Divas and Scholars

by Philip Gossett

  EXAMPLE 12.4. GIOACHINO ROSSINI, OTELLO, SCENA E DUETTINO DESDEMONA–EMILIA (N. 4), MM. 7–30.

  When we contemplate this music today, we are faced with a paradox: there is no doubt that Rossini canceled two measures of his horn solo, and yet the musical fabric suffers considerably from their absence. Given the problems of performing the music on a natural horn in an Italian theater during the 1810s and 1820s, Rossini’s decision seems justified, but modern instruments render the passage more accessible. To which deity should we pay homage? Or, to put the matter in more practical terms, would it be better to perform this solo on a natural horn and cut the two measures or use a modern instrument and include them?

  The Otello solo is an extreme example, to be sure, but looking carefully at the use composers made of horns and trumpets during the first half of the nineteenth century, one becomes ever more aware that they were working around quite precise technical limitations. For example, the original happy ending of Tancredi comes to a conclusion in D major, with the entire orchestra playing the tonic for five measures.40 For four measures we hear the full D major triad, but in the very last measure Rossini asked all his instruments to play just the tonic note, except for the trumpets. Tuned “in A,” these early nineteenth-century trumpets had no way to play that pitch in an appropriate register. Yet Rossini did not want trumpets to disappear from the orchestral sonority. Instead, he allowed them to play a written c, producing a sounding octave on a, the fifth degree of the D major scale. Given the overall orchestral sonority, this fifth degree is heard not as a distinct pitch but as a reinforcement of one of the overtones of the tonic.41 Faced with an instrumentalist using a modern, valved trumpet, however, which can easily produce the tonic degree, a modern conductor may feel it appropriate to intervene.42

  A composer writing for natural horns can overcome many technical limitations by asking his players to insert different crooks in their instruments, thereby modifying the length of the tubing and hence the fundamental notes of the instrument; but this action takes time. As a result, some nineteenth-century horn parts can seem strange, as in the second-act finale of Rossini’s L’Italiana in Algieri. The scene in which Mustafà is inducted into the order of “Pappataci” (complaisant husbands and lovers) is in B major, and Rossini’s two horns “in B” play until the very end of the scene. As the ship on which the Italians plan to escape appears, the music jumps to D major. Rossini wants to use horns at the beginning of this barcarolle in 6/8, but there is no time for them to change crooks. Instead, in the opening D major orchestral ritornello, the horns “in B” are instructed to play an octave on e, sounding d, the new tonic. Only after six measures is there sufficient time (eleven measures) for the horns to change their crooks, and at that point they are instructed to prepare “in D.” When the opening passage returns with the chorus, the orchestral music is unchanged except for the horns, which are now “in D” (where they will remain for the rest of the finale) and therefore can play the notated octave on c in order to produce the same d. The differences between natural horns “in B” playing an octave on e and horns “in D” playing on c is trivial, even in nineteenth-century terms, but the example suggests how long it took a player to insert a new crook into his instrument. Rossini’s frequent use of phrases such as “in Do subito” (suddenly in C) to announce a crook change suggests that he sometimes pressed his musicians hard.43

  Verdi, too, frequently wrote horn parts in tutti sections that seem strange to modern players. Near the end of the trio for Elvira, Ernani, and Silva in the second act of Ernani, there is a cadential passage within a fortissimo concluding section, in which every instrument (apart from the Cassa) plays on each chord, except for the second horn.44 In this B major section, the first pair of natural horns is “in F,” the second pair “in B ,” a combination that works well most of the time, permitting all four horns to participate in most sonorities.

  EXAMPLE 12.5. GIUSEPPE VERDI, ERNANI, RECITATIVO E TERZETTO (N. 7), HORN PARTS, MM. 336–340.

  But in the cadential phrase at mm. 336 –340, Verdi introduced an unusual diminished seventh chord in the second half of 336, and no pitch in the chord is available in the lower register for a natural horn “in F.” Hence, Verdi wrote a peculiar part for the first pair of horns (example 12.5). For a natural horn “in F,” the written d above middle c (sounding the g below middle c) is not an easily available pitch. To produce it, the player must place his hand in the bell, significantly changing the timbre of the note. While an able soloist might in part compensate for this problem, Berlioz opines that this stopped note is “very difficult and muffled.”45 In fact, there is not a single example of this pitch for a horn in the entire score of Ernani. With a modern valve horn, the problem disappears.

  Does that end the discussion? Should we always adopt modern instruments and add these missing pitches? Berlioz is once again a helpful witness. Since he lived in a period when the valve horn was gradually replacing the natural horn in orchestras, especially north of the Alps, his words in the Orchestration treatise are noteworthy. In the definitive version of his treatise he writes, “The tone of the piston horn is a little different from that of the usual horn; it would not do to substitute it on any occasion.” In an earlier version of the passage, he was more strident: “I am firmly convinced that one should never consider it an improvement on the horn, from which it differs in quality of tone.” 46 Indeed, if we move from the problem of technical limitations to those of sonority and balance, a largely different set of issues presents itself.

  MODERN INSTRUMENTS: THE PROBLEM

  OF SONORITY AND BALANCE

  As we saw in the discussion of the Metropolitan Opera’s production of Semiramide (chapter 6), balancing orchestral sonorities is never simple, although the problem becomes more tractable when we take into account the significant differences in the actual instruments used by most modern musicians and those for which nineteenth-century Italian composers prepared their scores. Rossini, for example, will often join together two horns and two bassoons to form a chordal sonority in four voices: their combined sound, easier to balance with older instruments, is much more difficult with modern ones.47 The following reflections touch on only a few of the instrumental problems that modern performers must often face, but they offer a framework within which similar questions can be addressed.

  Three Trombones and a Tuba

  Three modern, valved trombones and a tuba playing fortissimo can overpower the rest of Verdi’s orchestra; the three narrow-bore trombones and cimbasso for which the composer wrote his scores do not have that effect. Even when we recognize that the problem is in part historical, it is difficult to address it successfully, given the realities of available rehearsal time in today’s theaters and the stylistic diversity demanded from performers night after night. The Semiramide situation was hardly an isolated instance. In the winter of 2004, Eve Queler conducted a concert performance of the critical edition of Verdi’s Il corsaro with her Opera Orchestra of New York. Working closely with her in the orchestral preparation was Italo Marchini, a sensitive conductor in his own right. After the performance, Marchini vented some of his frustrations to me in a note about dynamic levels. If you mark the three trombone parts fortissimo in the score, he said, the instinct of the players will be to follow the direction literally, and the resulting sound will blast you, the rest of the orchestra, and the singers out of the theater. The problem is real, even though I’m not sure I understand what a “literal” fortissimo would be: after all, dynamic markings are not accompanied by a scientific equivalent in decibels. Such markings are always relative: a brass fortissimo in a Bruckner symphony and a Verdi opera are not the same.

  Those who prepare scores of this repertory constantly face the problem. Should a printed edition of a nineteenth-century opera attempt to foresee and overcome it by recommending different dynamic levels in various orchestral parts, thereby “tuning” orchestral sonorities? There are two issues to be addressed: (1) Did Italian composers themselves m
ake those differentiations? (2) Even if they did not, should modern editors introduce them? Despite Denis Vaughan’s famous polemic on the subject of dynamic levels, discussed in chapter 4,48 there are no grounds for asserting that Verdi or other nineteenth-century Italian composers made such differentations in all but the most unusual circumstances. Before we published the critical edition of Rigoletto in 1983, I charted every place in the autograph manuscript of the opera where Verdi wrote a specific dynamic level, sorting them according to the instrumental parts in which these levels appeared. When there were internal discrepancies (such as f in some parts and ff in others, within a single chord played by the entire orchestra), I compared similar passages throughout the opera. I severely tried the patience of a graduate seminar, asking participants first to go over my work and then to make parallel examinations of other scores. I even delivered a boring paper on the subject to an international Verdi conference. All evidence pointed in the same direction. For Verdi an orchestral chord could be very soft (ppp or pp), soft (p), of moderate volume (mp or mf ), loud (f ), or very loud (ff or fff ). He usually wrote these indications on every other staff and made no effort to tell the trombones to play softer and the flutes to play louder. He did not, in short, “tune” an orchestral sonority by providing different dynamic levels for different instruments, relying instead on good musicianship and stylistic acuity, as well as on the prowess of the orchestral director, whether he did his job with a violin in hand, as was typical during the first half of the century, or with a baton, the emerging practice from the 1850s onward.49

  To assign editors the task of providing differentiated dynamic levels is unrealistic. Not only do such levels need to be different from performing space to performing space, from orchestra to orchestra, and from country to country,50 but an invented practice of this kind would assign to notation more precision than any nineteenth-century composer or performer would have deemed reasonable. That dynamic levels in Italian opera scores refer to an overall effect, and are not instructions to particular instrumentalists, puts a special responsibility on modern performers to interpret the notation appropriately. Part of that responsibility includes a sensitivity to the way instruments have changed over the past two centuries.

  In some cases the changes are largely of degree. Although early nineteenth-century practice in France still specified three differently tuned trombones (alto, tenor, and bass), in accordance with the practice of eighteenth-century composers such as Gluck, by the 1840s all three trombone parts were almost always played by tenor trombones (“incontestably the best of all the trombones”), and composers could not expect to find three different trombones in orchestras, either in France or elsewhere.51 Modern trombones are similar to those used by Verdi and his immediate predecessors, but the use of valves changes the way in which pitches are produced, while the widening of the bore, characteristic of instruments by the beginning of the twentieth century, considerably increases the quantity of sound that emerges, a reasonable requirement for the later Wagner operas, for Bruckner, or for Mahler. Finding the proper interpretation for notated dynamic levels in earlier scores requires sensitivity to these physical modifications of the instruments.

  When a fourth, lower brass instrument is called for by Italian composers, the situation becomes even more complicated. Terminological differences in composers’ manuscripts, while common, may not always be meaningful: sometimes the terms simply suggest “whatever low brass the theater uses.” In the few cases where Rossini requires a lower brass instrument, he refers to it as a “serpentone” (a wooden instrument) in Italy. Yet when the identical music is printed in a French score, Rossini assigns the same instrumental line to an ophicleide (a brass instrument).52 Bellini and Donizetti do not regularly require a fourth brass instrument in Italy; when they do, they tend to use the terms “serpentone” (e.g., in Bellini’s Il pirata) or “cimbasso” (in Bellini’s Norma). After Donizetti moved to Paris in 1838, he wrote for the ophicleide (in La Favorite). Verdi, on the other hand, regularly requires a fourth instrument, which in Italy he usually refers to as a “cimbasso,” at least through the 1870s. Only in a French score, such as Jérusalem, do we find him explicitly requesting an “ophicléide.”53

  Relating these terminological distinctions to actual instruments and performance practice remains difficult; it is highly unlikely that the terms were used with any precision. Nor must we assume that a solution appropriate for Rossini would also be appropriate for Verdi, or, for that matter, that a solution for early Verdi can be applied to Aida.54 Indeed, as late as 21 December 1871, in preparing for the first performance of Aida at the Teatro alla Scala (20 January 1872), Verdi was trying to find an appropriate instrument to serve as the bass of the brass section:

  I still insist on the fourth trombone. That bombardone is impossible. Tell Faccio and, if you wish, consult the first trombonist as well to see what should be done. I would like a trombone basso from the same family as the others, but if it is too tiring or too difficult to play, get one of the usual oficleidi that go to low B. In other words, do whatever you want, but not that devil of a bombardone that does not blend with the others.55

  All of this for an instrument Verdi normally referred to as a cimbasso.

  Even for those who wish to adopt a low brass instrument consonant with documented historical practice, then, there are no simple answers. Should we use a wooden serpentone for Rossini’s Neapolitan operas, but replace it with a brass ophicleide when performing the operas in their French versions? Do we accept Meucci’s contention that cimbasso for Verdi did not mean a wooden “corno basso” (a successor to the serpentone), the instrument to which the word was first applied earlier in the century, but rather a brass ophicleide of the kind invented in Paris toward the beginning of the 1820s, which became the European standard by the 1840s?56 Do we simply exclude later “improvements,” making the ophicleide more sonorous and even less able to balance with the trombones? And if we accept Verdi’s desire to have a bass trombone for his later operas, an instrument that blends better with the three tenor trombones, do we push this practice back to operas from the 1840s through the 1860s?

  Simon Wills, a professional trombonist and composer, describes the “noisy folly” of modern brass in these terms: “The sound of brass has become denser and it projects more effectively. In many orchestras a tutti has become an assault by the brass: it is exciting for the first few sallies but ultimately exhausting and inimical to the expressiveness of the orchestra as a whole.”57 His practical solution, inherently sensible, is to follow those orchestras, such as the Chamber Orchestra of Europe, who “are adopting smaller instruments without entertaining any desire to be authentic [...] because the greater transparency of their sound allows the orchestra to play co-operatively, instead of doing battle with a monster.” Even the venerable Chicago Symphony Orchestra has banished the tuba from performances of Berlioz’s Symphonie fantastique, substituting a modern brass instrument of the kind that is being called a “cimbasso.”58 Whatever decision is taken in each particular case, Verdi expressed it best: the instrument must “blend with the others.” Given the changes that took place both chronologically and geographically in the low brass during the nineteenth century, there is not likely to be much agreement about an ideal modern instrument for this repertory, although Meucci recommends the use of either a half-size tuba or a bass trombone.59 But no matter what instrument is used, one shouldn’t have to tell a brass player in a performance of Norma that a marked fortissimo in his part doesn’t mean the same thing for him as it does for an oboe.

  How Many Strings?

  What is an “ideal” number of strings for the performance of an opera written by an Italian composer during the first six decades of the nineteenth century, and where should they be placed?60 While these are questions I am frequently asked by theaters and conductors, I cannot pretend to offer precise guidance. The matter is of great importance, to be sure, in a repertory for which there is normally a full complement of winds (2 fl
utes, doubling piccolos, 2 oboes, 2 clarinets, 2 bassoons), brass (4 horns, 2 trumpets, 3 trombones, cimbasso/ ophicleide), and percussion, not to mention the occasional English horn, harps, guitars, and so forth. If you search for historical evidence about the size of string sections, however, you find that opera houses in different cities had orchestras of different sizes and seating arrangements, and even within a single city the number of players and their arrangement varied over time.61 Nor would it make sense to generalize for a Rossini comic opera intended for a small Neapolitan theater (such as the 1816 La gazzetta for the Teatro dei Fiorentini), a Donizetti French work aimed at the Paris Opéra (such as the 1840 Les Martyrs), and a Verdi opera written for Rome’s Teatro Apollo (the 1853 Il trovatore). Even were we able to establish firm historical principles, their present applicability would remain uncertain. How, for example, should we use the knowledge that there were usually more double basses than violoncellos in an Italian opera orchestra during the 1810s and 1820s, since today’s instruments are no longer the same: Rossini’s three-string double bass was very different from the modern four-string double bass, the latter being an instrument of greater range and sonority.

  To address such questions, we must both reconstruct accurately what composers wrote and trace the performance history of their works. Thus, while Rossini always notated his viola parts on a single staff and never specified “divise” (divided), an indication he used for first violins (as “divisi,” since violins are gendered masculine in Italian and violas feminine), he regularly thought of violas as being divided into two sections. This is confirmed by contemporary performing materials, which often have separate parts for first violas and second violas. In the opening section of the first-act finale of Tancredi, for example, the violins, cellos, and double basses all play the same staccato figuration in their respective registers. Only violas sustain the harmony and later, along with the upper winds, double the melodic material of the four women’s voices.62 But it would have been unthinkable in the passage shown in example 12.6 for the violas to have been executed as double-stops, with each instrument playing both melodic parts. While the viola part could be played with one instrument on each line, it is unlikely that it would have balanced effectively the remainder of the ensemble. In short, Rossini’s score would seem to require a minimum of four violas in the orchestra. Yet only a few Italian theaters during the 1810s could accommodate this need.63