Tracing the transmission of an early-nineteenth-century Italian opera, in short, is not always simple, yet the issues involved are fundamental to the music we have before us as we attempt to perform an opera. L’Italiana in Algeri, in fact, provides a useful case study.
THE TRANSMISSION OF L’ITALIANA IN ALGERI
Opera houses during the first decades of the nineteenth century based their productions on manuscript copies, from which orchestral parts were drawn. But the copies in circulation were of many different types, and this diversity reveals how fraught with uncertainty was the transmission of an opera, especially one as popular as L’Italiana in Algeri. The problem is well exemplified by four different manuscripts of the opera preserved in the library of the Naples Conservatory. They demonstrate how scores still in use during the first three-quarters of the twentieth century acquired their particular characteristics.
Copying the Warts and All
The manuscript copy identified as “Rossini 22–1–59, 60” of the Naples Conservatory was made directly from Rossini’s autograph of L’Italiana in Algeri (or from another copy equally strict in reproducing its contents). It preserves every mistake of the original, every peculiarity of layout, every wart. Here is an example. Since the opera was produced under enormous time pressure, its libretto was derived from a preexisting one by Angelo Anelli, prepared for the Teatro alla Scala in 1808 (where it was set to music by Luigi Mosca) and revised in 1813 to suit Rossini’s requirements.41 To assist him in meeting his deadline, Rossini employed an anonymous collaborator, who composed Haly’s aria “Le femmine d’Italia,” perhaps Lindoro’s second-act “Ah come il cor di giubilo,” and all the secco recitative. This collaborator worked mostly from the printed libretto of 1808, using a separate bifolio (or bifolios) of music paper for each individual passage of recitative. These separate bifolios could then be placed in the right order within the complete score, between the musical numbers Rossini was preparing.
Here is the text of the final section of the first lyrical number of Rossini’s opera (its “introduction”) and the beginning of the ensuing recitative. The text is identical to the libretto prepared by Anelli in 1808. The introduction concludes with a quatrain in ottonari, whereas the recitative, as always, is a mixture of settenari and endecasillabi.
Rossini’s collaborator failed to notice that scene 1 concludes with two lines of recitative, “Ritiratevi tutti. Haly, t’arresta. / (Che fiero cor!) (Che dura legge è questa!),” before continuing—still in recitative—with scene 2 for Mustafà and Haly. Thus, he set this recitative only from the beginning of scene 2. After the manuscript of the opera was assembled, with Rossini’s introduction followed by the copyist’s incomplete recitative, someone became aware that the two lines of recitative at the close of scene 1 had inadvertently been omitted. Since Rossini had completed his introduction on the front side of the paper (its recto), the maestro responsible for the recitative entered a setting of the missing lines on the back side of that page (its verso), which had previously been blank. This peculiarity in the structure of Rossini’s autograph is altogether without musical importance, but it offers a revealing glimpse of how the (not entirely) autograph manuscript of the opera was put together. The faithful copyist of “Rossini 22–1–59, 60” repeated precisely this physical organization, an absurdity: every other manuscript source keeps all the measures of recitative together. But throughout the score, “Rossini 22–1–59, 60” similarly preserved every quirk and error of the original. Identifying such copies can be of great importance when Rossini’s own manuscript no longer survives. They give us access to a missing autograph through the eyes of a contemporary musician who tried not to exercise personal intelligence but aimed simply to reproduce what was in front of him. In the case of L’Italiana in Algeri, however, for which the autograph has survived, a copy of this kind is wholly without interest.
Hackwork
Manuscripts that reveal such a strict attitude toward copying are rare. More typical is the manuscript “Rossini 22–1–57, 58,” preserved at the Naples Conservatory library but prepared “in Florence in the music copisteria of Francesco Minati da Badja.” While this score incorporates essential features of L’Italiana in Algeri and makes no fundamental alterations to the sequence of musical numbers or their internal structure, it fails to respect the details of Rossini’s notation. Concerned primarily with getting down on paper the pitches, text, and rhythms, most copyists worked quickly. When rhythms proved complex or contradictory, especially in ensembles, copyists would simplify the notation, replacing a double-dotted rhythm by a single-dotted one: it is extremely rare to find a copyist who transforms a simple rhythm into a more complex one. Once a copyist had falsified Rossini’s notation, of course, others followed along or introduced further simplifications. Articulation signs (staccatos, accents, slurs) or dynamic signs (dynamic levels, crescendos, diminuendos) suffered a similar fate. Autograph manuscripts are hardly meticulous in their presentation of such signs, but examples entered by the composer are sufficiently extensive to provide necessary information for editors and performers. Manuscript copies, on the other hand, rarely preserve more than 50 percent of the signs of articulation or dynamics present in an autograph. Even those they do preserve are too often misread or misinterpreted.
Early in the nineteenth century, for example, “closed” crescendos, diminuendos, or accents were used extensively in northern Italy. I have seen them in autographs of Simone Mayr and Rossini, but there is every reason to believe they were fairly widespread (example 3.2).42 That a “closed” diminuendo or crescendo is sometimes present in one part while a sf (sforzato, or sudden and intense accent) followed by a diminuendo, or a crescendo leading to a sf, is present in another helps us interpret the signs: a “closed” crescendo indicates an increase in volume, leading to a sf; a “closed” diminuendo indicates a sf attack, followed by a diminuendo; a “closed” accent suggests a more intense accent than normal. Copyists, particularly in Rome or Naples, unfamiliar with this notation, inevitably substituted regular crescendos, diminuendos, and accents, abandoning a group of signs highly characteristic of Rossini’s style.
EXAMPLE 3.2. “CLOSED” ACCENTS, DIMINUENDOS, AND CRESCENDOS, ALL SIGNS CHARACTERISTIC OF ROSSINI’S AUTOGRAPH MANUSCRIPTS.
The presence of closed signs in manuscript copies, on the other hand, suggests that these copies are particularly close to the autograph. In a work from Rossini’s youth such as L’equivoco stravagante, for which no autograph survives, the one manuscript containing these signs immediately acquires authority. Although this manuscript is preserved in Paris, it comes from Bologna, where the opera was first performed in 1811, and could easily have been copied from Rossini’s original. Since no other surviving manuscript contains closed signs, the Paris copy could not derive from them: no copyist would have added such signs. This is only one of the reasons why the Paris manuscript of L’equivoco stravagante turns out to be the best surviving source. In it, for example, we find many distinctly off-color expressions, forbidden by the Bolognese authorities and absent in most other sources. Then again, one could well ask what Rossini was thinking when he set to music a libretto in which the heroine pretends to be a eunuch in disguise to avoid an unacceptable suitor!43
Copies of L’Italiana in Algeri such as “Rossini 22–1–57, 58,” in short, represent an opera’s bare outlines. This is the kind of hackwork found in the vast majority of manuscript copies through which Rossini’s operas normally circulated during the nineteenth century.
L’Italiana Transformed “in Farsa”
A worse fate faced Rossini’s score in Naples. In the manuscript “Rossini 22–1–62” of the Naples Conservatory library, L’Italiana in Algeri was “transformed in farsa,” i.e., arranged according to Neapolitan practice in a single act with spoken dialogue replacing the secco recitative. This combination of musical numbers and spoken dialogue was well known in the Naples comic theater: Donizetti’s one-act farsa, La romanziera e l’uomo nero, first pe
rformed at the Teatro del Fondo on 18 June 1831, was a work of just this type. From printed librettos, we also know that Rossini’s Il barbiere di Siviglia and La Cenerentola were often performed in Naples with spoken dialogue. Preparing the farsa version of L’Italiana, the reviser used a hackwork manuscript, eliminating many musical numbers and all the recitative. What remained was a series of musical numbers to be linked by spoken dialogue. In one case the copyist actually specified “Segue Prosa” at the end of a musical number. The Neapolitan provenance is further confirmed by the substitution for the original “Taddeo” of the typical Neapolitan name “Pompeo,” a practice followed in Naples even when the opera was performed complete—with recitatives—under Rossini’s direction in 1815.44
What is most amusing about the manuscript, though, is that bound at its beginning, in an entirely different hand, is a complete set of secco recitatives, not to the original libretto but clearly to the text that had been spoken in this one-act farsa. Since the dialogue had been heavily cut and manipulated, a later musician provided a new setting of this modified text, using quite a different musical style from Rossini’s original collaborator. In two stages within “Rossini 22–1–62,” then, L’Italiana in Algeri was first transformed in farsa with spoken dialogue, in the Neapolitan manner, then transformed into a one-act opera buffa with new secco recitative.
While one might complain that this sort of manipulation is disastrous for Rossini’s opera, it reflects well-defined objectives, inherently practical and specific. Those musical numbers preserved, furthermore, are indeed Rossini’s, and they constitute many of the finest moments in the opera. “Highlights from L’ltaliana in Algeri,” “rossini’s greatest hits”: a similar urge still motivates those who produce CDs and those who purchase them.
Transforming L’Italiana: The Censor
The most insidious aspects of transmission are those that transform a work from within, responding to pressures exerted by censors, altering the original structure or changing the sound to suit the perceived needs of a later generation. They are insidious because they are invisible, assuming an aura of authority to which they have no right. The fourth manuscript of L’ltaliana in Algeri in the Neapolitan collection “Rossini 22–1–61” exemplifies all these problems. Originally prepared in the Roman copisteria of Giulio Cesare Martorelli, the score is entitled Il naufragio felice [The Fortunate Shipwreck] ovvero L’ltaliana in Algeri, as the opera was known in its first Roman performances during the 1814–15 season.45 In this form the manuscript represented the hackwork standard among contemporary copies. But it goes a step further, incorporating alterations and accretions that reflect censorship, the pretensions of singers, and shifts in musical taste.
We are still today performing operas with their librettos modified by nineteenth-century censors, even though in many cases we could easily recapture the original text. Since censored versions of L’ltaliana in Algeri did not circulate widely, the opera avoided that fate, but more than one manuscript was affected. The Martorelli copy reflects practices in Rome, the city in which theatrical censors brought the most exasperating criteria to bear. Anything related to “la patria” had to be suppressed, and so Isabella’s famous rondò, encouraging the Italian slaves to flight, became “Pensa allo scampo” (Think of escape) instead of the original “Pensa alla patria” (Think of your country). Another Roman substitution, attested to in a libretto from 1819, was “Pensa alla sposa” (Think of your wife).46 But in Rome the word “patria,” in whatever context, was regularly replaced by “dover” (duty) or “affetto” (feelings).47
The effect on Italian opera of political and religious censorship throughout the first half of the nineteenth century was often devastating. The refusal of Neapolitan censors to permit the staging of Donizetti’s Poliuto, in which the life of a saint became material for a melodrama, led to the composer’s departure from Naples (and contributed to the depression that led the tenor scheduled to create the title role, Adolphe Nourrit, to commit suicide by throwing himself from the window of his Neapolitan apartment).48 In order to have I puritani produced in Naples, Bellini felt obliged to omit its most successful single number, the patriotic duet “Suoni la tromba.”49 Verdi’s experiences with Stiffelio, Rigoletto, and Un ballo in maschera (written, respectively, for Trieste, Venice, and—originally—Naples) are only the better-known cases of a long-smoldering problem.
Despite its buoyant exterior, L’Italiana in Algeri was more problematic to the censors than many other Rossini operas. One manuscript of the period shows a fascinating alteration for political motives. Announcing themselves ready to fight for freedom, the chorus of Italian slaves sings:50
Pronti abbiamo e ferri, e mani
Per fuggir con voi di qua,
Quanto vaglian gl’Italiani
Al cimento si vedrà.
Our weapons and hands are ready
To fly with you from here,
What Italians are worth
You’ll see in the moment of trial.
These lines, which were already present in the 1808 libretto for Luigi Mosca, might have been considered dangerous in the best of circumstances. Rossini made them much worse by accompanying the last two verses with a melody prominently featured in the orchestra that is an unmistakable reference to the French Revolutionary anthem, La Marseillaise (example 3.3). The irony was pungent: by 1813 the French had overstayed their welcome in Italy and hardly constituted a model of patriotic virtues. Nonetheless, two early manuscripts of L’Italiana, prepared in Venice and its surrounding region, recast the music of this chorus to avoid this potentially subversive quotation.51
EXAMPLE 3.3. GIOACHINO ROSSINI, L’ITALIANA IN ALGERI, CORO, RECITATIVO E RONDÒ ISABELLA (N. 15), MM. 26–27, ORCHESTRAL MELODY, AND A PHRASE FROM LA MARSEILLAISE.
If politics was taboo, so was sex, especially in Papal Rome. Describing how she can have her way with any man (“So a domar gli uomini come si fa”), Isabella in the original version of L’Italiana in Algeri concludes with these quinari verses:
Tutti la chiedono,
Tutti la bramano
Da vaga femmina
Felicità.
All men ask for it,
All men yearn for it
From a beautiful woman
Happiness.
But not even this suggestive snicker (which in Marilyn Horne’s interpretation could become pretty raunchy) was acceptable to Roman censors in 1819, who transformed the verses into:
Per noi sospirano,
Per noi delirano,
Sol da noi sperano
Felicità.
They sigh for us,
They are mad for us,
From us alone they hope for
Happiness.
Much more moral.
The locus classicus of censorial morality, of course, is Rigoletto, where censors had much to preoccupy them, as Verdi knew.52 At one time he and Piave planned to incorporate the scene in Hugo’s play (itself banned from the Parisian stage after a single performance on 22 November 1832) in which the abducted Blanche (Gilda) seeks protection from the courtiers by locking herself in a room, which turns out to be the King’s bedchamber. The third act concludes with the King’s producing the key, opening the door, and disappearing within. That was hardly going to pass muster in Italy any more than in France, and Verdi himself found it in questionable taste. As he wrote in a ribald letter of 25 November 1850:
I carefully examined the second act, and I think that for us too it would be better to find a [different] place for an aria for Francesco [King François I / the Duke of Mantua]: think about it, and I’ll do the same, and write to me about it. We ought to find something more chaste and get rid of that much too obvious fotisterio [brothel]. And remove the chiave [key] which suggests the idea chiavare [to fuck] etc. etc. Oh Heavens! they are simple things, natural, but the patriarca [patriarch or religious official] can’t take delight in this idea any more!!53
This is a problem that Piave and Verdi never did fully resolve, and if
Rigoletto has a weakness, it lies in the ambivalent feelings of the Duke for Gilda, which the Duke’s second-act aria only succeeds in sentimentalizing.
When the husband of the singer Teresa De Giuli Borsi wrote to Verdi in 1852 requesting an additional aria for Gilda, the composer responded by insisting that his “miserable talent” had done its best. Where, in any case, could the piece be added? “One place there would be, but Heaven help us! We would be whipped. We would have to show Gilda with the Duke in his bedchamber!! You understand? In any case it would be a duet. Magnificent duet!! But the priests, the monks, and the hypocrites would cry scandal.”54 But Verdi did not believe other elements of his opera required modification. He could not imagine why censors would care if the body of Gilda was deposited in a sack. It might not work on stage, of course, but he alone should be the judge of that, not the censors.
Divas and Scholars Page 14