4

Discipline

Opera is extravagant, absurd, loud and above all created by human voices. As we’ve already seen, it has often been thought in need of discipline, and sometimes such thoughts have turned to action. There is scarcely a decade in the eighteenth century (the golden age of the polemical pamphlet) that avoided some philosophical debate about opera’s ills – often, human nature being what it is, the ills of other people’s opera rather than one’s own.

The year 1762 saw such an earthquake, one more powerful and long-lasting than most. It marked the premiere of yet another Italian Orpheus opera, this one performed in Vienna. After more than a century of comparatively hard times for Orpheus, an Italian poet called Ranieri de’ Calzabigi (1714–96) wrote a new libretto about him; it was set to music by a German composer, Christoph Willibald Gluck (1714–87). Calzabigi and Gluck’s Orfeo ed Euridice was widely agreed to be astonishing, revolutionary both in form and style. Gluck went on to write further highly original works in Italian and in French, including Alceste (1767), Iphigénie en Aulide (Iphigenia in Aulis, 1774) and Iphigénie en Tauride (Iphigenia in Tauris, 1779). The subjects were all derived from Greek mythology, and that is significant – Gluck and his librettists demonstrated that neo-classical fervour could go hand in hand with a desire to curb operatic excesses, elaborate singing above all. A generation later, and in the same way, Greek-influenced gowns and simple, natural curls modelled on images from antique amphorae would replace onstage the reeking powdered wigs and huge panniers of eighteenth-century court dress.

Gluck’s new kind of opera was enormously influential: he is one of Mozart’s operatic father figures, and his name was a talisman well into the nineteenth century. One need only read Berlioz’s Traité d’instrumentation (Orchestration Treatise, 1843), or his Gluck-obsessed Mémoires (1870), or the polemics that swirled around his revival of Orfeo in Paris in 1859, or Liszt’s Orfeo in Weimar some years earlier, to sense how the names of both composer and work continued to be influential. To assume Gluck’s mantle was to make a play for opera’s moral high ground, to claim a level of musical purity and abstraction that messy, quotidian theatrical practice seemed never to achieve. Staging Orfeo with high-minded intent and re-discovered classical restraint has always signalled a reaction against theatrical extravagance. In Germany, Greek-revival productions of the opera in the first decades of the twentieth century gave voice to exasperation with decadent expressionism in opera. Richard Strauss’s Elektra may have shrieked, raved and jumped up and down in dirty rags, struggling to make herself heard over an enormous, blaring orchestra; but never mind, Gluck’s Orfeo, brought back on stage to sing out his grief in sunny C major, restored much-needed restraint. How did this strange new Orfeo come about? What currents shaped it?

To answer these questions, we need to go back several generations. The high aspirations that created Orfeo in 1762 had been prepared by what some intellectuals saw as decades of offences against audiences. The sources of difficulty in Handelian opera seria discussed in our last chapter were obvious, and continued with the next generation of serious opera composers. According to the aesthetic attitudes that led to Gluck’s new way with opera, libretto poetry had become formulaic and its emotions were made distant by endless metaphor – endless ways in which human feelings were not only likened to natural phenomena but in which those natural phenomena became the focus of the musical expression. What’s more, the plots tended to revisit the same archetypes: the troubled but ultimately enlightened king, the confused hero, the spurned lover, the evil plotter; politics and public duty were endlessly set up against love and family ties; mistaken identity was facilitated by pan-soprano sounds from male as well as female characters. The plot flew by in simple recitativo, barely musical recitation that few seemed to care about or even listen to. The drama – the sense of suspense and release, the emotional investments that were the essential preamble to operatic pleasure – was lodged exclusively in an unyielding succession of solo arias. And each aria repeated the same formal shape (the so-called ‘da capo’ form), one whose raison d’être was realized only in performance – in the surplus brought by improvised ornamentation when the initial ‘A’ section was repeated. It was not just that the astonishing costumes and flat sets tended to place on stage a succession of fabulous, feathered, glittering characters; it was that the singers impersonating these characters all produced singing of such immediate and spectacular virtuosity – whether in a slow, pathetic aria or in a show-stopper with cascades of rapid notes – that all other aspects receded.

The intellectuals and men of letters grumbled about this, in particular because the singers and their audiences seemed unfailingly to know where tragedy and triumph were situated. There is a famous story about the two pre-eminent castrati of the pre-1750 period, Senesino and Farinelli, told by the great British music historian of the late eighteenth century, Charles Burney:

Senesino and Farinelli, when in England together, being engaged at different theatres on the same night, had not an opportunity of hearing each other, till, by one of those sudden stage-revolutions which frequently happen, yet are always unexpected, they were both employed to sing on the same stage. Senesino had the part of a furious tyrant to represent, and Farinelli that of an unfortunate hero in chains: but in the course of the first song, he so softened the obdurate heart of the enraged tyrant, that Senesino, forgetting his stage-character, ran to Farinelli and embraced him in his own.¹

Senesino’s embrace of Farinelli is a wonderfully stagy (indeed, if we are to believe Burney’s story, staged) demonstration of how listeners can be transfigured by a moment of live singing. It also suggests, along with many other accounts of early eighteenth-century opera in performance, that ‘breaking out of character’ was commonplace and not necessarily viewed as the theatrical sin it would be today, amid our darkened auditoriums and constant calls for silent absorption in the theatrical unfolding. John Rosselli, one of opera’s best recent historians, retells an event in Ferrara in 1722: a cardinal is seated in a stage box (that is, he attends the opera from a seat literally on the stage, and thus in full view of most of the rest of the audience); just as a soprano sings the words ‘Give alms to a poor pilgrim’ he reaches out and hands her a purse of gold. As Rosselli comments, in making this gesture the cardinal ‘combined two favourite baroque devices, a public show of beneficence and playing about with theatrical illusion’.² But the point of both this and the Farinelli/Senesino anecdote is also that some large part of the essence of opera seria was ephemeral and cannot now be re-created. This is, of course, true of all opera; but its extent in the early eighteenth century is exceptional. Much that was most important musically and dramatically was improvised on the spot: it couldn’t be written down, still less recorded; we have no access to it. All we can do when we re-create the operas today is be aware of the loss. Now we sit in darkness; only the stage can be seen. We are isolated from each other and each other’s reactions. We do not feel free to comment, and can boo or cheer only at fixed moments. Nor do we feel free to ignore the spectacle entirely if we choose; we cannot put up the shutters of our box and play cards when the recitative starts, or buy oranges in the aisles, or reach across the stage to hand a purse of gold to a singer whose words momentarily and magically give meaning to our magnanimous gesture.

As theatrical archaeologists, we might say that early eighteenth-century opera seria is now relatively mute, since what counterpoised for those stylized plots, that relentlessly well-behaved poetry and nothing-but-arias structure was present in performances that have now permanently vanished. What has gone are not simply the obvious items like the castrato voice, which became such an important source of fantasy. In Samuel Richardson’s great epistolary novel of the 1740s, Pamela, or Virtue Rewarded, which appeared just as Handel’s opera career was coming to an end, one character despairingly says: ‘But what have I said, what can I say, of an Italian opera? – Only, little to the Purpose as it is, I wonder how I have been able to say so much: For who can describe Sound? Or what Words shall be found to embody Air?’³ It is a lament all operatic archaeologists can echo. In the Internet age, we call the workers who produce text and ideas the ‘content people’; they are not very high on the pay scale, no matter how much respect they garner as ‘author’. Eighteenth-century composers knew how this felt. There’s a famous and probably invented story about Handel threatening to throw the renowned prima donna Francesca Cuzzoni (1696–1778) out of the window when she refused to sing an aria from Ottone (Otto, 1723), one that had not been expressly written for her.⁴ Endlessly repeated, this tale caters to our modern-day sense of hierarchy: composers should be thus manly and thus in control. But the economic and practical realities tell a different story; Handel, for all his fame, was at base a ‘content person’; whatever his occasional outbursts, he was usually the one dangling out of the window, at the whim of the sopranos and castratos his music was created to showcase.

In the eighteenth century, the rumblings of revolt against this state of affairs came from both literary-critical and practical perspectives. It happened in two phases. The first concerned words, the second concerned words and music – the kind of music that singers should be encouraged or forced to favour or to avoid. A good part of the impetus for such changes, although it seemed to originate with pamphleteers and others, came from what was happening operatically elsewhere. For example, the relative decorum of state-controlled French tragédie lyrique, both in its libretti and its music, played a quiet role in disciplining Italian opera seria. With a strong tradition of high-classical spoken drama, France had theatrical rules that crossed over to and contained French libretti from their beginnings in the 1670s. And French compositional styles around the turn of the eighteenth century, in instrumental as well as vocal music, depended more on instrumentation and rich harmony – at least, so the French saw it – than on the constantly unfurling melody and unbridled coloratura of the Italian style. There has never been much point in trying to close off one operatic tradition from the alternative languages that feed it and are fed from it, whether those languages are other operatic genres or the shapes and sounds of contemporary instrumental music. There are ways in which Italian opera seria and French tragédie lyrique have sibling traits, just as there are ways in which da capo arias resemble concertos and other instrumental forms.

Exchanges were reciprocal: mid-century French musical reformers looked towards the relative simplicity and lyricism of Italian comic opera. In his Confessions (1770), Jean-Jacques Rousseau ascribes revelatory powers to Italian opera buffa. In Venice in 1744 he heard an opera called La finta schiava (The Feigned Slave, a so-called pasticcio, meaning an opera made by cobbling together arias from a string of previously existing works). The noisy, brilliant arias had little effect on him, but one simple comic tune brought him to an awakening: his ‘ears and at the same time eyes’ were opened. He repeatedly alludes to sleep and waking, to being brought into consciousness by the aria. Anxious to recapture the experience, he gets hold of the music, only to find that the notes on the page are not the same.⁵ Describing how in 1752 he was moved to write his own comic intermezzo, Le Devin du village (The Village Soothsayer), Rousseau once more makes clear that Italian comic opera unlocked his spirit. After a long night talking to a friend about opera buffa – just talking about it – Rousseau could not sleep and the day after, ‘in a kind of vaulted chamber that was at the end of the garden’, scribbled down two arias and the final duet of Le Devin du village.⁶ Whatever mythmaking is at work here – and there is no lack of it – the link was clear. Le Devin du village and similar operas that followed in its wake, with their Italianate manners, in turn ricocheted back to French opera; its easy musical candour lent alternative flavours both to opéra comique and to tragédie lyrique after the 1750s. And then, to complete the circle, opéra comique – ostensibly Italianate but nonetheless a French genre with its own set of musical conventions – re-inflected opera seria through Gluck.

THE METASTASIAN CODE

In the previous chapter we saw one important phase of operatic reform, stemming from the Arcadians in Rome; but even as Handel and his star singers were beguiling audiences in London, further waves of reform were washing over Italy. As ever, ancient Greece was routinely called upon: Pier Jacopo Martello’s 1715 treatise Della tragedia antica e moderna (On Ancient and Modern Tragedy), written by a second-generation Arcadian, constructed an elaborate fantasy in which Aristotle appears before him on the road to Paris and instructs him (at length) on how to tackle the libretto problem. We have already read ‘Aristotle’s’ advice about recitatives and arias, but there was much more. The philosopher’s stream of friendly observations satirized most of opera’s presumed ills: ‘not too many forests, for tree trunks and leafy branches are not subjects for the theatre painters … you should take care to choose a fabulous story composed of a mixture of gods and heroes or a true history of heroes’. Castratos must always be ‘elegant, not uncouth’. ‘Let the means by which the events take place lack verisimilitude’, the philosopher dryly recommended, ‘let there be recognitions and reversals of fortune. In recognitions let us be easily deceived by a sudden costume change, by certain objects found in the cradle of a character when he was an infant.’ Poets, he concluded, are schooled by opera to ‘conquer themselves and renounce their own wishes’.⁷ Five years later another satirist, the opera composer Benedetto Marcello (1686–1739), wrote a fantastic, at times surreal pamphlet called Il teatro alla moda (Fashionable Theatre, 1720) cataloguing chapter by chapter what he considered the excesses of the Italian scene: composers and impresarios toadying before singers (and, more degrading still, before singers’ mothers and other ‘protectors’), absurd plots, needlessly lavish stage sets, the slave status of poets. These two reformers give us an immediate sense of the impulse to shift opera away from giddy excess, spectacular machines and even more spectacular virtuosi; and, in tandem with this, to introduce notions of restraint and balance characteristic of the greatest spoken theatre. Martello’s Aristotle explicitly cites Greek tragedy as a model, as one might expect, but also evokes French classical drama, a tribute to the author’s Francophilia.

Pietro Metastasio (1698–1782) was the poet who took this second wave of reformist critique most successfully to heart. He was a close friend of the castrato Farinelli (they started letters to each other with the greeting ‘Caro gemello’ – Dear Twin), and became easily the most famous Italian poet of the eighteenth century – his thirty-odd opera seria libretti were the most frequently set in all operatic history. As late as the early nineteenth century, composers still occasionally resorted to his dramas, Rossini among them. His libretti were austere and balanced enough to fill a particular cultural black hole – the absence of serious spoken-theatre works in Italian to rival the great spoken tragedies of France. As Charles de Brosses quipped in 1739, ‘for tragedies in the form of opera, they [the Italians] have an excellent author still living, Metastasio, whose plays are full of wit, intrigue, dramatic turns of events, and of interest, and would no doubt work to great effect if they were played as simple spoken tragedies, leaving aside all the little arietta business and operatic devices, which would be easy to remove’.⁸

Metastasio had refined and built on Zeno’s work. Complicated subjects – all those foundlings later recognized as princes, or zany subplots required to supply minor characters with aria fodder – were once again purged; the number of characters was further reduced. Crucially, he followed Zeno in the elimination of comic characters. Purity of genre – not mixing tragedy and farce – was an Aristotelian ideal and chimed well with general eighteenth-century obsessions about classification. The irony was that the outcast comic characters – all those gardeners and servant girls, bourgeois lechers, old maids, rustic lovers and tipsy functionaries – received in recompense an operatic home of their own, first in comic intermezzi and then in fully fledged opera buffa. In other words, the first step towards a work such as Mozart’s Le nozze di Figaro (1786) was an act of banishment. More ironic still, the reformers’ self-important adherence to codes and high ideals would eventually place opera seria on a large, imposing, classically styled pedestal, one that audiences, singers and opera houses over the coming decades could feel free to ignore or treat with serene indifference.

So what was the situation of Italian opera seria around 1750? Libretti had become stately. Metastasio, child of the rational eighteenth century, prescribed a moral code in which virtue would be eternally rewarded and sin eternally punished or, better, magnanimously forgiven after tremendous expiation. Fewer people die, and they all get up at the end to sing a chorus together. No one commits suicide, or murders his or her offspring, with exceptions always made for Medea. What is more, operatic poetry begins to engage seriously, philosophically even, with matters of political significance. One such matter is the nature of kingship, of absolute monarchs troubled by their crowns or performing sudden, desperately last-minute acts of mercy (Metastasio’s La clemenza di Tito – The Clemency of Titus, 1734, later set by Mozart, is a classic example). Opera seria libretti of this kind have been seen as important social allegories, even as prescriptions for the true role of the absolute ruler. In this sense it was entirely fitting that Metastasio became court poet to the Austrian Emperor (Charles VI), as had Zeno before him. What is more, this position – in effect that of a liveried court servant – was one Metastasio fully accepted. His epistolary description of his first meeting with the Emperor gives a perfect expression of the social hierarchies his libretti endlessly celebrate:

I spoke with a voice that I fear was not too firm, expressing these sentiments: ‘I do not know whether it is my contentment or my confusion that is greater on finding myself at the feet of Your Imperial Majesty. This is a moment I have sighed for from my earliest days, and now I find myself not only before the greatest monarch on earth but here in the glorious title of his present servant.’⁹

While the spirit of these Metastasian libretti, in particular the directness of their celebration of the political status quo, was new, certain structural features stubbornly remained from earlier, more unruly libretti. The waves of reform were, in other words, primarily literary, not centrally addressed to the ways in which music is changed by alterations in poetry. To some extent, this reflected the greater prestige and longevity of the literary component of opera. Metastasio’s libretti were famous throughout Europe at a time when their numerous musical settings were mostly tailored to a particular cast of singers. It was easier to commission a new setting than revive an old one when a new cast was assembled. Metastasio himself had no doubts about the relative hierarchy among opera’s constituent parts. In a letter to a fellow reformer, Francesco Algarotti, he wrote:

Those parts of opera that require only the spectators’ eyes and ears to plead for them always garner more votes than the other parts, whose merits can be measured only with intelligence and ratiocination. Everyone can see, everyone can hear, but not everyone understands and not everyone reasons.¹⁰

But the Imperial poet’s disdain for those who merely look and listen cannot hide the fact that audiences, the paying public, continued to be principally concerned with their ‘eyes and ears’, with the spectacle and the music. And although their prestige among intellectuals may have been low, composers of the post-Handel, Metastasio-bred generation continued to make musical adjustments, albeit ones far less trumpeted than those of their librettist companions. The most famous of them, Leonardo Vinci (1696–1730) and Johann Adolf Hasse (1699–1783), perhaps responding to the new sense of literary restraint but also reflecting broader shifts in musical taste, fashioned operas with the same basic formal outlines as the Handelian generation (strings of da capo arias), but with a less elaborate musical surface. Their melodies became simpler and more predictable rhythmically, the bass lines more functional (repeated notes in the bass, a sort of harmonic marking-time, were a commonplace of the style). They produced, in short, the beginnings of what was first called galant music and later became known as the Classical Style. Charles Burney (who very much approved of these modern features) praised Vinci in particular for thus simplifying music and ‘disentangling it from fugue, complication, and laboured contrivance’.¹¹

This galant style was what Gluck inherited, and it was not to everyone’s taste. Some of the old guard took exception. Burney reported Handel as saying, when Gluck visited England in 1745, that ‘He knows no more of contrapunto than mein cook’ (the cook in question, though, was one Gustavus Waltz, who was a cellist and singer, and may have been quite capable of fugal excursions).¹² But in spite of local resistance, these musical-stylistic changes were of lasting significance, and had great impact on the instrumental forms then emerging, in particular the string quartet and the symphony. At least initially, though, Metastasian opera broke no structural rules. If we look at the corpus of his libretti, set by numerous composers (including Gluck) between 1730 and 1760, resulting in literally thousands of operatic numbers, we find little more than a festive parade of da capo arias – some exquisite, some tedious – brought to life more or less well by the singers on whom their fate mostly rested.

Pamphleteering discontent with opera’s ills stems not just from asceticism – a philosophical bludgeon that can seem ludicrous when aimed at singers, scenery, orchestras and public entertainment. Operatic poetry and music, and even the singing, was at perpetual risk of being ignored by the audiences who flocked to attend the performances: in the later eighteenth century, especially in Italy, behaviour was still magnificently informal, despite reforms already undertaken. The unspoken wish inhabiting reform writings in the eighteenth century may have been a desire to see the audience absorbed and moved by art, a desire for the myth of Orpheus to come true.

In Naples, thirty years after de Brosses played chess during the recitatives, the audience remained alarmingly unruly. An English visitor, Samuel Sharp, was irritated by the chaos:

There are some who contend, that the singers might be very well heard, if the audience were more silent, but it is so much the fashion at Naples, and indeed, through all Italy, to consider the Opera as a place of rendezvous and visiting, that they do not seem in the least to attend to the musick, but laugh and talk through the whole performance, without any restraint, and it might be imagined, that an assembly of so many hundreds conversing together so loudly, must entirely cover the voices of the singers. I was prepossessed of this custom before I left England, but had no idea it was carried to such an extreme … not withstanding the noisiness of the audience, during the whole performance of the Opera, the moment the dances begin there is a universal and dead silence, which continues as long as the dances continue.¹³

The comparatives begin to give a sense that audiences were not the same everywhere; in this case, that the English public is more absorbed in the performances than in Italy. Sharp was savaged by an Italian riposte, from Giuseppe Baretti in An Account of the Manners and Customs of Italy (1768), who complained about:

your solemnity of scolding, as if we were committing murder when we are talkative in the pit, or form ourselves into card parties in the boxes. Our singers, then, though we be unwilling to listen, would be very impertinent, if they did not sing their best, since they are very well paid for what they are doing; and Cafarello was soon taught better manners when he took it into his head not to do his duty on the stage of Turin on pretense that the audience was not attentive to his singing. He was taken to jail in his Macedonian accoutrements for several nights as soon as the opera was over, and brought from jail to the stage every evening, until by repeated efforts he deserved universal acclamation.

Mr Sharp wonders also that it is not the fashion in Italy, as it is in England, to take a small wax-light to the opera, in order to read the book. A very acute remark as usual; to which I have nothing to say, but the Italians are not so good-natured as the English, who have patience enough to run carefully over a stupid piece of nonsense while a silly eunuch is mincing a vowel into a thousand invisible particles.¹⁴

It was against this backdrop that the Gluckian revolution took place. In a polemical preface to a score of Gluck’s Alceste published in 1769, Calzabigi referred to Metastasio’s libretti as ‘saddles for all horses’, the horses in question being male and female sopranos, denounced in equal measure. According to Calzabigi, Metastasio’s kings and queens may have been newly serious and newly restrained; but their passions were still frozen into too-convenient metaphors, which then too easily translated into vocal ornaments when the ‘A’ section of the da capo aria started up again. The singers, undefeated, were winning once more, and winning on ever more flamboyant terms. The hysteria that surrounded figures such as Farinelli was unprecedented; rhapsodies to silver trumpets, divine voices, nightingales, gods and goddesses in mortal form grew if anything more intense. Metastasio may have fluttered with pride in his position as servant to the greatest monarch on earth, but his wages in comparison with those of his ‘dear twin’ Farinelli were a pittance.

LYRIC TRAGEDY

It is well to recall that French tragic opera (tragédie lyrique or tragédie en musique) had seen little of these Italianate excesses. Elaborate, slow recitatives accompanied by full orchestra passed into brief formal arias; ornament was thought vulgar, at least those roulades of the extravagant, free-flowing Italian kind (there was plenty of small-scale decoration to melodic lines). Castrati were regarded with a shudder, and never got in the door. Acoustic sensuality – which is opera’s fundamental note – was shared between vocal melody and instrumental sonority, the latter meaning both the sound of the accompaniment and the combinations of harmonies surrounding the singing. Orchestral playing in eighteenth-century France was the ne plus ultra of instrumental accomplishment. France’s pre-eminence in this domain lasted into the nineteenth century, with instrumentation in French opera showing from the outset far more variety than its Italian counterparts. The operas of Jean-Philippe Rameau (1683–1764), also a renowned theorist of harmony, figured as a straw man in the Querelle des Bouffons, the great pamphlet battle of the early 1750s in Paris, in which tragédie lyrique (which had fallen on hard times) was set against the emerging Italian comic opera. Rameau’s works tended to be classed, depending on allegiance, either as the ultimate in Gallic nobility and dignity or as a national burden to be suffered with groans of ennui. Some salvos in the Querelle took more nuanced positions: as one pamphleteer put it, Rameau as ‘sacrilegious innovator’ brought ‘unknown harmonies’ to Parisian ears, thus presaging the ‘desecration’ of an august institution by Italian clowns and unbearable lightness; ‘the fateful event with which [Rameau] threatened us has finally come to pass’.¹⁵ In retrospect, Rameau’s music, although plainly an advance on French operatic predecessors, seems an unusual suspect for crimes such as sacrilege and iconoclasm; but words such as ‘gibberish’ and ‘monstrosity’ were flung around freely by his critics during the Querelle.

There is no doubt that tragédie lyrique was at a low ebb when Rameau began composing: the court-sustained repertoire of Lully operas had at last (after half a century) proved out of date musically, and Rameau was generally seen as offering revolutionary change. His first serious opera, a tragédie en musique called Hippolyte et Aricie (1733), is typical of what he had to offer. Although it retained the outer trappings of the Lully prototype (the five-act structure, the ballets, the obsequious Prologue, etc.), in almost every musical area it added layers of interest. The recitatives tend to increased orchestral elaboration, the harmonies are denser and more complex, and – in particular – the monologues of individual characters can become extremely elaborate. Voltaire reported Rameau putting it all in a nutshell: ‘Lully needs actors but I need singers.’¹⁶ We see this kind of opera at its best in the celebrated finale of Act 4 of Hippolyte, in which Phèdre (a character tellingly related to Racine’s famous heroine) laments what she supposes is the death of Hippolyte (a monster has just emerged from the sea and carried him off). Phèdre’s lament, couched in a highly charged recitative, is punctuated by choral laments of great poignancy – a celebration of orchestral effect, declamation and harmony that is as far as can be imagined from the endless melodic hegemony of opera seria.

The late twentieth century has seen the warring parties of the eighteenth-century reform movements find a peaceful co-existence. Rameau’s operas began to be staged in greater numbers in the 1990s, more or less in parallel with the Handel revival, as if some French versus Italian querelle had been reincarnated in another age. As mentioned earlier in connection with Handel, the necessary accompaniment of this Rameau renaissance has, up to now, been that directors and choreographers will present tragédie lyrique as a post-modern spectacle of glorious estrangements. Although usually underpinned by historically informed musical performance, the now-unfamiliar operatic aesthetics become the basis for a stage festival of alien-looking delights and outlandish fashion statements. What is generally missing is a sense that the Italian and French genres were thought so different in the eighteenth century, their partisans as irreconcilable as Callas and Tebaldi fans in the 1950s. These differences are erased by the amiable bricolage of contemporary opera production – by the pleasurable sense of alienation postmodern directors so reliably provide. In the eighteenth century, the perceptible differences, the philosophical passion they inspired, explain why cross-fertilizations were so refreshing. What French tragédie lyrique infused into opera seria was an alternative sound: not just the pace and potential of a more thoughtful recitative, but the massed sound of choruses and ensembles and, of course, the vacations from singing that are provided by dance numbers.

A word more on vacations. At one time or another, everyone who has attended an opera feels the need for relief from singing. There is no use denying this. That is why all the gambling, eating, chatting and chess-playing endured for so long. Composers were sometimes wise enough to provide the relief, and were at other times constrained so to do by tradition and custom. In the eighteenth century, ballet interludes were a universal feature of serious opera in French, a genre that began life in Molière and Lully’s comédies-ballets. Indeed, the tradition demanding respite through dance was a recurring refrain in French opera production as late as the late nineteenth century. In 1861, when Wagner’s revised Tannhäuser (premiered in Dresden, 1845) appeared at the Paris Opéra, the composer was asked for a bigger and better ballet scene, and did not fully comply, merely expanding a scene in Act 1 rather than supplying the usual Act 2 blockbuster. His opera was duly howled down by members of the influential Jockey Club (lovers of traditional ballets) and withdrawn after three disrupted evenings.

In French opera danced interludes occurred within the boundaries of the drama; in Italian opera seria dances were entr’acte or post-opera entertainment, by and large dramatically unconnected to the opera that surrounded or preceded them. Dances also pepper opera buffa in the eighteenth century and beyond: sometimes, as with Mozart’s Don Giovanni, they are worked into finales by making dancing part of the plot, with the characters whispering asides or screaming for help over the strains of an onstage orchestra. Fundamentally, though, dance in opera is almost always a ‘divertissement’ – a diversion and distraction. Like any elaborate visual tableaux (and like today’s CGI special effects or car-chase sequences in films), they are a place to admire the view without the burden of dialogue. And dance was one of several aspects of French opera that Gluck – also an experienced ballet composer – imported with enthusiasm into his Italian reform operas.

In accordance with the French traditions, the dance scenes in Orfeo ed Euridice had some dramatic relevance, consisting of funereal rites for Euridice or demonic motions for the Furies in Act 2. Their choreographer, Gasparo Angiolini, was himself a reformer of dance; he claimed to have reconstructed the funereal rites from Virgil, thus adding a patina of historicity to this already self-consciously neo-classical project. To this day, Gluck’s Orfeo attracts choreographers, now not just as wranglers for the dances but as opera directors in charge of the whole. The German choreographer Pina Bausch staged Orfeo at Wuppertal in 1975, with solo dancers enacting Orfeo and Euridice. The unfortunate singers were frequently relegated to moping at extreme stage right or left, singing far from the spotlight. In 2007, Mark Morris directed the Metropolitan Opera’s new production, foregoing dancing doubles for the principal singing roles – thus keeping singers at the centre where they belong – while morphing the chorus, who are after all those most likely to be thrown off-kilter by having to move, into a phantasmic collective that looks down from above on the stage action. The allure of Orfeo for choreographers is not hard to understand, since Gluck composed brilliant dramatic music for the opera’s danced portions, from the almost painfully chromatic sighs of the first act’s minuet-lament to the melody-annihilating repeated scales of the Furies. It is hard to overestimate the force of this alternative soundscape in an Italian operatic realm dominated for so many generations by the monochrome tones of the single, high voice.

BETTER BEHAVIOUR

Calzabigi wrote with disciplinarian certainty – in Gluck’s name – about what had gone wrong with Italian opera seria, and what Gluck had done to put it right:

I decided to divest it wholly of all the abuses which, introduced either by the ill-considered vanity of the singers or by the expressive indulgence of the composers, have for so long disfigured Italian opera. … I thought I would restrict the music to its true function of serving the poetry in the expression and situations of the story, without chilling it with useless and superfluous ornaments. … I decided not to stop an actor in the heat of the dialogue, forcing him to wait out a tedious instrumental introduction, nor to stop him in mid-sentence over a favourable vowel, nor to display the agility of his fine voice with a lengthy ornamental passage, nor to let the orchestra give him the time to catch his breath for a cadenza. I did not feel obliged to hurry through the second part of an aria, though it was the more impassioned and significant, in order to be able to repeat four times the words of the first part, finishing the aria where the sense was left unfinished, all so the singer might have the leisure to show the many ways in which he can vary a ornamental passage at will.¹⁷

So much passion directed against the singers! This excerpt again comes from the preface to the score of Alceste, published in 1769, but it represents precepts that were also behind Orfeo in 1762, an opera which embodies the ‘beautiful simplicity’, ‘clarity’ and natural expression Calzabigi mentions as Gluck’s goals.

Calzabigi’s preface to Alceste both echoes and prefigures the groans of many. For example, Francesco Algarotti, in his famous Essay on Opera (1755):

arias are overwhelmed and disfigured by the ornaments with which they are increasingly embellished. The ritornellos that precede them are much too long and often superfluous. In arias expressing rage, for example, verisimilitude is stretched to the breaking point: how can a man in a fit of rage wait with his hands in his belt until the aria’s ritornello is concluded before venting the passion seething within his heart?¹⁸

Or Antonio Planelli, in Opera (1772):

If any theatre music rich in pathos is examined, it will be found to contain fewer notes than even one of those deadly trills that are so fashionable today. Furthermore, it can never happen that a song made up of many notes will produce pathos in the theatre, or strengthen the emotional charge of the words.¹⁹

Or Rousseau, in his Confessions (1770):

One day at the Teatro San Giovanni Crisostomo I fell asleep more deeply than if I had been in my own bed. The noisy and brilliant arias did not wake me up.²⁰

As we shall see, serious opera in Italian would fail to live up to these austere anti-virtuosity prescriptions, either at the time or in the decades to follow. The appeal for restraint and the resentment of vocalism has an anti-Italian flavour, and few Italian composers rushed to the barricades in its wake: Rossini and his constant outpourings of vocal excess, an all-conquering European vogue, followed Gluck by no more than fifty years. However, in their project for Orfeo, Gluck and Calzabigi were fortunate in being able to recruit a famous singer to their cause, in some senses making Orfeo a genuine collaborative effort. The singer who created the role of Orfeo in 1762 was Gaetano Guadagni (1729–92), an alto castrato and much-feted virtuoso whose voice had the strong low range of many castrati. But Guadagni was said to look more normally male, something affirmed (at least in story) by his many romantic conquests of women. Although we know little about the circumstances of Orfeo’s composition and first performances, one enduring story is that Guadagni was willing to co-operate with Gluck because he could thereby lend his genius to a vocal idiom that was atypical for a castrato: a lyric, controlled style with, as Calzabigi/Gluck noted in their later Preface, very few opportunities for improvised ornamentation. Burney described Guadagni’s gestures as ‘so full of grace and propriety, that they would have been excellent studies for a statuary’;²¹ and tells us that he eventually fell out of favour with the British public because (uncastrato-like to a fault) he refused to bow and repeat arias, feeling that such behaviour interrupted the seriousness of the onstage drama. Guadagni had in fact been trained in theatrical gesture and expression in London, and by no less a personage than David Garrick; his bona fides as a singing actor were impeccable.²²

Many myths were made by means of such anecdotes, with cultural uneasiness about opera seria and its legitimacy or frivolity added to the mix. A famous rhapsody to the historical importance of Orfeo, and to Guadagni’s role in its success, was written by the German music theorist and critic A. B. Marx in 1863, a century after the opera’s first performance. This passage, from Marx’s book on Gluck’s operas, is notable because it is one of the few sources for a now-popular fantasy that castrati had particularly forceful – as opposed to skilled or flexible – voices:

By means of the operation performed on castrati, the development of the vocal organs – more accurately speaking, the larynx – was arrested. It is this development, beginning in puberty, that converts the discant or alto voice of a boy into a tenor or bass voice. But the further maturation of the body … went on as usual. The chest and lungs of the boy attained the power and flexibility that they have in men; the vocal cavity attained the size and resonance of the mature male. Male force resounds with utmost violence within vocal organs that have remained boyish. In comparison to a boy’s voice, a superior increase of volume is obtained when this sound enters the mature vocal cavity. This is the reason for the greater power of the castrato voice, and for its violently penetrating quality. … No female alto is, in terms of voice quality, capable of replacing the castrato.²³

Again, so much anxiety! So many references to men, male, manly, mature, strength, power, size, violence and penetration. But Marx could not have heard Guadagni (who died in the 1790s); like everyone of his era, he had little opportunity to experience castrati except perhaps in Italian church choirs, which were their last refuge. More importantly, this polemic had a precise point. Marx, a virulent misogynist, was protesting without saying so explicitly against the mezzo-soprano Pauline Viardot’s recent, spectacular success in the title role of Orphée (Gluck’s 1774 French version of Orfeo), which had been revived for her by Berlioz in Paris in 1859.

What were the musical results of the Orfeo experiment? For one thing, Calzabigi and Gluck's aesthetic reclaimed declamation – recitar cantando – as a source of expression, both in recitatives and (more radically still) in arias. In the extended third-act recitative in which Euridice finally persuades the tormented Orfeo to turn, and then falls dead to the ground, Orfeo’s intense prose is vividly depicted in the orchestra (string tremolos, rapid changes of texture), while Euridice’s sudden fall makes barely a dent in the musical fabric. When employed within arias, the declamatory ideal here is very far from the coloratura flights of traditional opera seria heroes and heroines, in which outpourings of pure vocal noise constantly threaten to rupture verbal meaning. Just as Calzabigi’s libretto turned away from the metaphor-laden poetry of Metastasio in favour of more direct syntax, so Gluck aimed throughout for a more speech-like style in which individual syllables rarely receive more than a single note. The hero’s two famous arias – ‘Chiamo il mio ben così’ (Act 1) and ‘Che farò senza Euridice?’ (Act 3) – both embrace this simpler idiom. These reforms did not of course occur in a vacuum. Gluck had been a jobbing opera composer for nearly twenty years before he wrote Orfeo, and while his earliest works were firmly in the Metastasian mould, he had also – as musical director of the French theatre in Vienna – written several opéras comiques, in which such simple, almost folk-like declamation was the norm. Indeed, ‘Che farò’ was adapted from one of Gluck’s opéras comiques, and ‘Chiamo il mio ben’ was so beloved by the opéra comique composer Philidor that he stole it for one of his own productions in Paris.

The primal scene in Orfeo is a modest showpiece for the lead singer. The confrontation in Act 2 between Orfeo and the Furies who guard the entrance to the underworld is perpetually astonishing for its effect of terror – the earliest example of real uncanniness in opera. Hell’s denizens dance wildly, first very slowly, then for a few bars at breakneck speed, before singing a chorus that repeatedly demands an answer to the same miserable question: what mortal dares come to this terrible place? They sing every part of this question in the same triple-metre rhythm, with three long, even strokes followed by the snap of three short ones:

Chi mai dell’Erebo

Fra le caligini

Sull’orme d’Ercole

E di Piritoo

Conduce il piè?

[Who amid the mists / of Erebus, / in the footsteps of Hercules / and Pirithous / treads here?]

After dancing again, they repeat the question. The same rhythm recurs, with added emphasis concerning Hell’s unpleasant details, the wrath of the Furies, the howling of Cerberus – instrumental howls arising in concord in the orchestra, with the cellos playing loud, emphatic slurs. The message is simple. Hell is noisy. It is also repetitive, with sung speech reduced to a single rhythmic tic, as if the hopelessness of those condemned to toxic and eternal labour has been given musical shape. When Orfeo begins to sing against this noise, a second orchestra, made up of harp and plucked strings, becomes the sound of the lyre he carries. He sings his plea twice over in its entirety, then repeats lines 2–4 a third time:

Deh! placatevi con me,

Furie, larve, ombre sdegnose,

Vi renda almen pietose

Il mio barbaro dolor!

[Alas! Placate with me / furies, ghosts, / indignant shades, / at least make them feel compassion at / my cruel sorrow!]

Unlike the underworld guardians, his gift is to be able to effect change, so when the poetic text is repeated (the Furies keep yelling: ‘No!’) the music takes off in different directions: there is a melodic and harmonic variation on the first pass, but never its literal repetition. Gluck has composed a solo that mutates, putting into notes the idea that Orfeo is improvising, trying different flavours of the melody and, with its high-to-low sweeps, gauging which is most effective. As so often with onstage music making, the onstage audience response tells us what our reaction should be. The Furies’ repeated, loud, single-pitch ‘No!’ finally softens into a four-part harmony ‘No’, voices joined in a consonant, major-mode chord that could, for an instant, be the friendly hum of a barbershop quartet. It is a first sign that the Furies will yield. Orfeo’s vocal line avoids elaborate melismas – no syllable of text is allotted more than two notes. But if we consult recorded performances of the scene, we will hear that most Orfeos add some personal embellishments, at least to the final cadence in this section. It’s as though, despite Calzabigi’s polemics against singers, the illusion of improvisation and autonomy written into Orfeo’s melody encourages free-spirited additions to the letter of the text. After this first salvo, with the Furies still not convinced, Orfeo goes on with a second poetic verse and entirely new music. After yet another refusal, there is a third: Orfeo seems to dispose of limitless imagination. What is saddest about the scene is that, even when they are finally moved by his singing, the Furies never break free of their rhythmic shackles, but are condemned to sameness. ‘Let the Gate open, groaning on its black hinges!’ Orfeo is allowed to pass, with this final reference to Hell’s loud noise taking the form of quietness, a long fade in C minor, piano to pianissimo.

This scene was famous from the outset, and its sounds resonated through operatic and instrumental compositions for decades to follow. The orchestral beginning to ‘Deh! placatevi con me’ – a C-minor colour, arpeggiated triplets beginning in medias res on B-D-G, then C-E♭-G before moving on from there – that beginning is echoed in the tragic chorus ‘O voto tremendo’ in Act 3 of Mozart’s Idomeneo. It also haunts the first number of Don Giovanni, when what starts unequivocally as an opera buffa abruptly mutates into tragedy. The Commendatore is dying, Don Giovanni and Leporello lament in shock, and we get the same minor mode and triplets, the same beginning in medias res. And then, via that Mozart trio, it migrates into the first movement of Beethoven’s ‘Moonlight’ Sonata, Op. 27, No. 2, whose triplets seem to echo Mozart’s number and Gluck’s, and weave them (a semitone higher) into a famously atmospheric piano piece. Nor was this the only occasion on which Mozart remembered Orfeo. In Act 3, Euridice’s fatal moment is written as an elaborate orchestral recitative, in which she berates Orfeo for his refusal to look back at her. Gluck’s Euridice, passionate harridan that she is, having attracted his glance ‘rises up with tremendous force, then once more sinks to the ground’, saying ‘io manco, io moro’ (I faint, I die). In Act 1 of Don Giovanni, when Donna Anna discovers her dead father, the Commendatore, she faints to almost identical sounds.

Did Orfeo transfix people other than Mozart? Were audiences struck silent, and did this new kind of opera endure? Burney in 1772 reports Gluck lamenting how much labour had been involved in producing Orfeo in Vienna, how resistant the entrenched tastes and habits had been.²⁴ Gluck was indeed a polarizing force. One of the great musical pamphlet wars of the 1770s, Paris-based and utterly contrived, fed on a supposed rivalry between Gluck and Niccolò Piccinni (1728–1800). Yet this conflict had little to do with Orfeo, still less with old-style high-wire coloratura in opposition to Gluck’s austerity. Piccinni had written a single comic-opera hit, La Cecchina, ossia la buona figliuola (Cecchina, or The Good-Natured Girl, 1760), and his preferred style was melodious and simple. It was Piccinni’s acoustic lightness that formed the antithesis to Gluck’s gravitas, to Gluck as the composer of whom the historian Marmontel would quip, ‘one must admit that no one has ever made the trumpets rumble, the strings whirr, or the voices bellow as he’.²⁵ From what we know of audience reactions to Gluck’s Orfeo, however, it would appear that there was indeed a significant difference between riotous, indifferent Italian spectators and audiences elsewhere. Calzabigi wrote of Italian theatregoers in 1778, ‘how could one want to present a Greek tragedy in front of such a deranged audience?’²⁶ But in Vienna and France the situation had been different. There Gluck had accomplished a revolution in operatic form, a new way of writing opera that had the power to evoke unprecedented rapture and absorption in listeners. In the 1770s, in Paris, as evinced by weeping and solitary and silent devotion to the stage, a new ‘depth and intensity’ of experience ‘inconceivable to earlier audiences’ was being laid at Gluck’s door.²⁷

By the first decades of the nineteenth century, in Austria and the German states, Gluck’s mature operas enjoyed a reputation for genius that Mozart’s were only beginning to match. E. T. A. Hoffmann’s story ‘Ritter Gluck’ (Chevalier Gluck, 1809) is a succinct miniature within this landscape, and gives a sense for the reverence in which Gluck was held among those most repulsed by the Italianate in opera. In Berlin, the narrator recounts, Gluck’s operas are constantly being staged, and an eccentric, old-fashioned figure wanders the streets complaining about their performance:

Once I went to hear Iphigenia in Tauris. As I entered the theatre, I heard the orchestra playing the overture to Iphigenia in Aulis. I thought, hmm, so they’re playing this Iphigenia. But then I was astonished when the Andante and the subsequent Storm began [the first choral number from Iphigénie en Tauride]. A quiet ocean, then a storm, then the Greeks are washed up onto the beach – that’s the opera! What! Did the composer write in the score that you could do anything you want with it, like some bagatelle for trumpet?²⁸

The eccentric, in typical Hoffmannesque fashion, turns out to be the ghost of Gluck himself, condemned to wander unshriven for reasons mysteriously unexplained, since the real Gluck committed no greater sin (we are told) than overindulging in liqueurs towards the end of his life. What’s interesting about the story is the way in which Gluck’s reformist polemic, particularly his assertions about compositional authority against the pressures and exigencies of performance, are converted here into the pseudo-Gluck’s fury against an opera house that mixes and matches overtures. But such mixing and matching, substituting numbers or re-arranging them, was a practical fact of operatic life (Gluck’s included). Although Hoffmann didn’t know it, Gluck himself had reused large stretches of his ballet music for Semiramis (1765) in Iphigénie en Tauride, and recycled other music from his ballets and opéras comiques in his serious operas. Mozart, too, was happy to provide substitute arias for his (and other composers’) operas to please new singers at revivals. High-minded civilians and amateurs – and this includes many of the philosophers who debated about opera during the Enlightenment – have complained about what practising musicians do in realizing opera for a very long time. The same offended screams, for example, greeted Cecilia Bartoli when she sang two of Mozart’s substitute arias in Le nozze di Figaro at the Metropolitan Opera in 1999. Critics were determined that she should be chastised for supposed sins that endured an ephemeral few minutes on stage. The spirit of Gluck, if not his actual practice, was among them.

HIGH PASSION AND EFFECT

In Act 1 of Iphigénie en Tauride, Iphigénie tells her assembled maidens about a terrible dream: she is back in her father Agamemnon’s palace, and sees him fleeing from his murderer, who is her mother Clytemnestra. She is then impelled to kill her brother Orestes. The House of Atreus has its dreadful history summarized in a recitative that begins with an amazing sound, a repeated F♯, pianissimo in the brass and winds: a claxon going off loudly, but very far away or very long ago. As Iphigénie recounts meteorological details – ‘fire burned in the air, and lightning fell on the palace in bursts, embracing and devouring it!’ – the orchestra loses its musical mind; tonalities succeed one another according to their disquieting effect rather than any functional sense; with the F♯ alarm bell, now loud and clear, returning for the palace’s demise.

Orchestral activity like this is typical in French tragic recitative, and forms a contrast to the held-back arias, where the very fact that there was only one character – played by a potentially wayward soloist – seemed to inspire composers to keep characters under a firm grip. Gluck unbuttoned things for duets – the railing and remorse in Orfeo’s duet with Euridice, or the extraordinary scene for Oreste and his friend Pylade in Iphigénie en Tauride, with its funereal C minor. The two men declare their willingness to die for one another, with reproaches that speak of a ‘passionate friendship’ that is now, in the twenty-first century, perhaps unavoidably staged as homoeroticism (the 2003 New York City Opera production was famous for its semi-nude Oreste–Pylade duo). Passionate duets for the ‘wrong’ pairing are one of opera’s great complications, numbers in which voice-characters declaring love via passionate music are laid on to plot-characters whom taboos must keep rigidly separate.

That last formulation sounds ultra-modern for Gluck but will (as we shall see in the next chapter) become almost routine for Mozart. It will be mostly in opera buffa that Mozart’s propensity for such confusion reaches its height. Opera seria, with which Mozart was principally involved in his youth, tended to see him at his most conventional. The breakthrough came with Idomeneo (1781), a tale of crossed royal destinies during the Trojan Wars, in which a Trojan princess, Ilia (soprano), has been captured by Idomeneo, king of Crete (tenor), and has fallen in love with his son, Idamante (soprano castrato). There is no doubt that this opera displays Mozart’s greatest debt to Gluck: numerous moments plainly adopt the principles of reform opera, with a prominent use of orchestral recitative, frequent scenic marvels, elaborate ensemble pieces (especially the famous Act 3 Quartet, ‘Andrò, ramingo e solo’) and a dynamic use of the chorus. This debt is so obvious that it has become something of a cliché, also carrying with it a suggestion that the composer came of age operatically with Idomeneo precisely because he imbibed a good dose of Gluckian discipline. There’s some truth in this, but Idomeneo has another story to tell. The opera is also full of solo arias: numbers whose musical elaboration is unprecedented; numbers which celebrate, unashamedly so, the glories of untrammelled singing that Gluck’s reforms were so eager to banish.

Given the profusion of arias, it is no surprise to learn that Mozart went to some pains to ensure that Idomeneo was carefully tuned to the capabilities of the first cast, whether or not the results might make ideal dramatic sense. The most obvious case is with Idomeneo himself, a part written for an ageing tenor of the old school, Anton Raff, which Mozart tailored carefully so that Raff’s diminishing capabilities and old-style sensibilities would not be stretched too far. So at the centre of this reform drama stands (and sings) a relic of the past, a bewigged Metastasian presence who, in Mozart’s words, ‘stands around like a statue’,²⁹ and sings like one too. This is doubly disconcerting in that many of the other arias tend in quite the opposite direction – offering a kind of musical elaboration that looks forward rather than back.

One of the best, and most complicated of these new-style arias is Ilia’s ‘Se il padre perdei’ from Act 2, in which Ilia, with uncommon, unsettling sensuality, tells Idomeneo that he must now be her adopted father. We are lucky enough to have a Mozart letter about this aria, with instructions to be passed on to the librettist, Giovanni Battista Varesco. Mozart is extremely clear about his priorities. The original text he had been offered contained an ‘aside’ (a whispered comment to the audience). Mozart wants this removed. Asides, he says, are fine for dialogue (he means recitative of some kind), ‘but in an aria – where the words have to be repeated – it makes for a poor effect’. Another way of saying this would be that arias are intended for music: verbal changes of register shouldn’t intrude. What is also clear is that Mozart already had music in mind for this aria: he wants it to be ‘natural and flowing’, and what’s more to have elaborate instrumental accompaniment, so that the piece will also be suitable for concert performance.³⁰ We are already some considerable distance from the musical severities of Gluck’s style.

The aria itself bears this out to a remarkable degree. The words that eventually emerged would suit an old-fashioned da capo form. One can imagine a gentle, perhaps pastoral ‘A’ section (the first verse); some musical contrast for the ‘B’ section in which Ilia remembers her past problems (the second verse), and then an ornamented reprise:

Se il padre perdei,

La patria, il riposo,

Tu padre mi sei,

Soggiorno amoroso

È Creta per me.

Or più non rammento

Le angoscie, gli affanni.

Or gioia, e contento,

Compenso a miei danni

Il cielo mi dié.

Parte

[If I have lost my father, / my homeland, my rest, / you are a father to me, / an amorous sojourn / is Crete for me. / Now I no longer remember / the anguish, the pains. / Now joy, and happiness, / consolation for my injuries / Heaven gives me. / She leaves]

Mozart’s musical realization of this conventional text adds new complications. It has, for a start, four so-called ‘obbligato’ wind instruments (flute, oboe, bassoon and horn), all of which require their own musical space, trilling away individually and in ensemble as the singer prepares herself. This instrumental elaboration is then grafted on to an aria that has little trace of the da capo form. Instead, the entire text is stated twice, and treated with great fluidity. For example, both musical stanzas are interrupted by a new, questioning woodwind figure just before the ‘soggiorno amoroso’ (amorous sojourn) line, and this new idea seems to precipitate the sudden, disconcerting dive into the minor mode and an injection of vocal sighing figures as Ilia recalls her past ‘angoscie’ (anguish) and ‘affanni’ (pains). What emerges from such sudden and unpredictable changes of mood is what we might now call emotional complexity, a sense in which the pastoral surface so simply depicted at the start can have complex undercurrents. As if to stress this further, Idomeneo’s recitative that follows is launched by a shadowy, minor-mode repetition in the strings of that questioning woodwind figure, as if the aria did not, after all, completely succeed in quelling its own inner doubts.

Mozart had been in Paris at the height of the pamphlet war between Gluck’s supporters and those of his mild-tempered Italian rival, Piccinni – a conflict that one venerable musicologist characterized as that between an agate and a sponge.³¹ Although Gluck’s influence on Idomeneo is unmistakable, Mozart’s simultaneous embrace of expansive arias and untrammelled lyricism would ultimately place him among the Piccinnistes – a fact often overlooked by opera historians who wish to build robust, reform-led narratives peopled by strong figures. For this and other reasons it can be difficult to know where to place Gluck’s operatic reform. The composer’s association with an aggressively articulate and self-promoting avant-garde, Calzabigi at its head, may lead us to overlook the fact that his revolution was not much imitated. During his final decade (he died in 1787) he was a venerated figure, but one who had few immediate followers. His last Parisian opera (Echo et Narcisse, 1779) was a failure; he left the capital, dogged by attacks from the Piccinnistes, to finish his days in Vienna. To understand the importance of Gluck as a symbol of operatic restraint we must look to the nineteenth century and away from Italian opera, above all to Berlioz and Wagner. As we shall see in the next chapter, the crowning glory of the late-eighteenth-century opera did indeed centre on Vienna, and precisely during those years in the 1780s when Gluck was a resident elderly celebrity. But it came in a very different brand of opera, one hardly touched by Gluck’s great musical manifestos.