1

Introduction

Opera is a type of theatre in which most or all of the characters sing most or all of the time. In that very obvious sense it is not realistic, and has, through most of its 400-year history, often been thought exotic and strange. What’s more, it is almost always ridiculously expensive to stage and to attend. At no time in history has society at large managed to sustain easily opera’s outrageous cost. Why, then, do so many people love it so deeply? Why do they devote their lives to performing it, writing about it, watching it? Why do some opera fans travel the world to watch a new production or hear a favourite singer, paying huge sums for the fleeting privilege? And why is opera the one form of classical music still developing significant new audiences in spite of the fact that, in the last century or so, the stream of new works that was once its lifeblood has dried to a trickle?

These questions are mostly about opera as it is today: about what opera has become at the start of the twenty-first century. In what follows we will have much to say about the history of our subject, about the ways opera has developed during its 400-year journey towards us; but our emphasis will also be on the present, on the effect opera continues to have on audiences around the world. Our aim is to get to grips with an art form whose most popular and enduring works were almost all written in a distant European past, and so were created in cultures very different from our own, but whose hold over many of us – whose meaning in our lives today – is still palpable. Opera can change us: physically, emotionally, intellectually. We want to explore why.

WORDS AND MUSIC

It is often said that opera, being in essence sung theatre, involves a battle between words and music. Whole operas have been written about this supposed battle. One of the more famous (at least in the history books) is Antonio Salieri’s little comic work, Prima la musica, dopo le parole (First the Music, Then the Words), which was premiered in 1786 amid the opulence of the Orangerie, a deluxe greenhouse-cum-conservatory at the Schönbrunn Palace in Vienna. A poet and a composer must complete an opera within four days. The poet complains about the indignity of being asked to concoct words to fit pre-existing music; the composer replies that the poet’s angst is trivial – no one cares about the words anyway. The basic terms of this tussle – the battle-lines between words and music – are drawn up again and again in the history of opera: Richard Strauss’s Capriccio, premiered at the Munich Staatsoper during one of Germany’s darkest years, 1942, deals with the same theme in an understandably more world-weary manner.

On the surface, this sense of a continuing competition might seem strange: the words, after all, supply the basic story; the music then gives that story added impact and aura. While it’s little wonder that operatic poets and composers occasionally get into arguments (for one thing, the relative prestige of their professions has changed a great deal through the centuries), they need each other and always have done. As soon as we press a little, though, the reason why the words/music opposition is often so tense and anxious becomes clearer. In a libretto there are at least two separate domains. The first involves the narrative element, basically the plot and its characters; the second involves the representation of this narrative in text, in specific (almost always poetic) words. While the first domain has tended to remain stable when operas are performed, the second is often highly variable. Operatic poetry, the libretto text, is rarely thought so exquisite as literature that it commands reverent handling at every turn. There is, for example, a continuing and heated contemporary debate about whether we should hear opera in translation, in the language of the audience rather than the language of the original work. Those who want translations are in effect arguing that the first of the libretto’s domains, the one involving plot and characters, is more important than the second, in which individual words reside. Muddying the waters of this debate, though, is a further circumstance, one that Salieri’s composer points out rather cruelly: that words set to music tend to lose a good deal of their semantic force, and that this loss can be at its most extreme in opera.

The reasons for this loss are many. The musical surroundings – the instrumental forces, as well as the kind of music they play – can overwhelm as well as accompany the human voice. Composers can also use words as a kind of roughage: in coloratura singing, cascades of vocal ornament on an open vowel reduce the verbal element to mere sound, almost like an instrumental tone colour. But the voice itself, specifically the operatic voice, contributes in its own way to the disappearance of linguistic sense. Opera singers, given the requirements for volume and pitch, must at times produce sound in ways that will force verbal articulation to take a back seat. The words will be obscured no matter what the language of the libretto. Even in languages we know well, listening to opera’s verbal text can be frustrating: single words or phrases – ‘la vendetta’, ‘das Schwert’, ‘j’ai peur’, ‘I am bad’ – will swim briefly to the surface of comprehension, but what follows will again be drowned in musical activity. Avid listeners may memorize long stretches of libretto poetry and then, during the act of listening, supply from memory the words that lie underneath the babble. But even this doesn’t mean that they necessarily understand the words they have by heart, just as singers, in some cases, learn their parts phonetically without having more than a general sense of what they’re actually saying when they sing.

Some singers are much better at verbal articulation than others – in German, Franz Mazura comes to mind, in Italian, Giuseppe di Stefano – but the counterexamples, singers who are famous for their wooliness in this department, are probably more numerous and include revered superstars such as Joan Sutherland. (It’s no accident that our examples of clear-articulators are both men and the non-articulator is a woman renowned for her elaborate feats of coloratura singing: the higher the voice and the more elaborate the ornamental flights, the less chance there is of understanding the words.) Some listeners don’t care if words are fuzzy; but for others the sense that language is being enunciated with intent, passion and conviction is crucial even if they don’t always recognize or understand what they hear. The latter position is powerfully articulated by the historian Paul Robinson, who points out that while the specific words of the libretto often seem of little moment among the pleasures opera can bring, it is still true that three hours of drama in which the characters sang ‘la, la, la, la’ would be intolerable.¹ In other words, it matters a great deal that verbal meaning is embedded somewhere in opera, even if the listener can’t always hear exactly what the words are at the moment the singing takes place.

The debate about the role of words in the experience of opera has taken a further, specifically contemporary form in today’s endless debates about super- or surtitles. When opera houses perform works in the original language, should they provide audiences with translations (displayed above the stage or in the back of the next row’s seat) so that all the words can be clearly understood? Many think this is plainly an advance, giving us back the words’ second domain, their specific meaning. But others remain passionately opposed, arguing that too detailed an exposure to the semantic aspect of words will take attention away from what is most important in the operatic experience. As the British critic Rodney Milnes put it, ‘You go to opera to listen and to watch, not to read.’ Before surtitles a neophyte was often expected to work: knowing something of both the story and the poetry beforehand was deemed a necessary preparation for taking pleasure in the operatic experience (as Milnes wrote, ‘You read beforehand’). But Milnes represents an historically limited attitude. For one thing, reading the libretto during the performance was standard practice in many theatres from the seventeenth to the nineteenth centuries. As Samuel Sharp wrote from Naples in 1767, complaining of a lack of candles in the auditorium, ‘dark as the boxes are, they would be still darker, if those who sit in them did not, at their expense, put up a couple of candles, without which it would be impossible to read the opera’.² This applied even more forcibly to the experience of foreign works: when Handel’s Italian operas were performed in eighteenth-century London, dual-language libretti were published in conjunction with productions. What is more, reading the libretto was just one of the alternative activities available to the operagoer of the past; there was also gambling, chess-playing, eating, talking, ogling one’s fellow attendees and, in the so-called loges grillées (shuttered boxes), quite possibly things a gentleman never mentions. In other words, an all-absorbing devotion to what was happening onstage – the attitude represented by ‘you go to opera to listen and to watch’ – was not the historical norm and, for much of opera’s history, the experience of the event was not exclusively focused on the stage world, or indeed even on documents related to that world such as a libretto translation in a book.

This debate about understanding the words and the plot in opera can even take on ethical overtones. The notion that pleasure is based on or at least enhanced by knowledge (and hence a prior expenditure of labour) crops up again and again in opera’s history. When Carl Maria von Weber’s Der Freischütz was first performed in French, in Paris in the early 1840s, Richard Wagner thought the Parisians so incapable of uninformed delight in the work that he wrote an elaborate essay instructing them on the opera’s background, plot and cultural significance.³ In other words, the fact that the audience in Paris were hearing an opera in French was, to Wagner, no guarantee that they would understand the words, thus understand the story, thus understand the work’s meaning and thus take delight in the whole package. Surtitles seem to short-circuit preparatory labour of the kind Wagner prescribed; they provide – in the form of verbal comprehensibility – easy access to the story, and hence a kind of instant cultural connection, a sense of belonging.

Perhaps, then, understanding specific words is not as critical as it might seem in the experience of opera. Opponents of surtitling claim that such understanding could even become a distortion, counterproductive of the theatrical absorption or attention to music and voice that are a specifically operatic ideal. They may have a point; but again operatic history will not always support them. In what follows, for example, we will have a great deal to say about absorption and detachment, but it will soon be clear that, historically, theatre audiences have been prone to these experiences in different ways at different times. Certain aspects of opera are now routinely – and uncritically – assumed to transfix the soul (usually the musical components), while others are assumed to produce disenchantment or distraction (‘bad poetry’, or an unpleasing performance). Hence, perhaps, the persistence in different periods of the idea of opera as a Gesamtkunstwerk – Wagner’s famous word – meaning a total work of art, multi-media (words, music, scenery), the simultaneous experience of several registers and kinds of art-work.

We can examine the case further by looking at some examples of real violence to the integrity of the word–music bond. We’re not talking here about opera performed in translation – even though the very practice goes some way to demonstrating that the words of the original libretto often count for little. There are, though, well-known instances of substitute aria texts being written for existing music, and succeeding indecently well in their transposed state. This shows that at least some operatic music is potentially mutable and non-specific in its content, and that music which would seem to correspond to an aria text about (let’s say) a lost beloved – to express the emotion to perfection, to set the specific words with subtlety and art – may work just as well with a different aria text about something else. There are notorious examples in Rossini’s operas, perhaps the most famous being his revision of his Italian Mosè in Egitto (1818) for the Paris stage, as Moïse et Pharaon (1827). In Mosè the soprano, Elcia, sings a cabaletta ‘Tormenti! affanni! e smanie!’ about the torments that afflict her wounded heart. But when Rossini recycled the piece for Moïse (admittedly with some changes) he gave it to another character, one joyfully celebrating a happy turn in the plot: the opening words are ‘Qu’entends-je! ô douce ivresse!’ (What do I hear! oh sweet bliss!). Those tempted to argue that this kind of switch would only be possible in Italian opera d’un certain âge need to explain how music that Wagner sketched in 1856–7, which ended up in Siegfried Act 3, was labelled in the sketch ‘Act 3. Or Tristan’, indicating that this particular music might do perfectly well for either piece.⁴ But even substitute aria texts for identical music do not go as far as the most radical cases. In the early nineteenth century, the libretto of Mozart’s German opera Die Zauberflöte (1791) went through an unpopular phase, being thought inane or ludicrous. The music, though, was standing the test of time: indeed, Mozart was already beginning to be canonized. The solution was to graft on to the music an entirely new libretto: new narrative, characters and words. Anton Wilhelm Florentin von Zuccalmaglio (1803–69), a literary polymath whose work included several volumes of poetry and folk song collections, wrote new libretti for several Mozart operas. In 1834, he turned Die Zauberflöte into Der Kederich (the ‘Kederich’ is a cliff high above the Rhine). The opera now takes place on the Rhine and involves water nixies, Nibelungenlied (The Song of the Nibelung) references and ‘Rudhelm’ (he used to be called Tamino), a Crusader back from the Holy Land. The Speaker (the Priest who enlightens Tamino in Act 1) becomes ‘Sibo, Lord of Lorch’, and Pamina, now renamed ‘Garlina’, is his daughter rather than the Queen of the Night’s. The Queen becomes ‘Lore von Lurlei’, a sylph who floats around in white veils.⁵ On the whole, the alternate libretto works very well. Are those who thought up or those who enjoyed this new opera to be condemned, or did they know something that we’ve now lost about how opera can communicate?

That question is interesting most of all because it invokes an historical era, now almost 200 years past, whose cultural attitudes retain the power to surprise us. We need to ask why we are so shocked: why our present fear of loss, our cultural pessimism, has turned operatic performance into an activity policed by a reverence for the work as a well nigh sacred object – a reverence in almost all cases not present at the time it was created. Addressing such issues may in turn encourage us to ask radical questions about the supposed fusion of, or perfect correspondence between, the components of opera (the Gesamtkunstwerk issue again), even in canonic masterpieces. It may ultimately suggest that historically informed performances of opera could include far more creative licence with the published text – far more irreverence – than anything seen nowadays. For example, almost all eighteenth-century opera composers spent time writing substitute arias for their own or other composers’ operas: when a work was revived with new singers, what would be more natural than to throw out some of the old arias and write some new ones, better adapted to the skills of the new cast? But who would dare take such an attitude today; even – or, in the case of Mozart, especially – to operas written during that precise period?

The more specifically musical side of opera is also split in a way we will often return to during this book. On the one hand is what we might call the ‘composer’s music’, what’s notated in the score, a document that could be reproduced on this page or, more important, that could form the basis of a keyboard-and-voice performance in our living room. Before recordings, such living-room performances were the most common way people enjoyed opera when not in the theatre, which was most of the time. This score, this composer’s music, is an essential blueprint, but it doesn’t constitute what most people understand as, and love about, opera: it gives us something more like a reminder of the operatic experience. To get at the real thing, we need a lot more – the sound of a live orchestra, the view of the stage, and so on. Most important of all, what’s missing is voice, the specific, irreplaceable quality of human vocal cords, membrane stretched across the larynx, vibrating in song. Voice is much harder to talk about than ‘composer’s music’, partly because it’s not reproducible in symbols and thus fixable on the page. But that mustn’t deter us. We need to keep in mind that the human voice has almost always been at the centre of operatic experience. And so, while we’ll be talking often – as historians should – about ‘composer’s music’ in this book, we’ll also keep reminding ourselves that the voices carrying the music are quite rightly essential to most people’s operatic experience.

An example may help here. At the end of Act 3 of Verdi’s Il trovatore (1853), Manrico’s aria ‘Di quella pira’ features within it and then closes with high Cs from the tenor, notes quite out of keeping with the rest of the role, and ones long known to have had nothing to do with the composer. No score near to the time of the first performance has a trace of them. Despite these circumstances, those high Cs are the most famous utterances in the entire work. When the conductor Riccardo Muti, who is famed for his adherence to the exact musical text and nothing more, opened the 2000–2001 La Scala season with Il trovatore, he instructed the tenor on no account to sing those offending high Cs. The tenor trembled and obeyed. The loggionisti, opera fans who haunt the upper reaches of the theatre and who know every recording intimately, went mad with rage, cries of ‘Vergogna’ (shame) rained down on the stage.⁶ Why was there so much passion? One waggish critic had earlier tried to defend the note, suggesting that if it wasn’t by Verdi, then it was best thought of as a gift to Verdi from the Italian people. It would be easy to dismiss this as sentimentality, or even as wilful disregard for the composer’s intentions. But the formulation could also say something basic about the operatic experience, suggesting that opera fans can grow to feel that they have a right to particular notes, or rather to the particular extremes of voice occasioned by those notes.

Precisely because it is an experience so difficult to put into words, the emotions aroused by hearing operatic voices can be enormously powerful and can lead to feelings of devotion that will seem irrational to others, those untouched by opera. It’s important to bear this extreme devotion in mind, because the story of opera and how it communicates can otherwise seem inexplicable. Important aspects of the operatic experience may matter very little for those in the grip of voice-generated emotion. They may not understand the plot. They may not even understand the words (and, as mentioned a moment ago, at extreme vocal moments the words almost always disappear, as if they have been consumed). Yet the power of the human voice still holds them.

A potent exploration of this irrational devotion to voice occurs in a 1981 French-language film by Jean-Jacques Beineix called, simply, Diva. The film has a strange mixture of styles, part comedy, part thriller – it’s famous to some primarily for its spectacular motorcycle chase. An excerpt near the start shows a young Parisian mail courier, Jules, attending the concert of an American opera star whose peculiarity is that she refuses to make recordings. His devotion to her voice is so profound that he smuggles a tape recorder into the concert hall. The star appears, and once she starts singing Jules enters a state of trance-like bliss, forgetting what he’s doing: the voice washes over him. What’s interesting from our point of view is how this act of singing is treated in the film. We hear an aria from a late nineteenth-century Italian opera (‘Ebben? ne andrò lontana’ from Alfredo Catalani’s La Wally, 1892). The aria obviously has a text, but the film doesn’t bother to subtitle the words; we have no idea what she’s singing about, and may suspect that Jules is equally ignorant – that he doesn’t care, and neither do we. What’s more, this isn’t an operatic performance: the aria is out of context, out of its plot environment, sung by someone not in costume, not pretending to play a part. None of this seems to matter. What captivates Jules is pure operatic singing, devoid of narrative and perhaps even of language. It’s an extreme case, of course, the far end of the spectrum; but the scene serves as a reminder of how this enormously powerful aspect of opera’s music can function, of what voice doesn’t need in order to make its impact.

The word impact is justified. Typically, operatic heroines will be sung by a high soprano voice. The singer may be required to make herself heard above an orchestra of up to 100 players, including (in the nineteenth century and beyond) many instruments that have steadily evolved with a view to producing an increased volume of sound, not least a formidable phalanx of newly invented, acoustically aggressive brass. Microphones and discreet amplification are sometimes used, but this is almost always regarded as shameful, a crutch that real opera singers would never need. So the singer, unaided, may have to make herself heard in all corners of an auditorium seating more than 3000 people – that is, in a cavernous place where the back balcony is many, many metres from the stage. At the Metropolitan Opera in New York, the seats farthest from the stage are about 150ft away, a distance one expert on theatrical acoustics has called ‘staggering’.⁷

Admittedly, since the nineteenth century the orchestra has tended to play in a sunken area in front of (and partly underneath) the stage, a position that serves to dull the sound somewhat. But the presence of this ‘pit’, and the economic need to cram more and more seats into modern theatres, has also meant that the stage now rarely comes out into the auditorium, as it typically did in earlier times. The singer thus has to project from a recessed space whose proscenium arch tends implacably to block the sound. Depending on theatre size, that feat of projection can approach the miraculous.

When a new Metropolitan Opera house opened in New York in 1883, the architectural critic for the New York Times took the occasion to marvel at its ‘vast auditorium’ and to compare the sizes of contemporary opera houses worldwide, giving us a snapshot of the building’s dimensions alongside those of its biggest European relatives in the late nineteenth century.⁸ As the Times reported, the auditorium at Covent Garden was 78ft deep and 62ft wide; at the ‘new’ Paris Opera House (this meant the Palais Garnier on the Place de l’Opéra), it was 90 by 67ft. The auditorium at the 1883 Met was described as 95 by 89ft. At Munich (the Nationaltheater, destroyed in 1943 and reconstructed in the early 1960s) the auditorium space was smaller, 68 by 61ft; at Vienna the proportions were (and are) 83 by 67ft. What stands out at the 1883 Met is also the height of the auditorium. At 82ft, it was 15–20ft higher than any other mentioned, with the exception of San Carlo in Naples, a notorious barn. This height makes for an increase in capacity and a larger distance from the stage in the highest regions: on the hypotenuse of the triangle, about 100ft away for Covent Garden, as opposed to about 125ft at the 1883 Met. Covent Garden seats a little over 2000; the 1883 Met seated about 3000. These differences in size will of course have acoustic consequences, no matter how the interior might be sculpted. What is more they have perceptual consequences for one’s experience of the event: as near and overwhelming, or as faraway and imbued with (perhaps magical) distance.

The new Metropolitan Opera, which opened in 1966, is bigger still, although its seating capacity at about 3800 is less than that of the Chicago Lyric Opera, which can hold 4300. But things can, it seems, still be extremely loud, even far away from the stage. In December 2006, the Metropolitan put on special performances of Mozart’s Die Zauberflöte in an English-language abridged version for young people. When the New York Times music critic Anthony Tommasini interviewed these junior listeners about their reactions, he was told repeatedly that the voices were loud, unpleasantly so.⁹ Struck by such comments from an audience accustomed to powerful stage amplifiers and high-volume headphones, he reasoned that the perception of loudness wasn’t a matter of decibels per se but of raw human decibels: of loudness without crutches; of a sense of what the sound must be like close to. The physiology of this feat of human decibels is worth pausing over. The singer uses her diaphragm to draw air into the lungs and then push it out, projecting her voice (produced by the larynx) into various resonating chambers in the skull. At the same time, she uses her throat muscles, jaw, lips and tongue to vary the sound, which then radiates out through the mouth and nose. This process is common to the production of all human vocal sound; but people who sing opera generate it on an unparalleled scale, letting loose huge acoustic forces; if they turn their voices on you at close range, you have to retreat and cover your ears. There are very few people who have the basic bodily equipment – the sheer muscular strength and flexibility – to do this; there are fewer still who succeed in mastering such tasks well enough to give others musical pleasure.

It’s a commonplace that those who can are by no means always able to impersonate an operatic heroine convincingly from a visual point of view. They may, for example, be large in all directions. Indeed, another common caricature of opera (one that goes back long into its history) is of a grotesque mismatch between the singer and the physical image intended dramatically. The surprising thing is that such mismatches don't seem to matter very much; or at least they haven’t for large stretches of operatic history, even at the times when the caricatures were in circulation. Opera is the one spectacle – the one ‘something you look at’ – in which conventional physical attractiveness counts for relatively little. The comparison with theatre or film is obvious; and it’s also a different story in ballet, which is as artificial as opera in other ways, but in which an unusually tall female dancer, or an unusually short male one, would not be tolerated. Opera alone can trade in imperfect faces and unfashionable forms, such offences against our visual expectations being overlooked or even celebrated in the roar of singing. Of course, there have been fluctuations in tolerance. It may be that our contemporary culture, so dominated by the visual, and in which a particular image of physical perfection is so obsessively valued, is among the most intolerant of the ‘wrong’ bodily shapes. But it’s still a matter of degree. Today’s opera houses are places where the look of a character is rarely dominant. In that sense, opera may provide a space in which we can appreciate an alternate and valuable truth.

TELLING THE STORY, THE REALITY ISSUE

We wrote at the start of this chapter that opera was non-realistic. This is as much true of its narrative capabilities as of its performers. There is, for example, often a lack of verisimilitude even within the rational element of the action, the part taken directly from a literary source (very often a pre-existing play). Because characters spend their time singing, they get through relatively few words, so plots and sentiments have to be condensed: the source text is sometimes so compressed that it seems ridiculous when removed from its operatic environment. A good question for an opera buffs’ party game might be: narrate the plot of Verdi’s La forza del destino (first version in 1862; Verdi then revised it in 1869) in one coherent sentence. Denis Forman’s attempt in The Good Opera Guide goes like this: it’s ‘the one where a marquis is killed by an exploding pistol and his daughter in monk’s costume finds her lover has murdered her brother just outside the front door of her cave’.¹⁰ Admittedly, this passes over large swathes in the middle (most of the opera, in fact). It also omits the fact that the daughter dies too, murdered by her brother just before he expires; and that, in the first version, the lover responds to all this carnage by calling down a curse of mankind and throwing himself off a nearby precipice. But the missing bits aren’t really essential. La forza del destino is what Verdi called ‘an opera of ideas’,¹¹ one in which crowd-pleasing musical numbers are less important than impressive human subjects and abstract themes such as ‘fate’ (the destino of the title). What emerged was close to a sprawling historical novel put on stage, with all sorts of extravagant gestures that lead nowhere. It’s an extreme case of narrative wildness, but by no means the most improbable. Indeed, among the essentials of an operatic drama seem to be exaggerated coincidence, obscure motivations and (if it’s tragic) multiple deaths. If you take away the music, the plots are ripe for parody or kitsch; and the poetic text is usually second-class, formulaic at best. Thus denuded, there’s very little left to explain why opera is worth loving, a sentiment that forms the basis for a famous comic animated film by Kim Thompson, All the Great Operas in Ten Minutes (1992). The essence of Wagner’s epic, Der Ring des Nibelungen (1876), which lasts around sixteen hours in performance over four nights, is covered in less than two minutes. Here are just the first two operas:

The Ring of the Nibelungs is actually more of a miniseries ’cos it’s four full operas. These girls in a river have this gold and a dwarf steals it and makes a ring from it. The head guy Wotan gets the gold back but there’s a death curse on the ring and these two giants get the gold and one kills the other and then turns into a dragon to guard it. Wotan wants the ring so he decides that his illegitimate human son Siegmund will get it for him. Siegmund meanwhile falls in love with his long-lost sister Sieglinde. He picks a fight with her husband so Wotan gets Brünnhilde to protect him, then changes his mind and orders her to kill Siegmund instead, but she doesn’t want to so Wotan gets Sieglinde’s husband to kill him and then Wotan kills the husband too just for the hell of it. He punishes Brünnhilde by putting her to sleep on a mountain in a ring of fire.

At the end of the film, the narrator says: ‘That’s it. That’s opera. Nothing else you need to know. Just a lot of people in costumes falling in love and dying. Yup, that’s pretty much it. Oh, except for the music. There’s singing too? See, they don’t actually say anything, so they sing everything. Lots of music. Yup. Some of it’s kind of nice, too.’

But before we laugh and move on, we should acknowledge that there’s a problem here. All lovers of opera will agree that the story, the narrative element, can often be ludicrous; but it’s also essential. We can’t seem to live comfortably with it; but we can’t live without it. Narrative and opera are, in other words, locked together in an uneasy embrace. Neither wants divorce, but untroubled cohabitation is almost always out of the question. In this sense, the narrative dimension is rather like the verbal dimension as a whole. It’s another example of the fact that opera can make us forget: forget that we don’t understand the language being sung; forget the size of the singer, and the fact that he’s impersonating an ardent and athletic young troubadour although patently nearer sixty-two than twenty-six; forget the fact we have only a shaky idea of what on earth is happening on stage. Opera communicates in strange, unpredictable ways; it appeals to something beyond the narrow cognitive dimension.

All of which brings us to what is probably the central improbability about opera: the fact that most of the characters sing most of the time. We have to be careful not to exaggerate here: after all, drama that is sung or chanted is probably more the rule than the exception throughout world history. But the brand of sung drama that emerged in Western Europe around 1600 was nevertheless a peculiarly dense and radical one, and it caused consternation from the very start. These debates about opera, although often intense, never managed to curb opera’s excesses, or at least not for long. Pretty soon opera was too widespread for philosophical scruples to prevail. Pretty soon everyone could be impersonated on the operatic stage: gods and coachmen, pirates and vestal virgins all sang away, happily accepted or at least tolerated. Yet elements of unease about singing recur throughout opera’s history. This discomfort accounts for the fact that operatic characters throughout history – starting with Orpheus, one of the first – have readily embraced moments that would have been sung even in a spoken drama or, if it’s possible to imagine, in real life. Opera is bestrewn with serenades to the distant beloved and with drinking songs shared by confederates in crime or love; any tenor or baritone worth his fee will know how to enter the stage winningly with guitar in hand. The fact that these moments are operatic commonplaces is significant, suggesting that we all need reality checks – times during which we can relax our (willing) suspension of disbelief, for an instant or two ceasing to maintain the fiction of all these characters singing at each other. But such moments of pause shouldn’t deflect us from confronting what, in the end, opera asks us to believe.

We might make an imaginative leap and think for a moment about what it would be like to inhabit a world that is operatic; a world in which everyday life takes place and ordinary time passes, but in which everything – every action, every thought, every utterance – is geared to never-ending music. It’s similar to the kind of thought-experiment explored in Peter Weir’s film The Truman Show (1998), in which Jim Carrey gradually becomes aware that his entire life is being played out for the benefit of a TV reality-show. How would we figure in another kind of man-made fantasy, one in which everything surrounding us was opera?

Think of the metaphysical questions that this state-of-opera would raise. First, and most important: who is making the music? Do we know we are singing, that we are constantly swimming in musical sound? Or do we maintain the illusion that the stage world is really quite ordinary and that we hear music in some magical way? When the music seems omniscient or prescient, whose mind is telling us its secrets? Such speculations may seem fanciful, or even idle; but pondering their significance turns out to be important to our understanding of opera. Filmmakers have constantly made artistic capital from playing with audience expectations about the origins of sound in film: does it spring from the world of the characters or is it a commentary on that world? Can our attitude to the narrative be made more interestingly complex if these two sources are sometimes confused? In just the same way, opera composers (and not just those of the recent past) have played with the idea of a mutable musical source, of exploring and enriching the operatic experience by making us ponder that basic question: where does the music come from?

Thinking about these issues also stresses the existence of the ‘marvellous’ in opera: the fact that the whole business is in so many ways fundamentally unrealistic, and can’t be presented as a sensible model for leading one’s life or understanding human behaviour. No one conducts her life in song and continuously hears fabulous music around her. This is a critical point, since this basic lack of verisimilitude, while it didn’t end up restricting opera to certain exceptional character types (gods or elves or water-nymphs), in broad ways determined opera’s liking for narrative extremes: moments of extraordinary passion and impossible dilemma; magic and the irrational; world-historical events. Spoken theatre can also be extreme in its plot material, populated by ghosts and improbable turns of fate. But it can never match the central unrealism of opera, the fact of continual singing. And this foundational trait has tended to guarantee that opera libretti almost never deal with the ordinary, while spoken theatre can thrive on it. Even when libretti (briefly) marched under the banner of ‘realism’ in the nineteenth century, their reality was an immensely charged affair. Modernist works like Ernst Krenek’s Jonny spielt auf (1927) or Francis Poulenc’s La Voix humaine (1959) don’t contradict this, instead playing against opera’s historical taste for the extraordinary by including scenes of self-conscious banality, everyday life made into art: La Voix humaine, after all, stages nothing more than a telephone call. The first recitative in Stravinsky’s The Rake’s Progress (1951) tries something similar. It begins with the (sung) words ‘Anne, my dear, your advice is needed in the kitchen’, a sentence that is surely a deliberate ploy to estrange us, to tell us that The Rake will be a very knowing opera. A different case is Berg’s Lulu (1937), where the petty rituals of the high bourgeois co-exist in a dissociated fugue with acts of murder and betrayal. In Lulu, as in other examples of German modernism, the impulse to put mundane events on stage was aligned during the Weimar era (1918–33) with a much-theorized and politicized aesthetic populism: with the idea that art composers should produce Gebrauchsmusik, music for everyday use, or Zeitopern, operas on contemporary, topical themes.

But these are exceptions. For most of its history opera has disdained ordinary events and actions; or, as Lulu did famously, it has treated them as a surreal counterpoint to a more dramatic and violent hyper-existence. Hardly any operatic character makes a cup of tea, or reads the newspaper, or puts on his socks; those who did so would be embedded in a comic reversal of normality or would have been deliberately trivialized. In Wagner’s Ring, the dwarf Mime is made ludicrous when he cooks a good breakfast for the imperious young hero Siegfried, the implication being that cooks, like newspaper-readers, tea-drinkers and sock-wearers, are by their ordinariness excluded from the heights of true passion or mythic significance. The irony in Siegfried (hardly intended by Wagner) is that Siegfried’s own efforts to re-forge his father Siegmund’s magical sword, accompanied as they are by much filing, pouring, stoking of fires and measuring of temperatures, can in performance easily slide into a parody of domesticity: breakfast-making while high on excess testosterone.

Nowhere is opera’s affinity for the unreal more obvious than in its ways and means of manipulating time. That high-C-laden Act 3 aria for Manrico in Il trovatore again provides a classic instance. Manrico is about to marry his beloved Leonora when a messenger enters to inform him that his mother has been captured by a deadly enemy and is about to be burnt at the stake. He calls his companions in arms together and draws his sword; they must go to rescue her immediately. But then, instead of rushing off, Manrico faces the audience and sings two (quite long) verses of the cabaletta ‘Di quella pira’, prolonged high Cs and elaborate cadential writing for the chorus not excluded. Act 2 of Gilbert and Sullivan’s The Pirates of Penzance (1880), which like all their operettas uses Italian serious opera as the primary butt for its satire, has an apt parody. A small regiment of police gears up for battle in the chorus ‘When the foeman bares his steel’, a number that ends with endless repetitions of ‘Yes forward on the foe, yes forward on the foe’. The major-general steps out of the opera to comment caustically: ‘Yes, but you don’t go’.

Related but even more spectacular examples of opera’s cavalier way with the realistic often occur when its major characters are constrained to die onstage. A classic example here is the last scene of Verdi’s La traviata (1853), in which the courtesan Violetta succumbs to tuberculosis. In the plot, the character’s lungs have been destroyed; but operatic Violetta sings on without evident trouble, producing well-supported, beautiful sound. A soprano who dared to make Violetta’s affliction realistic by croaking and coughing would have left opera behind. Mortally wounded people who nevertheless sing on and on mellifluously are inhabitants of opera in its normal mode; we accept the fantasy without resistance; indeed, it is reality that would be shocking. Violetta’s case shows that opera involves perpetual mismatches between assumptions about real-world plausibility based on human characters, and demands that musical performance will make on both singers and audiences.

We can take this further. The idea that opera puts a dying, literary Violetta comfortably aside in favour of a flourishing, powerful operatic Violetta brings us to a critical distinction, one that will reappear several times in this book: the difference between the character as she exists in words, what we could call the plot-character, and the corresponding character that exists in music, the voice-character.

Plot-character could, for instance, be Wolfram von Eschenbach in Wagner’s Tannhäuser (1845), a tedious and uninteresting suitor who preaches chaste devotion. In Wagner’s libretto, he’s a sidekick and foil to Tannhäuser (tenor), the erotically sophisticated anti-hero whom the heroine cannot resist. But as voice-character Wolfram (baritone) is not simply parallel or equivalent; he is constructed by means of an amalgam of the music assigned to the plot-character and the feats of performance that music elicits from the singer. In the Act 2 singing contest between Tannhäuser and Wolfram, there’s an odd moment during Wolfram’s decorous serenade. Praising the grandeur of the scene, he sings ‘my gaze grows drunk from seeing’ to an exotic harmonic progression accompanying a long lyric fall in the baritone’s melody. This is the only truly sensual passage in the entire song contest, outdoing anything by Tannhäuser. It marks the appearance of a voice-character, an alternative Wolfram with attributes beyond, even at odds with, the plot-character. The same contrast appears in Wolfram’s famous ‘Evening Star’ aria in Act 3. The words are studiously poetic, about a character with a good soul, but the sensuality elicited from the voice tells another story entirely. Perhaps we should think of the strange verbal turn in Wolfram’s Act 2 serenade – the line about growing intoxicated through sensation – as a slip in the plot-character’s language, a place where the voice-character, for an instant, works his way into words.

One could quite rightly argue that this same distinction between plot-character and voice-character also occurs in spoken drama; anyone who has attended a Shakespeare soliloquy or a Hugolian tirade will know that the speaker in question is engaging a different register at these points. But in opera the distinction is so much more extreme, so much more spectacular. The Manrico who must rush off to rescue his mother is plot-Manrico; the one who remains centre-stage singing high Cs is voice-Manrico. Plot-Violetta is dying of tuberculosis and desperately fights for breath; voice-Violetta sings robustly, lamenting but also, in a sense, celebrating in song her imminent death. And the difference emphasizes once again that the tension between representations of human reality amid omnipresent music is a basic characteristic of opera. Concerns about verisimilitude are behind opera’s general tendency for libretti with mythic or divine characters, or ones that invoke magic or emotional overload or other extremes: the theory is that music belongs or is somehow more explicable in such cases. In the end, though, verisimilitude gets discarded no matter what. In the end, opera can’t ever be anything other than unreal.

POPULARITY AND CONVENTION

As mentioned already, opera has for most of its history seemed strange and exotic to most of its audience. It is important to realize that this was the case even in Italy, which was the home of opera, and where for this reason it is sometimes thought to be, or at least to have been, a natural form of expression. And although its financial basis has gradually broadened – from the court-sponsored private entertainments of the seventeenth century, to the (almost) free capitalist enterprises of the nineteenth century, to the often precarious state- or corporate-funded institutions of the present – opera in its primary, theatrical form has typically been the province of elite groups, only becoming more broadly popular when technologies of various kinds allowed it to spread from the opera house into the streets and into people’s homes. It is arguably only in the last thirty or so years, the era of the Three Tenors all singing ‘O sole mio’ together on the eve of a soccer World Cup, that opera has become popular in the modern sense, and it’s probably more accurate to say that, in this case, particles of opera have now become iconic (if also ironic) objects in mass culture.

In Hollywood, for example, opera-going has often stood for Italianness and emotional excitability or prone-ness, as in Moonstruck (1987), where Ronny Cammareri (Nicolas Cage) wins the love of fellow working-class Italian-American Loretta Castorini (Cher) by taking her to see Puccini’s La bohème. In some films, opera houses become places where different classes and characters can meet as if in a goldfish bowl, pursuing their many agendas and pleasures. In Billy Wilder’s Love in the Afternoon (1957) the major characters show up for a performance of Tristan und Isolde at the Paris Opéra: spoiled playboy Gary Cooper, dragged along by a socialite friend, clowns around with his programme book; earnest composition students from the Conservatoire conduct along with the Prelude while sneering at the common audience; only Audrey Hepburn, terribly in love, seems wholly lost in a sensual experience – but it’s less Tristan than Gary Cooper who is doing the business (see Figure 2). The irony is that no one is represented as properly or truly moved by the opera itself, something hardly unexpected given Wilder’s proclaimed disdain for Wagner.

Love in the Afternoon also shows how popular culture can make opera a sign for upper-class privilege. The opening scene of Martin Scorsese’s The Age of Innocence (1993) centres on a performance of Gounod’s Faust at the old Met in New York. This establishes a milieu for the plot, but with a camera that swoops drunkenly and supernaturally everywhere around the house: the upper classes, he tells us, get derailed by passions, just like everyone else. In the Marx Brothers’ A Night at the Opera (1935), the running joke is how a stiff, old-money audience can be outraged by low-life antics from the brothers. At one point the latter infiltrate the pit, switch the orchestra’s parts and cause the players (ostensibly against their will, of course) to strike up with ‘Take me out to the ball game’; the clash of high and low culture couldn’t be clearer. A more insidiously sentimental example comes in Garry Marshall’s film Pretty Woman (1990), in which Julia Roberts plays a prostitute picked up by Richard Gere (a ‘handsome corporate mogul’ according to the publicity flier) on Hollywood Boulevard, and is then given the Cinderella treatment. During their week-long affair Gere flies Roberts (private jet) up the coast to the San Francisco opera, where she makes all kinds of gaffes among fan-fluttering socialites, but eventually shows that she is more open than them, more sensitive to what she sees on stage (the opera in question is, of course, La traviata).

These scenes may also reinforce a contradiction in the way we approach opera today. On the one hand is its age-old identification with aristocracy, ostentatious wealth and exclusivity. Running parallel, though, is a sense that opera can appeal directly to our emotions. Both attitudes are given voice in A Night at the Opera: socialites go to the opera as a duty, and are outraged when decorum is assailed. But the movie’s hero and heroine – an underdog tenor and soprano helped by the brothers into their starring roles – are shown singing triumphantly in the final reel of the film, with immense joy and glorious smiles (and this even though it’s the ‘Miserere’ scene from Il trovatore). They are, quite properly, moved by opera: not by a silly operatic plot but by singing itself. This scene of underdogs and comic outsiders, transformed by singing, shows opera as having a powerful effect on people who are different in some way – on Julia Roberts’s character (who is stubbornly herself no matter what the opulence of the surroundings, and who sings Prince in the bath), or on large Italians with white handkerchiefs – people who have declared their diversity, their distance from the everyday, emotionally constricted life we think we now lead.

For at least its first 250 years, the opera industry (just like the film industry today) was sustained by a steady stream of new works. There was no repertory to speak of, no body of familiar, endlessly repeated works such as we find in every opera house today. Because of this, operas (like films) needed to make their impact immediately. One way they did this was through using a set of conventions: working practices common from opera to opera, readily understood and appreciated. And so emerged different operatic types, usually dictated by language and register (Italian opera seria or opera buffa, French grand opéra or opéra comique, German Singspiel or romantische Oper), each with its own peculiarities, its own codes of communication.

More basically still, though, operatic character types gradually evolved into voice types, the register and weight of the voice belonging to a particular character who could then pass from opera to opera, and eventually into operetta and musical theatre. A repertoire of operatic characters emerged: the female lead (soprano), the woman of advanced age or little virtue or touched by sorcery (mezzo-soprano or contralto), the heroic male lead (in the eighteenth century usually castrato, after that tenor), the villain or father or loyal-unto-death companion (baritone), the grandfather or priest or other symbol of patriarchal authority (bass). Although all these types were liable to variation (there have been soprano priests, but not many), it is often important to know what was conventional, in particular to appreciate how turning convention on its head is itself a form of drama. One instance of voice cast against character type is Mephistopheles (the devil in the Faust legend: he of the pact signed in blood, a favourite operatic subject). Most Mephistopheles characters of the nineteenth century (there are several) are basses, but in Busoni’s Doktor Faust (1925) Mephistopheles is a tenor; he is heard first calling Faust's name from offstage, and he does so on a high As – in a whisper. It is knowledge of the going-against-type that makes the effect so striking.

The question arises whether these choices of voice type are the product of nature or convention. Some (the bass authority figure) might be thought natural – both men and women are encouraged to lower their voices when they want to be deemed especially worthy of attention and respect – and it’s probably significant that this character/voice type has been around throughout the operatic centuries. For the most part, though, history suggests that simple convention is at work. For example, in early eighteenth-century serious opera (so-called opera seria), almost all characters, male or female, sang in high voices. It was quite common for male roles to be sung by females, or by castrated males whose voices remained ever high. There are various explanations for the demise of the castrato and the rise of the heroic tenor in the nineteenth century, and some plausibly invoke changes in ideas about human subjectivity (about what we think makes us who we are). But it’s likely that even tenor heroes were never experienced as normal. The notion of aberration, of opera’s strangeness, is always there.

Conventions track deep into opera. Through most of its long history (until at least the mid nineteenth century), composers tended to write several operas each year. They did so out of financial necessity. Works were rarely revived (and, if they were, were generally unprotected by copyright), and it was only by writing new works that composers could earn a living from their art. The consequent need for instant communication with an audience tended to make both libretto and music formulaic. The word ‘formulaic’ now sounds pejorative: in our culturally pessimistic state, we tend to think that only the startlingly original can be artistically worthy. We shouldn’t forget, though, that formulas can be a source of pleasure, not (as usual these days) from knowing a piece backwards and forwards through recordings, but from harbouring expectations about how operas usually work. Expectations can give comfort and furnish ready understanding, and can also be productive of a frisson when something happens against type, when a convention is creatively manipulated, as in Busoni’s upsetting of voice-type associations.

Because the structural formulas of opera are important, we should introduce some of the basic ones immediately. They didn’t emerge simultaneously, and some got much weaker in the nineteenth century. But, for all that, they have been surprisingly long-lived and malleable, some of them adapting to radical changes elsewhere in verbal and musical language. The two most basic terms involve both words and music: ‘recitative’ versus ‘aria’ or ‘number’. These distinguish between two kinds of poetry in the libretto and two kinds of music that sets that poetry. The dualism existed from opera’s beginnings, wasn’t broken down until the mid nineteenth century, and still to some extent exists in operas written today. A brief scene from the first act of Mozart’s Don Giovanni (1787) offers an illustration of the two at work. The scene involves the lecherous aristocrat Don Giovanni (baritone) and his servant Leporello (bass). The libretto starts with dialogue. Leporello is becoming exasperated with the various risky stratagems he is obliged to carry out in his master’s name:

          LEPORELLO

Io deggio ad ogni patto

Per sempre abbandonar questo bel matto!

Eccolo qui: guardate

Con qual indifferenza se ne viene!

          DON GIOVANNI

Oh, Leporello mio! va tutto bene.

          LEPORELLO

Don Giovannino mio! va tutto male!

[LEPORELLO: Whatever happens, I must / abandon for ever this fine madman! / Here he is: look / at the nonchalance with which he comes and goes! DON GIOVANNI: Oh, my Leporello! everything’s going well. LEPORELLO: My dear little Don Giovanni! everything’s going badly!]

This is intended as a typical passage of recitative. It’s written in the Italian equivalent of blank verse (so-called versi sciolti), a free mixture of seven- and eleven-syllable lines, with only a few, sporadic rhymes, and with no fixed accents within the lines. Meant to approximate ordinary speech rhythms, it’s also quite ordinary in the lexicon it uses, with few poetic words or conceits. This conversation continues for some time, as Don Giovanni and Leporello prepare for a banquet that night in which Don Giovanni will (he hopes) be able to have his wicked way with a peasant girl called Zerlina. He ends the recitative in jubilant mood and then launches into an aria, ‘Fin ch’han dal vino’ (‘So long as there’s wine’). This number is often called ‘the champagne aria’, but although wine is mentioned in the very first line, Don Giovanni is overwhelmingly concerned with the women who will attend the banquet. In a wild fantasy, he imagines how he will vary the dances so that they succeed each other chaotically; in the mêlée thus created he will, he hopes, add ‘at least ten’ to his already considerable list of conquests. Here are the last five lines of recitative and then, indented, the aria text:

       DON GIOVANNI

Bravo, bravo, arcibravo!

L’affar non può andar meglio. Incominciasti,

io saprò terminar. Troppo mi premono

queste contadinotte;

le voglio divertir finché vien notte.

        Finch’han dal vino

        Calda la testa

        Una gran festa

        Fa preparar.

        Se trovi in piazza

        Qualche ragazza,

        Teco ancor quella

        Cerca menar.

        Senza alcun ordine

        La danza sia;

        Chi ’l minuetto,

        Chi la follia,

        Chi l’alemanna

        Farai ballar.

        Ed io fra tanto

        Dall’altro canto

        Con questa, e quella

        Vo’ amoreggiar.

        Ah! la mia lista

        Doman mattina

        D’una decina

        Devi aumentar!

        (They leave.)

[DON GIOVANNI: Bravo, bravo, bravissimo! / This business couldn’t go better. You begin it, / and I’ll finish it off. These peasant girls / urge me on too much. / I want to enjoy them as night falls. / While they have their heads / warm with wine / you must prepare / a grand party. / If you find around the place / any girls, / try to get them / to come along with you. / The dances should be / without any order; / one a minuet, / one a follia, / one an allemande / I’ll make dance. / And in the meantime, / on the side / I’ll make love / to this one and that one. / Ah! By tomorrow morning / my list / must be longer / by at least ten!]

It’s easy to see that the start of the aria is marked by a change in the poetry. Now all the lines are of equal length and have a common stress; there is also a fixed rhyme scheme outlining a four-line verse structure; in other words, the poetic register has made a decisive appearance.

These are the most elementary differences between ‘recitative’ and ‘aria’ as we see them from the libretto; but embedded within such differences are important distinctions having to do with content. The most basic is that recitative is the place for narrative, informal dialogue, stage action: moments in which the plot moves forward. The aria, on the other hand, is a static mode. It is fundamentally about contemplation, and through contemplation the communication of mood to the audience; it is what the poets call ‘musing’. In this aria, as in most others, nothing external, nothing plot-related happens; what the aria does instead is characterize Don Giovanni, making us aware of his desire to mix dances ‘without any order’, to create chaos, which is the medium in which he, an anarchic force, can flourish. In some sense, then, arias stop time – they let nothing else happen while they unfold, allowing us to sample a kind of internal time, one in which the character’s mind reveals itself. And what is said here about arias goes equally for all the contemplative parts of the opera: the duets, trios and bigger ensembles. Admittedly, one of the great departures of nineteenth-century opera is that all these fixed forms may be liable to injections of outside action; but even then there remain frequent occasions during which time seems to be arrested. The presence of singing of a particular kind will brook no competition. Singing that would not take place in the fictional world still has the capacity to exclude other events from that world.

The musical distinctions between recitative, on the one hand, and arias as well as ‘numbers’ for more than one voice, on the other, are obvious and immediately audible. In the simplest kinds of recitative, found mostly in the late seventeenth and eighteenth centuries, the vocal line is ‘recited’, usually with great rapidity, accompanied or punctuated by a simple succession of chords on what is often called the ‘continuo’: usually a harpsichord, with its bass strengthened by a lower string instrument, usually a cello. Although such simple recitative (sometimes called by its Italian name, secco recitativo) tends to be notated in a fixed time signature, it is typically performed very freely, with no sense of a regular beat. The melodic lines are rudimentary, there are many repeated notes; the pitches tend to follow the intonations of speech. In many ways this ‘recitative’ is nearer speech than song, and given the more studied, musical way in which actors of past centuries declaimed speeches in a spoken drama, it may have seemed even less like music when it was first performed. The aria is very different. It is dominated by musical expression, with obvious (and usually recurring) musical ideas, with the orchestra to introduce and accompany, and with the voice singing elaborate melody. The regular rhythms of the poetry create (almost demand) regular rhythms in the music. Unless the aria is very short, it will also have an internal form of its own: the most common was a simple ternary form (schematically ABA). In eighteenth-century opera these ternary forms crystallized into what were called ‘da capo’ arias because they returned to the initial section after a period of musical digression or contrast; in later centuries the forms typically became more complex.

What is also obvious is a significant difference between recitative and aria in terms of the way words are treated. In the recitative, words are consumed at a great rate and hardly ever recur. In the aria, on the other hand, fragments of text are often repeated (in this example from Don Giovanni, almost obsessively). As we would expect, the music then responds to the sense of the words in an individual way. Don Giovanni’s manic fantasy comes in the form of relentlessly driven music, an aria in which there are hardly pauses for breath. However, particularly after the first exposition of the musical material, words are repeated ad infinitum to fill out the demands of musical elaboration and then closure. In this, and in very many arias, the literary text melts away: it partly or even completely loses its meaning. Music speaks beyond the text, whose meaning has been drastically diminished.

Of course, there are many more points one could make about Don Giovanni’s ‘champagne’ aria. In one sense (and like almost all arias), it is there partly to show off the voice, and individual singers will often add improvisations of their own: extra high notes, ornaments, frenzied laughter at the end. But even in the printed score the breakup of the text, the repetitions, the way melodic fragments obsessively return, seem excessive by any rhetorical measure. Perhaps what we’re hearing is a musical–linguistic symbol for the whirling confusion Don Giovanni wants to create. There may also be a sense in which the character is caught up in this whirl of action, rather than (as he thinks) controlling it – a sense in which Don Giovanni is at the mercy of these obsessive musical rhythms. It’s a very rare singer who doesn’t, by the end of the aria, sound breathless and almost incoherent, as if the music is now driving him. If we know the rest of Mozart’s opera, know the way in which Don Giovanni finally meets his fate, the sense of lost control that is created by the music – and its performance – is an acoustic harbinger of the protagonist’s end.

EXPRESSION AND BOREDOM

The aria and its recitative were the basic building blocks out of which opera was constructed for much of its history. From the seventeenth century onwards, in Italy and elsewhere, recitative carried action and narrative, and was used overwhelmingly for conversations between characters. Recitative had a generally poor reputation as music; one contemporary of Mozart lamented the necessity of ‘the recitative, which is both dull-sounding and neglected by composers and singers alike, and which no one thinks of listening to any longer. In truth its insipidity and monotony are insufferable.’¹² Once vernacular opera (Singspiel) began to develop in Germany in the earlier years of the eighteenth century, spoken dialogue served the analogous purpose. But operatic music involves so much more than just a distinction between various kinds of expression, between recitative and aria or other set number. Here we return to an aspect of opera that is harder to pin down and yet central to the experience: what, exactly, does music do when it becomes an aura around or beyond the action, beyond the emotions represented by the libretto, characters and situations? How might music embody the biggest things, such as an atmosphere or a worldview; how does it give life to the particular fictional world up there on the stage? These questions will have almost as many answers as there are operas; and for the best operas each answer will be elaborate.

One place to make a very small response to the most elementary version of the question is with the notion of musical signs: musical ideas that are understood within a particular opera, or even widely within a particular culture, as standing for or giving expression to an idea. Opera relies on small signs to make small points, and knowing how these work helps extrapolation to the bigger issues. One of the most familiar musical signs in Western popular culture comes from an opera: the Wedding March from Act 3 of Wagner’s Lohengrin (1848) – known nowadays to countless millions as ‘Here Comes the Bride’. In Lohengrin the march is played as an accompaniment to Lohengrin and Elsa’s wedding, with the chorus singing along. In Victorian England, the march – sober, andante and optimistic – was quickly adopted as the processional of choice at middle-class weddings; the more ecstatic march from Mendelssohn’s incidental music to Ein Sommernachtstraum (A Midsummer Night’s Dream) became the traditional recessional. The fact that Lohengrin and Elsa’s marriage turns out to be one of opera’s great romantic disasters is conveniently placed to one side.

The melody from Lohengrin, we might say, symbolizes a wedding ceremony in our minds. Imagine for a moment, though, that something happens to that melody, something that never actually happens in Wagner’s opera. Imagine that the melody is changed to the minor mode, is punctuated by solemn drumbeats, and is enwrapped in a dense chromatic accompaniment. Now we are on our guard: predictions for this marriage become quite a bit more pessimistic. There is a complicated social and musical system at work here. The melody itself is a sign that is reliant on very specific terms – music equals wedding. But the shift from major to minor – from the mode that suggests joy, optimism, contentment and triumph to the one that suggests gloom and tragedy – has changed our understanding of the music, and has conveyed, in an instant, a newly complicated message. What is important, though, is that this message is for our ears only: it is surplus to what is happening in the church, or on stage, and it is surplus to what is being said by any of the characters on stage. This very procedure was used by early accompanists in silent cinema, in full realization that a fundamentally operatic trick could now be used to say something, in music, to a cinema audience. As one such musician reported, ‘Wagner’s and Mendelssohn’s Wedding Marches were used … for fights between husbands and wives, and divorce scenes: we just had them played out of tune, a treatment known in the profession as “souring up the aisle”.’¹³

The combination of specific association and audible musical distortion, which enables us to appreciate ‘souring up the aisle’, is a two-pronged technique that composers replay in countless operatic circumstances. One could reverse the trick, say, with the ‘Jealousy’ theme in Giuseppe Verdi’s Otello (1887) – a sinuous minor-key idea first sung (in Act 2) by Iago, to a line about the ‘green-eyed monster’. The theme recurs instrumentally several times in the opera, always dark, and always haunted by its original verbal sidecar. You could of course dress the theme musically in ways Verdi never did, by re-harmonizing it into a sort of saccharine cadence. The new message might be that the jealousy problem is resolved. And it is a message that needs no words whatsoever.

But creating aura via a musical sign need not be so specific, or rely on decipherable associations between a theme and an idea. For instance, in a famous Act 1 trio from Mozart’s Così fan tutte (1790) the sisters Fiordiligi and Dorabella (two sopranos), along with a sardonic philosopher called Don Alfonso (bass), bid a sad farewell to their fiancés, who have just embarked on a sea journey (the fiancés will soon return in disguise to try to seduce each other’s bride; they do this because of a secret wager with Don Alfonso, who has no faith in female virtue; but that gets us ahead of the story). The text of the trio is extremely short and disarming in its simplicity:

Soave sia il vento

Tranquilla sia l’onda

Ed ogni elemento

Beningo responde

Ai nostri desir.

[Gentle be the winds / tranquil be the waves / and may every element / answer benignly / to our desire.]

During the music’s first pass through these words, the string oscillations (to an eighteenth-century audience, they would have been an obvious sign for waves), the tranquil harmony and the close, slow circling of the three voices evoke a musical garden of Eden before the fall. But in a second section, as the text is repeated, Mozart creates a special effect when the characters reach the word ‘desir’: there is a strange harmony, a sudden, subtle loudness and odd voicing in the woodwind instruments, as if they want to be noticed; and, at this exact place, the oscillating strings, which had disappeared from the texture, return. It’s like a black underline made musical, drawn under the word ‘desire’ to make it alien and suggest (quite rightly, as it turns out) that desires neither tranquil nor officially sanctioned may be circulating subconsciously among those present.

Emphasis of this sort is common in opera, but is not the only or even the principal way in which operatic music works as expression: we will see many others. Music can, for example, contribute an added layer to the words, illustrating them through its shape and contour. Usually called ‘word painting’, this type of expression was widely accepted as a theory of operatic expression through long periods of the eighteenth century. Yet at other times there were theories and practices opposed to the notion of individual emphasis. Rossini, for instance, espoused the conviction that operatic music should be in some sense ideal, self-contained within itself, not expressive of any individual word or poetic text. That exchange of texts between Mosè and Moïse, mentioned earlier, is a good illustration. Indeed, there’s a sense in which the elaborate ornamental singing that predominates in Rossini’s operas is radically non-symbolic: it does not express anything that can be verbalized; it does not tell you much about the character or his or her state of mind; it is simply there to be beautiful.

When today we go to see, say, a nineteenth-century French grand opéra or an opera from the early eighteenth century, both of which tend to be very long, we may get bored. Boredom can even overcome us at excellent performances of a canonic work. Rossini had a nice quip about Wagner, saying that he was a composer who had ‘beautiful moments but bad quarters of an hour’.¹⁴ Opera often lasts for hours and hours, and there is no operatic work, not even the greatest, without its moments of tedium. What is more, we are nowadays further encouraged to be bored by the conditions under which opera is performed: we are forced to sit in the dark, without interacting with our friends and neighbours; we are forbidden to leave the auditorium during a performance (if we do, we are barred from getting back in); rapt and above all silent attention is demanded as a courtesy to the performers and fellow attendees – and, strangely because they are almost always dead, as a courtesy to composers. Through most of opera’s history, in most national traditions, this was not the case.

It is to Wagner, to his theatrical innovations and artistic demands, that we are traditionally thought to owe this model of complete absorption (at his home theatre of Bayreuth, attendants to this day lock the auditorium doors once the lights go down). Lighting is a good indicator. Illumination in candle-lit eighteenth-century theatres may have been dim, but it was dim both on and off the stage. In the mid nineteenth century gas lighting allowed for more intense stage illumination, but the same technology was also used in the auditoriums, meaning that spectators could see each other as vividly as they could the stage. Wagner was the first theatre producer to call for complete darkness in the auditorium; and in his writings on performance he stressed again and again that spectators should be drawn overwhelmingly into the fictional world on stage, losing their anchor in reality. The introduction of electric lighting in theatres in the last decades of the nineteenth century finally allowed this darkness to be achieved. Before then, attendance at the opera was first and foremost a social occasion; audiences interacted with each other, and with the performers and the performance, sometimes in unruly ways. Another Wagnerian innovation aiding this sense of absorption was taking the orchestra out of public view, indeed dispatching many of the noisier instruments to a ‘pit’ largely beneath the stage. Other composers took up Wagner’s innovations enthusiastically. In the early 1870s Verdi wrote to his publisher, Giulio Ricordi, about performance conditions for his new opera, Aida (1871), recommending that:

they get rid of those stage boxes, taking the curtain right to the footlights; and also make the orchestra invisible. This idea isn’t mine, it’s Wagner’s: and it’s excellent. It seems impossible that in this day and age people tolerate seeing tired evening dress and white ties mixed up with, for example, Egyptian, Assyrian or Druidic costumes; and, what is more, seeing the massed ranks of the orchestra, which is part of the fictional world, almost in the middle of the stalls, amongst the whistlers or the applauders.¹⁵

The sentiments are, as Verdi was the first to recognize, remarkably similar to Wagner’s: above all one must preserve the special quality of opera’s ‘fictional world’. After the mid nineteenth century, silent attention to this world became more the norm, although Italian audiences in particular have often resisted the requirement. We now ask all operas to sustain our focused, unwavering attention, even though most operatic works were not designed to bear this burden. In the eighteenth century, few worried if there were too many arias or tedious stretches. During those artistically more generous, culturally more confident times, the audience was permitted to ignore what it found dull, and occupy its mind in other ways. We are in this sense fortunate that recordings have now made possible a non-exclusive experience of opera, one that, although belonging to a technological present, recreates in an alternative domain the freedoms enjoyed by live audiences in opera’s past. Listeners can (and do) fast forward through the DVD or pre-select their tracks on a CD or iPod; they come and go as they please, dropping into the experience and keeping the lights switched on.

The question of opera and boredom also arises as a broad historical conundrum: why did the composing of opera, which for so long belonged to the present, become around the time of the Second World War a gesture to the past? Until about 1800, most operas were written for a specific season in a specific theatre (only French tragédie lyrique had something approaching a repertoire); some works might be repeated in other cities a year or two later, and might remain in performance or be revived a little longer than that, but they were generally expected to be supplanted by next year’s crop. In the middle years of the nineteenth century, this began to change: in parallel with increasing gravitas accorded to the very idea of musical works, and with a new historicism that sought out and preserved past music, operatic repertories began to emerge and then harden. By 1850, it was the norm for an opera from the previous decade to be revived and performed alongside recent revivals and new, specially commissioned pieces.

This situation continued into the earlier twentieth century, with the proportions of old to new shifting inexorably towards the former; and then at a certain point – although the point is not exact, different operatic cultures negotiating the turn in different ways – the operatic repertory became primarily a matter of revivals of canonized works. One sees, in parallel to this shift, evidence of another kind of boredom: a gathering sense that the body of known works is finite and too well-known, but that refreshment is more likely to come from the past than the present. Early examples of this process emerged more than a hundred years ago now, and in more than one country almost simultaneously: the championing of Monteverdi’s Orfeo by Paul Dukas in 1893, or Gustav Mahler’s completion of Carl Maria von Weber’s unfinished opera Die drei Pintos in 1888. The so-called ‘Verdi renaissance’ in 1920s Germany, the re-emergence of Mozart operas in the 1930s, the bel canto revival of the 1950s, the Rossini revival of the 1970s and 1980s, and the Handel revival of the past two decades are all more recent forms of this historicist passion.

There is an interesting convergence here: just as the commissioning of new operas dwindled to a trickle in the years leading up to the Second World War, the staging of canonic operas became an art in its own right. Before that time, staging had often been regarded as an important part of the operatic occasion, but one that aged along with the work in question. With staging now thrown open as an interpretative extra, as a ‘reading’ of a well-known text, new questions emerged about the role of directors, who seemed now to bear much of the burden for renewing works that could not be new in any other way. A classic instance is Wieland Wagner in postwar Bayreuth. Richard Wagner’s grandson was given the task of purging both the Bayreuth festival and Wagner’s operas of their visually embodied past, in particular the so-called realistic stagings that, although updated in the 1930s, had barely mutated since Wagner’s day, and had become strongly associated with the Nazi regime. Wieland Wagner’s visually stunning minimalism seemed to make the works themselves new; and this form of magic has been practised both well and badly ever since. Directors took on the responsibility for making canonic operas seem fresh or contemporary – if you will, for assuaging boredom. Staging became more assertive and visible in the experience of opera. This in turn raises questions about the role of performance in the operatic experience, and the limits (or infinite possibilities) of opera production in the twenty-first century.

Imagine you are fortunate enough to attend revivals of Verdi’s Un ballo in maschera (1859) in London, Milan and New York. The performances are, as you might expect, all different – different casts, conductors, orchestras, staging teams – but you have expectations of both sameness and difference. You will trust the singers, despite their individual quirks, to sing a more or less identical musical text. True, a phrase here or there may be interpolated to display what used to be called the ‘money notes’; and – more often than one might think – there may be a discreet transposition to ease a singer afflicted by the depredations of age or nerves; but these are matters of small detail. You will also look forward to a roughly identical literary text, although again there may be wider variation than you might think for such a classic of the repertoire. Some performances may choose to substitute aspects of an original Swedish setting, thus replacing an occasional place or character name; and there is a notorious line in the first scene (the Judge describes Ulrica as ‘dell’immondo sangue de’ negri’ – of filthy negro blood) which is often censored. You will trust, nevertheless, that the conductor and orchestral musicians have more or less identical parts before them – although again conductors routinely adjust details to suit the theatrical space and aid audibility.

But what of the staging of these three productions? Your expectations of sameness in that department will, if you are a seasoned campaigner, be of the most modest imaginable, and just as well. In London there is no scenery except a naked light bulb (which swings incessantly), a crooked doll’s chair and an equally crooked bed suspended halfway up a bare wall – a bed from which the soprano precariously hangs to sing her entrance aria. In Milan, eighteenth-century Stockholm has been re-created with no expense spared: real horses pull real carriages, stately ships pass in the brilliantly lit background; a many-Euroed hum is heard everywhere. And in New York the staging is what die-hards proudly call ‘traditional’, which means it resembles productions that took place when the people bankrolling the present show were young enough to enjoy newness: a kind of mid-twentieth-century pastiche or – more accurately – a mid twentieth century frozen in aspic.

Now imagine a different world, one in which – through some unimaginable warp in our civilization – the staging world is as fixed as the musical or literary one. In just the same way that there is a libretto and a musical score, there is now also a ‘book’ that instructs us about how to stage Un ballo in maschera: what the sets and costumes should look like, who stands where, what gestures the characters are permitted to make. As with the musical performance, tiny variations or inflections can now give pleasure or cause indignation. How expressive and daring! Riccardo was wearing mauve tights instead of red ones! I’m not sure what it meant exactly, but somehow it seemed to fit marvellously with that slightly off-colour door at the back of the Act 1 set. At the end of his Act 3 aria, though, Renato raised both hands when he turned to Riccardo’s picture. That simply goes too far. He looked like he was parking an aeroplane. The text stipulates quite clearly that his other hand should be resting on his sword, and there’s a very good reason why it should. These modern directors! Sharpen my quill: I must write a letter to Opera magazine.

This landscape, one in which ‘same-staging practice’ is the norm, is of course fantastic: a kind of Groundhog Day made opera, in which the same actions will scroll past us again and again – more like a movie rerun than live theatre. But we conjure it up because it may add a new perspective to the endless modern debates about ‘contemporary’ operatic staging, in particular about whether directors should be encouraged or vilified when they offer audiences a sharply modern – and often sharply contradictory – stage-reading of operatic classics. The point is this: that the existence of visual extravagance has almost always been a critical issue in opera, and visual splendour has sometimes been more important than either the words or the music; but the precise nature of that splendour in any individual work has always changed with changing technologies.

To return to the beginning: opera is a form of theatre in which most or all of the characters sing most or all of the time. This continuous singing is an odd state of affairs, and we should above all hang on to that idea of opera’s strangeness, its special qualities. One of the most common ways in which writers have tried to make sense of opera is to domesticate it, to talk about it in ways that encourage comparisons with other art forms, ways that make it less strange. That approach can work, and in what follows we occasionally do something similar. But there will always be a lingering question about how an art form that by any other standards is almost bound to seem preposterous can inspire such intense emotions. We need to keep that question, and those emotions, before us in our journey through opera’s 400-year history.