Divas and Scholars

by Philip Gossett

  In La Fille du régiment Donizetti wrote four full sets of couplets, each with two strophes: (1) for the terror-stricken Marquise within the introduction of the opera, “Pour une femme de mon nom”; (2) for Marie, “Chacun le sait,” with its refrain praising her regiment, “Il est là... Le beau Vingt-et-unième”; (3) for Marie, the romance “Il faut partir,” within the first-act Finale, as she prepares to follow the Marquise; (4) for Tonio, a romance near the end of the opera, “Pour me rapprocher de Marie,” in which he sings of his love for her. In the Italian version the couplets for the Marquise disappear altogether (in fact, they were often cut on the French stage) as do those for Tonio. Tonio gets compensated for his loss through the addition of a comic cavatina in Italian style, borrowed by Donizetti from an earlier opera, Gianni di Calais. Changes of this kind show a composer trying to position an opera born in one theatrical culture for a new public.

  We have only sporadic evidence, however, about how much work Donizetti actually did on the Italian translation: we know neither whether he reviewed and corrected Calisto Bassi’s words nor whether he himself adapted his music to the new text. One of the most popular numbers of the opera, for example, has always been Marie’s air with chorus in the second act, “Par le rang et par l’opulence” (By rank and wealth) which concludes with the joyous “Salut à la France!” (Hail to France!). In the first section Marie voices the sadness she feels in her new social position as the niece (unbeknownst to her, she is actually the illegitimate daughter) of the Marquise of Berkenfield, as well as her despair at giving up her beloved Tonio for a noble marriage she feels compelled to accept. When the arrival of the French regiment that found her as a baby and has raised her (they are all her “fathers”) changes her mood, she launches into the most famous melody in the opera, one that soon became an unofficial French national song. The entire text, which is in two quatrains, reads:

  Salut à la France!

  À mes beaux jours!

  À l’espérance!

  À mes amours!

  Salut à la gloire!

  Voilà pour mon cœur,

  Avec la victoire,

  L’instant de bonheur!

  [Hail to France! To my happy days! To hope! To my loves! Hail to glory! Here for my heart, with victory, is the instant of happiness!]

  Notice that the second through the fourth verses of the first quatrain are shorter than the other verses: such metric anomalies were acceptable in French poetry for music, but frowned upon in Italy.

  Donizetti set the first quatrain as shown in example 11.13.64 The melody is conceived in two-measure units: the note on the downbeat of each unit serves as a point from which to leap to the syncopated and accented higher note on the second beat. The declamation of the text, although unusual, supports this musical structure perfectly.65 Indeed, at the very end of the opera, when the Marquise, who now has admitted that she is Marie’s mother and has blessed her marriage with Tonio, it is a reprise of “Salut à la France”—declaimed in precisely the same way—that brings down the curtain.

  On 15 August 1840, from Milan, Donizetti wrote his childhood friend from Bergamo, Antonio Dolci, “I am adjusting, cutting, etc., the Fille du régiment for La Scala.”66 By early September, already back in Paris, he wrote to the publisher Francesco Lucca in Milan on the subject of the Italian translation of Maria’s “Salut à la France!” He cited the following senari verses from Marie’s Aria:

  EXAMPLE 11.13. GAETANO DONIZETTI, LA FILLE DU RÉGIMENT, AIR ET CHOEUR MARIE (N. 9), IN THE ORIGINAL FRENCH.

  Salvezza alla Francia etc.

  Tomate bei giorni

  Di vita guerriera

  La gloria ritorni

  Ritorni I’amor.

  [Salvation to France, etc. Return happy days of the life of a soldier, let glory return, let love return]

  He did not like this strophe because “that last ritorni won’t do, since Maria has always loved.” Therefore he wants the translator, the same Calisto Bassi who worked on Rossini’s Guglielmo Tell, to find something in its place. Donizetti himself suggested:

  Chi nacque al rimbombo

  Del bronzo guerriero

  Disprezza l’Impero

  D’un vano splendor.

  [He who is born to the noise of the bronze cannon scorns the realm of vain splendor.]

  “They are doggerel of my own, but the idea will come across better if Bassi converts them into good verses.” Afterwards, an employee of Lucca’s, one Mandanici, was to take responsibility for underlaying the new words.67

  What we don’t know at this point is whether Bassi had written a first quatrain beginning with the words “Salvezza alla Francia,” as suggested by Donizetti’s citation, in which case the letter is specifically about the translation of the second quatrain. But in the Bassi translation published by Lucca, which persists in the performing tradition today, there is no reference to France in these quatrains at all. They become:

  Di gioja bramata,

  Di tenero affetto

  Già sento nel petto

  L’arcano poter.

  E l’ira calmata

  Degl’astri nemici,

  A giorni felici

  Ritorna il pensier.

  [I already feel in my breast the secret power of desired joy, of tender emotion. With the anger of unfriendly stars calmed, my thoughts return to happy days.]

  Whose decision was it to remove all references to France and the French regiment in these quatrains? We don’t know. It is not as if France disappears altogether, for in the cadences of the Italian translation there are several “evviva la Francia” and “viva la Francia.” Still, the translation apparently spares the Italian audience a powerful reminder that this is an opera written to please the French.

  The new words fit without difficulty under the original music, and even the regularization of the length of the verses in the Italian poetry requires only minor adjustments in the music: an added upbeat, small shifts of syllables to produce a more regular declamation in the new language, and so forth. The musical setting of the first two verses is shown in example 11.14. If one is willing to accept the decision (whether made by Donizetti or by his Italian collaborators) to decrease the dramatic emphasis on France, there is no reason not to continue using this perfectly adequate translation of Donizetti’s original.

  The situation is very different, however, at the end of the opera. In fact, Donizetti modified the original conclusion for the Milanese performances, removing its reprise of the opera’s hit tune “Salut à la France” and substituting a concluding duet for Maria and Tonio, “In questo sen riposati.” For reasons that remain obscure, however, Lucca decided not to print these new final pages, but instead to return to the French score. There is one important dramaturgical shift. In the French original, after Sergeant Supplice embraces the Marquise, the aristocratic guests leave the stage, so that only the soliders and soloists intone the final “Salut à la France.” In the Italian translation, the guests remain on stage (a female chorus) to add a contrasting “andiam, partiam” through the end of the score.

  For the first time, the translation for this final reprise of the melody returns to the French original:

  Salvezza alla Francia

  A suoi lieti dì!

  Vivan le gioje

  Che amor nudrì!

  [Salvation to France! To her happy days! Long live the joy that love nourishes!]

  EXAMPLE 11.14. GAETANO DONIZETTI, LA FIGLIA DEL REGGIMENTO, SCENA ED ARIA MARIA, IN THE CONTEMPORARY ITALIAN TRANSLATION.

  EXAMPLE 11.15. GAETANO DONIZETTI, LA FIGLIA DEL REGGIMENTO, FINALE SECONDO, IN THE CONTEMPORARY ITALIAN TRANSLATION, WITH A SUGGESTION FOR FIXING THE DECLAMATION.

  This is a fascinating text, because it is metrically irregular: the first two verses are senari (with accents on the fifth syllables), the last two quinari (with accents on the fourth syllable). Thus, whoever made the translation actually attempted to reproduce in part the metric irregularities of the original French. But whoever put t
he words under the music committed an unpardonable sin against Donizetti “[Fran]-cia” (example 11.15): oh that “-cia” on the downbeat! What a horror, even if it is deemphasized by the singer! What a falsification of the melody, with its carefully wrought two-measure units! What a denial of the symmetry and élan that gives the melody its character! And how useless, too, for the problem could easily have been solved by following the French, anticipating the “a’ ” to the downbeat and the “suoi” to the second beat, as in the bracketed text in example 11.15, and slurring the last two notes of the measure together. We do not know who was responsible for restoring the original finale and arranging the translation (perhaps they were verses that Bassi had prepared for the aria and later rejected, although the metrical irregularity makes that somewhat dubious). In any event, “-cia” is not a pretty sound with which to bring down the curtain on one of Donizetti’s most delightful operas.

  “IT IS IMPOSSIBLE TO MAKE A GOOD ONE”

  Verdi’s experience with Les Vêpres siciliennes powerfully exemplifies the translation problem. Given the significant literature on the subject (including an important study by the Italian critic Massimo Mila),68 it is unnecessary to rehearse the issues at length. The most important difference between this example and the others we have discussed thus far is that Verdi was directly and fully involved with the translation of this opera. Even so, the experience was an unhappy one for the composer. However careful and conscientious he may have been, the rules of the game in Italy before political unification (and before the loosening up of the metric rules that exercised near-tyrannical sway during the first three-quarters of the nineteenth century) made it impossible to make an adequate translation. Political and moral censorship had a devastating impact on the libretto, and the formal requirements of Italian poetry clashed with a musical style developed to suit the freer verse of the French.

  The problems with the plot of Les Vêpres siciliennes, prepared for the Opéra of Paris, where it had its premiere on 13 June 1855, are self-evident. A rebellion in which Sicilians rise up to massacre the French rulers of their country was hardly a theme that could pass muster in Italy under the Austrians during the 1850s. The entire plot was summarily removed to the Iberian peninsula, to Portugal under the rule of the Spanish, with a new title, Giovanna de Guzman. Yet even when the unification of Italy made it possible to return the action to Sicily, by and large the same translation was employed, with modifications in the names and corrections in isolated expressions to suit the new geography. Never was the libretto carefully emended to reflect Verdi’s original French text.

  It could be argued, nonetheless, that these complaints are pedantic. Verdi himself, after all, was responsible for underlaying the Italian words under the original French of Les Vêpres siciliennes. Already on 29 April 1855, a month and a half before the premiere, he had proposed to sell rights to the opera (excluding France, Britain, Belgium, and Holland) to Ricordi, informing the publisher specifically, “I obligate myself to send you a copy of the entire orchestral score with French text and with the ballet, which constitutes a self-contained action entitled The Four Seasons; also an Italian translation made under my direction by an able poet, changing the subject so as to render it acceptable to Italian theaters.”69 Soon after the premiere, on 6 July, he sent a copy of the full score from Paris to Milan. The text, as Giovanna de Guzman, was prepared by Ettore Caimi, but Verdi himself took responsibility for placing the words under the music. As he wrote to Ricordi,

  Today I sent you the score, in six big fascicles, in one of which is the overture and the music for the ballet. Have the overture separated and placed (as is obvious) at the beginning. The words written in red are the Italian words I have entered. In the ensembles and repetitions, I avoided a chore that anyone can do; but that anyone needs to be a knowledgeable person who is at least as patient as I have been in this tedious work.

  Now I know what translation means, and I have sympathy for all the awful translations that exist, because it is impossible to make a good one. The story chosen is a historical event; there’ll still be some words that the censor may not allow; but I think the subject can be permitted.

  The title we first chose was Maria di Braganza, but for the name of Maria I thought to substitute Giovanna; thus, in the score you’ll find Maria and in the libretto Giovanna. It’s all the same, and you can put whatever you like; just be sure to have the score adjusted according to the name you choose... I imagine you’ll have a copy of the score made with all the Italian words written out in full in the repeats etc. etc.; on my way home, I will be able to review it.70

  We don’t know whether he ever did review a score of the opera, but it would not have helped very much. He could do nothing about the problem of the censors. Moreover, he could do nothing about the insistence that French poetry be transformed into Italian verse forms, with all the attendant damage.

  Many verse forms translated easily, of course, and provided fluent Italian analogues, but it is remarkable how many key moments in the music drama are seriously compromised. Here are two examples from one of the most passionate moments in the score: the duet in the fourth act, in which Hélène reproaches Henri for having interrupted her act of striking down the tyrant, Monfort; this reproach compels Henri to admit that Monfort is his father. Example 11.16 gives the first part of the passage that follows, one of the most powerful in the entire opera, in which Henri laments the “horrible knot” that separates him from his beloved and leads him to perdition.71 In French this poetry is essentially in alexandrines (the first verse opens with Hélène’s previous “Ton père!”), but in Italian it has been rendered instead into settenari. Notice how this increases significantly the number of syllables that need to be accommodated by the music. Thirty-four notes bear the requisite syllables in French, but in the Italian translation forty-one notes are needed (example 11.17).72 Some of the resulting musical modifications are harmless: in the first six measures, for example, extra syllables are accommodated musically within the basic shape of the line through the addition of notes to upbeats or through declamation with two quarter notes in place of a half. The last two measures before the final note, however, are a musical disaster. In the original, Verdi sent the voice ascending to a powerful high g on “[révé]-lé,” held for a full measure, then tied to the first note of the triplets in the next measure, reserving a final syllabic declamation for “pour me perd-re à ja-[mais].” The necessity to accommodate at this point two verses in settenario (“Che eternamente a perdermi / Mi rivelava il ciel”), the first of which is even a sdrucciolo, proved daunting for Verdi, even impossible. That octave leap for “perdermi” and the attack on the high g with “mi” are much weaker than the original French, but accommodating the extra syllables was not optional.73 In 1855, an honest Italian poet could do nothing less.

  EXAMPLE 11.16. GIUSEPPE VERDI, LES VÊPRES SICILIENNES, DUO [HE´LE`NE–HENRI] (N. 12), IN THE ORIGINAL FRENCH.

  EXAMPLE 11.17. GIUSEPPE VERDI, I VESPRI SICILIANI OR GIOVANNA DI GUZMAN, GRAN DUETTO [ELENA-ARRIGO] (N. 12), IN THE CONTEMPORARY ITALIAN TRANSLATION.

  At the end of his soliloquy, Henri compares Hélène’s willingness to give her life to avenge her brother (killed by Monfort) to his own situation, and the poetry emphasizes the parallelisms: “Tu donnais tes jours pour venger ton frère” (You gave your days to avenge your brother) is set off against “J’ai donné mon honneur pour mon père” (I gave my honor for my father”). Henri’s anguished final phrase is shown in example 11.18). It is elegant, with the strong “plus” repeated on two successive downbeats and an upbeat of two sixteenth notes in three successive measures. Verdi reserved the arrival at the high note for the crucial word of the passage, “[hon]-neur,” while “pè-[re]” occupies the entire penultimate measure, with a fermata. Here again, the Italian version, hampered by its poetry, is very poor: the two syllables of “feci” on the downbeat make nothing like the same effect as “plus”; the upbeat is eliminated in the phrase “al crudel padre”; an
d in a measure where the French has only one syllable (“pè-[re]”), four must be accommodated (example 11.19). I would differ somewhat from Verdi’s opinion, for I do not agree that it is impossible to make a good translation. But it was impossible given the constraints within which mid-nineteenth-century translators and musicians were working.74

  EXAMPLE 11.18. GIUSEPPE VERDI, LES VÊPRES SICILIENNES, DUO [HE´LE`NE–HENRI] (N. 12), IN THE ORIGINAL FRENCH.

  EXAMPLE 11.19. GIUSEPPE VERDI, I VESPRI SICILIANI OR GIOVANNA DI GUZMAN, GRAN DUETTO [ELENA–ARRIGO] (N. 12), IN THE CONTEMPORARY ITALIAN TRANSLATION.

  Opera has always been an art of compromise, but rarely are the issues so perversely incompatible as in translation between Italian and French. I believe that we should hear operas written by Italian composers in their original language, whether that language be Italian or French. I do not see any justification for the belief that Italian composers did not set the French language well. On the other hand, if performers wish to continue using Italian translations, the nineteenth-century models need to be overhauled. Sometimes they need to be abandoned altogether; more often what is required is a strong editorial hand. The translations must be reviewed to be certain that the words mean in Italian what they originally meant in French (rather than producing a travesty of that meaning in concession to nineteenth-century censorship). And we must abandon the impossible expectation that regular Italian verses in the forms common during the first half of the nineteenth century can be underlaid to music written for different metric patterns. Small shifts in rhythm are certainly acceptable, but the musical meaning of a passage, the rhythmic matching of words and music at the heart of Italian opera, must not be compromised. Just as this chapter started with Verdi, let it end with Verdi. These are his words to Giulio Ricordi of 16 January 1883, referring to problems in the translation into Italian of the duet for King Philip and Posa at the end of the second act of Don Carlos: “Oh, translations are a horrendous thing! I would prefer them to be in prose (at least for the most part) in order to say everything that needs to be said and to respect more the accent and the meaning.”75 Nineteenth-century translations worked on the principle “Prima le parole, poi la musica”; those required for today’s theaters must follow Verdi’s advice and reverse the equation: “Prima la musica, poi le parole.”