Here is one example from the opening scene of Rigoletto. In Verdi’s autograph the Count of Ceprano, after feeling the sting of Rigoletto’s tongue, vows revenge. He tells the other courtiers, “In armi chi ha core / Doman sia da me” (Let him who has courage come to my house, armed, tomorrow). But a moment later the same Ceprano turns to the same courtiers and exhorts them, “Stanotte chi ha core / Sia in armi da me” (Let him who has courage come to my house, armed, tonight). Did Ceprano intentionally change the appointment from one day to the other?
From Verdi’s letter to Piave of 14 January 1851, we know that the composer was troubled about the time sequence of the opera.13 All events of the first two acts were originally intended to take place in a single night. But Verdi complained that “after the party Triboletto [sic] changes, sings a duet with the assassin, an endless scene with Gilda, [who has] a duet with the Duke, and an aria, and finally this abduction. This cannot all happen in a single night, since if the party finishes toward dawn, Triboletto [sic] cannot meet the assassin toward evening, and it isn’t very likely that Bianca [sic] would remain awake all night.” Thus, Verdi asked Piave to alter the verses of the chorus in the middle section of the Duke’s aria:
Poiché la festa cessò di corte
Moviamo uniti prima del dì
Once the party at Court was finished
We set out together before dawn.
Piave agreed and gave the courtiers these verses instead:
Scorrendo uniti remota via
Brev’ora dopo caduto il dì
Passing together on a remote street
Just after nightfall,
Verdi was perfectly happy with these new verses, which he entered directly into his autograph manuscript.
At the same time, the composer pointed out to his librettist that the time sequence also had to be changed in the first act, where Ceprano (at that time still called “Cavriano”) originally had the following interchange with the courtiers:
In place of this text, Piave provided Verdi with new words:
Verdi dutifully scratched out of his autograph manuscript the offending words (which he had used in the skeleton score) and replaced them with the new text.
But the composer forgot that the text is repeated a few measures later, and hence failed to correct the second appearance of the text.14 Thus, Ceprano first invites the courtiers to meet him “tomorrow,” then two minutes later asks them instead to come “tonight.” As incredible as it may seem, no printed edition of the opera ever corrected this blatant error, and countless performances made before preparation of the critical edition blithely employed the faulty text. Does the time sequence of Rigoletto really matter? Well, it mattered enough to the composer and his librettist for them, after a careful correspondence, to correct it. Does that mean that every production of Rigoletto must endeavor to render this time sequence palpable? Not necessarily: the relationship between a written score and a performance is a complex one, which forms the subject of the second part of this book. But no matter how performers may wish to manipulate stage time for the purposes of a particular production, it is clear that “Verdi’s Rigoletto,” as preserved in his autograph manuscript, is incoherent at this point.
Of a different order are the many places in the autograph manuscript of Rigoletto in which the composer omitted signs, the kinds of problems on which Denis Vaughan’s polemic foundered. The initial orchestral presentation of the melody of “La donna è mobile,” for example, is played simultaneously (in varying registers) by flute, piccolo, oboe, clarinet, bassoon, first violins, and cellos. The first five notes are normally articulated with three accents and a diminuendo. All printed editions before the critical edition, it is worth pointing out, print these signs as accents over the first four notes, a very different effect and not reflecting at all the notation of the autograph manuscript, which is unusually precise (example 5.1). But in notating the various instrumental parts playing the melody, Verdi wrote a diminuendo at m. 39 only for flute, first violins, and cellos, omitting it in piccolo, oboe, clarinet, and bassoon, whereas at m. 41 he wrote it explicitly in each part.15 Are we supposed to believe that “Verdi’s Rigoletto” was meant to be performed with some instruments playing a diminuendo and the others remaining unchanged? At the courtier’s entrance in the middle section of the Duke’s aria (“Duca, Duca! L’amante fu rapita a Rigoletto!”), Verdi wrote ff in the first violins to signal a loud chord. In the next measure, however, where the instrumentation is vastly reduced (to oboes, bassoons, and strings), there is a pp for bassoons, cellos, and double basses, but no additional marking for the first violins. Are we expected to believe that in “Verdi’s Rigoletto” these violins must continue to play ff while the bassoons, cellos, and double basses assume the lower volume?16
EXAMPLE 5.1. GIUSEPPE VERDI, RIGOLETTO, SCENA E CANZONE [DUCA] (N. 11), MM. 38–41.
And so on through a range of many trivial and a significant number of nontrivial matters, affecting almost every measure of the autograph of Rigoletto. Some tangible intervention is required to transform “Verdi’s Rigoletto,” as incorporated in his autograph, into a musical score that can be used to perform the opera. Such intervention is necessarily the work of some later musician or musicians. The musical score commonly known as Rigoletto in whatever edition, whether “critical” or not, as opposed to “Verdi’s Rigoletto,” results from a series of editorial interventions. The Rigoletto most commonly available before publication of the critical edition in 1983 was the product of interventions that began with the first preparation by the Ricordi firm of a printed full orchestral score for rental during the 1890s, continued with the first publication of a full orchestral score for purchase early in the twentieth century, and was followed by further interventions each time it was reprinted during the twentieth century. If we believe that performers should play the work “as Verdi conceived it,” to which Rigoletto should they turn?
Editorial interventions deemed essential by one generation of musicians are quite different from those deemed essential by another. Nowhere is this more keenly felt than in music written before 1830, where the pendulum of editorial technique has swung from editions adding a full panoply of expressive markings and slurs (in music whose sources had few), to those supplying so little information as to be practically diplomatic transcriptions of an old source, to those accepting again modest editorial interventions. Although the finest scholars of the nineteenth century provided Gesamtausgaben (editions of the complete works) of Bach, Handel, Mozart, Beethoven, and Schubert, over the past fifty years the music of these composers has been (or is being) reedited in new complete editions. Certainly we know more today about their life and works than did nineteenth-century editors, and access to better musical sources can lead to new editions strikingly different from the old. Often, though, the same sources are simply being employed in different ways by a new generation of scholars, committed to a new set of editorial principles.17
Yet it is extremely unlikely that two musicians asked to prepare a printed edition of Rigoletto today would agree in every case about where the score requires editorial intervention and what this intervention should be. Editors make those adjustments they consider necessary on the basis of the historical moment at which they are working, the musical culture by which they are surrounded, their specific knowledge of the music of Verdi and its sources, their inherent musicality, and their insight and intelligence. A critical edition of Rigoletto, then, is necessarily an interpretation of “Verdi’s Rigoletto”—as is any edition whatsoever. The difference is that a critical edition makes its substantive interventions graphically explicit, and explains them in ample critical notes. Users of the score can ascertain where “Verdi’s Rigoletto” ends and editorial intervention begins. A critical edition also differs from other editions in its insistence that criteria for editorial interventions be clear and that such interventions be restricted to those which derive from Verdi’s explicit indications or meet levels of consistency and logic that r
eflect Verdi’s notational practice. But it would be naive to assume that universal agreement could ever be reached as to what those levels should be.
FINDING THE SOURCES
Although a scholar’s work is often accomplished in the library or in his or her study, the preparation of critical editions depends fundamentally on locating and using sources, particularly autograph sources, hidden in nooks and crannies of public or private libraries, in venerated collections of noble families, in bank vaults in Switzerland. Some library collections are well catalogued and preserved, although particular items may go unmentioned or be unknown for decades. Sometimes identifications are faulty, even in responsible libraries. The Library of the Conservatory of Milan long thought it possessed the autograph manuscript of Rossini’s La gazzetta, while the real autograph was (and still is) in the library of the Naples Conservatory. The copy that ended up in Milan, part of the “Noseda” collection, was “authenticated” by the nineteenth-century librarian of the Naples Conservatory, Francesco Florimo, who was thoroughly knowledgeable about Rossini’s handwriting and surely knew that his identification was fraudulent.18 Then again, the library of the Bologna Conservatory believed for years—against all appearances—that it owned the autograph manuscript of Rossini’s Stabat Mater, when the composer’s original was sitting in the British Library in London, where it remains.19
Entire library collections can be in sad disarray. Among Rossini sources at the Naples Conservatory in the mid-1960s were autograph manuscripts that had gone unidentified, including several works presumed lost. Only by asking to see every item related to Rossini in the library (and extending a projected two-week stay to six weeks) was I able to sort them out. Among the unknown works was Rossini’s 1820 Messa di Gloria, of which the original performing parts (with autograph annotations by Rossini) turned out to be in the collection.20 The problem is not restricted to Italy. Edmond Michotte was a close friend of Rossini’s in his youth and later director of the Brussels Conservatory, to which he left his important collection of manuscripts.21 Visiting the conservatory for the first time in 1965, I found a pile of uncatalogued manuscript pages, few of which had ever been examined. As I worked through that pile, a treasure trove appeared: alongside the autograph of an entire Rossini opera, Matilde di Shabran, were autograph manuscripts of arias from operas, songs, and sketches; copyist’s manuscripts belonging to Isabella Colbran and Giuditta Pasta (sometimes with ornamentation in Rossini’s hand); and other sources. There was also the one surviving page of the autograph manuscript of the Messa di Gloria, which Rossini had given to Gustave Vaëz in 1846.22 Not until the 1970s did the Music Department of the Bibliothèque Nationale in Paris admit grudgingly, to Elizabeth Bartlet (a Canadian scholar working on the operas of Méhul) that they owned a considerable collection of manuscript parts from the archive of the Théâtre Italien. That collection ultimately yielded the first musical sources to surface in the twentieth century pertaining to Rossini’s Il viaggio a Reims. I will have more to say about that work and its discovery later in this chapter.
The quest for manuscripts in private collections can be especially satisfying or especially frustrating. A fleeting glimpse of an important source in an auction or a dealer’s catalogue or a casual reference in a book can lead to an extended search, sometimes crowned with success, sometimes not. Since the publication of the first volume of Julian Budden’s magisterial The Operas of Verdi, for example, it has been known that Verdi prepared a new romanza, “Sventurato! Alla mia vita,” for his opera Attila when the Russian tenor Nicola Ivanoff sang the role of Foresto in Trieste during the autumn of 1846 (shortly after the opera’s Venetian premiere on 17 March of the same year).23 Ivanoff was a great friend of Rossini’s, and it was Rossini himself, in a letter of 21 July, who commissioned Verdi to write the piece for his protégé. Verdi was not fond of “substitute” arias, but he could hardly refuse Rossini. And so Verdi, in turn, penned this wonderful note to Piave on 10 August:
I need a favor: a romanza with recitative and two quatrains; the subject will be a lover who is moaning about the infidelity of his beloved (old hat!). Write me 5 or 6 lines of recitative, then two quatrains of ottonari; there should be a masculine ending every other line, because it’s easier to set that way...
Make sure they’re pathetic and tearful: have that imbecile of a lover say that he would have given up his share of paradise and that she rewarded him with... Horns... Long live those horns: bless them!... If I could, I’d like to give them out myself all the time!24
By early September, Ivanoff had his romanza.
Ivanoff sang the romanza for the final time in Turin in 1849, after which all traces of it disappeared. Then, more than thirty years ago, Verdi’s autograph manuscript was acquired by a distinguished London antiquarian music dealer, Albi Rosenthal, who nevertheless refused to show it to anyone. After he sold the manuscript, its new owner would not even allow his identity to be revealed. Finally, in the spring of 1992, I was asked to write an article for a memorial book in honor of the collector Hans Moldenhauer, who had recently died, leaving an extensive group of manuscripts to the Library of Congress. The library wanted me to describe a Verdi aria in the collection, which they believed to have been written for Ernani. It turned out to be “Sventurato! Alla mia vita.” This beautiful romanza, acquired by Moldenhauer from Rosenthal and then willed to the Library of Congress, will be published in an appendix to the critical edition of Attila. Its first modern performance in the context of Attila took place at Lyric Opera of Chicago during its 2000–2001 season, where it was well received.
Many quests are less successful. The autograph manuscript of the principal soprano aria (“Quelle horrible destinée”) that Rossini added to his Italian opera Mosè in Egitto when he transformed it as Moïse for the Opéra of Paris in 1827, sits inaccessible in a private collection. It appeared fleetingly in an exhibit of musical manuscripts owned by collectors in Basel, but all subsequent efforts to see it have failed.25 In an 1864 biography of Rossini, the French critic Aléxis Azevedo mentions that he owns the autograph manuscript of a chorus from the same opera, but no one has ever found it.26 In 1916 a manuscript of Verdi’s sketches for a major scene in Jérusalem was sold in New York: its subsequent whereabouts are unknown.27 Donizetti’s manuscript for La Favorite was owned by a wealthy Milanese family, the Treccani degli Alfieri, in the 1940s, but has since disappeared. Although a microfilm copy is deposited in the New York Public Library, the manuscript is so complex (much of it began life as a different, unperformed work, L’Ange de Nisida) that access to the original would be an enormous boon.28 Verdi letters fundamental to our understanding of his middle-period operas are regularly bought anonymously at auction and disappear for decades. Most auction houses will send requests for information to new owners, to be sure, but when the owners disregard those requests, there is no recourse.
Even if we grant collectors their right to privacy, musical manuscripts are sometimes bought and sold with noticeable avidity and even duplicity. In 1979, Christie’s in London attempted, unsuccessfully, to auction what they claimed to be the autograph manuscript of a wedding Cantatina by Rossini. From a reproduction in the auction catalogue, it was perfectly clear to Rossini scholars that the manuscript was not in the composer’s hand.29 Nor could it have been a copy of a piece by Rossini, since the score was obviously a composing manuscript, not a copyist’s rendering. I immediately wrote to Christie’s, alerting them to their error; they ignored my letter. But when the manuscript remained unsold, it was duly returned to the owner, who later brought it to be seen by scholars at the Fondazione Rossini. We informed him in no uncertain terms that the manuscript was not in Rossini’s hand. But, the owner insisted, the musical style is similar to Rossini’s (and whose style in Italy in 1832 wasn’t?); further, the manuscript was certainly written early in the nineteenth century (and what does that have to do with Rossini’s handwriting?). Many years later he was still trying to convince us. Finally, in 2003, he persuaded the artistic direct
or of the “Hampstead & Highgate Festival” to give the Cantatina its “world premiere,” listing it as “Rossini (attrib.).” The printed program included two pictures: part of Rossini’s autograph manuscript for the Stabat mater and a snippet from the Cantatina. Look at the capital “E,” the program notes suggested, ignoring the fact that the clefs and time signature, absolutely characteristic of Rossini’s handwriting, were entirely different from those in the Cantatina. And so disinformation continues to spread.
On another occasion a manuscript dealer from Los Angeles, Scriptorium, claimed that an authentic Rossini manuscript was a fraud. The story pertains to one of the first volumes published in the Rossini critical edition, a volume of piano music from his Péchés de vieillesse (Sins of Old Age), as Rossini called his late Parisian music, 1855–68. This volume was entitled Quelques Riens pour album (Several Nothings for Album), in that ironically self-deprecating manner typical of late Rossini.30 These twenty-four piano pieces, complex and often fascinating, are anything but traditional album leaves. Not surprisingly, Respighi used several of their themes in the ballet score he arranged in 1919 from Rossini’s late piano music, La Boutique fantasque, for the Ballets russes. All but one of the twenty-four autograph manuscripts are in the collection of the Fondazione Rossini; the remaining autograph is in the private collection of a generous American music lover, Mario Valente, who kindly shared it with us.
Divas and Scholars Page 21