CHOOSING A VERSION OF BELLINI’S I PURITANI
There are many Italian operas for which the problem of choosing a version to perform is perplexing. It may be necessary and sufficient for a critical edition to make available the score in what we can call its authorial versions, but those versions may be neither sufficient nor even satisfactory when we seek to present the work in the theater. One opera whose many textual problems embody in a particularly thorny way many of the matters discussed in this chapter is Bellini’s I puritani, the composer’s last opera and musically one of the finest works to be written by an Italian composer during the 1830s. It is also an opera with a history that has long been considered murky, even faintly mysterious, with a version prepared for a famous prima donna but never performed. Cholera in Marseilles and subsequently the death of both the composer and the prima donna intervened. After a century and a half of uncertainty, much of the history has finally been unraveled, but the mystery and romance of the multiple versions of I puritani remain.94
Bellini had met Maria Malibran in the spring of 1833 in London, where she was singing in an English rendering of La sonnambula. The composer was captivated by her talent and charmed by her character (“quella diavoletta della Malibran,” that little devil of a Malibran, he called her, remarking that she could learn an entire opera in a single day).95 Engaged by the Théâtre Italien of Paris to compose what proved to be his final opera, I puritani, Bellini agreed simultaneously to tailor the work for a projected Neapolitan performance with Malibran. He had hoped actually to be present to stage the work in Naples, but it gradually became clear that the timing of the Parisian premiere would make that direct engagement in Naples impossible. Thus, the composer was forced to compromise: he decided instead to ship his revised version for Malibran to Naples.96
But first there was the original version for Paris to complete. Bellini prepared his opera for four of the greatest singers that the European theater had known during the first half of the century, singers who would become known as the “Puritani quartet”: the soprano Giulia Grisi (Elvira), the tenor Giovanni Battista Rubini (Arturo), the baritone Antonio Tamburini (Riccardo), and the bass Luigi Lablache (Raimondo). The libretto, written by a Bolognese political exile in Paris, Count Carlo Pepoli, tells of the love between a Puritan woman, Elvira, and a cavalier, Arturo.97 Elvira’s father objects to the marriage on political grounds, but is gradually brought around by her uncle, Raimondo, even though another Puritan, Riccardo, is also in love with Elvira. An unknown prisoner at the Puritan castle turns out to be Henrietta Maria, the queen of the executed Charles I. Orders have arrived that she is to be turned over to Parliament, and it is assumed that the Puritans will execute her. Arturo arrives for his wedding with Elvira. When he realizes that the prisoner at the castle is the queen, he vows to help her escape. Aided by the wedding veil that Elvira, in her joy, has innocently draped over the head of the prisoner, Arturo escapes with Henrietta Maria, who is presumed to be the bride. Riccardo realizes what is happening, but, hoping thereby to gain Elvira’s hand himself, he allows the escape. Believing that Arturo has been unfaithful to her, Elvira goes mad, as the curtain falls on the first act.
In the remainder of the work, Elvira’s mind alternately clears or clouds over as she gains hope that Arturo will return to her or fears that she has lost him forever. Her state (particularly as expressed in one of the great “mad scenes” in the operatic literature) rouses pity in all, including Riccardo, but the Puritan Parliament has voted a death sentence on Arturo for having assisted in the escape of Henrietta Maria. Nonetheless, he returns to the castle in search of Elvira. After a brief moment with her, during which her reason returns, he is captured and is to be executed (while Elvira again slips into madness) when a messenger arrives from Cromwell himself: the wars are over and all political prisoners are to be freed. Elvira and Arturo are finally reunited as the opera ends, not without some discomfort for those of us brought up on Bertolt Brecht’s ferocious commentary on this particular brand of deus ex machina in The Threepenny Opera (in the translation and adaptation by Marc Blitzstein: “But in real life the ending it is not so fine, Victoria’s messenger does not come riding often”).
As best we can tell, Bellini had largely completed the Parisian version of I puritani during the first part of December 1834, although he was still making important decisions about the second act several weeks later. In order to meet the deadline he had set for himself with the Neapolitan theater, however, he was obliged to complete and ship to Naples the revised version of the opera before the Parisian rehearsals had advanced very far. The first act and part of the second were sent before the end of 1834; the rest of the second act (at this stage the opera was divided into only two acts) followed on 5 January 1835.98 For Naples there were to be two important shifts in vocal range among the four principal characters. Not only was the role of Elvira to be sung by Maria Malibran (a mezzo-soprano, rather than by the original soprano), but Riccardo was to be sung by a tenor, Francesco Pedrazzi, since there was no appropriate baritone on the Neapolitan roster.99 Bellini accomplished this revision in four ways: (1) he had a copyist prepare a manuscript of the pieces for which no changes (or only the most modest changes) were necessary; (2) he had a copyist prepare a manuscript of the pieces for which the Neapolitan version was simply a transposition of the original; (3) he entered small corrections and modifications on these manuscript copies; (4) he wrote out entirely those pieces or sections of pieces for which the changes he chose to introduce were significant and extensive. As fate would have it, there was a cholera outbreak in Marseilles, as a result of which (and through no fault of the composer’s) the score’s arrival in Naples was delayed; in fact, the second act (which went by land) arrived several days before the first act (which went by sea from Marseilles).100 Because of the uncertainty that had surrounded the arrival of I puritani, the theater decided not to wait more than a few days before proceeding with other plans. Despite an intervention in Bellini’s favor from Malibran, the theater invoked its legal right to break the contract, claiming that the score had not arrived as specified in the contract. So there was no performance of I puritani in the “Malibran” version, and the score remained, temporarily at least, in the hands of Bellini’s Neapolitan friend, Francesco Florimo.101
Meanwhile, rehearsals of the original version proceeded apace in Paris at the Théâtre Italien, where I puritani had its premiere on 24 January 1835 to enormous popular and critical acclaim. During the rehearsal period and between the first two performances, Bellini made many changes in his score, cutting sections of music, modifying his orchestration, replacing important parts, varying the order of compositions and the number of acts. Some of these changes may have been suggested by Rossini, who was serving in fact (if not in name) as musical director of the theater and as the point of reference for all Italian music in Paris, but they were adopted and executed in all details by Bellini, who was flattered by Rossini’s attention and overjoyed at the success of his opera.102 It is the version of I puritani Bellini himself presented at the Théâtre Italien of Paris on 24 January 1835, as further modified before the second performance, that was generally known thenceforth in the nineteenth century.103 It was published in all printed editions of the opera and was normally performed throughout the nineteenth and twentieth centuries (often with further cuts).
Thus, leaving apart those temporary versions that appeared and disappeared from day to day during the rehearsal period (all of which came into and out of existence between the time Bellini shipped the Neapolitan version to Naples and the time of the Parisian performances), I puritani existed in three principal authorial versions: (1) the original version prepared for Paris; (2) the modified version of this original Parisian score, prepared for Naples; (3) the version performed in Paris. Since Bellini altered physically the autograph manuscript of the original Parisian score during the rehearsal period in Paris, adding and subtracting pages, crossing out elements of orchestration, moving sections around fr
om one place to another, the entire original version for Paris no longer exists: only the second version (Naples) and third version (ultimately performed in Paris) can be completely reconstructed. The task of a critical edition of I puritani seems clear: it must produce complete texts of both the second and the third versions of the opera in a form that permits either of them to be readily used for performances. As for the first version, the edition must provide all surviving information about it. Given the fragmentary surviving materials and the evidence of Bellini’s having altered the score during the rehearsal period, a critical edition is probably not the place to attempt a reconstruction of that version.
The verities of a printed edition and those of a performance, however, are not necessarily the same. What follow are the kinds of questions pertaining to the principal differences between the versions that might give us cause to reflect:
1. In some of the numbers reproduced from the original Parisian version without change for Naples or simply transposed for Naples to a new key (such as the aria for Riccardo in the first act), Bellini made important changes in the instrumentation during the Paris rehearsals. In the cabaletta of the Riccardo aria, for example, “Bel sogno beato,” the melody was originally accompanied in the same rhythm by flute, two clarinets, and two horns. That orchestration was dutifully copied out for Naples, in the new key. During Parisian rehearsals, Bellini lightened the orchestration, removing the flute and horn parts (as well as other accompanimental parts for oboes and bassoons). Does it make sense to perform the Naples version with this heavier orchestration, when Bellini himself removed the extra parts during his rehearsals in Paris?104
2. Bellini made the Neapolitan version to favor Maria Malibran, a mezzosoprano. He had no special interest in modifying the role of Riccardo from a baritone to a tenor. In fact, the change creates a decided imbalance in the male vocal forces (two tenors and a bass instead of a tenor, a baritone, and a bass), not at all typical of Bellini’s or Donizetti’s operas. Even if we might wish to perform the opera with a mezzo-soprano, might it not be appropriate to keep Riccardo a baritone?
3. I puritani was conceived in two acts, and both the original Parisian version and the Neapolitan version have that structure. Only when Bellini manipulated the score during rehearsals, placing “Suoni la tromba” at the end of what then became the second act, did the three-act structure emerge. Does his motivation still hold valid? Or should we return the Parisian version to its original two acts?
4. Bellini announced to Florimo that he would not send a chorus to Naples because it had patriotic references that could not be performed there. This was the piece that Bellini replaced in Paris with “Suoni la tromba,” which—while sent by Bellini to Florimo—was replaced in the Neapolitan version by a recitative. Since the exclusion of this highly popular number in Naples can be attributed primarily to its political content, why should we not restore “Suoni la tromba” to the score today when we perform the Malibran version?105 If we use a baritone Riccardo, it could be added without any change; even if we use a tenor Riccardo, only minimal adjustments (of the kind that Bellini himself made on multiple occasions) would be needed.
5. During rehearsals and after the premiere, Bellini cut a number of sections from the original Parisian score because the division into three acts and the encores of two numbers demanded by the audience lengthened the evening.106 Since most theaters today do not allow encores by audience demand, should we not restore some of this music? If not, what justification is there to leave it in the Neapolitan version if we know that Bellini omitted it later in Paris?
6. Bellini originally ended the opera with a lyrical cabaletta for Elvira and Arturo together (“Ah senti, o mio bel angelo”). For Naples he turned it into a solo for Malibran, transposing the piece down by a third. During the course of the Parisian rehearsals, however, he substituted a choral conclusion, tearing the original ending “a 2” from his autograph manuscript of the Paris version. Fortunately, the cabaletta “a 2” is preserved in a manuscript copy of the opera at the Naples Conservatory (3.8.21–22), so it is possible to reconstruct the music as originally written for Paris. Sopranos, however, love that solo cabaletta every bit as much as mezzo-sopranos do. Joan Sutherland regularly sang this conclusion in her performances of the opera during the 1960s and 1970s, with the music transposed back up into the original register.107 Is there any reason not to do this when an appropriate performer is available? Or should we insist that only a mezzo-soprano can sing the cabaletta alone, while a soprano must sing it together with her tenor?
I raise these issues as a series of questions, because establishing a performing text of an opera with a history as complex as that of I puritani demands that we consider the particular parameters of the performance being prepared. In the performance based on the Malibran version of I puritani, given by Boston Lyric Opera in autumn 1993, my colleagues and I followed the shape of the score sent by Bellini to Naples and used a tenor Riccardo, as in that score. While we did not stray from Bellini’s organization, we did allow ourselves to emend the orchestration following changes introduced by Bellini during the Parisian rehearsals. It seemed absurd to insist on accompanying a passage in the first-act finale (transposed to accommodate a tenor Riccardo) with tremolo chords that interfered with the intelligibility of the text, when Bellini himself modified this accompaniment in Paris.108 However tempted we were to introduce “Suoni la tromba” into the score, we resisted, and were happily surprised that what we lost in not performing the single most popular number from the opera was repaid by a clarity and strength in the dramaturgy that in some ways is superior to the final version. Finally, we performed essentially all the music Bellini sent to Naples, even sections later omitted in Paris, but allowed ourselves to make some cuts, adopting principles discussed in chapter 8.
Should these same decisions be taken elsewhere? Probably not. Each performance of I puritani will need to consider the multiple versions of the opera in terms of its own requirements. That certain procedures are appropriate for an edition does not mean that the same procedures are appropriate for a performance. The critical edition of a complex opera, though, if well prepared, provides all the musical materials a theater should need to attempt to realize that opera onstage.
Modern performers are in much the same position that prevailed in the nineteenth century, where decisions about versions were made by local musicians responsible for performances in individual theaters. These local musicians worked with the same three-dimensional grid set forth here: aesthetic criteria, historical circumstances, and the practical conditions of modern performance. That our present-day aesthetic criteria may be different, our knowledge of history both greater and lesser, and our performance conditions further from the stylistic paradigms of early nineteenth-century opera does not affect the force of these structural parallels.
In discussing the historical circumstances of nineteenth-century Italian opera, however, I have emphasized performances in which the original composer of a work participated, rather than placing all contemporary versions on the same plane. Critical editions of these works behave in much the same way. In a world that has increasingly called into question the concept of the “author,” are we justified in privileging the composer? Or should contemporary musicians intervene as freely in modifying operas, making substitutions, composing new music, rearranging scores, as did some of their nineteenth-century ancestors?
At one extreme, we are under no obligation to stage these operas in a form the composer would have recognized at all. We could follow the lead of Peter Brook and invent a series of works entitled The Comedy of Figaro, Violetta Goes to the Ball, or Lucy on the Stage with Daggers. The advantage of doing so on a regular basis, however, is by no means obvious. The public does not appear to be clamoring for such adaptations, and few contemporary writers or musicians would find satisfaction in preparing them.
Any effort to gauge what constitutes acceptable or unacceptable intervention in the text of nineteenth-century I
talian operas must take into account the very different social and historical position modern theaters have with respect to these works. When Rossini became director of the Théâtre Italien in 1824, he was under tremendous pressure to produce in Paris some of his legendary Neapolitan operas, word of which had been spread by travelers, critics, and other musicians. Similarly, no self-respecting theater in Italy in 1853 or 1854 could have failed to produce Il trovatore. It is as if a movie theater today refused to show the latest blockbuster hit. But a modern opera house is under no such obligation to perform Rossini’s La donna del lago or Verdi’s Il trovatore if such a performance would be possible only by having another musician heavily revise the score.
Singers in the nineteenth century, furthermore, were hired for entire operatic seasons, as we have seen. That means they were obliged to sing parts in each opera presented during the season. Although impresarios sought to avoid incompatibilities, these were inevitable. The excellent artists for whom Simone Mayr wrote Fedra at La Scala in 1821 were anything but ideal for Rossini’s La donna del lago. Hence, a local musician made revisions, which the printed libretto reveals to have been numerous indeed. Modern opera houses working on the “stagione” system or producing individual operas under that system are not in the same position: they hire singers for particular roles. Unless their artistic administrators have done their job poorly, there should be no need for a modern opera house to perpetrate what were already considered abuses in the nineteenth century.
Divas and Scholars Page 33