cover set changes, or facilitate the exit of choral masses, on the other hand, modern theatrical stagecraft can eliminate the need for such passages, thereby justifying their omission.

  2. When making internal cuts in secco recitative, consider both the significance of the dialogue and its intelligibility (whether through direct comprehension, supertitles, or physical action). Actual dialogue must clarify significant events occurring on stage. Past events can be consigned to the synopsis that forms part of operatic programs (functioning as did the antefatto [background] of nineteenth-century librettos). Indeed, in these librettos, verses useful for understanding the story but omitted in performance were printed with virgoletti (quotation marks) to indicate that they would not be sung. In some cases, passages were cut only after the composer set them to music. In Il Turco in Italia, where the secco recitative of Rossini’s collaborator is often longer than the text printed in the original libretto, the decision to abbreviate the recitative in performance apparently preceded the printing of the libretto.22

  3. Lyrical passages in accompanied recitative, which become more prevalent as the century advances, help establish the emotional tone of an opera and should rarely be eliminated. Neither, however, should the surrounding recitative be so severely trimmed that these passages stand alone, falsifying their dramaturgical function. Recitative sustained throughout by a lyrical impulse (such as the music Verdi introduced in the prologue of the revised Simon Boccanegra of 1881) is a later practice in Italian opera and cannot be imposed on earlier works. Although Verdi could recast the first scene of his 1857 Simon Boccanegra to create a new musical dramaturgy, a pair of scissors will not allow us to achieve a similar result.

  4. When cuts are made in recitative, music must be modified to guarantee an acceptable harmonic and melodic flow. The stylistic context needs to be understood: in music of the early nineteenth century a D-seventh chord cannot resolve to an F-major chord in root position. When I arranged the secco recitative to allow Dario Fo to make the cuts he wished to introduce into La gazzetta, I began by studying attentively the progressions used throughout the opera. With this point of reference against which to measure my creative instincts, I gave myself up to the pleasure of modifying the recitative in the most musically and dramatically effective way I could.

  5. Finally, composers used recitative to bridge disparate tonalities between closed numbers. There are, to be sure, occasions in which these adjacencies are faulty. Particularly disturbing are cases in which a secco recitative points to a key, but the piece that follows the recitative turns out to be in an unrelated tonality. This frequently happens when the recitative is by one musician and the ensuing musical number by another (there are pungent examples in Il Turco in Italia, as well as in Mosè in Egitto). While a critical edition cannot atone for breakdowns of communication between those who prepared the original score, performers disturbed by jarring adjacencies should feel free to make modifications to avoid them.23 Modern performers, on the other hand, normally have no need to fret about places where the elimination of a recitative between numbers results in an ungracious tonal adjacency. Twentieth-century social custom, after all, encourages audiences to applaud after numbers, and such applause interrupts the succession of keys.

  To those who reflect about the practical problems of operatic performance, these guidelines will come as no surprise, but there has been little reflection on such issues. Indeed, the number of occasions on which similar principles are routinely flouted is appalling. In the context of festival performances, it is not inappropriate to avoid cuts in recitative; for most theatrical circumstances, however, judicious cutting of recitative, planned for the needs of a particular performance, is justified and can be accomplished in a musically responsible fashion.

  MAKING CUTS: ENTIRE NUMBERS

  In 1974, Lyric Opera of Chicago sponsored a week-long international congress of Verdi studies. Lyric Opera seemed an unlikely venue for the congress, since the general manager at that time, the late Carol Fox, disliked scholars intensely: “Scholars,” she said to me during the week, “I know all about scholars: my [ex-]husband was a scholar.” She certainly was not pleased by the participants’ criticisms of the new Lyric production of Simon Boccanegra by Giorgio De Lullo, whose failure to pay attention to details of the libretto turned an opera somewhat difficult to follow in the best of circumstances into a total mystery. I recall an invented procession in the first scene, in which the body of Simone’s beloved Maria was carried from her house at stage right, across the front of the stage, to the church at stage left. Thus, when Simone entered the house, there was no corpse for him to find. Not only did the procession make no dramaturgical sense, it made no liturgical sense: but those were the days (before supertitles) when many Italian stage directors simply presumed an American audience wouldn’t know or care.24 In sessions of the congress where the scholars were not being actively disagreeable, Fox found the discussion arcane. But she had prepared a secret weapon to humble her academic guests. For a panel on the subject of performance practice, she produced a singer of whom everyone was in awe, Maria Callas.

  As “La Divina” listened quietly, one professor after another recited the virtues of uncut performances, analyzing tonal schemes that emerged only from complete operas, demonstrating hidden melodic continuities lost when passages were omitted, and so on. When La traviata was discussed, every speaker deplored the practice of cutting the cabaletta “No, non udrai rimproveri” after Germont’s famous baritone cantabile “Di Provenza il mar, il suol”; without this cabaletta, according to the scholars, the act lost its shape and the dramaturgy suffered considerably. By then Callas had had enough. The real problem with the end of the scene, she informed us, was that the baritone sang at all. Violetta was the principal character of the opera, and the drama’s emotional center was Violetta’s “Amami, Alfredo, amami quant’io t’amo... Addio!” After it, the curtain should fall and the scene come to an end. There was no baritone present to register his opinion of cutting the most famous baritone aria Verdi ever wrote, and the scholars were too cowed to take on the diva.

  From the point of view of Violetta, of course, Germont may be expendable, but it is less certain that from the point of view of Verdi’s opera he can be summarily dismissed. La traviata is not the drama of a single character, and the interaction of Germont and Alfredo, father and son, is essential for our understanding of and emotional reaction to the entire work. Nonetheless, Callas neatly raised a crucial issue: what is the justification for eliminating entire numbers from an operatic score?25 These omissions include, in Mozart’s Le nozze di Figaro (to extend our range slightly), the arias for Marcellina and Basilio in the fourth act; in Rossini’s Il barbiere di Siviglia, the Count’s aria just before the finale; in Donizetti’s Lucia di Lammermoor, the “Wolf’s Crag” scene, a duet for Edgardo and Enrico that precedes Lucia’s mad scene; in Donizetti’s Don Pasquale, the servants’ chorus in the third act. In the less standard repertory, similar cuts are often taken: in Donizetti’s Anna Bolena, the great tenor aria “Vivi tu” disappears; in Rossini’s Tancredi, the aria for Argirio at the start of the second act is snipped away, as is Lucia’s aria in the second act of La gazza ladra.

  Whether aware of it or not, upholders of such traditions in the performance of Italian opera are invoking—explicitly or implicitly—the work mentioned above that appeared in 1958 (in Italian only) under the names of Tullio Serafin and Alceo Toni: Style, Traditions,
and Conventions of Italian Opera of the Eighteenth and Nineteenth Centuries. Their book was part of a series of defensive maneuvers undertaken by Casa Ricordi to respond to the charges, spearheaded by Denis Vaughan, that their printed scores of Verdi’s operas were filled with mistakes. By invoking the myth of a continuous performance tradition, Serafin and Toni sought to justify then-current (1950s) practice as both authentic and ideal. Although the volume appeared under both names, a footnote specified that these observations were “transcribed directly from the precise words of Maestro Serafin” by Toni.26 Because they reflect much of Serafin’s own performance practice, we will refer to Serafin as the author. It is not my purpose to trace the roots of Serafin’s ideas or to explore their relationship to practices from earlier in the twentieth century,27 nor do I claim that he invented the procedures he embraces. Because of his prestige, however, and because he left both written and recorded testimony, he is a convenient point of reference.

  Let us hear what Serafin has to say about the tradition of omitting the Count’s aria, “Cessa di più resistere” from near the end of the second act of Il barbiere di Siviglia:

  The omitted scena and aria are really superfluous to the action, and Rossini himself may not have written them with complete conviction: perhaps, as usual, they merely accommodated a tenor’s request for a show of bravura. In any event, this piece was omitted already starting with the first performances and was never again taken up. Rossini inserted part of it wholesale into Cenerentola.28

  Notice that Serafin (1) makes an aesthetic judgment (the piece is “superfluous to the action”); (2) invokes an “authorial intention” (“Rossini may not have written it with conviction”); (3) posits a singer’s preference (“a tenor’s request for a show of bravura”); (4) claims a historical truth (the piece “was omitted already starting with the first performances and was never again taken up”); and (5) invokes another historical circumstance (Rossini reused part of the piece in La Cenerentola). Let us consider further each of these statements.

  Is the piece “superfluous to the action?” It depends on how we perceive that action. Certainly the absence of any solo singing by a principal character from the lesson scene through the end of the opera (absent this aria) is most unusual, and the revelation delivered in secco recitative, “Il Conte d’Almaviva io sono” ( I am the Count of Almaviva), only to be followed by more secco recitative, inevitably falls flat in performance. Dramaturgically, furthermore, what should be a crucial theatrical event, the public revelation of the Count’s identity (with its parallelism to the moment he reveals himself to the officers in the first-act finale) is reduced to insignificance. Stage directors flounder helplessly to identify a nonmusical device that will give this moment weight, having the Count flash a medal or remove his cloak to reveal aristocratic clothes. Then again, even if one were to grant the superfluity of the aria to the action, in the strictest sense, a great deal of solo music in Italian opera is superfluous to the action, which often takes place in recitative or ensembles. Would the action of Il barbiere di Siviglia be different if Figaro didn’t sing “Largo al factotum” or if Berta omitted “Il vecchiotto cerca moglie?” Would the action ofNorma be different if the priestess omitted “Casta Diva?”

  Did Rossini not write the piece “with conviction”? There is no evidence to that effect. Not only is the autograph score beautifully written in every detail, but a most effective alteration in the score shows that he was paying close attention to its phrase structure. The Larghetto originally concluded with a four-measure cadential phrase, repeated with a variation.29 Rossini subsequently determined that the repetition gave the cadence too much weight, and so he decided to present the phrase only once, sensitively altering the melodic line so that it would include elements of both the unadorned and the varied versions within a single four-measure phrase (example 8.1). Considering how few structural alterations there are in the autograph manuscript of Il barbiere, it is impossible to assert that Rossini approached the Count’s aria without conviction.

  EXAMPLE 8.1. GIOACHINO ROSSINI, IL BARBIERE DI SIVIGLIA, RECITATIVO ED ARIA CONTE (N. 17), THE CONCLUSION OF THE CANTABILE, MM. 100–103.

  Did he write the aria simply to tickle the fancy of his tenor? There is no evidence whatsoever that Rossini felt his compositional interests to be at odds with the successful utilization of the best singers of his day. It was Rossini himself, after all, who not only brought the great Spanish tenor Manuel García from Naples to Rome to take the role of the Conte d’Almaviva (the opera’s original title was Almaviva, not Il barbiere di Siviglia), but also imposed this expensive choice on the management of the Teatro Argentina.30 For Rossini, the presence of García was a guarantee of artistic quality and an extraordinary compositional opportunity, not a trial to be overcome. Were modern conductors to omit every piece an Italian composer wrote to suit the talents of a star performer, they would be better advised to make yearly pilgrimages to Bayreuth.

  What about the purported historical evidence: that the piece disappeared soon after its first performance and was never taken up again in Il barbiere di Siviglia, while Rossini soon reused most of it for his Neapolitan wedding cantata of April 1816, Le nozze di Teti, e di Peleo, and some of it for La Cenerentola in January 1817? Although the music develops somewhat differently in the three compositions, the first phrase is essentially identical in the two operas and cantata, except for its key (example 8.2).31 while it is false to affirm that the aria was never sung again in Il barbiere di Siviglia (we have librettos attesting to its use well into the 1820s, and it never disappeared from Italian printed editions of the opera), there is an interesting historical angle to explore. After its Roman premiere in February 1816, Il barbiere was revived twice that same year: at the end of the summer in Bologna and in the fall in Florence. All 1816 productions featured Geltrude Righetti-Giorgi as Rosina. After the Roman premiere, Manuel García disappeared from the cast, but “Cessa di più resistere” was sung in Bologna, though not by the Count. Instead, Righetti-Giorgi, clearly fond of the piece, appropriated it. (Another Rosina, Catterina Lipparini, did the same for performances in Venice in 1822.) Although “Cessa di più resistere” was omitted from the Florentine revival of 1816 (as were several other pieces)—these were the performances for which Pietro Romani wrote “Manca un foglio”—it was no accident that it reappeared in a revised form as the concluding number of La Cenerentola, first performed in Rome on 25 January 1817, in which the title role was assumed by none other than Righetti-Giorgi. What makes the story even more piquant is that, for a performance of Il barbiere di Siviglia in Trieste in 1821, another prima donna, Fanny Ekerlin, inserted the Cenerentola aria where “Cessa di più resistere” was originally sung.32

  EXAMPLE 8.2. GIOACHINO ROSSINI, IL BARBIERE DI SIVIGLIA, RECITATIVO ED ARIA CONTE (N. 17), FIRST PHRASE OF THE CABALETTA, MM. 127–130; LE NOZZE DI TETI, E DI PELEO, ARIA DI CERERE (N. 9), MM. 125–128; AND LA CENERENTOLA, FINALE SECONDO: CORO, E SCENA CENERENTOLA (N. 16), MM. 154–157.

  An attempt to discuss these issues rationally, though, fails to take Serafin’s approach into account. His book is not a thoughtful effort to come to grips with historical problems but an apologia for his own artistic instincts (themselves a part of history), as sustained by an appeal to that limited part of the performance tradition with which he was familiar. Because he considered the omission of the aria traditional, Serafin saw no reason to understand its history. Further discussion existed merely to justify the tradition. That his statements are at best debatable and often false becomes irrelevant. Yet listen to the comments of an older generation of conductors when questions of this kind arise, fine and thoughtful musicians, such as Bruno Bartoletti, or those ruled by instincts developed in the 1960s, such as Zubin Mehta: they often mimic the sentiments of Serafin. A surprisingly analogous situation is found in the monumental, three-volume Rossini biography by Giuseppe Radiciotti, written between 1927 and 1929. This superb scholar was unable to recognize that his work betrayed all the prejudic
es of someone who had grown up knowing the later music of Verdi and the operas of Puccini, but very little Rossini. Reading Radiciotti’s diatribes against the composer’s vocal style, one wonders why he bothered writing about Rossini at all.33 Someone so evidently upset by florid vocal lines should have devoted his studies to the operas of Giordano.

  There are excellent reasons for eliminating the Count’s aria from Il barbiere di Siviglia, but they are not the ones Serafin invoked. The single most important reason for cutting “Cessa di più resistere” is that few tenors can do it justice, and even those who might succeed are often exhausted by this point.34 The evening is getting long, and it is surely not worth prolonging for an indifferent performance of that aria. Although its disappearance creates an emotional gap in the drama, Serafin is correct that no significant element of the plot is sacrificed. With the right combination of circumstances, including a distinguished and resilient tenor, a stylish and well-paced production, and an audience not running to catch the last train home, “Cessa di più resistere” can be effective and meaningful. And a theater, in order to obtain the services of a Rockwell Blake or a Juan Diego Flórez, may find itself with contractual obligations to include the piece. Ask the audience at the Metropolitan Opera who gave Flórez a standing ovation after his debut performance as Count Almaviva on 10 January 2002 whether or not the piece should be cut. On the other hand, when the circumstances are not right for the aria’s inclusion, a cut may be appropriate. The question should be decided by analyzing the circumstances of a particular performance.