9—
Gran dicerie e canzonette :
Recitative and Aria

As the geographical and creative separation between composers and librettists increased, the language of their communication with one another became, by necessity, more explicit and efficient. Librettists' texts differentiated more clearly between aria and recitative verse and composers' responses became more predictable. Aria and recitative attained their own distinct, and mutually exclusive, identities.

The Florentine Background

At the beginning of its development, operatic poetry was relatively undifferentiated. The first librettists, borrowing from the pastoral, chose versi sciolti for their dramas, a meter characterized by freely alternating settenario and endecasillabo lines, unrhymed or rhymed irregularly. Lacking formal restraints or conditions, the pastoral meter suited the open-ended stile recitativo devised by the Florentine composers for the clear and immediate communication of dramatic poetry in music. The stile recitativo was relatively unencumbered by textural, harmonic, or melodic responsibilities. It could thus aspire to the condition of speech, ebbing and flowing in response to the changing emotional temperature of a dramatic text. Governed by the form and sense of poetry rather than by principles of musical structure, it was an effective means of communication that did not strain verisimilitude.

The musico-dramatic continuity promoted by versi sciolti was occasionally interrupted by an unusual textual passage that attracted special attention to itself by virtue of its structure: a succession of tightly rhymed lines, perhaps, or a passage in a new meter. Inspired by some unusual event or emotion in the drama, such passages frequently gained further emphasis (and greater differentiation from their context) through immediate repetition of their distinctive verse structure, thereby producing a strophic form. Self-contained forms like

these suggested a contrast in musical setting, a more closed, structured treatment in which the freedom of speech would yield to the control of musical form. But any kind of musical structure, especially strophic repetition, interrupted the dramatic continuity and undermined the verisimilitude that the stile recitativo had been developed to sustain. The use of the same music for two different passages of text was inimical to a style in which music sought to respond directly, newly, and uniquely to specific words. It broke the illusion of spontaneity: people did not speak in strophes, or in rhyme, or in meter, let alone in melody—at least not under ordinary circumstances.^[1]

But not all circumstances in opera were ordinary. Some explicitly called for music. In addition to dialogue, operas (like plays) contained songs, choruses, and dances. For such occasions, musical organization was not merely possible; it was necessary. There were other instances, too, not so obvious, in which organized music was more effective than ordinary speech for projecting character and situation: moments of great emotional intensity—elation or despair—often needed a special outlet, the heightened emphasis provided by song.^[2]

Formal music, then, might temporarily slow the dramatic action by stepping outside its flow, but by that very separateness it could also serve the action by summarizing, punctuating, or commenting on it; or it could carry the action on a new level, that of music rather than speech, as songs and choruses. Early librettists reserved textual formality for just such occasions: where it could contribute directly to the drama without seriously disrupting or undermining its verisimilitude.^[3]

The composer's response to the librettist's formal hints depended, of course, on his interpretation of their function. If actual songs or choruses were indicated, he usually chose aria style, a measured, lyrical setting that unfolded according to musical rules. In more ambiguous cases he might try to minimize, undermine, or counteract their static implications by using a declamatory vocal style or writing a strophic recitative only loosely structured by means of a repeated bass; or else he could use a combination of styles, reserving lyricism perhaps just for the conclusion of such a passage. Finally, he might choose to ignore the librettist's formal hints altogether, or, conversely, to impose a formal structure where the libretto had none, in response to his own interpretation of the drama.

The composer's confirmations of the librettist's formal structures, even if they resulted in more highly organized music, were as consonant with the precepts of the stile rappresentativo as was open-ended recitative. Both kinds of response were governed by the words; the aim of both was the same: to interpret, reflect, or project the drama through its poetry. The stile rappresentativo of the Florentines, then, actually covered a wide spectrum of music, ranging from free, speech-imitating recitative to more formal, self-contained, and musical moments. This wide spectrum eventually separated into recitative and aria.

Closed forms, for solo as well as chorus, internal as well as external to the drama, were built into opera from the very start. Orpheus, the quintessential operatic hero, sang his way into Hades, proving that music could promote action. To be sure, singing was his natural mode of expression. Yet even for him librettists and composers distinguished between speech and song. He achieved his goal specifically through a formally organized song: Rinuccini's "Funeste piaggie" comprises three irregular stanzas linked by a refrain set lyrically by Peri; Striggio's "Possente spirto" is a six-strophe text in terza rima , which Monteverdi cast as an increasingly elaborate series of strophic variations. Despite—or actually because of—their form, Orfeo's songs satisfied the requirements of verisimilitude.

Even in operas not featuring musicians, however, music per se could serve the drama. Songs were a frequent ingredient of early opera, as they had been in the pastoral. Nymphs and shepherds could sing in musical plays because they sang in nature. Likewise, divinities could express themselves in any form they chose; their songs actually helped to distinguish them from ordinary mortals.

Verisimilitude was not an issue for the inhabitants of Olympus or Arcadia.^[4] Yet, even more realistic opera plots concerned with human characters, such as Arianna in Mantua and those involving biblical and chivalric heroes introduced in Rome during the 1620s and 1630s, contained a variety of closed forms. These were mostly choruses and dances, though there were some solo songs as well— especially for servants and other minor characters, but occasionally also for protagonists. If they did not always serve specific dramatic functions as songs, they at least did not interrupt the action.^[5]

Venetian Conservatism

By 1640, Roman librettos distinguished quite clearly between aria and recitative verse; they contained numerous closed forms (suggesting musical interruptions), especially strophic ones, in a wide variety of meters. By comparison, early Venetian librettos are remarkably undifferentiated in poetic style and exceptionally poor in arias. The first two Venetian librettos, Ferrari's Andromeda and La maga fulminata , contain very few hints of anything but recitative. Interruptions of the open-ended recitative style in Andromeda are limited to five strophic texts, all but one for chorus (i.e., structurally external to the dramatic narrative), and a few rhymed quatrains or otherwise patterned metric arrangements for Venere and Astrea that suggest some kind of closed setting. In La maga fulminata the number of such indications is somewhat greater, particularly in connection with the comic nurse, Scarabea. Her speeches are quite highly organized in meter and rhyme, lending them a sing-song quality that tends to emphasize their humor. In addition, strophic texts are provided for several characters, including the god Mercurio and the enchanted mortal Pallante. While strophic arrangements of settenario and endecasillabo lines are the most common formal signals, both librettos also contain individually closed sections marked by unusual metric organizations involving regular successions or alternations of short and long lines, many of them in versi misurati .^[6]

In light of the progress toward integration of music and drama in the works of his predecessors to the south, Ferrari's reticence toward closed forms is somewhat surprising. Lacking Manelli's settings, we cannot be sure, but the musical highlights of both operas seem to have been the choruses—or "madrigali," as they were called in the descriptions published soon after the works were performed. Apparently, the question of verisimilitude, rendered less urgent through experience in Florence and Rome, was still an issue in Venice. Perhaps this was because of the wider social range of the opera audience there, an audience that may have been less willing and able than its courtly counterparts in central Italy to suspend disbelief for entertainment's sake. ^[7]

Elaborate scenery and, of course, mythological plots should have helped to diffuse the need for strict verisimilitude or at least have expanded its limits. But, as I suggested earlier, perhaps even more than for the aristocratic Medici and Gonzaga audiences, satisfied by implicit connections between opera and classical drama, for Venetian hoi polloi identification with the actions on stage was a prerequisite for the appreciation—and success—of a work. Plays with songs, commedie , had been the usual fare of Venetian audiences during the early years of the century, and they provided the standard for verisimilitude in its early operas.

The absence of firm poetic distinctions between recitative and aria (and the limited number of arias) in early Venetian librettos may also be the result of another, quite opposite, influence, that of the academic librettists, who, as we have seen, took control of opera in Venice during its crucial formative period— the first decade. It was they who felt the pressures of classical precedent most strongly; and, aside from Ferrari's first two librettos, it is their texts that adhere most closely to notions of propriety that frequent arias would have violated. To be sure, whereas the problem of verisimilitude was attenuated considerably in mythological or magical contexts, it was especially troublesome in librettos that dealt with more realistic, historical subjects—subjects that the academic librettists introduced quite deliberately in their attempts to emulate the classical dramatists.^[8]

Monteverdi and His Collaborators

Among the academics, perhaps the most ascetic of all early Venetian librettists when it came to arias was Giacomo Badoaro. His initial entry into the operatic field, Il ritorno d'Ulisse , was written for Monteverdi and performed in 1640.^[9] It is cast primarily in versi sciolti and provides few clues as to where a composer might have halted the recitative flow to write an aria. We are at an advantage here, however, with respect to Ferrari's librettos, because we have Monteverdi's score, which, while fundamentally in recitative, is surprisingly rich in lyrical moments—though these do not always qualify as full arias. The inspiration for these moments can usually be traced to subtle formal hints in the libretto, passages of text that are more highly organized than those around them. Quatrains or sestets may be structured by means of regular rhyme or repeated meter: a group of four settenari , perhaps, or a succession of six regularly rhymed alternating settenari and endecasillabi . While such metric organization suggests musical closure, it does not insist upon it—in fact, the rhyme is never that tight; sometimes a group of endecasillabi remains ambiguous because the rhymes are so far apart. Badoaro almost never used versi misurati , which would automatically distinguish themselves from recitative poetry. However subtle the librettist's formal suggestions might have been, Monteverdi nearly always capitalized on them.

Only on three occasions are Badoaro's formal hints unambiguous. His libretto contains three strophic texts, which Monteverdi set as strophic variation arias. Badoaro's conservative use of strophic form, in only these three instances, confirms his belief in the need for verisimilitude. Each of the three texts—and, by extension, Monteverdi's aria settings—is dramatically justified. In each case, the formal structure enhances the drama rather than undermining it. Melanto, a simple servant girl flirting with her lover, sings the first of them; the second is sung by Minerva, a goddess in disguise—and therefore doubly exempt from normal rules of behavior; and Iro, the parasite, a ridiculous comic character whose appetite is as peculiar as his manner of speech, sings the third.

Although Monteverdi's score is filled with lyrical, or arioso, expansions, the composer nevertheless seems to have wanted more structure than the libretto provided, and often edited Badoaro's text accordingly. His intervention is notable in three of the most intensely emotional moments of the drama, all of them marked by the presence of refrains. In one of them, in act 1, scene 9, where Ulisse finally recognizes that he has returned to Ithaca, Monteverdi converted

an exceedingly amorphous text into what almost amounts to a strophic aria, utilizing two irregularly spaced refrain lines (italicized) to mark the opening and closing of each strophe. Badoaro's text reads as follows (example I):

Monteverdi expanded the two refrain lines enormously through textual and musical repetition so that they take up most of the aria; then, despite their unequal length, he treated the penultimate line of each "strophe" similarly, using the same extended melisma for "lieto" and "guerra." Inspired by Badoaro's refrain as well as by the expressive content of Ulisse's words, which actually invite song ("dolce canto . . . disserra"), Monteverdi's lyrical setting effectively changes not only the weight but the form of the text.

In another highly dramatic moment toward the end of the opera (3.8), where Ericlea wrestles with herself about revealing Ulisse's identity to Penelope, Monteverdi turned an irregular, 24-line text into a refrain form comprising four unequal sections of nine, four, five, and six lines, each closing with a sententia of self-justification (italicized) (example 2):

By adding a ritornello after sections 1, 3, and 4, and setting each final sententious line to the same highly expanded music, the composer intensified the formal implications—and the affect—of Ericlea's monologue. The music con-cretely marks her progress from her initial vow of silence (sections 1 and 2), through ambivalence (section 3), to her decision to speak. As in Ulisse's aria, but by different means, Monteverdi superimposed a kind of strophic structure on the text and, far from sacrificing affective intensity, he increased it. The absence of a confirming ritornello after the second "refrain" (line 13) and consequent telescoping of sections 2 and 3 creates a sense of urgency that matches Ericlea's ambivalence.

Monteverdi's most impressive intervention, one that involved extensive text editing as well as reweighting, occurs in the very first scene of the opera, where he completely restructured Penelope's opening lament. He took a diffuse text of no fewer than 125 lines, in four uneven sections, that wandered rather aimlessly from topic to topic, and transformed it into a tripartite recitative of slowly building intensity, utilizing two irregular refrains provided by the librettist as structural pillars of the powerfully dramatic form. Intensified by the composer's restructuring, Penelope's torment resonates throughout the entire opera. ^[13]

Although written expressly for Monteverdi, Badoaro's libretto clearly did not completely satisfy the composer's lyrical impulse, his yearning for texts to set lyrically. In fact, as we have already noted, Monteverdi's alterations—

which, in addition to those already mentioned, involved a variety of repetitions, single-line expansions, the creation of refrains, cuts, and so on—rendered the text virtually unrecognizable to its original author, or so Badoaro admiringly reported in a letter attached to one of the contemporary manuscript copies of the libretto.

To the Most Illustrious and most Reverend Signor Claudio Monte Verde Great Master of Music: Not in order to compete with those talented men who, in recent years have publicized their compositions in the Venetian theaters, but to stimulate the imagination of Your Lordship to make known to this city that in warming the affections there is a great difference between a real sun and a painted one, I initially dedicated myself to compose the Return of Ulysses . . . . Now, having seen the opera performed ten times, always before the same [large] audience of the city, I can positively and heartily affirm that my Ulysses is more obligated to Your Lordship than the real Ulysses was to the always charming Minerva . . . We admire with the greatest astonishment those rich ideas of yours, not without some perturbation, because I can no longer recognize this work as mine. (Appendix 1.7a, c)

Monteverdi's powerful brand of editing is set into relief by comparison with an opera contemporary with Il ritorno d'Ulisse , Sacrati's setting of Strozzi's La finta pazza . To be sure, of the latter two collaborators, the librettist rather than the composer was the more experienced and more self-confident; certainly La finta pazza is a more effective and skillful text than Il ritorno d'Ulisse . It is also considerably richer in explicit invitations to lyricism, containing eleven formal texts, all but two of them strophic, many utilizing versi misurati , and nearly all of them dramatically explicable as actual songs.^[14]

Whatever the reason, Sacrati seems to have accepted Strozzi's text quite willingly, satisfied to follow the libretto's lyrical implications without creating his own closed forms. He did sidestep Strozzi's structural directive in one case, however, setting each of the two quatrains (strophes) of Deidamia's aria "Verga tiranna ignobile" to different music, thereby creating an AB aria instead of a strophic one; and in two other instances (Acchille's aria "Felicissimo giorno" in 1.3 and the Eunuch's "Serva, serva chi vuole" in 2.10) he restructured Strozzi's text slightly by bringing back the opening line (or lines) later in the form, creating a miniature ABA in the first instance and a rondo in the second (the composer's repetitions are in italics):

As for Strozzi's recitative verse, Sacrati emphasized a fair number of lines and couplets by arioso treatment, many of them significant with respect to the meaning of the work. In act 2, scene 2, for example, a particularly pregnant speech within an exchange between Ulisse and Acchille, although formally rather neutral, is nevertheless set lyrically. In response to Acchille's question as to whether he thinks that a young lover can change his affections and his beloved when he wishes, Ulisse responds:

The exchange involves Ulisse's fidelity, and is probably an allusion to his role in Il ritorno d'Ulisse of the previous season.

In another instance, a six-line passage of Deidamia's in 3.2 that refers pointedly to matters outside the drama itself, namely to the theater management, is also set lyrically:

Aria style, finally, is used to set a particular passage directed to the audience by the Eunuch in 3.3 (example 3):

Other, briefer passages elicit lyrical treatment because of their emotional content. These include Acchille's plea to Deidamia for forgiveness at the end of a speech in 3.4 ("Perdona, tu, perdona"), and Deidamia's acceptance of his hand later in the same scene ("Caro pegno di fede").

Such passages are neither as elaborate nor as frequent as Monteverdi's arioso expansions in Il ritorno d'Ulisse ; furthermore, Sacrati ignores a number of formal hints—such as the four sestets at the end of act 1, and various rhymed couplets and quatrains, and lengthy sequences of settenari — that Monteverdi would have pounced on as excuses for musical elaboration or structure. La finta pazza contains extended passages of straight recitative setting of versi sciolti uninterrupted by lyricism that nevertheless reveal in Sacrati a powerful musical imagination at work. It must be said that poet and composer were more compatible in La finta pazza than in Il ritorno d'Ulisse .

Monteverdi's next librettist, Busenello, less conservative than Badoaro, as well as more experienced in the art of libretto-writing, seems to have produced a more satisfactory text. ^[17] Although L'incoronazione di Poppea also needed many alterations, to judge from the printed libretto, it seems to have provided Monteverdi with what he lacked in Il ritorno d'Ulisse —namely , multiple occasions for lyrical expansion. In addition to thirteen strophic texts, most of them for secondary characters, and all of them arias in Monteverdi's score, the libretto of Poppea contains a large number of prominent couplets and quatrains. Monteverdi almost always set these lyrically, sometimes splitting them line by line, sometimes treating them as a whole ("Poppea sta di buon core," end of 1. 10), and sometimes turning them into miniature ABA arias by repeating the first line at the end as a refrain ("E pur io torno" [1.1]; "O felice Drusilla" [3. 1]). In

L'incoronazione di Poppea , as in Il ritorno d'Ulisse , Monteverdi left almost no suggestion for musical structuring or lyrical expansion unexploited; but in the libretto of Poppea there were more of them.^[18]

Poppea , even more than Il ritorno , depends on lyricism; it owes its affective impact to distinctions between speech and song. Whether fleeting emotional outburst or fully considered pleading, song lies at the heart of the work, touching all the characters and all the situations. Nino Pirrotta regards the unusual abundance of song in Poppea as evidence of relaxed standards of verisimilitude, which he ascribes to the fact that the characters are carried away by love.^[19] In fact, very little of the lyrical expansion in Poppea is actually formal, and thus "unnatural." Predictably, most of the strophic arias are songs sung by comic characters, the repetition and patterning enhancing the humorous effect; those that are not (and even some of those that are) are treated as quite free strophic variations, which minimizes their repetitiveness. And in most cases, whether comic or serious, the structure contributes to the development of the drama.^[20] Monteverdi's song, a correlative of heightened passion, emerges from and fades back into speech quite naturally, feelingly. Poppea is especially lyrical, airy (arioso ), but not especially formal.

Busenello and Cavalli

Busenello's distinctions between aria and recitative are not totally unambiguous; but Monteverdi could read them, or at least he did read them, the way he wanted to. Beyond the arias, Busenello's text provided Monteverdi with the kind of structure and imagery that stimulated his musical imagination. Given what we know of the composer, and in the context of the deficiencies of Il ritorno d'Ulisse , it is likely that the text of Poppea was constructed to Montever-

di's specifications, or at least with his tendency toward mimetic musical expansion in mind. This is particularly clear from a comparison with Busenello's two previous librettos, written for Cavalli, Gli amori d'Apollo e di Dafne (1640) and, especially, Didone (1641). The feature that chiefly distinguishes Didone from Poppea is its many strophic texts and few independent quatrains. In practically every scene of the earlier libretto, the free succession of versi sciolti yields to organization in the form of strophes, usually three or four, but sometimes as many as eight (eleven in one case!), comprising from three to eight lines each, which are set off from the surrounding text by their more regular meter and/or rhyme scheme.

In Didone , then, Busenello established strophicism as the clearest and most frequent closed form for opera. But its musical implications remained ambiguous for a number of years. The distinctions between the strophic forms and the surrounding poetry are not all equally sharp. While a number of strophes utilize meters that contrast strongly with the predominant versi sciolti , such as quinari or ottonari —versi misurati —others are cast in the characteristic meters of recitative—settenario and endecasillabo —though in more structured patterns: typically a quatrain of settenari followed by a hendecasyllabic couplet. In some cases the strophes fail to create a strong metrical effect because their rhymes lack prominence. They may be too far apart, as in a succession of rhyming hendecasyllabic verses, or they may be counteracted by an opposing or simply non-confirming metrical structure; or else the rhyme may be restricted to the closing couplet, with the remaining text unrhymed, much like the standard recitative verse.

The formal units created by these strophes, some of them more insistent than others, always suggested some change in musical language, a closed rather than open-ended setting that would reflect, through musical repetition, the specially structured text. But since such musical repetition, whether in aria, recitative, or mixed style, generally threatened the illusion of dramatic continuity and verisimilitude, it could be used only rarely, and in specific circumstances.

Busenello's sensitivity to the implications of strophic form is revealed by the occasions on which he employed it. Nearly all of the twenty-six strophic texts in Didone are dramatically justified in one of the standard ways. A number of them are for gods, traditionally exempted from the laws governing human behavior; several others are for essentially comic or ironic characters (Sinon Greco, the ladies-in-waiting), who speak in clichés, and whose unwanted advice falls quite naturally into rhyme; and three of them are for the boy Ascanio, who may be thought of as not yet having learned the rules of adult behavior or whose youth is projected in his rhythmic, singsong speech. Busenello's serious

characters fall into patterned rhythm and rhyme only in spite of themselves, when self-control has failed owing to some kind of extraordinary pressure: fear of death, abandonment, or madness. In general, then, the librettist is quite conservative in his demands on the composer. His strophic forms suggest some special musical structuring, but not indiscriminately, and not necessarily in the form of an aria, as we can see from Cavalli's score.

If Busenello is conservative, Cavalli is downright reactionary in his respect for verisimilitude. To compensate, however, he penetrates more deeply into the drama, finding justification for musical expansion on psychological grounds. Of the twenty-six strophic texts in Busenello's libretto, Cavalli set only twelve as strophic arias; he set eight as strophic recitative, four with mixed treatment—part recitative, part aria—and ignored or changed the strophic form altogether in two. The libretto has only one non-strophic closed form, a series of quinari sdruccioli for Ecuba's invocation, which Cavalli set as an aria.^[21] With his thirteen arias in the whole opera, Cavalli exploited fewer than half of the opportunities provided in the libretto for formal music.

Perhaps the most eloquent illustration of Cavalli's attitude toward verisimilitude and toward Busenello's textual directives—his basic disinclination to confirm strophic structure in music—is provided by an example in which he changed compositional styles, from recitative to aria, within the same strophic form. In act 3, scene 6, the gods have just informed Enea that he must leave Carthage. He knows that he is obliged to obey them, although he is reluctant to abandon his beloved Didone. Busenello provides him with a text consisting of seven strophes of six hendecasyllabic lines each, with the rhyme scheme abbacc , in which Enea expresses his bitter conflict between his love for Didone and the realization that the gods' command is law (example 4). He articulates that conflict most explicitly in the third strophe:

Cavalli set the first four strophes, in which Enea gives voice to his torment, in expressive recitative style. Although they share general harmonic shape, the four strophes vary in length, are supported by different bass lines, move to different internal cadential goals, and exhibit remarkably different rhythmic and melodic profiles.^[22] The most abrupt change, however, occurs at the fifth stro-

phe, where recitative yields to aria. It is at this point that Enea's thoughts shift from himself and his own conflict to his beloved Didone, whom he addresses gently:

Cavalli set this strophe and the next as a kind of lullaby in lyrical aria style; clearly responding to a change in the mood of Busenello's text, he himself supplies, through music, an emotional climax to the turmoil expressed in the previous recitative strophes. Even the two aria strophes are not treated in a strictly strophic manner, however, although they share many features. Nor is aria style maintained consistently throughout them; the penultimate line of each is set as recitative, which overflows into the final lyrical line. In fact, the second strophe builds on the first in many ways, but its whole structure is more continuous and its phrases are more closely related, producing a greater sense of growth from one to the next. This second strophe is also more final than the first, containing a lengthy ritardando and cadence in the subdominant just before the recitative line. Melodic restraint and balance in the first strophe rather fittingly convey Enea's feelings, his tender sorrow, his hesitancy, his fear of disturbing Didone's repose and of her awakening, perhaps to thwart his resolution. Full of self-doubt, he is most concerned with reassuring her. He speaks louder and more forcefully in the next strophe as he becomes stronger in his resolve. Now he is more passionately concerned with himself, with expressing his own suffering. The more sustained buildup of musical tension and affect of the second strophe heightens the opposition between desire and destiny. ^[23] In the final strophe, he returns once more to the basic conflict—and to recitative style—as he bids his beloved addio .

The form of Busenello's text alone, then, does not allow us to predict Cavalli's response. Indeed, text meaning—or, rather, dramatic context—far outweighs rhyme scheme, meter, or any other formal device as the chief determinant of his style. Nowhere is this clearer than in the recitative itself, from which Cavalli occasionally extracted lines for lyrical setting purely on the basis of their meaning: emotional outbursts, resolute conclusions or summaries, statements of intention could all incite his lyrical imagination. In many cases

these passages take on a culminative structural function when they occur at the ends of scenes or action segments. Unlike Monteverdi, however, Cavalli does not seem to have been seeking excuses for lyricism; he flatly rejected some obvious opportunities, maintaining a powerful commitment to recitative even within closed forms.

Indeed, despite Didone's much larger number of strophic texts, its center of gravity—unlike that of Poppea —lies not in the arias or in arioso but in the recitative. The first act in particular, arguably the highpoint of Cavalli's accomplishments in this style, provides a lesson in recitative expression that other composers, even Monteverdi, rarely matched.^[24] Only infrequently interrupted by arias—and then less interrupted than intensified—the dramatic thread is sustained essentially by recitative, now narrative, now heightened by feeling, which even intrudes, as we have seen, into the arias themselves. The mutuality of text and music in the recitative of Didone is almost matched in the arias, which rarely move beyond straightforward text presentation into musical elaboration. Typically, the setting is line for line, with any repetition or musical expansion saved for the end. In fact, there is often more "music" in the recitative than in the arias, more of the composer's art.

The small proportion of Busenello's strophic texts set completely as arias is a reflection of Cavalli's attitude toward verisimilitude. It also confirms the fact that, in 1641, strophic texts did not necessarily require aria setting. Indeed, aside from distinctions in meter and rhyme, librettists had not yet developed specific, unambiguous poetic signs for arias—as opposed to strophic recitative or mixed style; nor were such signs yet necessary. In opera of the early 1640s, strophic recitative or a mixed recitative-aria style was still a viable response to a strophic text, as it had been since the beginning of the century—a strategy designed to limit or minimize undramatic stasis. The choice was still primarily the composer's. It was only with the increased emphasis on the singer—and the aria— that the librettist's language gradually acquired greater specificity.

Cavalli and Faustini

An important step in that direction was taken by Cavalli's next—and probably most important—librettist, Giovanni Faustini. Undoubtedly encouraged by the composer, with whom he collaborated steadily and exclusively for a decade, Faustini developed more explicit and more varied ways of indicating closed forms and suggesting a change from recitative to aria style. In addition to

strophic arrangements, Faustini's texts included numerous passages marked off by refrains as well as individual sections of highly metrical, rhymed texts forming individual poetic stanzas. Cavalli's response was generally more predictable than it had been in Didone .

Faustini's formal signals were not only clearer and more varied than Busenello's; they were also more numerous, but not because he was any less bound by verisimilitude than his colleague. On the contrary, he expressed his commitment to verisimilitude by stretching its boundaries, developing additional pretexts, new ways of justifying formal music. His librettos are constructed with a view to rendering song more natural. If his lack of academic background proved an advantage in this connection, it was because it permitted him to move beyond classical and mythological sources for his librettos, to create characters and situations that were not weighted down by responsibilities inherited from the past. He was free to create an imaginary new world in which fictional behavior—speaking in song—was more plausible.

Faustini exercised his freedom of invention by imbuing his characters with qualities, and his plots with incidents, that translated well into formal music—comic servants pontificating or spewing clichès, expansive, self-indulgent heroes and heroines easily carried away by love (and grief), and plots revolving around disguise, which encouraged, even required, participants to behave or think unrealistically—unlike themselves. Furthermore, to stimulate their inclination to musical expression, Faustini presented many of his characters in solo scenes, thereby releasing them from the necessity of realistic communication with their fellow actors. It was easier for an audience to accept the singing of soliloquies, of inner thoughts, than to accept sung communication between two characters. Finally, Faustini constructed texts that were formal and dramatic at the same time, that served the needs of music and action simultaneously.

His librettos contain far fewer strophic forms than Busenello's; but these are at once more standardized and more clearly differentiated from their recitative surroundings—and this by virtue not of meter but of rhyme. They usually consist of three six-line strophes that, although utilizing the preferred recitative meter—the standard seven- and eleven-syllable lines—are tightly rhymed, normally closing with a couplet. Most significant for their translation into music, successive strophes frequently share a concluding refrain (it occasionally opens the strophe as well) that emphasizes even further their isolation from their context. Cavalli, in response to their greater formal clarity and standardization, was much more consistent than before in setting them as lyrical arias. In Ormindo (1644), for example, Cavalli set twelve of the fourteen strophic texts as arias and only two as recitative. But the score has many more than twelve arias. Indeed, strophic texts comprise only about half of the total number of closed

forms in a typical Faustini libretto. Most of the others are articulated by means of refrains.

Refrains were an important component of Faustini's attempt to stretch verisimilitude. Comprising either single or multiple lines, they could promote continuity as well as closure: a refrain might recur within a recitative text for dramatic reasons, for emphasis, only temporarily interrupting the recitative flow; or it might enclose a static form. Refrains recur effectively and affectively in a wide range of situations throughout the libretto of Ormindo . While Cavalli invariably marked their recurrence musically, his treatment varied, depending on their form or dramatic context. If his response to strophic texts had become conventionalized, with refrains he continued to exercise his composer's freedom. The drama was still in his hands.

In Ormindo , Faustini often used a single-line refrain to isolate a tight rhyming quatrain or cinquain—abba, abaa , or abbaa —within a lengthy section of recitative verse. In Cavalli's setting, some of these brief texts become miniature tripartite arias, while others shade into the recitative background. In act 1, scene 8, for example, he makes a little ABA aria out of Erisbe's simple quatrain responding to Ormindo's protestations of love (refrains italicized) (example 5):

He gives additional prominence to the single-line refrain through extensive repetition of its text and music and by the addition of strings to the continuo accompaniment. Similar quatrains for Melide (a lady-in-waiting) and Erice (the nurse) in act 1, scene 5, however, are treated as simple recitative, with just a hint of extra musical expression given to the refrain line (example 6a, b):

Cavalli evidently regarded these formal hints as an excuse rather than a command for lyrical emphasis. His settings clearly depended on larger dramatic considerations. In the first example, lyrical expansion of Erisbe's protestation

of love to Ormindo is particularly appropriate because it needs to be overheard by her other lover, Amida; in the second and third examples, Melide and Erice are merely encouraging their mistress, Sicle, not to lose hope.

In the case of lengthier refrains—of four and five lines—Faustini's message may be louder, but Cavalli's response is just as independent. Indeed, although in these cases he invariably set the refrain lyrically—to impart the emphasis Faustini called for—Cavalli's treatment still depends primarily on his own interpretation of the drama. In act I, scene 7, for example, a five-line refrain encloses a highly structured seven-line text for Erisbe (example 7):

Cavalli expanded the refrain considerably by means of text repetition and melismatic extension, dividing it into two distinct sections with contrasting music for its final couplet, music that is then repeated and further expanded in a lengthy ritornello. Alone, Cavalli's setting of this five-line text has all the earmarks of a bipartite aria. It recurs intact as a refrain, complete with ritornello, creating a fully rounded ABA structure with recitative B section; lyrical expansion weights the refrain more heavily than the B section, whose text is heard only once. This emphasis is particularly appropriate since the refrain effectively encapsulates Erisbe's predicament, the incompatibility of her youthful yearnings with her marriage to an aged husband, a conflict that lends plausibility to her subsequent behavior.

But the composer does not always emphasize the refrain at the expense of the enclosed text. In act 1, scene 5, a four-line refrain, itself including a refrain, encloses five lines of equally structured text for Sicle (example 8):

Cavalli's setting is much more continuous; he distinguishes his refrain from the enclosed text chiefly by meter, not by affective intensity, which continues to build in the central section. The return of the refrain thus becomes the climax of the whole passage, imparting a sense of continuity rather than contrast, and creating a form that, despite—because of?—its closure, convincingly portrays Sicle's increasing desperation over her betrayal by Amida. Closure is reinforced in two different ways in these two examples: in the first by expansion of the refrain itself, in the second by the building of the middle section toward culmination in the refrain. In both cases formal closure suited or enhanced the dramatic situation. In neither was it the form of the text alone that determined the musical treatment.

Musical closure was not a necessary consequence of all refrains, however. For a text for Erisbe's aged husband Hariadeno in act 1, scene 9, in which a two-line refrain encloses nine rhyming settenari , Cavalli provided an undifferentiated recitative setting, emphasizing neither the refrain nor the highly structured central section. An aria was possible here, but the composer did not deem lyricism appropriate to the dramatic situation. Perhaps he wished to minimize sympathy for Hariadeno, to enhance his characterization as a cold old man. Drama rather than form must have been the chief motivating factor, since Cavalli set a number of texts less highly structured than this one as arias. In act 3, scene 13, he wrote an aria to a simple seven-line text, five settenari followed by a hendecasyllabic couplet, without refrain, obviously because of its dramatic position: it serves as an emotional release for Ormindo, who joyfully recognizes Hariadeno as his long-lost father.

Cavalli was more likely to accept Faustini's formal cues under certain conditions. Monologue scenes, for example, which often end with strophic texts, frequently display other hints for closure such as a refrain or a metrical passage of text. Lyricism and formality are justified in these instances, or at least mitigated, as we have already said, by the understanding that the character is voicing his thoughts to himself—and only incidentally to the audience—rather than to his fellow actors. The monologue situation seems to have encouraged Cavalli to set such texts lyrically, thereby producing a lyrical crescendo that culminated in the strophic aria.

Melide's monologue in 2.5, for example, comprises two sections, a six-line passage of versi sciolti and three six-line strophes of ottonari with refrain, calling for recitative and aria setting, respectively. Indeed, the recitative text prepares that of the aria. In it Melide reflects that she, too, wishes to love, but has decided not to, since Cupid is so cruel (example 9):

She then addresses her aria to the god of love himself:

Rather than treating the opening six lines as recitative, however, Cavalli, taking advantage of their closed rhyme scheme, ababcc , set them in duple-meter aria style, actually rounding out the form by bringing back the musical material developed in lines 1-2 for line 6 and for a concluding sinfonia. The lyrical setting of this section is clearly not required by the meaning or form of the text, but it provides an effective musical springboard for the following strophic aria, which, although in a different meter, is in the same key.

Although quite rudimentary in form, this little scene prefigures the later operatic scena , which, after a lengthy period of expansion, ultimately solidified in the cantabile-cabaletta convention of nineteenth-century Italian opera. A position is taken at the beginning of the scene and expounded or amplified at the end: an opening aria launches a topic that the closing aria discharges. Paired arias like these, usually separated by a recitative passage that developed the argument of the first aria and precipitated the second, were particularly common for Faustini's secondary characters; they became a staple for the protagonists only later in the century.^[26]

Like Monteverdi, Cavalli would change or reorganize text when it suited his purposes. Nowhere in Ormindo is the composer's independence of the libretto more powerfully demonstrated than in the dialogue aria at the climax of the opera, just before the resolution of the plot (in act 3, scene 12).^[27] The two

characters involved are the protagonists Ormindo and Erisbe, illicit lovers who, while attempting to flee the kingdom of Erisbe's husband Hariadeno (the fact that he is also Ormindo's father is as yet unknown), have been captured, imprisoned, and poisoned by the king's soldiers, and are now awaiting the poison's fatal effect. No aria is signaled here by the librettist, who has provided an irregular text, structured only by a single recurrence of Ormindo's two-line refrain, "Non ti doler d'Amore, / Non l'oltraggiar mio Core" (example 10):

Cavalli molded this text into a lengthy, three-part ostinato aria, each part containing a statement by Erisbe followed by Ormindo's refrain. While the librettist repeats Ormindo's refrain only once, within Ormindo's single speech, interspersed with five lines of recitative, Cavalli uses it twice again (bracketed here) to comfort Erisbe during her speech, integrating it into her lament by using the same ostinato figure to accompany it. Moreover, he expands these lines extensively by means of internal text repetitions. The ostinato, which has acquired enormous momentum through its persistence, almost uninterrupted, through thirty-two lines of text, terminates abruptly as Erisbe feels the first effects of the poison, and Cavalli shifts to recitative for her three final lines—and from a diatonic to a chromatic accompaniment. Extra word repetitions and aria setting were not suggested by this text. Cavalli the dramatist modified the libretto to create a much larger and more highly structured form than that provided by Faustini, a form that expresses with great intensity the lovers' increasing closeness as death approaches.

Despite the greater specificity and range of Faustini's formal cues, in particular the standardized strophic forms that evoked consistent aria treatment from the composer, Ormindo is remarkably rich in instances of the composer exercising his stylistic prerogative: in his treatment of refrains, in his use of lyricism within recitative, and in his expansion and rearrangement of text. Recitative and aria styles still mix freely, and some of the most expressive passages in the score are those particularly Cavallian efflorescences, the arioso expansions of single lines or couplets within passages of text that were clearly intended as recitative. The freedom granted by the librettist and exercised by the composer here was reduced considerably by the middle of the century, as the distinction between aria and recitative became more absolute.^[28] That distinction was strengthened as the formal structure of aria texts became increasingly confirmed by their meaning and dramatic function, and eventually their position within the scene. A new stage in this development is represented by Cicognini's and Cavalli's Giasone of 1649, one of the most popular and, consequently, most influential operas of the entire century.

Cavalli and Cicognini

Giacinto Andrea Cicognini's background and career differed as much from those of his contemporaries as did his librettos. Educated in Florence rather than Venice, he was well versed in the tradition of Spanish comedias , of which he produced several adaptations and translations.^[29] And he was fully established as

the author of a number of prose dramas before entering the world of opera toward the end of his literary career. Two of his librettos, Giasone and Gli amori d'Alessandro magno e di Rossane , even led an independent existence as prose dramas; in the former case, the libretto was the source of the play, while in the latter the relationship was reversed.^[30] Cicognini's librettos are more varied, more individualized, and poetically more sophisticated than those of his Venetian contemporaries. They stand out especially for their mixture of comic and serious characters—even the permeation of serious elements by comedy—and for the dramatic impact of the poetry itself. Although versi sciolti still form the basis of his poetic language, Cicognini employed a much greater variety of meters throughout his text, primarily in the arias, but also in recitative for dramatic purposes.

The most famous scene in Giasone is Medea's invocation of the spirits of the Underworld at the end of the first act. Standing as the prototype of all subsequent operatic incantations, it illustrates the ways in which Cicognini used poetry for dramatic contrast. The variety of accent produced by the changing meters (two symmetrical groups of quinari sdruccioli , each closing with a tronco [A: 12 lines], followed by a group of mixed settenari and endecasillabi [B: 11 lines], then an irregular mixture of two-, three-, four-, five-, seven-, and eleven-syllable lines, variously accented [C], and finally ten quaternari tronchi [D]) combined with free rhyme irregularly interspersed with couplets, and the contrast between sdruccioli, piano , and tronco verse-endings creates a scene of remarkable energy (example 11):

Such verses imperiously demanded to be matched in the musical setting, and Cavalli responded effectively to the poet's cues with music that reinforces the metric individuality of the scene. In restricting the melody to repeated notes and simple triadic figures following precisely the accentuation of the text, the composer powerfully projects the intensity of Medea's invocation. It is worth noting here that the composer distinguishes individual metric sections of the text

from one another and that part of his setting—the first twelve lines—clearly qualifies as a closed form or aria, though it is not particularly lyrical.

More generally speaking, Cicognini controls meter and rhyme with particular skill to distinguish clearly between recitative and aria verse. Indeed, the contrast between the two types of poetry is much cleaner in Giasone than in any of Faustini's (or Busenello's) librettos. Strophicism is by far Cicognini's preferred method of indicating an aria, characterizing more than two-thirds (seventeen of twenty-four) of the closed forms in the work; and Cavalli set virtually all of Cicognini's strophic texts as arias. All seven of the non-strophic arias,like those of Faustini, are based upon sections of text that stand out because of their special meter and rhyme—like the opening section of Medea's incantation scene already mentioned. But their independent setting is assured—or at least strongly suggested—by their highly individualized meter and unusual length: several of them comprise as many as ten or eleven lines.

Cicognini's strophic texts as well are unusually long. Rather than the characteristic Faustinian six lines, most of Cicognini's strophes range between eight and ten lines. Furthermore, they are extremely varied: hardly any two forms share the same meter or rhyme.^[31] The variety is achieved not only through the choice of a different meter for virtually every text but through the metric changes within individual texts themselves. Whereas some arias are in a single meter, relying on the rhyme scheme to articulate their form, others combine lines of very different meters, sometimes as many as five or six in a single strophe. Delfa's aria in 3.10, for example, utilizes four meters and a variety of accentuation patterns and verse endings, which produces a distinctly offbalance (comic?) effect (example 12):^[32]

Cavalli's setting underscores the irregular structure of the text by sticking quite closely to it, creating an aria that is comically erratic and unpredictable, particularly in its phrase structure. The pattering musical rhythm exactly translates that of the text, from which the unassuming melody, consisting primarily of repeated notes and mincing half-steps, does not detract. Only at the final line does the voice finally cease its patter, the bass now taking over for a characteristic conclusive flourish. Although Cavalli has followed the text precisely, almost slavishly, until this point, he expands the aria at the end, slowing the vocal part down for the last line, which contains the main message of the aria.

In contrast to Faustini's exploitation of refrain forms, only three of Cicognini's arias—one of them strophic—are marked by a refrain—apparently too few for Cavalli, who created several refrain forms himself by repeating first lines of arias at the end.^[33] Cicognini hardly used refrains in his recitative either; but when he did their message seems clear: Cavalli regarded them all as invitations to lyrical setting. And he found plenty of other opportunities for arioso expansion—within and increasingly at the ends of speeches. Significantly, the most extended arioso passages are reserved for Egeo and, especially, Isifile, the two principals whose love remains unrequited for most of the opera and who therefore have less to sing about and are more prone to emotional excess seeking or requiring an outlet than the other characters.

In keeping with a rather old-fashioned conception of verisimilitude, Cicognini distributed his arias quite unevenly among the cast. The four principals, Giasone, Medea, Isifile, and Egeo, sing very few—only one or at the most two each—all of them specially justified by the dramatic circumstances. (And this despite the fact that established legend was left far behind in this plot, the

traditional names of the characters notwithstanding). This contrasts markedly with the large number of arias given to the secondary characters: five for the old nurse Delfa, and three each for Alinda and Oreste. Confirmation, perhaps, of a kind of abstract taste for arias is provided by Rosminda, a gardener, who seems to have been introduced solely for the purpose of singing an aria in 1.3—and possibly to provide cover for the set change required for Medea's appearance in her throne room in 1.4.^[34]

Cicognini's decisions were determined by considerations of verisimilitude not only in the distribution of arias, but in their dramatic function and placement within scenes. With the exception of three or four, all the arias take place out of earshot of the other characters—either in solo scenes or when the other character in the scene is asleep;^[35] or else they are specifically used, as arias, to enhance a dramatic situation or a characterization. For example, Giasone's first appearance in the opera is marked by an expansive aria that conveys, economically and operatically—both to the audience and to his lieutenant Ercole—Medea's power over him. Ercole, responding to his captain's "aria as aria" as evidence of irrationality, urges him to return to his senses.

In placing most of his arias at the beginnings of scenes, Cicognini exploits their natural potential for emphasis: to set up a situation against which other characters (or the same character in a monologue) can react. In the few instances, usually comic monologues, where he places them at the ends of scenes, they are more static, summarizing the action that has occurred and marking the singer's exit.^[36] Of course the dynamics of the two kinds of scenes are very different from one another, and the variety of aria placement corresponds to larger dramatic considerations. Arias can be propulsive at beginnings of scenes but not at ends. Giasone's scene-opener, in addition to characterizing his position within the plot, sets the whole drama in motion, beginning to build anticipation for the appearance of the legendary Medea two scenes later. Monologue scenes that close with arias are usually external to the plot; they are intended not to further the action but to develop characterization, to let the audience in on the character's thoughts.^[37]

Cicognini distinguished more clearly between aria and recitative than any previous librettist, reinforcing that distinction by comfortable, appropriate

placement within the drama, and provided the opportunity for greater musical variety in his arias. But Cavalli still asserted his privileges as a musical dramatist in the usual ways: by imposing his own form on arias and, even more characteristically, by exercising his option of setting recitative text lyrically for affective purposes.

The composer's ultimate control over the librettist's form is especially striking in the opening scene of act 3, for Oreste and Delfa, confidant and nurse of Isifile and Medea respectively, who represent opposite sides in their mistresses' tug-of-war for Giasone. Their dialogue is organized in three strophes of four successive ottonari sdruccioli , two strophes for Oreste enclosing one for Delfa. Although formally parallel, the three stanzas do not share the same mood. Oreste's opening strophe, in which he comments lyrically on the beauty of the shadowy night, is countered by Delfa, who punctures his lyrical effusion by denigrating the shadow as temporary and fleeting, offering the non-sequitur that she would find the embraces of a lovable husband more delightful. Her response causes an abrupt change in Oreste's tone as he defensively strikes back at her, hoping to disqualify himself as an object of her desires (example 13):

Despite their different moods, Cavalli set both of Oreste's strophes to the same music; at least in the first strophe that music hews closely to the meaning of the text. Three sequential phrases of ascending eighth-notes in 5/4 meter set the first three parallel lines; but for the fourth line, appropriate to its meaning—rest after action ("A posar l'ombre ne chiamano")—Cavalli counteracts the parallelism, almost wilfully doubling the note values as the melody descends from its peak to cadence on its low point. The result is a rather asymmetrical, but nicely shaped, strophe that ignores the text form as it follows, literally, its meaning.

Delfa's strophe, though it is in the same form, is set to different music, but with equal attention to text meaning. Here Cavalli's treatment is again sequential, with three of the four parallel lines set syllabically to repeated quarter notes; it is the second line of this strophe that evoked a contrasting setting from him, in this case an acceleration rather than deceleration, to elaborate, wide-ranging melismatic motion in sixteenth-notes for literal portrayal of the words: "Ch'è fugace, e vana, è instabile." These are the subtle adjustments that contribute so much to the effective matching of music and text that characterizes this opera.

To serve the drama, Cavalli not only suppressed or overrode textual regularity; for the same purpose he occasionally did the opposite and regularized Cicognini's poetry. One of the best examples of this procedure occurs in the duet between Giasone and Medea in 3.2. The scene, entirely in recitative meter, is laid out almost symmetrically for the two characters. Following an opening quatrain for Medea, set in rather lyrical, well-shaped recitative style by the composer, each of the lovers has a three-line stanza, which Cavalli set in a parallel, if not exactly strophic, manner, again in a kind of arioso style. The lovers then join voices for a brief duet, which is followed by two passages of text, the first for Medea, the second for Giasone. These, although they begin similarly, are of unequal length and dissimilar form (example 14):

Cavalli transforms this text into a kind of reciprocal refrain-enclosed lullaby, setting the two lovers' statements to similar ("strophic") music, despite their different texts. Taking advantage of the parallelism of the opening lines, he turns them into refrains, which he brings back at the end of each strophe, compensating for the differences in their texts—essentially Giasone's extra line—in a freer, strongly text-interpretative central section where the formal disparity passes virtually unnoticed. Cavalli's molding of this lullaby into the climax of a wonderfully symmetrical scene owes its impetus, perhaps, to Cicognini's word choice; but it is the composer who increased the mutuality of the lovers by making them closer in music than they are in words.

Cicognini's Legacy

For the Arcadian critics bent on the reform of Italian literature at the end of the seventeenth century, Cicognini's Giasone was a crucial work. According to Giovanni Maria Crescimbeni, one of their chief spokesmen, the author of Giasone was worthy of both praise and blame: praise for having created the first and most perfect drama in existence ("il primo, e il più perfetto Dramma, che si trovi"), and blame for having opened the floodgates to all kinds of abuses, the mixing of genres, the abandonment of linguistic elegance and purity, and, through the introduction of arias, the destruction of verisimilitude in drama.

Around the middle of that century, Giacinto Andrea Cicognini. . . introduced drama [as opposed to favole pastorali ] with his Giasone , which, to tell the truth, is the first and the most perfect drama there is; and with it he brought the end of acting, and consequently, of true and good comedy as well as tragedy. Since to stimulate to a greater degree with novelty the jaded taste of the spectators, equally nauseated by the vileness of comic things and the seriousness of tragic ones, the inventor of drama united them, mixing kings and heroes and other illustrious personages with buffoons and servants and the lowest men with unheard of monstrousness. This concoction of characters was the reason for the complete ruin of the rules of poetry, which went so far into disuse that not even locution was considered, which, forced to serve music, lost its purity, and became filled with idiocies. The careful deployment of figures that ennobles oratory was neglected, and language was restricted to terms of common speech, which is more appropriate for music; and finally the series of those short meters, commonly called ariette , which with a generous hand are sprinkled over the scenes, and the overwhelming impropriety of having characters speak in song completely removed from the compositions the power of the affections, and the means of moving them in the listeners. (Appendix II.7)^[38]

Although Crescimbeni admits his ignorance of the period immediately preceding Cicognini's work and fails to appreciate Cicognini's connections to an already burgeoning operatic tradition, the focus of his attention on Giasone is symptomatic of its historical position. He was not wrong in ascribing special importance to the work: even from our vantage point it appears to stand at an important crossroads in the history of opera. It is, of course, unlikely that any single work (out of so many) could have had the impact Crescimbeni ascribes to Giasone . But the opera was clearly a symbol of the times; and its extraordinary popularity allowed it to represent those times quite legitimately.^[39]

In Giasone the definitive separation of aria and recitative was finally achieved. Cicognini's standard means of distinguishing them persisted until the end of the century: strophicism and/or versi misurati meant aria; versi sciolti , recitative. The distinction was reinforced by clarified dramatic functions for arias: to promote action (the incantation aria; the lullaby), to comment on action and philosophize on life (the comic arias), and to express intense feeling (Giasone's opening love song; Egeo's lament).

Despite these general features, however, Cicognini's arias do not seem predictable either in form or function because they arise so naturally out of the drama. Clear as his signals are, the only formal feature shared by most of his arias is their strophicism. Otherwise, each one of them is unique: in a different meter, with a different rhyme scheme, and an altogether different shape, conferred by highly irregular line lengths. And, clear as their dramatic function is, each emerges from its context in its individual way, for its own purpose, strictly in accordance with verisimilitude. No two characters are presented in the same way. Each scene, each action segment, each act unfolds musically in its own particular fashion.

The special strengths of Giasone , and its significance as a model, lie in the balance it embodies. The clear signals of the librettist are perfectly matched by the composer's response, achieved both without recourse to rigid formula and without excessive strain on verisimilitude. The musical drama is shaped by an appreciation, shared by the poet and composer, of the distinctions between speech and song. Giasone is an ideal dramma per musica , in which both elements of the now-historic compromise have equal weight—mutually justifying each other. Giasone also offered a model for operatic conventions of a more general kind, presenting traditional scene-types with a naturalness rarely matched in its successors. We shall have occasion to consider this aspect of the opera in detail in chapters 11 and 12. Literally, then, Giasone represented a brief moment of equilibrium in the history of opera: at once the endpoint of a process of generic maturation and the beginning of a new stage in which, now fully legitimized, and aided and abetted by the rising influence of the singer, musica would eventually subjugate dramma .

Paradoxically, perhaps, the very inventive freshness of Giasone was both the source of its popular success and the cause of its eventual indictment by the Arcadians. Although it was itself carefully constructed so as not to disrupt verisimilitude, either in the arias or in the mixing of comic and serious elements,

it spawned imitations that were less observant. The decadence deplored by Crescimbeni is in fact much better exemplified by one of those imitations, Alessandro vincitor di se stesso (Venice, 1651), which was specifically modeled on Giasone , by two operatic neophytes, Francesco Sbarra and Antonio Cesti.^[40] Often cited by modern historians as marking the definitive breakdown of operatic verisimilitude by having initiated the invasion of the aria, Alessandro is far better known for the preface to its libretto than for anything else. In fact, its score was misattributed to Cavalli until relatively recently.^[41] In confronting the crucial issue of verisimilitude in his preface, Sbarra acknowledged that arias are unsuitable for serious characters such as Alessandro and Aristotile, but he justified them on the basis of operatic necessity: "if recitative were not interrupted by such jokes, opera would cause more annoyance than delight" (Appendix I.29f).

But, as the libretto itself illustrates, abuse of verisimilitude runs much deeper than the mere misbehavior of Sbarra's heroes. Admittedly, the distribution of the arias is atypical. Not only does Alessandro sing five "ariette" and Aristotile two—out of a total of some thirty closed forms—but another principal, Efestione, sings seven, while Apelle, a character who could easily—and humorously—have sung more, has only one. By focusing on Aristotile's and Alessandro's few "ariette," however, Sbarra obscured the real abuses of verisimilitude in his work, abuses that reveal his lack of experience as a librettist and his misunderstanding of the model represented by Giasone . Sbarra's misconception has to do with the function of arias in the drama. For it is not so much that Alessandro and Aristotile sing arias, but when they do so, and why . Most of their arias, as well as some sung by other characters, are wholly unjustified dramatically.

Alessandro bears signs of inconsistency, both in the librettist's method of signaling closed forms and in the composer's response, an inconsistency born of confusion over the purpose of such forms. Indeed, it is that confusion, dressed as purposeful, that is described in Sbarra's preface. Although it has a somewhat greater number of arias than other operas of the same time (some thirty-odd), many more of their texts are formally ambiguous. Only eight—fewer than one-fourth—are strophic, a much smaller proportion than in Cicognini's librettos, and the others fall into surprisingly many patterns, hardly any two of them alike. They range from as few as four to as many as thirteen lines in a variety of meters and rhyme schemes, and only a few of them utilize

refrains. Particularly in the case of the shorter texts, it is often difficult to know whether to regard them as aria signals at all, although the composer obviously took them that way.

The dramatic function of these texts does little to reinforce their formal significance or clarify their message to the composer. Indeed, they often seem to contradict the implicit conventions of verisimilitude altogether. Invitations to lyrical expansion are reserved for neither commentary, contemplation, nor highly emotional expression, but occur almost indiscriminately, even in the midst of conversations.

For example, the dialogue between Efestione and Alessandro in act 1, scene 5, is cast in a succession of tightly rhymed, highly metrical passages; the text form suggests aria setting but the dramatic situation does not.

And then later in the same scene:

Cesti set the first dialogue and the last passage in aria style in response to clear signals from the poet. But such treatment renders the interaction between the

characters extremely unnatural and stilted. Sbarra's lack of discrimination here is exacerbated by Cesti's faithful setting (example 15). The libretto contains a number of other instances of one character addressing another in aria. They include Cina's midscene strophic address to Alessandro in 1.3; Aristotile's, also strophic, to Alessandro in 1.4; and his non-strophic closing address at the end of the same scene, which, although it might have served to mark his exit, fails to do so because he remains on stage.

Sbarra's blurring of textual-formal distinctions between action and contemplation licensed Cesti to exploit the slightest formal cue to justify lyrical expansion, regardless of the dramatic situation. In act 1, scene 6, for example, the inappropriateness of Alessandro's aria is more the fault of the composer than the librettist. The text is a quatrain addressed to Campaspe:

Although this is not as unequivocally formal as the previously quoted text—its only regularity is a rhyming final couplet—Cesti confirms closure by setting it as an elaborate, florid, and highly expanded aria (example 16).

Cesti set most of Sbarra's metrical texts lyrically, no matter how short, but he weighted some more heavily than others through the use of string accompaniment, instrumental ritornellos, and repetition of music and text, thereby emphasizing their separation from the recitative context, their "aria" character. In other instances, however, the lyricism is more fleeting—and more acceptable from the point of view of verisimilitude. Emerging suddenly from the recitative context, it disappears back into it with minimum impact; the composer runs straight through the text, only once, as if it were recitative, with no musical elaboration at all.

Even when separation from the dramatic context is specifically legitimized, both by the particular situation and the text form, as in most of the strophic arias, Cesti adhered closely to the structure of the poetry. Such adherence often yields an effective mixing and juxtaposition of styles, "alla Cavalli." In Fidalpa's aria in act 2, scene 7, for instance, the composer breaks the nine-line text into three sections—aria-recitative-aria—thereby retaining flexibility of text portrayal, even within an aria. The variety of Cesti's responses to Sbarra's signals for closure would seem to require a richer descriptive terminology; aria alone does not suffice. Such terms as arioso, mezz'aria , and arietta are useful here to distinguish between lyrical passages that are integrated within recitative, short

arias that are musically undeveloped, and light "singsong" arias that are based on strongly metrical texts.^[42]

Sbarra's numerous metrically distinct texts, almost all of which Cesti set lyrically, may have assured sufficient relief from musical tedium, but at great cost to the drama. By depriving arias of their traditional—and purposely limited—functions as songs or as vehicles of emotional release, and thus flattening the distinctions between them and recitative, Sbarra and Cesti actually deprived dramma per musica of one of the chief sources of its strength. Aristotile and Alessandro indeed did not act like heroes. They could not behave properly in public because they did not know the difference between speech and song, knowledge that any comic servant or Venetian audience was fully privy to.

In marked contrast to Giasone, Alessandro seems to have failed at its first performance.^[43] It would be reassuring to be able to ascribe that failure to its shortcomings as a music drama, to the tastes of a discriminating audience; but unfortunately evidence for such discrimination is lacking. Indeed, despite its problematic character, the work was revived at least six times during the 1650s.^[44] But it remained an anomaly. The librettists and composers of the second half of the century—including Cesti, and even Sbarra—eventually found more effective ways of incorporating additional arias into their operas: they did so not by abusing verisimilitude but by expanding the opportunities for justified song. Rather than depriving arias of their dramatic function, Cicognini's heirs altered the dramaturgy of their librettos to accommodate more of them, developing structural as well as dramatic conventions to shield them—in monologues and dreams, at entrances and exits. It was the crystallization of these conventions, originally inspired by the requirements of verisimilitude, and the attempts to circumvent or vary them, that eventually led to the decline lamented by Crescimbeni.