Divas and Scholars
Page 9
With a libretto (or at least part of one) in hand, the composer set to work. Time was short even under the best of circumstances. Under the worst, the need for disciplined creativity must have been overwhelming. While extant documents in the composer’s hand, sketches or autograph manuscripts, are an inadequate guide to the complex mental processes (conscious or intuitive) through which a composer produced a new work, they are all we have. Understanding these documents is crucial, for the dissemination—or what I will refer to as the transmission—of nineteenth-century Italian operas in handwritten or printed form begins with them. Many of the difficulties in producing an edition of an opera or performing that opera derive directly from problems inherent in these earliest autograph sources.
Sketches were strictly private, not intended for any eye other than that of the composer. The very few known sketches in Rossini’s hand pertain to some of his most complex music: the terzettone in Maometto II, two extensive numbers from Semiramide (the first-act finale and the great scena and aria for Arsace in the second act), the trio “A la faveur de cette nuit obscure” from Le Comte Ory, and the second-act finale added to the French Moïse. What the composer chose to write down, however, often involves details: only the sketches for Semiramide and Le Comte Ory establish the musical content for long stretches of a composition. Still, nothing suggests that Rossini habitually had recourse to extensive sketches during the preparation of most of his operas.35 He probably adopted them largely for the final works of his Italian career (after 1820) and then again in preparing music for the Opéra (1826–1829). A similar pattern seems to be true for Donizetti. Bellini, on the other hand, often worked quite differently, as we have seen, writing down page after page of melodies before he had a libretto in hand. Later, he would return to those pages as a bank of ideas from which to draw the lyrical capital he would transform into the vocal lines of an opera.
Of the four major Italian opera composers working during the first half of the nineteenth century, Verdi alone employed independent sketches extensively. Since 1941, when the Verdi family permitted Carlo Gatti to publish the sketches for Rigoletto in facsimile, scholars have been intrigued by this material.36 While the Rigoletto sketches contain a few miscellaneous jottings pertaining to individual melodic ideas, they mostly consist of a draft of the entire work, beginning in the festive atmosphere at the court of the Duke of Mantua and concluding with the tragic death of Gilda and desperation of Rigoletto. Verdi worked out the opera in order, number by number, in a form that included the principal vocal lines with a bass part, as well as significant orchestral melodies. This type of sketch is commonly referred to as a “continuity draft” of the opera. According to Gatti, similar sketches exist at Verdi’s home in Sant’Agata for every opera from the 1849 Luisa Miller through the 1893 Falstaff, as well as for the major nonoperatic works such as the Messa da Requiem of 1874. In his Verdi nelle immagini Gatti provided a few barely legible samples from other operas, whetting the appetites of all those who study Verdi’s music.37
Study of the Rigoletto sketches makes it clear that Verdi’s repeated requests for changes in the libretto were almost all formulated during the process of sketching, which is what he called the “creative work” for an opera. Employing the text supplied by his librettist, Verdi used the continuity draft to set out the basic musical structure of each piece in the opera. When problems arose with the words, he would improvise a solution or even draft a passage without text, immediately shooting off a letter to the librettist requesting modifications or even entirely new poetry in a specific meter or with specific accentual patterns.38
Our knowledge of the Verdi sketches has increased significantly since Gatti’s publications. Recently the family, through the good offices of the Istituto Nazionale di Studi Verdiani, has made available to editors of The Works of Giuseppe Verdi the sketches for Stiffelio, La traviata, Un ballo in maschera, and La forza del destino. These manuscripts usually have continuity drafts of individual numbers or even entire acts, similar to those we know from the sketches for Rigoletto. To a greater extent than in the case of Rigoletto, however, they also provide preliminary notations for an opera or significantly different earlier versions. The sketches for Stiffelio offer many early ideas for solo compositions, ideas independent of any continuity draft.39 The Ballo continuity draft has an entirely different orchestral introduction for the Amelia aria that opens the second act. For La traviata Verdi laid out in an abbreviated form the basic structure of the entire first act before receiving any poetry from Piave, even before knowing what the characters’ names would be in Italian. He set down in words and with a few selected themes his sense of how the drama would be shaped.40 Among the wordless themes staring at us from this intriguing page are the brindisi or toast (“Libiamo ne’ lieti calici”), “Ah forse è lui,” and “Sempre libera.” We do not yet know enough about Verdi’s working habits, however, to gauge whether this page is an anomaly or whether he regularly wrote down ideas for his operas before having the libretto in hand.
By using extensive independent sketches (usually continuity drafts) for all his mature operas, Verdi interposed a level of advance planning in the compositional process that earlier nineteenth-century Italian opera composers rarely employed. Only after preparing these continuity drafts did he begin to lay out the autograph manuscript of an opera, the public document that would subsequently be used by copyists and editors. First he would prepare what we call a “skeleton score,” basically entering the same material he had included in his continuity draft (vocal lines, bass, important instrumental ideas), now in a more complete form and with each element in its proper position in what would become the full score; subsequently he would fill out the skeleton, so to speak, by completing the orchestration.
Sometimes there are quite significant differences between the continuity drafts and the final product. This is particularly true for operas such as Un ballo in maschera, where fifteen months of battles (from November 1857 through January 1859) with the censors of two cities (Naples, then Rome) gave the composer a longer time than usual to ponder the music he had earlier sketched.41 It is also true for La forza del destino, where the illness of a prima donna caused the postponement of the first performance in St. Petersburg from the beginning of 1862 to November of the same year.42 On most occasions, however, Verdi’s continuity draft for a musical number is quite similar to the final version, and the task of copying the vocal lines, bass, and important orchestral melodies from that draft into his autograph manuscript was largely mechanical. No composer of the period, of course, would ever have “copied” his own music without touching up details of articulation, rhythm, even pitches, and confusions resulting from this process remain visible throughout the autograph manuscripts. Access to the sketches, then, turns out to be of great assistance to scholars and performers who wish as complete a view as possible of the written sources for a Verdi opera.
What Verdi prepared as an independent continuity draft and then copied over as a “skeleton score,” Rossini, Bellini, and Donizetti usually entered straightaway into their autograph manuscripts. They certainly worked ideas out at the keyboard, and they sometimes jotted thoughts down on paper, but for these composers the “skeleton score” was their continuity draft. They wrote their music in its skeletal form (vocal lines, bass, and significant orchestral ideas) directly into their autograph manuscripts. This difference in the way they composed explains many of the physical differences between the autograph manuscripts of Verdi, on the one hand, and those of Rossini, Bellini, and Donizetti, on the other. It also explains many of the contradictions and problems that emerge when one studies or performs their completed operas.
A COMPOSER’S PUBLIC DOCUMENTS: AUTOGRAPH MANUSCRIPTS
Paper. A composer of Italian opera could not begin to prepare the autograph manuscript of a work until he had procured music paper. This was not a trivial matter. Four issues, some interrelated, needed to be considered: the quality of the paper, its physical dimensions and format
(whether oblong—wider than it is high—or vertical), the number of staves ruled on each page (already present when the paper was purchased), and its structure (single pages, bifolios, or gatherings). The choices a composer made at this stage were important for the state in which his opera would be transmitted to the musical world.
Most composers during the first half of the nineteenth century employed paper of excellent quality, and their autograph manuscripts have survived in exemplary condition. It was not until the second half of the century that the quality of the materials regularly used in papermaking deteriorated to the point that many documents from that period are today disintegrating, to the dismay of librarians, collectors, and scholars. But even earlier in the century there could be severe and unfortuante discrepancies in the paper. In composing Elisabetta, regina d’Inghilterra (1815), Mosè in Egitto (1818), and La donna del lago (1819), Rossini used paper of excellent quality for the musical numbers, but of poorer quality for the intervening recitatives, even though these are all accompanied by the orchestra. The ink has bled through the paper, often making his notation difficult to read. This bleeding must have begun early in the history of the documents, for pages within passages of recitative surely composed by Rossini or entire recitatives perhaps composed by him were recopied by another hand and inserted into the autograph manuscripts.43
In La donna del lago these recitatives in the hand of a collaborator are particularly unfortunate, since Rossini also employed an anonymous assistant to compose the aria for Duglas (“Taci, lo voglio”) in the first act. As a result, we cannot always be certain which recitatives were composed by Rossini and which by his assistant. That knowledge is of more than theoretical significance. When Houston Grand Opera staged La donna del lago in 1981, the general manager, David Gockley, realized during the dress rehearsal that he was facing the costly specter of double overtime for the orchestra. Five minutes needed to be cut from the score, he told me, and five minutes were indeed cut. But in making those eminently practical decisions, it would have been nice to know which recitatives were actually by Rossini himself.44
As operatic scores during the first half of the nineteenth century became more complex orchestrally and more often employed large ensemble scenes, the size of manuscript paper used for those scores gradually increased. The change was by no means uniform. Rossini’s paper for his Neapolitan works during the 1810s is in oblong format and measured around 28 cm in width by 23 cm in height, but his Guillaume Tell of 1829 required much larger paper, measuring 33 cm by 25 cm, hardly different from that used by Verdi for most of his operas in the 1840s. The Verdian scores, however, were mostly written on paper in vertical format, approximately 24 cm in width by 33 cm in height. There could occasionally be extreme variations in these proportions. The composer Saverio Mercadante, a figure to be reckoned with in the world of Italian opera from the 1820s through the 1840s, began to lose his sight toward 1840, and he tended to require ever larger manuscript paper, sometimes involving two sheets pasted together.
More significant than changes in the size of the paper was the transformation in its format. Early in the century composers wrote their operas on music paper in oblong format; by the 1840s vertical format had become standard. Vertical paper allowed a composer to write fewer measures per page, but provided enough staves for complex ensembles, using an expanded orchestra, stage band, soloists, and chorus. In Rigoletto, for example, Verdi employed upright paper with thirty staves per page for his massive “Introduzione,” the Duke of Mantua’s festa that comprises the entire first scene. Oblong paper allowed composers to write more measures per page, but limited the number of staves available. For the operas of Cimarosa and Paisiello, oblong paper had been adequate, as it was for Rossini’s earlier operas. At a certain point, however, the number of staves that could comfortably be drawn on standard oblong paper was insufficient for many composers’ needs, and it led them into various compromises.
In the ensemble sections of the first-act finale of Il pirata, for example, the sheer size of the forces led Bellini, who was still using oblong paper, to write out the vocal parts in one section of the manuscript and the instrumental parts elsewhere. If this did not promise enough confusion, the vocal parts of the final ensemble (“Ah! partiamo, i miei tormenti”) are notated in one key (C minor/ major) and the instrumental parts in another (B minor/major). At some point between preparing the vocal parts and completing the orchestration, the composer changed his mind, but this history remains to be unraveled.45
Even without going to such extraordinary lengths, most large ensembles by Rossini cannot be accommodated on the oblong music paper he chose for his autograph manuscripts, so that he was obliged to employ what composers referred to as spartitini, “little scores,” in which were gathered instrumental parts that could not fit into the main score because of space limitations. They are almost always mentioned in the main score, where Rossini wrote phrases such as: “Tromboni, Timpani e Gran Cassa in fine” (Trombones, Timpani, and Bass Drum at the end). But often these original spartitini have disappeared, and secondary sources preserve multiple versions. The autograph manuscript of Rossini’s Semiramide, for example, refers to a number of spartitini, but none are present in the manuscript. The confused and contradictory readings of secondary sources suggest that these spartitini were misplaced almost at once. Only in 1990, during preparation of the critical edition of the opera for its first performance at the Metropolitan Opera, when the original orchestral parts were closely examined by my associates at the Fondazione Rossini, Patricia Brauner and Mauro Bucarelli, did the autograph spartitini of Semiramide emerge, hidden among thousands and thousands of pages in the hands of various copyists. For the first time since 1823 it was possible to restore Rossini’s original orchestration.46
That the paper used throughout an opera was generally of the same size and format is hardly surprising; it is not difficult to imagine the problems that could (and did) arise in binding and preserving an autograph manuscript whose papers were of varying sizes. But there was no reason for every composition in an opera to use paper pre-lined with the same number of staves. Indeed, to ensure maximum legibility, it was advantageous for a composer to choose manuscript paper for each composition with the fewest possible staves, consonant with vocal and orchestral requirements. The larger the staves and the more room between them, the easier it was for a composer to write clearly his signs of articulation, notes, and text. Thus, in Verdi’s Alzira, the vertical paper ranges from sixteen to twenty-four staves, with spartitini added for the finales to the second and third acts.47 In Rossini’s Otello, the oblong paper ranges from ten to sixteen staves, with a spartitino added for the finale to the first act; in Guillaume Tell, on the other hand, the larger oblong paper used by Rossini ranges from twelve to twenty-two staves, with spartitini for the finales to the first, third, and fourth acts.48
This physical description of the paper has exquisitely musical implications. In Rossini’s Tancredi, the composer chose ten-stave paper for Amenaide’s prayerful aria in the second act, “Giusto Dio, che umile adoro.”49 These staves are assigned to the following instruments:
First Violins
Second Violins
Violas
1 Flute
2 Oboes
2 Clarinets
2 Horns
2 Trumpets
Amenaide
Violoncellos and Double basses
(Notice the typical nineteenth-century layout of the score, with upper strings at the top, followed by winds and brass, vocal parts, and then lower strings.) But Rossini also wanted two bassoons to play, for in the margin he wrote “I Fagotti nella riga del basso” (the bassoons with the bass line), suggesting that the bassoons should join with the violoncellos and double basses. Yet, the situation is more complex: in six measures within the aria, explicit bassoon parts are notated on the staff normally assigned to trumpets (silent at this point). Did Rossini really want the bassoons to play everywhere else with the bass of the strings? E
ven in passages scored for pizzicato strings alone, where the presence of the bassoons would create a dubious effect? Had Rossini chosen twelve-stave paper, we would not have to guess. As it is, his choice of ten-stave paper created uncertainties with which every generation of copyists, performers, and editors has had to cope. Perhaps when he began drafting the number he did not know there was going to be a problem; realizing too late that there was one, he improvised an unsatisfactory solution rather than rewrite the entire score or provide a spartitino for the bassoons.
Nor are the Verdi autographs immune to such problems. Elvira’s cavatina in Ernani (“Ernani, Ernani, involami”) is written on sixteen-stave vertical paper, organized as follows:50
First Violins
Second Violins
Violas
1 Flute
1 Piccolo
2 Oboes
2 Clarinets
2 Horns in F
2 Horns in B
2 Trumpets in E
2 Bassoons
3 Trombones and Cimbasso
Elvira
Chorus of Women
Violoncellos
Double basses
Whenever a composer places the parts for two instruments on the same staff (even identical instruments, such as two oboes), problems arise, unless he is explicit—and few of the composers under discussion were—about how many instruments should be playing at any moment in those measures where only a single melodic line is present. But the problems become truly daunting in this Ernani example, where Verdi notated three trombones and a cimbasso on a single staff. When Verdi wrote four notes, the matter is clear; when he wrote three, it becomes ambiguous; when he wrote two or one, it is incomprehensible. In Elvira’s cavatina at one point he indicated “cimbasso solo.” A later passage is marked simply “solo.” But which instrument should play that solo, a trombone or the cimbasso? Verdi’s notation provides no further information. The old Ricordi performing parts assigned the part to the first trombone, but the logic of the musical situation suggests it should be played by the cimbasso. Had the composer chosen eighteen-stave paper, we would not have to guess.51