We ultimately produced that more accurate score we promised, and it required not only exhaustive but also exhausting scholarship. Nonetheless, the critical edition of Semiramide uses that scholarship (and more than a little imagination) to make Rossini’s opera available to performers in as complete and correct a form as modern textual criticism allows, recognizing fully that this edition (as with every edition of a musical or literary text) is a product of our era and its complex relationship to the past.
PERFORMING FROM THE CRITICAL EDITION
OF SEMIRAMIDE
What did the New York Times critic mean when he said that Semiramide “was played in a new edition that put exhaustive scholarship before operatic effectiveness”? If by “edition” he was referring to the new critical edition of Semiramide, the meaning of the sentence is obscure: by whatever criteria Rossini’s opera may be judged effective or ineffective, an edition of Semiramide (as opposed to an adaptation) cannot transform it into something else. Semiramide is not Les Contes d’Hoffman or Carmen or even Don Carlos, operas in which there are significant doubts about what music should be present in the score, either because the composers (or others) revised the works so many times that the situation is unclear or because they never lived to see their operas performed at all. Rossini wrote an opera called Semiramide with a certain musical-dramatic structure: the critical edition presents that opera as the composer conceived it, to the best of our knowledge. Punto e basta, as the Italians say.
If “edition” here means only production, on the other hand, the phrase is even harder to understand, since there was nothing “scholarly” about the Metropolitan Opera’s production in any commonly understood sense of that term. Although there are many surviving drawings of sets and costumes for this once highly popular opera, no effort was made to employ them. At most, the production followed nineteenth-century practice by changing sets without dropping the curtain, thereby sparing the audience those interminable pauses in a dimly lit theater typical of opera productions of thirty years ago.28 Although a great deal of ornamentation used in Semiramide by early nineteenth-century singers survives, none was adopted by the cast, all of whom developed, either by themselves or with the help of coaches, ornamentation they considered suitable for their own voices.29 Neither the audience nor the critics judged ineffective the ornamentation used by Lella Cuberli, June Anderson, Marilyn Horne, or Samuel Ramey (each of whom ornamented his or her music in different ways). Although scholars know a great deal about the instruments Rossini would have had at his disposal in 1823, no member of the Metropolitan orchestra was asked to leave his or her modern instrument at home. At most, dynamic levels in their parts were changed to reflect, for example, the different weight of a modern brass instrument or the different ways instrumentalists were arranged physically during nineteenth-century performances. And although Rossini had only a single pair of tuned drums available to him, the fine Metropolitan timpanist (who employed four individually tuned drums) followed modern practice in substituting chordal notes when the composer, to keep the timpani playing in an ensemble passage, resorted to a nonharmonic tone.
Once the critical edition of Semiramide was in the hands of the performers, in short, no participant in the production had any thought other than creating an operatically effective performance for a modern New York City audience. Yet the question remains: What happens at the point of intersection between scholarship, with its effort to develop accurate texts and to provide precise historical knowledge, and performance? What kinds of questions are asked? What kinds of answers are deemed acceptable? What are the limits beyond which the performer feels constrained or the scholar feels compromised? And what happens when those limits are crossed?
Of the four categories of differences between the new critical edition and the earlier score, two are relatively unproblematic: no one has suggested it would have been better as a matter of principle not to use Rossini’s orchestration; no one has suggested any changes in the basic editorial procedures employed. That does not mean that every editorial decision is unimpeachable, either in theory or as it affected tangibly this particular production. But the critical editions currently in progress of the operas of Rossini, Donizetti, and Verdi take as a point of pride that, wherever possible, first performances of a new edition use proofs and not a published score. Critiques of editorial decisions by fine musicians help scholars reassess their solutions and, if necessary, produce a more accurate and nuanced text. In addition, nothing reveals a simple mistake faster than when a clarinettist plays a b in a B-major chord.
The other two categories of differences between the new critical edition of Semiramide and the earlier score, the use and orchestration of the stage band and the problem of cuts, are more controversial and therefore more interesting. They raise problems that arise at the point of intersection between scholarship and performance, problems to be explored in the second part of this book. We will also look briefly at some specific issues faced by the performers of Semiramide when employing vocal variations and ornamentation neither specified by the composer nor deriving, whether directly or indirectly, from his intervention.
Orchestration and Editorial Procedures
The computer wars did not end with the production of the full score. Performance materials derived from the score sported errors we could not have anticipated, which emerged only during rehearsals for the orchestra alone, without the singers. On several occasions the second violin part was printed a second or a third too low, even though the line was correct in the orchestral score. There is almost nothing as infuriating as watching costly rehearsal time tick away in order to correct nonsense beyond one’s control. It was even worse for me to realize that we had occasionally failed to correct notes in the clarinet and horn parts—transposing instruments (that is, what you see is not what you get, and a written c actually sounds d, e,f, g, a, a, or b, depending on the transposition). These were places where Rossini’s accidentals (sharps and flats) were wrong, yet we had neglected to intervene. There were only a handful of errors in the whole 1400-page orchestral score of Semiramide, but each one sent a dagger through me. Nonetheless, we somehow got through orchestral readings, errors in the parts were corrected, and we were able to begin rehearsals with the singers.30
The Metropolitan Opera orchestra consists of players who are serious professionals and fine musicians, but rehearsal conditions were not ideal. For repertory operas the players are so good that lengthy rehearsal is counterproductive, but for a work new to them the rehearsal time available was inadequate. On many days after the readings for orchestra alone, the players went right into the pit for run-throughs of entire acts or the whole opera. There was not even a Sitzprobe scheduled, a musical rehearsal in which the conductor works on musical details with orchestra, soloists, and chorus. Some of the fault for this scheduling lies in the common misperception that Rossini operas are relatively simple for the orchestra to play. If that means that the orchestral score is less dense or contrapuntally complex than a score by Wagner or Berg’s Lulu, fair enough. But if it means that the style can be grasped and the orchestral lines brought to life by musicians not accustomed to playing similar music, it is totally false. Indeed, American orchestras need considerably more rehearsal time to play Rossini well than to play Wagner. They need to understand how the accompaniments, ostensibly simple, must be shaped, phrased, and articulated in order to sustain, envelop, and give radiance to the vocal lines. More than in most repertories, players also have to deal with fermatas, freedom in tempo, and modifications in the literal values of notes, all elements that cannot simply be inferred from the notes in the parts on their music stands.
Furthermore, the personnel making up the orchestra were highly variable. Because the orchestra has so many calls for rehearsals and performances each week, there is no “single” Metropolitan Opera orchestra but rather a group of musicians who belong to the orchestral family, as well as a group of “stringers” who come in when called. Some musicians were assigned
permanently to Semiramide rehearsals and performances; others came and went without warning. As long as a body with the right instrument filled the requisite seat, Maestro Conlon was expected to be satisfied. That system might work for a composer whose style was well known to the players, but it was a nightmare for Rossini.
Never have I seen a set of parts marked up with such a surfeit of additional signs as were these Semiramide materials.31 Rossini normally writes his famous orchestral crescendos beginning at a dynamic level of pp. After eight measures there might be a “cresc.,” after another eight a f, and finally a ff. Editions of Rossini’s music, whether critical or not, follow his notation, as they should, intervening or regularizing only where there are inconsistencies or unusual problems. Given the nature of modern instruments, the size of the modern opera orchestra, and today’s theater acoustics, however, it is standard for a conductor to tell the orchestra to start a Rossini crescendo as softly as it can play, to keep it at that same level even after the word “cresc.” appears, letting the increase in the number of instruments take care of the crescendo effect at first, and to save increases in dynamic level for the very end of the passage. Stated clearly once, such an explanation covers many similar situations throughout an opera.32 But not at the Met. Since Maestro Conlon could never know who would be sitting at each desk during a rehearsal or even at a performance, it was decided to mark each and every crescendo in the performance materials in a way that would have been considered ludicrous in most other opera houses. At the start of each crescendo all of Rossini’s pp were turned into ppp; all indications of “cresc.” were crossed out and postponed for four or eight measures, and so on. In a few cases, it was so difficult to get the message across to players who were showing up for one rehearsal or another that we actually removed some of Rossini’s doublings in order to be certain that the vocal lines would be heard. Once we did the opposite, reinforcing a line that Rossini assigned to first violins alone by having all the second violins play along in order to give it sufficient weight.
Throughout these manipulations of details, I could not help wondering what the orchestral balances would have been like had we been using early nineteenth-century trombones, horns, and trumpets, instead of our “new and improved” modern variety. How absurd to change every ff in the trombones and tuba to f or even mf, so that they would not overpower the remainder of the orchestra. With a stable orchestral contingent, the players could have been told once and for all that their instruments are more powerful than those Rossini had in mind; hence they should never play louder than mf or f, even in a tutti passage. But in the real world of the Metropolitan Opera, with modern instruments, we were compelled physically to alter every single dynamic marking in Rossini’s score so that a musician reading his or her part for the first time would know what was expected.
The Banda sul Palco
Rossini projected the use of a band during three numbers of Semiramide, sometimes onstage (with its members in costume), sometimes in the wings. As we have seen, the original scoring of this band, for twenty-two instruments, functioned well in rehearsals for band alone (with only minor modifications). Both scholarship and good sense would suggest, then, that the original band parts be used.33
In the nineteenth century, the local militia provided players for the band. In today’s world, unionized theaters do not have the option of calling on the fire department, and few theaters are prepared to engage twenty-two additional union musicians for a few minutes of music. None has been willing to put a brass band onstage, where Rossini often expected it to be, whether because of the cost of costumes, the necessity to pay supplements to the members of the band, or a feeling of discomfort at the convention, which has long since disappeared from opera houses around the world. For most performances of Semiramide and similar operas, the music written for the banda sul palco is assigned to instruments in the pit, a most unwelcome compromise since the stage band is usually meant to represent music “actually heard” by the protagonists, whereas music emerging from the pit normally (but not always) exists in a different realm. Eliminating the stage band altogether, then, disturbs the dramaturgical and musical structure of the work. Important theaters, such as the Metropolitan, compromise: they hire a band (often with reduced numbers) and put it in the wings, where, of course, problems of coordination between the band and the conductor become serious, even in the presence of television monitors.
Given the physical production of Semiramide at the Met, there were two places in the wings where the band could be placed: upstage right or downstage left. We began upstage right, a site from which the sounds arriving in the theater, which had seemed perfectly balanced during rehearsals of the band alone, were so distorted that the treble instruments, piccolos and high clarinets, completely overpowered the bass, giving the impression of a handful of pennywhistles. We then moved the band to downstage left: better, but the acoustics still distorted the sound. And so, working together, the scholar and the bandmaster modified the orchestration to suit the acoustic conditions, removing some of the higher-pitched instruments and strengthening the bass. Those modifications resolved the orchestration problem for the intervention of the stage band in the second act, where the band is intended to play alone offstage: by quoting music heard earlier in the opera the band announces impending festivities that directly bear on the dramatic confrontation taking place onstage.
More difficult to resolve were the two interventions of the stage band in the first act, where the band must play together or in close coordination with the pit orchestra. Rossini begins his first-act finale with an important choral movement. The orchestra plays a lengthy passage alone, allowing the chorus and the band time to enter and take their positions onstage. The music is then repeated with the chorus and the band joining the orchestra. It is not difficult to imagine a splendid effect achieved by this combination of musical forces. When the band is in the wings, however, two problems arise: first, it is difficult for it to produce a sound weighty enough to balance the fortissimo of the orchestra and the singing of the chorus; second, the problem of coordination becomes difficult, and in a passage where band and orchestra must often play the same music, the effect can be ragged. We rehearsed the passage numerous times, but ultimately all participants in the production agreed that the benefits to be gained by employing the offstage band did not offset the dangers of failed coordination between musical forces. Thus, the band was omitted from the first-act finale.
To the chagrin of everyone involved in the production, scholars and performers alike, the band parts for the Allegretto chorus of the introduction, “Di plausi qual clamor,” were also omitted, with the notes assigned to instruments in the pit. The band here has a specific dramaturgical function: it announces the imminent arrival of Semiramide and her court, representing in sound the offstage “clamor” described in the text. The effect Rossini was seeking depended upon integrating the band within the entire musical context. The band plays only a few notes, but those notes either introduce and give rhythmic impulse to each musical phrase or effect tonal modulations between phrases (example 6.1). The nineteenth-century scoring was unproblematic, and given the dynamic context in the orchestra, piano and pianissimo, there was no problem hearing the band. Because the coordination of band and orchestra requires absolute rhythmic precision, however, placing the band offstage turned out to be a severe handicap. With sufficient rehearsal time, this handicap might have been overcome, but in 1990 at the Met, that time was unavailable. Once rehearsals moved from the rehearsal rooms to the theater, the band entries were never rhythmically correct.
EXAMPLE 6.1. GIOACHINO ROSSINI, SEMIRAMIDE, INTRODUZIONE (N. 1), THE CHORUS “DI PLAUSI QUEL CLAMOR,” MM. 365–374.
The band interventions were therefore written into the orchestral parts, using those few instruments not already playing. The result was a pallid imitation of the original, with no musical force, no dramatic logic, and no sense. Exhaustive scholarship and operatic effectiveness were both sacrificed on
the altar of practical expedience.
Cuts
While only those thoroughly familiar with Semiramide or professionals directly involved in the production were aware of problems with the banda sul palco, everyone who attended the performances could talk knowingly about the length of Rossini’s opera. That cuts would have to be made in Rossini’s score was a principle accepted from the beginning by all. Although prepared to do whatever the performers felt necessary to present the work responsibly, the company’s judgment was that Semiramide would have greater public success if it could be brought in at just over four hours in length, including intermission and applause but not final curtain calls. In the modern world, all theaters try to hold overtime to a minimum, but the crucial element for the Metropolitan’s budget was to get the theater emptied before midnight. They were prepared to have a seven o’clock curtain if necessary, but their experience suggested that beginning an opera before 7:30 created undue difficulties for a New York audience, difficulties that could be overcome only for works that had already developed a place in a particular repertory, such as Götterdämmerung or Parsifal. (Do not think such constraints affect only modern theaters: practical considerations caused Rossini to make major cuts for the first performances of Guillaume Tell and Verdi to do likewise for Don Carlos.)34
Divas and Scholars Page 27