16

Realism and clamour

La traviata and Carmen: both are realistic in their fashion. Under the influence of the latter, realism became an operatic buzzword in the late nineteenth century – and in several countries almost simultaneously. But the large differences between the types of opera that sailed under a realist flag should make us wary of taking the term too literally. As we wrote in the very first chapter, opera is in a basic sense not realistic – operatic characters go about their business singing rather than speaking. To compound matters, the idea of realism in the arts is famously problematic even when applied to literature and fine art, in spite of its extensive history and obvious purchase in those forms. Linda Nochlin starts her classic survey of nineteenth-century artistic realism by identifying ‘a basic cause of confusion bedevilling the notion’ in the movement’s ‘ambiguous relationship to the highly problematic concept of reality’.¹ In opera – in any genre that involves music – these problems become more confusing still. But realism keeps emerging as a slogan in the history of opera, usually as a means to attack yesterday’s operatic practice, an agent in those continual attempts to reform opera, discipline it, rein it in, purge it of the excesses yesterday is deemed to have committed. The late nineteenth century was no exception; indeed, realism was summoned to the cause as never before. One reason was that late nineteenth-century operatic rebels needed all the slogans they could muster to distance themselves from the immediate past. The ‘yesterday’ they were confronting loomed before them, imposing as never before, in the form of three shadows cast by three puissant giants: Giacomo Meyerbeer, Giuseppe Verdi and – largest and darkest of all – Richard Wagner.

The idea of operatic realism is, then, historically important, but is always a relative term, potentially meaning many things. It could involve realism of acting, of a singer entering into the character rather than simply performing in costume. Both Wagner and Verdi fought strenuously for reforms in this direction. Singers, they argued, should lose themselves in their roles, should become the figures they impersonate, allowing the audience to absorb the illusion that the world on stage exists and matters. Again, though, we need caveats: calls for – or elaborate praise of – engaging operatic acting predate the later nineteenth century. They can be found, for instance, in critiques of the singing actresses who populated French tragédie lyrique during the eighteenth century.

There were other forms of realism through which the fundamental strangeness of opera came under challenge. One of these involves time – specifically the suspended time of operatic numbers or parts of numbers, in which nothing aside from singing is happening. In most opera before the later nineteenth century there are places in which action or discussion is set aside, in which all energies are devoted to sheer music. The performance keeps unfolding, but time in the stage world stops. During these freeze-frames, one might imagine a kind of psychological truth at work – at moments when bliss or shock is extreme, the world may seem to pause for an instant. Operatic numbers can take hold of that instant, making space for a large musical object that will occupy several minutes. Wagner was most radical in breaking this down. He had the insight that characters’ bliss or shock could be extended almost endlessly, meaning that one didn’t necessarily have to pause fictional time. Tristan und Isolde stages a continual unfolding of intense emotional states: as its composer famously said, it stages the art of transition.²

Wagner was not alone in attempting to reorganize operatic time by favouring continual dialogue and stage action, by reducing the libretto’s pauses for poetic roughage consumed by soloists or multiple characters simultaneously. To a greater or lesser extent, all late nineteenth-century composers felt this pull towards continuity. In the last chapter, we discussed how Verdi’s final operas, Otello and Falstaff, crossed an intangible divide, realigning the relationship between ‘set piece’ and ‘dialogue’, making the first an interruption of the second rather than an inevitable end-point. The change particularly called into question solo arias, which had been the staple of opera for two centuries and more but which now became exceptional events. Grand duets and larger ensembles, numbers in which a looser idea of musical conversation could occur, inexorably replaced them. And the shift from arias to duets, from opera as soliloquy to opera as dialogue, often went hand in hand with the gradual acceptance of prose as opposed to verse libretti. Wagner was yet again enterprising in this regard, especially in his four-opera extravaganza Der Ring des Nibelungen (1876). With its marching army of arrogant Gods, sword-brandishing heroes, cavorting water nixies and sweaty dwarves, the Ring could hardly be thought realistic in terms of its story. It is, though, far more realistic than, say, the human tragedy of Verdi’s Rigoletto (first performed in 1851, around the time Wagner started working on the music of the Ring) in letting time in the stage world flow relatively unimpeded within each scene. These innovations profoundly affected the musical shape of opera, changing the relationship between musical time and fictional time, bringing them into a more continual alignment.

Wagner was also a musical realist in his love of noise or, as he would have preferred to put it, of natural sound. Scenic music, music to accompany stage effects or stage pictures, had existed in opera for quite some time. But such music typically acquired conventions that assured it a readily understood relationship to the world depicted. Rossini’s pastoral music in Guillaume Tell (1829) draws on imitations of real sounds (the Swiss cowherd’s horn) to create local colour; but mostly such descriptive music works by behaving correctly as a musical code that resonates with emotions traditionally associated with pastoral scenes – peace, a sense of simplicity and timelessness, benign lassitude. Wagner often worked in the same way, by what we might call ‘mediation’. The orchestral interlude (known as Siegfried’s Rhine Journey) between scenes 1 and 2 of the first act of Götterdämmerung (1874) is understandable as geographical progress through a landscape mostly because its leitmotifs – Siegfried’s horn call, the ‘Fire’ motif connected to Brünnhilde’s place of exile, the Rhine music – were close to traditional musical representations of hunting scenes, fire and water.

However, there was also a more direct way to represent the locations inhabited by operatic characters, and that was by imitating natural sounds and adding almost nothing. This is the unruly music, at minimal distance from the thing it represents, that we saw in the Pilgrims’ Scene in Act 1 of Wagner’s Tannhäuser. Another example is from his Der fliegende Holländer (1843), in which musical storms are elemental as well as symbolic: they contain not just the conventional minor-key tremolos, but the pure shrilling of wind in rigging (flutes and piccolos) or claps of thunder (timpani with hard sticks). Verdi was equally adept. In Rigoletto, the storm scene in the final act uses distant human voices, singing with closed mouths, to create an uncanny effect of unruly wind; in Aida, Act 3, he invokes a kind of noisy silence (a starlit night on the banks of the Nile) by means of muted strings playing a single note in many octaves, pizzicato, tremolo and with harmonics. When it came to the human voice, Wagner was sometimes even more direct: he sometimes instructed his singers to scream – not to sing a scream written into the score as notes, but to scream out for real, to break the musical shell. This was as radical as any of his more famous innovations. Noise can reorder the balance of power in opera, changing both singing and the way operatic music is written.

BORIS AND DIVOS OF THE DEEPS

Sheer noise and a resistance to frozen moments are, then, the advance guard of operatic realism: before libretti shifted to prose, or adopted the urban poor as favoured subjects, natural sounds and continuous dialogue were signalling a more fundamental change. Boris Godunov (1869, revised 1874) by the Russian composer Modest Musorgsky (1839–81) is in this sense one of the first and most radical late-nineteenth-century realist experiments. Musorgsky’s formative background was the period between 1830 and 1850, which saw the establishment of a number of self-consciously national operatic traditions, in particular those in Russia, Poland and various parts of the Habsburg empire, notably Hungary. All these areas had seen vernacular opera during the eighteenth century, but the emergence of a ‘national opera’ was, as in Italy and Germany, intimately bound up with the process of cultural nation-building undertaken by the expanding middle classes. In several cases one can identify key operas that managed, more by dint of multiple performances or association with political events than by their occasional use of folk materials, to collect around them a miscellany of musical and literary motifs that could function as symbols of a nation. The process here is important, and is often misunderstood: rather than appropriating an already existing fund of national musical material, these operas typically tended to construct that material – becoming (as Verdi’s would in Italy) ‘national works’ through the cumulative acts of national reception they underwent.

A good case in point, and the earliest of these national operatic figures in Russia, is Mikhail Ivanovich Glinka (1804–57). His A Life for the Czar (1836), which describes itself rather grandly as a ‘patriotic heroic-tragic opera’, is in some ways a ‘rescue’ opera in the style of Cherubini’s Les Deux Journées (1800), and also shows more than a hint of Rossinian influence, doubtless deriving from Glinka’s Italian travels in the 1830s. Set around the seventeenth-century figure of Ivan Susanin, a peasant fighting against Polish invasion, the score makes one or two gestures towards folk material, but most of its Russianness derives from an urban tradition of salon music. The newness in Glinka’s work, however, comes through the way in which this material, which had been used often enough in earlier works as ‘local colour’, inhabits the core of the drama, in particular during climactic moments. The novelty and importance of A Life for the Czar was very quickly appreciated, and the opera is to this day regarded as a watershed in the development of Russian music. Glinka’s second opera, Ruslan and Lyudmila (1842), has never been as successful, although it was much imitated by later Russian composers, who developed its fairy-tale and Orientalist themes.

Against this background, Musorgsky will sound more radical than ever. The second scene of Boris begins with bells: not real bells but a fearless imitation, an alternation of functionally unrelated chords (based on a non-tonal collection of notes called the octatonic scale, a favourite device of advanced twentieth-century composers such as Bartók and Stravinsky). These dissonant masses seem extraordinarily alien for 1869; they make little sense within the musical grammar of the time. The sense they do make is by reference to the real sounds they imitate: those deep bells that peal forth on solemn ceremonial occasions, and which, more pertinently, create discordant overlapping pitches and overtones – Musorgsky does nothing to prettify the sound. Soon afterwards, real bells also peal forth, the point thus being made explicit. Those opening, octatonic bell-chords, with their shadowy kind of musical sense, are on the cusp between music and the bells one would hear in the everyday world. Musorgsky’s bells are, in other words, a portent of fin-de-siècle things to come. Wagner – without knowing a then-obscure Russian opera – made his own version of atonal bell noise in Act 3 of Parsifal: the cruel voices of the grail knights reproach Amfortas in tolling, dissonant waves that turn bell sounds into an uncanny form of communal song.

In Boris, the clamour of the bell music is one of many blunt moments that testify to Musorgsky’s drive towards frankness, his displeasure with opera’s fabulous and frivolous aspects. Although his choice of subject matter and – occasionally – musical idiom has reminiscences of Glinka, his attitudes to such material were very different. Rather than hammering some literary source into a proper libretto, with well-behaved poetry and ample opportunities for frozen moments, Musorgsky carved words directly from an earlier play by the poet Alexander Pushkin, itself based on events from Russian history. The explicit intention of both poet and composer was didactic – to instruct and educate by issuing warnings from the past. The setting is the late sixteenth century, and in that bell-resonant second scene Boris Godunov (bass) is about to be crowned Tsar. But he harbours a guilty secret – he has committed murder on the road to power. At the end of the drama, wracked by guilt, fearful for his young son and threatened by a Pretender to the throne, he succumbs to a terrible seizure.

The libretto in the original, 1869 version of Boris allows for only a few traditional operatic situations; there are instead many long conversations and some monologues, all cast as musical prose that ebbs and flows with the words. For the revised version, staged at St Petersburg’s Mariinsky Theatre in 1874 (the 1869 version was rejected by the authorities and not performed until the Soviet era), Musorgsky was encouraged to bulk the piece up with more conventional operatic fare, and did so with some enthusiasm. Along with other, more piecemeal changes, he added an entire act set in Poland. It has a mazurka-style aria for the heroine Marina (soprano), a brilliant formal Polonaise and a lengthy love-duet finale between Marina and the Pretender Dimitry (tenor), complete with harps and interlocking voices. In such passages, Boris sounds like grand opéra in its Russian incarnation. Sometimes the influence came via predecessors such as Glinka and his A Life for the Tsar, but at other times it is more general and direct: the historical/political subject matter, the determined variety of musical styles, or the Coronation Scene with massed tableau and scenic glitter. There is, though, one sense in which Boris (in whatever version) makes a deep commitment to an eternal operatic truth, and that is in its celebration of the alpha singer. Musorgsky manipulated Pushkin’s Boris in order to make his opera a showpiece for a star bass, a landmark in the repertoire of that rare creature, the divo of the deeps. The work’s performance history has been punctuated by a magnificent procession of these prodigies, from Feodor Chaliapin to Alexander Kipnis, from Boris Christoff to René Pape. Recalling its attractions for the star singer can caution against overemphasizing the radical aspects of Boris, a move that often goes hand in hand with exoticizing Russian music generally – keeping it separate by extolling its palpable difference from the Western European mainstream.

We could underline that caveat by noticing that Boris mixes up past traditions without reserve, putting experiments that were sui generis next to a splendid variety of opera-as-usual. For this reason, the work eludes easy categorization according to historical position or genre, but probably fits better in the fin de siècle than it does as an appendix to grand opéra. Partly this is because of a further aspect of Musorgsky’s style which could, like those bell chords, attract the realist label, and which could also be thought a Musorgskian lance into the operatic future. During the creation of Boris, its composer described his operatic aesthetic thus: ‘my characters speak onstage as living people speak, but so that the character and force of their intonation, supported by the orchestra, which is the musical background for their speech, hit the target head-on; that is, my music must be an artistic reproduction of human speech in all its most subtle windings’.³ Tracing the literary origins of this credo can uncover an interesting journey, one that (although Musorgsky cannot have known it) can lead back via a circuitous route to the ideals of those sixteenth-century Italians whose theories helped create the first operas.⁴ They too had insisted on fidelity to the flow of emotions and the rhythms of speech, and they dubbed the new style recitar cantando – singing recitation. But that had been three centuries ago, three centuries during which, despite various attempts to stem the tide, the cantando part of the operatic equation or, more broadly, the musical aspect taken as a whole, had been in almost continual ascendancy. The political, didactic aspirations of Musorgsky and some of his Russian contemporaries (a group of reformers who were dubbed the ‘Kuchka’ – literally the ‘little heap’ or ‘little group’) caused them to strive for a new kind of declamatory realism, one in which the rhythms and cadences of Russian speech would be etched on to their operatic style. Although these attempts had little immediate influence, their later reverberations were powerful and would reflect the operatic attitudes of many of the greatest late nineteenth- and twentieth-century composers.

Boris’s death scene, which ends the 1869 version and became the penultimate scene in 1874, is to be sure not just a musical version of speech. It is better thought of as a virtuoso demonstration of the powers that can be invested in realistic sound, where little of what we hear seems to involve musical artifice or clear formal construction. The first section of the monologue, as Boris addresses his son, Fyodor (mezzo-soprano), draws on all the lyric authority of the bass voice, in the form of short, individual melodic waves for virtually each line of the text. When Boris appeals to God for mercy, his almost Schubertian lyricism gives way and the voice climbs to its highest register – to near-whispers with high tremolo strings. But the strange bell chords from the Coronation Scene return and, from that point until shortly before the end, the music is nearly all meant to be audible to the characters onstage, including the bells and an offstage chorus of mourners. Boris starts singing with them, winding his voice through them and around them.

Whenever this happens in opera – whenever characters respond musically to sounds around them – there’s a sense in which borders have been erased and reinvented. It’s not just that the characters can hear music being produced nearby: that, after all, is true in all those let’s-sing-a-song-in-the-inn situations. But Boris responds to his ambient music by improvising a sung counterpoint. In one sense we are very far from realism. What human being, at a moment of extreme emotion, would hear a hymn in the background and respond by singing his words in loose concord? This part of the death monologue nevertheless marks an important aesthetic turn for opera, one in which an old certainty – that operatic characters don’t know they are singing – dissolves, giving rise to a fruitful confusion. And this basic compositional device – using background music or background sound as the anchor, with characters singing freely and conversationally around it – becomes a widespread technique in fin-de-siècle opera. It’s a kind of musical curtain against which sung conversations or monologues can be held, with the naturalness and only-half-sung quality of the vocal lines anchored the richer music in the (apparent) distance. Act 2 of Puccini’s Tosca proceeds in this way for a long stretch, when Tosca – who is offstage – sings a cantata for the Queen while Scarpia and Cavaradossi exchange rhythmically free insults and information in the foreground. There was one significant earlier model for the trick that all these composers would have known: Mozart’s comedies, in particular Don Giovanni, where in both finales dances or aria arrangements are played on stage while characters eat, whisper, conspire, joke and converse in singing over the musical background.

The quirks and habits that have arisen over more than a century of Boris performances reflect a continuing sense of the opera’s peculiar brand of truthfulness, its reality factor. As anyone who has seen the work will know, it has become de rigueur for the principal bass to give his spectacular all and (in excess of the stage directions) cause the stricken Tsar to keel over in front of his throne. Those massive bodies crashing to the floor, the very sound of it, the fearlessness of the gesture, is a perpetual shock. As the great Russian director Stanislavski advised his Boris in 1928, ‘do not tear at the collar of your shirt to show you are suffocating. That is what all the other singers in this part do and it is just a stale cliché. Lean forward and fall as an ox does when he is butchered.’⁵ So now they all do it, and gasps are always heard. But there is another performance tradition which, though less overtly shocking, reflects this fearlessness in another way. For his 1874 revision, Musorgsky added a local colour number right out of opéra comique, a rustic song for the blowsy Innkeeper (mezzo-soprano) in Act 1, scene 2. The Innkeeper’s role has become a favoured destination for great divas in their twilight years, and is often sung by genuinely ravaged voices, with a roughness appropriate to the character’s identity. Martha Mödl (1912–2001), for instance, whose roles at Bayreuth in the 1950s included Kundry, Isolde and Brünnhilde, appeared as a half-comic, half-tragic Innkeeper in performances of Boris in Munich in the 1970s and 1980s. Musorgsky’s unreserved attitudes towards realistic singing seem, in other words, to call up fearlessness in performers as well, attitudes that cause the aesthetic world of bel canto to seem an exceedingly distant memory.

ONEGIN: SALON OPERA

It is in some ways fitting that the old woman with the ravaged voice, sung by those not afraid to demonstrate the toll taken by years of service in large theatres, will recur frequently in twentieth-century opera. Clytemnestra in Richard Strauss’s Elektra (1909), opera’s Queen of Decadence, is also played by ex-Brünnhildes whose range and beauty of tone have disappeared for ever. Astrid Varnay (1918–2006), a star Wagnerian soprano of the 1950s and 1960s, can be seen and heard stealing the show as Clytemnestra in the 1984 German television version directed by Götz Friedrich. Two decades before Strauss, Pyotr Ilyich Tchaikovsky (1840–93) created a similar role in his penultimate opera, The Queen of Spades (1890). The plot, based on an 1833 novella by Pushkin, is full of phantoms, phantasms and other inexplicable phenomena. An obsessive hero, Hermann (tenor), is in love with Liza (soprano), the aristocrat granddaughter of a haughty old Countess (mezzo-soprano) who is a decayed remnant in paint and patches. Hermann believes that the Countess has a supernatural gift and knows a secret combination of three cards that will always recoup a gambler’s fortunes should he bet on them. He invades the Countess’s bedroom at night and frightens her into a fatal heart attack before she can reveal the secret; nevertheless, she later appears to him as a ghost and names three cards. But her final card is wrong, as Hermann discovers in a disastrous game that ends with his suicide. The Countess, a grotesque character with no redeeming virtues, is invariably performed either by a brave ex-diva or by a young mezzo pretending to have no voice left. The ugly sounds the singer makes are meant to be an indication not just of the Countess’s age, but of her rotten soul. Mödl first sang the Countess at the Nice opera in 1989, at the age of seventy-seven. Varnay, who added the Countess to her repertory in 1984 at the age of sixty-six, referred to this and other post-Brünnhilde character roles as ‘silencing the heavy artillery’.⁶

Tchaikovsky was never associated with the Kuchka, and had – unlike them – benefited from a thorough professional training and thus greater acquaintance with the music of Western Europe. He is rarely a point-blank composer, and showed little interest in brash representations of noise. But despite the aesthetic gap, we can trace Boris-like flavours in Tchaikovsky’s operas, flavours that involve the many alternative senses of realism and the paradoxes that could arise from their musical manifestations. In Act 2 of The Queen of Spades, for example, there is an extended ballroom scene whose music is largely presented as taking place onstage. Besides the obligatory dancing, the guests also assemble for a complete performance of a small pastoral opera called The Faithful Shepherdess, written in the style of Mozart. Tchaikovsky’s music-within-music involves disorientating chronological displacements. Mozart-like music would be the proper contemporary style for Hermann’s actual era, since the libretto, by Tchaikovsky’s brother Modest, shifts Pushkin’s novella to the eighteenth century. But Tchaikovsky doesn’t write imitation Mozart. Instead he creates something that sounds like Mozart through the filter of lost time, music that may have been performed long ago but somehow is heard only now, after a hundred years. The dramatic situation is clear. This is a ball and The Faithful Shepherdess is a divertissement being played for real, right there on stage. The characters hear the dance and operatic numbers just as the audience does. But the strangely refracted music tells another story, suggesting that the experience of this real performance is a dream or hallucination. It is a wonderful trick, in its way just as radical as Musorgsky’s dissonant bell chords.

Tchaikovsky’s other great opera, Yevgeny Onegin (Eugene Onegin, 1879), was also based on Pushkin and was consciously written against the grain of Kuchka ideology, with no overt nationalism and not a hint of grand historical scenes, tsars, battles or marches. Onegin is essentially a chamber opera and tends to creak under the pressure of lavish production values or the starry, larger-than-life singers that typically inhabit Boris. This much can be seen from Tchaikovsky’s diffident promotion of the opera: it was first performed by students of the Moscow Conservatory; even when it reached the professional stage at Moscow’s Bolshoy Theatre in 1881, the composer continued to insist on modest forces and discreet gestures – above all nothing too theatrical. The plot is at heart a drama of sentiment. An impressionable girl, Tatyana (soprano), becomes infatuated with a world-weary older man, Onegin (baritone). One night, alone in her bedroom, she impulsively writes him a love letter; he rejects her with gentleness but more than a hint of condescension. Some years later, with Tatyana now married to elderly, doting Prince Gremin (bass), Onegin reappears and declares to Tatyana that he is hopelessly in love with her. But Tatyana will not desert her husband and Onegin ends the opera in pathetic despair. There is a subplot in which Onegin needlessly provokes a duel with his friend Lensky (tenor) during a ball scene, and then kills him in a dawn duel; and there are various interpolated choruses, including a group of peasants in the first scene who sketch in a rare moment of folkishness. Elsewhere, especially in the two ballroom scenes, the tone is determinedly urban and urbane.

The realist aspects of Onegin certainly include the restraint of Tchaikovsky’s subject matter – its intimacy and domesticity – and the fact that the music is written for singers of relatively modest skills. At the time, these could be seen as defects of realism. As one critic had it:

it seems to be the custom nowadays to contend that the modern domestic or social element is best suited for the requirements of an operatic libretto … to associate music with colloquialisms and the conversation of the nineteenth century appears to us the height of the ridiculous.⁷

But the opera’s best claim to the realist label is the way in which its musical ideas and forms are embedded in everyday language. In La traviata, as we saw in Chapter 15, waltz rhythms continually underpin Violetta and her progress through the opera. In Onegin, Tchaikovsky does something similar (it was in this sense fitting that the first Bolshoy production recycled old La traviata sets in its final scene): he constructs his most intimate scenes from refractions and repetitions of the light, salon music of his day. It is as if salon music, with its predictable phrases and periodic repetitions, is sounding somewhere nearby, with the characters singing against its background, with it or around it. This is in one sense like the device in Boris’s death scene, the difference being that the background in Onegin is not music played within the stage world. It is instead music that sounds as if it could be from the stage world, a background inaudible to the characters on stage but nevertheless influencing profoundly the way they sing.

Tatyana’s famous Letter Scene, one of opera’s great monologues and Tchaikovsky’s first inspiration for Onegin, is a slightly different case. During the scene, in which Tatyana gradually convinces herself to write her declaration of love to Onegin, the heroine’s emotional journey takes the form of four loosely defined sections. Each is related in its melodic shapes, yet each could be musically self-sufficient and almost function as an independent orchestral number, a concert piece. Tatyana’s vocal delivery is mostly conversational; because of the orchestral background she is free to linger over thoughts and indulge in sudden hesitations, to weave through the sound in a quasi-natural rhythm or to become lyrical in tandem. Significantly, each of the sections carries strong reminiscences of the public music elsewhere in the opera, music that was actually performed and heard in the stage world. So radical was this innovation that the first reviewers missed it completely, accusing the scene of being recitative throughout. Later audiences learned to hear the unusual musical flow differently.

THE FRENCH CONNECTION

The forms of operatic realism in the later nineteenth and early twentieth centuries often transcend local operatic languages, and – as Musorgsky and Tchaikovsky show – were not necessarily devised at the Western European centre, to travel by osmosis to outlying places. Nor was writing for voices against a heard or presumed stage-musical background the sole province of Russian composers. Indeed, if anywhere would seem the natural home of realist opera it would be France, origin of the earlier realist innovations in the other arts such as the novels of Gustave Flaubert and the paintings of Gustave Courbet. Whatever realism these two achieved was by no means confined to subject matter. One of the outcomes of Flaubert’s notorious lack of empathy, disgust even, with the characters he created in his most famous novel, Madame Bovary (1857), was to abandon grand rhetorical gestures in treating their vicissitudes; a related technique can be seen in Courbet’s avoidance of conventional ways of making his chosen scenes picturesque. Together with later developments in the novels of Emile Zola (who even tried his hand at libretto writing), and of course the enormous impression made by Carmen, the French background might seem as rich as anywhere.

But so far as opera was concerned there were two large obstacles. One was the continuing influence of chorus-rich grand opéra at its most inflated. Camille Saint-Saëns’ Samson et Dalila (1877) is one of the strangest stylistic concoctions along these lines to emerge from the period. Much of the first act betrays all too clearly the work’s origins in oratorio: there’s even a Handelian fugue for the Hebrews. The Philistines, as one might expect, seem to be having more musical fun, although their quasi-exotic inspirations in Act 3, which are deliberately trivial, risk sounding merely trivial. The saving grace of the opera is the love music of Act 2, in particular Delila’s famous ‘Mon cœur s’ouvre à ta voix’ (My heart opens at your voice), which remains one of the stalwarts of the mezzo-soprano’s recital repertory. Much of the rest of the act wallows quite enjoyably in the other large obstacle, which was (again) the influence of Wagner; in this case the seemingly irresistible attraction of those distinctive harmonies that characterize the steamiest moments of Tristan.

A revealing illustration of the difficulties faced by French composers of the post-Carmen era can be gleaned from listing the manners and modes tried out by its most successful composer, Jules Massenet (1842–1912). Massenet himself seems to have sampled almost every libretto tradition available: Goethe adaptation (Werther, 1892), grand opéra historical extravaganza (Le Cid, 1885), femme fatale from the exotic East (Thaïs, 1894), comedy in powdered wigs (Chérubin, 1905), high-minded sentimental fable (Grisélidis, 1891) and even a Wagnerian amalgam with magic swords, an eroticized knight and teleportation (Esclarmonde, 1889). The cornucopia can on occasions seem rather desperate and perhaps resonates with Massenet’s notoriously chameleon-like personality, his penchant for pranks and for pretending to be a dog or a monkey at fashionable parties.⁸ Among them, Chérubin is an object lesson in the fact that libretti owing nothing to Wagner’s plot models could nonetheless become places where Wagnerian acoustic shadows were all too obvious. While the libretto is a sequel to Beaumarchais’s Figaro plays, the music is a flood of free, melodious dialogue, standard procedure for post-Wagnerian opera around 1900.

To say that Massenet’s operas were merely a litany of types is, though, to underestimate the appeal of his music, which in its time was understood to have a distinctly feminine genius in its celebration of the soprano voice. French grand opéra began in the 1830s with its eye fixed on men and male voices, with the great all-male ensembles of Meyerbeer as the classic examples. But by the end of the century, and especially in the hands of Massenet, vocal preference moves to the opposite pole. In Esclarmonde, the hero (Roland, tenor) is overwhelmed by the sound of Esclarmonde’s voice calling to him in cascades of soprano coloratura. The battle lines are clearly drawn: just as Roland is about to yield, Catholic priests come thundering in with ‘In the name of the Father, Son and Holy Spirit’. This patriarchal task-force obviously fights fire with fire, making enough sound to blow the trilling heroine clean away. Such victories were, perhaps, all the more impressive when the heroine in question was a difficult customer. Massenet wrote the part of Esclarmonde for the famous American soprano Sibyl Sanderson, a ‘fair Californian’ with whom, one correspondent reports in 1894, the composer was enraptured. ‘Some had seen Massenet dining in a restaurant in the Rue Daunou with an American girl, accompanied by a lady who … was probably her mother’, reports the observer; Massenet was heard to enthuse at the same restaurant, ‘this girl has an extraordinary voice, from the G below treble clef to the G in the fourth line above’. That G – a whole step above the Queen of the Night’s high F – was dubbed ‘the Eiffel Tower note’ by Parisians of the time.⁹

Massenet made so many stylistic experiments that ‘realism’ was bound to be among them. Manon, first performed at the Opéra-Comique in 1884, is loosely based on episodes from an eighteenth-century novel by the Abbé Prévost, and so might seem – on setting alone – an unlikely candidate for the realist label. Fifteen-year-old Manon (soprano) is destined for a convent, but en route is spirited away by handsome young Des Grieux (tenor), thus evading both her venal cousin Lescaut (baritone) and a lecherous old aristocrat called Guillot (tenor). They enjoy a blissful if impoverished life in Paris before Des Grieux is abducted by his anxious father and Manon becomes the mistress of a richer man. The lovers get back together in the church of St Sulpice, with Des Grieux about to take holy orders. Des Grieux then gambles recklessly in order to keep Manon in luxury. He is accused of cheating by Guillot, and he and Manon are arrested. The final scene takes place on the road to Le Havre. Manon is to be deported; Des Grieux fails to save her, and she dies of exhaustion. As the reviewer of The Musical Times put it in 1884, ‘the story is painful and its atmosphere unwholesome’, unwholesomeness having by then become a code word for realism amongst the disapproving.¹⁰

Manon had stock-in-trade opéra comique scenic favourites such as drama in church and drama in a gambling house, not to mention an eighteenth-century setting with attendant chances to write pastiche and so evade for a time the lure of Wagnerian sounds. However, and although Manon has a standard poetic text, Massenet’s self-confessed mode of composition – to memorize the words and repeat them endlessly until the perfect melodies emerged from the individual rhythm of each phrase – meant that his opera sounds very much as if it were mined directly from a prose libretto. One need only sample the soprano heroine’s first number, the famous ‘Je suis encore tout étourdie’ (I am still completely overcome), to hear the freedom of word-setting. Some lines, such as the first, are sung with breathless or lingering pauses (‘Je suis … en-core … tout é-tour-di … e’), others are rushed through as fast as possible, others furnish occasions for long-held high notes. When this unpredictable rhythmic surface is matched by a form that seems improvised on the spot – by sudden, seemingly unprepared repetitions and by harmonic progressions that often unfold in metrically unexpected ways – the overall effect is very close to musical prose. Such qualities may have been what inspired Thomas Beecham’s most famous provocation, ‘I would give the whole of Bach’s Brandenburg Concertos for Massenet’s Manon and … think I had vastly profited by the exchange’.¹¹

The overtly realist declamation might possibly gesture towards Wagner (in form if not in manner), but less ambiguous Wagnerian signs are also present. The hesitant melody of ‘Je suis encore’ is repeated twice during the aria (to give it some graspable form), and then reappears later, played by the orchestra and underpinning other vocal sentiments, suggesting that it is a subterranean voice with a message. Indeed, the entire opera is somewhat weighed down with leitmotifs, melodic ideas that recur out of duty rather than of necessity. Perhaps this is just another illustration of the extent to which French opera at the time had become enmeshed in Wagnerian fantasy.

Act 2 of Manon, which shows the lovers living together in Paris, contains two of Massenet’s most famous arias, moments in which he emerged from the shade of any looming influence to produce tiny gems of musical sentiment. In the first, Manon knows of the plot to take Des Grieux away from her, and also knows there are other, wealthier lovers waiting; left alone, she bids their room a wistful farewell in the aria ‘Adieu, notre petite table’ (Farewell, our little table). As in her earlier ‘Je suis encore’, there are repetitions of the opening phrase (in the middle and at the close); but these are not placed as formal markers (at the start of new verses) but give the impression of loose, almost unintentional returns to past expressions – repetitions that are, if you will, conversational rather than rhetorical.

This mood is then matched by Des Grieux, who summons up a dream-like evocation of their life together, ‘En fermant les yeux’ (On closing my eyes). Although this is one of the most famous arias in the tenor repertory, it is again restrained in its poetic language:

En fermant les yeux je vois

Là-bas une humble retraite,

Une maisonette

Toute blanche au fond des bois!

[On closing my eyes I see / down there a simple retreat, / A little house / all white in the depths of the wood!]

The aria is marked by an almost constant accompaniment figure, high oscillations in the muted violins, to which Massenet adds, for further brightness, flute and oboe. Unlike those over-determined leitmotifs discussed earlier, this figure is vague in specific meaning. It resonates with the simplicity of the house ‘Toute blanche au fond des bois’, and later in the aria with babbling streams and joyous birdsong. But part of the oscillating figure’s effect is that it resists firm assignment to a visual image. There are virtually no bass notes in the entire aria, and the tenor melody is also restrained, as if the number might float off into space at any moment. The sense of evanescence comes via harmonic means. The accompaniment constantly swings between two chords, and the vocal melody – strangely repetitive, almost narcotic – also sketches harmonic gestures; but the two rarely combine, which again gives the aria its airborne feel. All this is Massenet at his most persuasive; and also – not coincidentally – his least Wagnerian. ‘En fermant les yeux’ probably resembles most closely those local colour invocations so popular in the final acts of Meyerbeerian grand opéra – in the end a far less threatening past tradition. But in its minute depiction of small detail and its lingering over simple, evocative images, it bears the epithet ‘realist’ more plausibly than most. Over the years it has, as have all great operatic inspirations, proved itself remarkably adaptable. In 1904, Enrico Caruso recorded it as ‘Chiudo gli occhi’, with piano accompaniment, ringing high notes and even a sob or two; in 1929, Julius Patzak’s recorded version, sung in German as ‘Ich schloss die Augen’ but still sounding nothing like Wagner, is orchestrally accompanied, much slower, and extraordinarily moving in its restrained vocal intensity.

VERISMO

Late nineteenth-century Italy is conventionally seen as the true home of operatic realism. The Italian term, verismo, had been applied to literature since at least the 1870s, being connected with a taste for scientific objectivity in low-life situations, notable in the works of the Sicilian writer Giovanni Verga. But in the 1890s verismo became associated with a new kind of Italian opera, and the term has stuck. The starting point is usually said to be Cavalleria rusticana (Rustic Chivalry, 1890), a one-act opera by Pietro Mascagni (1863–1945) based on a short story (and then play) by Verga: a lurid tale of infidelity and murder among Sicilian peasants. A companion piece emerged two years later in the form of Pagliacci (Clowns, 1892) by Ruggiero Leoncavallo (1857–1919), a story that involves the same elements as Cavalleria, but is set among a troupe of travelling players, and thus with an added spice that the violence – a jealous husband kills his unfaithful wife – is staged as a play-within-a-play. Both operas offered something new, in particular a directness of melodic and orchestral effect, not to mention a liking for lurid dramatic shock, and both made great headway internationally, particularly in Germany, where they were a welcome antidote to Wagnerian domination.

Many found them distasteful. An American essayist writing in the 1890s sums it up thus:

Simple means shake the spectator. The march of events rasps his nerves. Dramatic touches are really blows in their directness … phrases are short. The rhythm frets. Dissonances scream. There is feverish unrest. … Examine the librettos of Cavalleria rusticana, Pagliacci, A Santa Lucia, Mala vita, A Basso Porto, La martire. You will find them to be tragic episodes in low life. The characters are peasants, mountebanks, stevedores, drunkards, punks. The tragedy is the outcome of illicit sexual relationships. Animal passions rage and cry out. The elements are squalor, lust, and blood. The life depicted is short, brutal, and nasty.¹²

Before we imagine this as mere Puritan prejudice, we should recall that many of the old school in Italy also found this repertory boorish and offensive. The aging Verdi, for example, was dismissive, saying that it was far better to be like Shakespeare and ‘invent the truth’;¹³ but what he disliked most was the raw immediacy. The plots of both Cavalleria and Pagliacci owed something to literary verismo; and so the term began to be applied to them and to other, similar operas. But several early critics were sceptical of the verismo label. Partly this was because any realistic elements in the source texts tended to get submerged in the stubbornly old-fashioned, high-flown libretto language. But a more basic reason concerned prevailing Italian ideas about music aesthetics: the view that music was in essence abstract, and thus simply incapable of realism in the sense of truthfully representing human situations. Yet again there is the sense of an impasse: the sense that the term realism (or verismo) will always be problematic when applied to opera.

Stranger still, at least at first glance, is that verismo might refer – as it sometimes does in the history books – to the works of Giacomo Puccini (1858–1924), Italy’s greatest opera composer of the post-Verdi generation. Certainly Puccini’s first international success was as far from gritty realism as could be imagined. This was Manon Lescaut (1893), another setting of the Prévost novel that Massenet had used, and so an opera taking place in a distant, bewigged eighteenth century. In homage to its historical setting, Puccini indulged in lovingly fashioned, antique-style madrigals and gavottes at the opening of Act 2 – as with The Queen of Spades, the rococo setting seemed to cry out for pastiche. What’s more, both Manon operas anticipate a twentieth-century vogue for the exotic eighteenth century, and for nostalgic allusions to its simpler, more self-contained music, as in Strauss’s Der Rosenkavalier (1911) and Ariadne auf Naxos (1912/16). As an example of verismo – in the literary sense of urban blight or muddy peasants – Manon Lescaut is, then, bizarrely inappropriate. Puccini’s opera is nevertheless a powerful example of what the composer would bring to post-Verdian musical theatre, in particular of his almost infallible instinct for devastating and concise emotion, a talent that would make him for decades the most successful living operatic composer in Europe and far beyond.

Late nineteenth-century Italian operas tended to experience tortuous births, and Manon was no exception. Puccini declared himself inspired by Prévost’s novel, and was in all likelihood equally enthused by the success of Massenet’s opera. In the end, though, the libretto took three years and at least three librettists to complete. The result, as with so many libretti of the time, looked far less conventionally poetic, and far more like prose, than had those earlier in the century. What’s more, Puccini continued to tinker with the opera for decades after the premiere: there is no definitive Manon Lescaut, performers simply making a choice from among the variants (far from minor) chaotically on offer in competing versions of the vocal and orchestral score. In this uncertainty Puccini was very much of his time, betraying a difficulty especially severe in Italy, which had boasted the most imposing operatic lineage. In the first half of the century, composers such as Bellini, Donizetti and the young Verdi had worked within this centuries-old tradition. They could kick against the formal conventions, the showy cabalettas, dispiriting cori d’introduzione and elephantine concertati, but the fixed forms were nevertheless there to fall back on when inspiration flagged. By the 1880s, however, the joint onslaught of French and German influences, and the vogue for Meyerbeer, Bizet and the theories of Wagner, had virtually destroyed this predictability. Each opera had to create its own formal world, define its musical and dramatic terms uniquely. An aria, far from merely freezing the action and taking on a well-tried form, was – ideally at least – expected to assume an individual shape, intimately suited to the particular situation; orchestral timbre and even harmonic language should likewise be governed by the dramatic ambience. The rate of Italian operatic production slowed as composers picked over subjects, searching for a plan that would be at once effective and – the new watchword – original. Puccini offers a vivid illustration of such creative struggle. His maturity is marked by repeated periods of stagnation, compositional blocks in mid-opera and obsessive searches for new subjects. But, uniquely among his rivals, he produced an unprecedented number of operas that have survived to the present day. His most famous contemporaries, Mascagni and Leoncavallo, managed no more than one each; Puccini wrote at least seven. What accounted for his success?

We might start with a seemingly negative point. The most initially striking aspect of Manon Lescaut, particularly in comparison with Massenet, is its blatant discontinuity. Act 1 ends as young Des Grieux (tenor) has persuaded Manon (soprano) to escape with him to Paris, under the noses of her venal brother (baritone) and a raddled old roué called Geronte (bass). Act 2, though, opens with Manon in Geronte’s household, already bored by her pampered existence. A few sketchy narratives tell us that life with Des Grieux had been wonderful but penniless, and that Geronte’s money lured her away. Similarly, Act 3 ends with Des Grieux and Manon once more reunited, this time boarding a ship deporting her to America (on quitting Geronte at the end of Act 2, she tried to take her new jewellery with her: a bad mistake). But the start of Act 4 reveals the lovers staggering about in a vast desert outside New Orleans (geography becomes approximate in exotic locales). Puccini routinely drove his librettists, anxious as they were about their literary reputations, to despair. But he bullied and overrode them because he knew instinctively that modern opera didn’t rely on such trivial narrative coherence. What mattered was that each of Manon’s four acts had its own powerful individuality and dramatic shape.

Equally remarkable about Manon Lescaut is how easily it lays to rest those looming shadows that so much troubled others of the period. Admittedly, Wagner’s presence is powerful during the Act 2 love duet between Manon and Des Grieux, especially when the hero finally succumbs to Manon’s pleas for reconciliation. The falling sevenths that end vocal phrases, the progressions by chromatic sequence, the liberal use of interrupted cadences, the frequent appearance of the Tristan chord, the general orchestral colour: all conjure up a distinctly Wagnerian world. This atmosphere then comes to an apotheosis during the instrumental Intermezzo between Acts 2 and 3, in which the aspiring string melody and concluding cadences sail very close to plagiarizing the end of Tristan, Act 3. On the other hand there are moments in which Puccini showed that he could appropriate rather than be submerged by Wagnerian musical language. Most significant is the ‘love theme’ in the Act 2 duet, first sung to Des Grieux’s words ‘Nell’occhio tuo profondo’ (Deep in your eyes). The Wagnerian ancestry is clear, but now it points to the heroic Siegfried of Götterdämmerung. Wagner’s rhythmically robust melodic idiom serves here as a triumphant affirmation of love, a clearing of the air after the earlier, Tristan-like murkiness.

Just as telling in terms of the dominant trends in late-nineteenth-century opera, Puccini skilfully avoids the stock-in-trade of the old Italian school. There is just one gesture towards the classic Verdian way of doing things: the Act 3 embarkation scene, in which Manon and other female prisoners are called forth and consigned to their convict ship. This scene centres on an ensemble that recalls obliquely a concertato, the imposing centrepiece of Verdian drama. According to convention, the concertato was the greatest of all freeze-frame moments: it would begin with an extended solo by one of the principals; other soloists would join in, to comment or offer conflicting views; finally the chorus would add weight in a grandiose climax. The Act 2 finale of Lucia di Lammermoor or the Act 3 finale of Otello are two fine examples separated by half a century. In Manon Lescaut, though, the suspension of time is abolished. Instead of a principal’s lamentation, the opening is given to a minor character (the Sergeant, bass), who begins a slow roll-call; bystanders offer desultory comments as the prisoners move slowly, one by one, across the stage; Lescaut then mixes with the crowd, trying to instigate a riot. Only later do Manon and Des Grieux make their emotions heard and, in spite of an impressive climax, they return to muteness at the quiet close of the ensemble. The roll-call has continued throughout: there has been no ‘frozen moment’ in which the passing of stage time is halted.

Puccini’s avoidance of leisurely musical unfolding goes further in his solo arias, which tend to be fleeting ghosts of the formal Italianate model, arising seamlessly out of the surrounding texture and then disappearing back into it, with all the more emotional power for their brevity and lack of ostentatious beginning and ending. Beautiful nodes of musical attraction – more musically formal than the flow of dialogue around them – tend to coalesce gradually, as with the roll-call in Manon Lescaut, making us realize we’re in a set piece without knowing exactly how we got there. A famous example occurs in Act 2 of Tosca (1900), in which a trio emerges from nowhere when Spoletta (tenor), henchman of the evil tyrant Scarpia (bass-baritone), bursts in to announce Napoleon’s victory at Marengo. Given that Spoletta’s audience – Scarpia, Cavaradossi (tenor, who has just been brutally tortured) and Tosca (soprano, his lover, trying to save him) – are at the apex of a private and traumatic impasse, this political reportage might seem beside the point. But Cavaradossi, being a staunch republican, draws vocal inspiration from the news. Shouting ‘Vittoria! Vittoria!’, he launches into music of martial exultation, around which Tosca adds injunctions to prudence and Scarpia gloats ferociously. A very old operatic gesture – shocking news elicits ensemble reaction – is being replayed, but only for a moment; the trio passes away almost as soon as it has begun. A second example comes towards the end of Act 1 of Puccini’s last opera, Turandot (1926). Prince Calaf (tenor) decides to wager his life in a Riddle Contest for the hand of Princess Turandot (soprano). Timur (bass), Calaf’s aged father, and Timur’s servant, Liù (soprano), attempt to dissuade him. Calaf has declared his passion in a circling, minor-key phrase that suddenly becomes the scaffold around which Liù adds her voice, and then come Timur’s grieving exclamations. As if by magic, the three voices are together, and the circling phrase starts to sound like a collective dirge. It is over almost too soon, leaving a sense of amazement that something so swift could be so devastating.

Lacking elements of literary verismo but toying with realist devices in other domains: Manon Lescaut has long stretches of stage music, including most of the first half of Act 2, before the love duet, done as an eighteenth-century pastiche. And it maintains stage action in large ensembles. Puccini’s next and most famous opera, La bohème (1896), is different. Its setting is bohemian Paris, and the main characters are far from aristocratic. The heroine, Mimì (soprano), is a simple seamstress who gazes out over the roofs and is afflicted by tuberculosis; the hero, Rodolfo (tenor), is a poetic dreamer, trying to write grandiose plays but scraping a living as a journalist. Poverty, disease, hunger and cold are constant companions, even though these trials are seen through a sentimental lens that makes the tone very different from that in Cavalleria or Pagliacci.

But La bohème has another aspect that became important to fin-de-siècle operas and can certainly be related to realism: its extended soundscapes, crowd scenes depicting Parisian street life that sample many varieties of ambient noise without caring in particular about musical coherence. The start of Act 2 is knit from the cries of street vendors, conflicting choruses of citizens and children who traipse in and out singing or just shouting. Against this babble, the conversations of the principals emerge from time to time but can at any moment be barged aside by the crowd’s urgent noise. There are frequent changes in tempo, metre, dynamics and texture, as if the music were eavesdropping freely at this or that location in the crowd. Puccini may have learned this technique from Massenet (Manon has some good examples), and such scenes also became something of a French speciality. Another French opera that persistently received the realist label was Louise (1900) by Gustave Charpentier (1860–1956), which has two soundscape scenes. At the beginning of Act 2, there is an urban aubade in which the cascading musical cries of street vendors and workers at dawn do not just hawk wares but consider existential questions about life and the future. And the end of Act 3 sees a chaotic parade featuring the King of Fools, that stalwart of Parisian bohemianism who goes all the way back to François Villon. The reign of the King of Fools is the reign of carnival, and the anything-can-happen licence that prevails seems to liberate the music to choose whatever sounds it wants to, no matter how brief or extended, or disconnected to what was heard just before.

La bohème seems almost completely free from Wagnerism, something that – as we shall see in the next chapter – could be said of very few operas of the 1890s. This partly has to do with the subject matter: as Puccini said, La bohème was an opera of ‘small things’,¹⁴ of tiny objects delicately sketched. (One thinks of Massenet’s Manon and the farewell she bids to her ‘petite table’; La bohème is much more influenced by Massenet than is Manon Lescaut, in which the shared subject matter probably discouraged any fleeting reference.) Whereas the Ring deals in spears, swords and mighty ash trees, La bohème presents hand-warming muffs, bonnets and an unreliable, smoky stove. Each of these objects is delicately attached to a musical motif, but Puccini’s treatment of recurring themes is again distant from the Wagnerian norm. The very first motif in the opera is an energetic idea identified with the Bohemians, and derived from a student composition of Puccini’s entitled Capriccio sinfonico. The original title is apt, as the motif dominates the drama’s exposition (the first part of Act 1) and is caught up in a quasi-symphonic process of tonal tension and release. But there’s an important difference from Wagnerian usage: the motif’s shape is unchanging and is rarely heard in connection with other themes. Rather than functioning primarily as a semantic marker, as something attached to an object, it is fundamentally connected to gesture; it’s not the ‘bohemians’ theme’ but an accompaniment to moments when their energy dominates the stage; it returns only when that kind of energy is (however briefly) repeated.

The same could be said, in an even more remarkable way, of the very last chords of the opera, which have presented many academic commentators (though few listeners) with a famous motivic problem. Earlier in the act, one of the minor bohemians, the philosopher Colline (bass), decides he must pawn his overcoat in order to buy a cordial for the dying Mimì. In a gesture typical of the opera’s ‘small things’, he then addresses a tiny farewell aria to his overcoat, a solo that ends with solemn orchestral chords in parallel motion. This aria is important in its context: it acts as a brief moment of stasis, of lyrical contemplation, before Mimì’s death scene begins. But what are we to make of the fact that these solemn chords then return to close the entire opera, underpinning the final, desolate tableau with Mimì faded away and Rodolfo shouting in despair? Efforts to connect the two citations by means of semantic identification risk deflating the pathos. It can hardly deepen the impact of Rodolfo’s mourning to connect it with Colline’s distress at losing his overcoat. On the gestural level, though – the sense in which this motif announces the final tragedy and then marks its conclusion – no such strain occurs. Indeed, it may be the absence of semantic connection that makes the recurrence so telling. Puccini allows a space to emerge between words and music, a space in which musical drama could reside.

The famous arias in La bohème again involve a technique mentioned earlier: the ghost number, in which a musically substantial statement is summoned forth out of ambient threads and textures. When Puccini’s arias are performed in concert, they always seem short compared to the set pieces of earlier generations, and often require retrofitted musical frames to help them stand alone – a new introduction that is not a transition, or a new ending that really is an ending, and not the beginning of what happens next. Mimì’s autobiographical aria near the end of Act 1, ‘Mi chiamano Mimì’ (They call me Mimì), is an immortal example. Puccini’s word-setting means that the formal poetry hardly survives from a rhythmic point of view; more surprising is that, until nearly the end, the language is disarmingly prosaic. Rodolfo has, in the moments before, outlined his poetic aspirations in grandiose terms. Mimì lowers the temperature in her answering narrative, which starts in the most basic of ways. ‘Yes, they call me Mimì, but my name’s Lucia.’ Puccini sets this as a kind of question-and-answer. ‘They call me Mimì’ is harmonically and melodically unstable, left hanging in the air suggestively, like a continuation, not a beginning. This is then countered by the cadences and melodic closure of ‘but my name’s Lucia’, which is an ending, not an intermediate phase. It’s as if Mimì has two characters as well as two names, one that is ‘poetic’ and potentially tragic, another that is determinedly ordinary.

This alternation between the poetic and the prosaic characterizes the aria as a whole: not just because the ‘Mi chiamano Mimì’ motif turns up twice more but because the entire piece is taken up with such alternations. The grandest occurs towards the end. First there are further details of her simple life: ‘I make dinner for myself, alone. I don’t always go to mass, but I pray often to the Lord. I live alone, all alone, there in a little white room; and I look over the roofs and into the sky.’ All this is set to predictable rhythms and simple harmonies, with frequent uncomplicated cadences. Then something different emerges. ‘But when the thaw comes, the first sun is mine, the first kiss of April is mine!’ The words become conventionally poetic, and the music floods into one of Puccini’s great lyrical inspirations. Strings and woodwind double the melodic line and a rising sequence takes shape: ‘ma quando vien lo sgelo’ (first phrase); ‘il primo sole è mio’ (repeated, higher); and the third statement explodes into the wonderful melodic climax of ‘il primo bacio dell’aprile è mio!’ A lesser composer might have capped this with repetition, a bang on the drum and solicitations for applause. But Puccini gradually leads us back to the humble ‘Lucia’ music once more. By the end, Mimì is again expressing herself in simple recitative, as unassuming as when she started. ‘Mi chiamano Mimì’ is what Puccini was wont to call a pezzo forte – something he knew would make an effect. It’s also La bohème in a lyrical nutshell. The constant oscillation between the ordinary and the sentimental is what makes the opera as a whole so effective.

Tosca, which followed La bohème, took Puccini four years to write and is very different from its predecessor. The libretto was based on a recent Grand Guignol play by the Frenchman Victorien Sardou and is set in 1800 during the Napoleonic Wars. It features passion, blackmail and murder among Rome’s political and artistic elite. Baron Scarpia, one of the great Bad Boys of opera, is the despotic head of a repressive regime and uses his position to feed two great intertwined enthusiasms: sadism and lechery. At the end of Act 1 this demon gets caught up in a grand religious ceremony, creating a scene that would remain the most radical of Puccini’s many operatic soundscapes. While chasing a runaway convict, Scarpia guesses that the tenor hero Cavaradossi has something to do with the escape; Scarpia is in the church of Sant’Andrea della Valle and has just interrogated Tosca, Cavaradossi’s lover and the present object of the baron’s carnal ambitions. As two bells toll slowly in the background (‘in the distance, but audible’, marks Puccini in his score), a religious procession starts up. Various choral collectives sing Latin texts in quasi-liturgical monotones.

One extraordinary feature of this scene is its obsessive harmonic repetition, the bells providing two low pitches, Ba and F, which alternate for long minutes. From offstage, cannon blasts timed to the beat of the music are heard, signalling the prisoner’s escape. Puccini devises harmonies that wind around the bells’ fundamental tones but cannot depart from them. The Latin chanting fits around them too, as does an orchestral melody – necessarily a circular one – that in turn will join and underpin the soliloquy Scarpia delivers over all this rising clamour. Once again there is a sonic background, with a sung peroration in front of it, but here the background grows louder all the time and the explosions from the cannon, like the anvils in Wagner’s Rheingold, break the boundaries that limit operatic noise to nature’s murmurs and bring it into a sterner age. The baritone singing Scarpia has to put all his power into delivering the soliloquy, in which he imagines converting Tosca’s jealous fire into the passion of a willing lover. Finally, coming out of his lascivious reverie and recalling that he is in a church, he shouts out, ‘Tosca, mi fai dimenticare Iddio!’ (Tosca, you make me forget God!). As if in response to an unheard blasphemy, the choir bursts forth with a unison statement of an authentic Te Deum melody, Scarpia joining in. Just when you imagine things couldn’t get any louder, the full orchestra then blares forth Scarpia’s leitmotif (full of evil tritones, rasping brass and cymbals). When the curtain comes down, you almost expect the stage fabric to fall with a crash.

Surprising as it might now seem, early performances of Tosca often confused audiences and were excoriated by critics. One of the latter wrote:

The sonatinas and cantatas from the wings, and the organ, and the Gregorian chant, and the drums that announce the march to the scaffold, and the bells, and the cow bells, and the rifle shots, and the cannon fire, which at times constitute essential elements in the development of the opera, are not enough to fill the holes left by the lack of music.¹⁵

Another was sure of the opera’s fate:

In thirty years … Tosca, together with all the other operas of its type, will be an obscure and uncertain memory of a time of confusion in which music was subtracted, by the logic of history, from its own dominion, from its own laws, and from common sense.¹⁶

Most mysterious of all is a review of the first London performance that once more takes issue with raw sounds:

Tosca is too artificial, and when the composer wishes to be most intense, there is little save irritating noise – much sound with little musical sense. This remark applies chiefly to the second act. There are some who say that they best enjoy Wagner’s music at the theatre by shutting their eyes, and not being worried by what is taking place on the stage. In the second act of La Tosca, on the contrary, it is the sound of the music which seems to interfere with the undoubtedly strong dramatic situation.¹⁷

The mystery, of course, is why the musical noise, which is one of the most profoundly realistic gestures in opera, is here dismissed as ‘too artificial’ – as if opera as a genre had so firmly established its unique blend of naturalness and artifice that common sense can no longer apply. Of course, we can now smile at these critics if we choose, just as we can dismiss those who found nothing but recitative in the letter scene in Onegin, or thought its libretto a domestic bore. It may, though, be more productive to take their complaints seriously. The various attempts at operatic realism towards the end of the nineteenth century were indeed radical and unsettling, challenging as they did long-held ideas about music’s proper place in the operatic spectacle. But our second critic’s confident prediction, that such innovations would soon fade away, could not have been further from the mark. In realism’s wake some of opera’s strangest and noisiest moments were about to break confusingly on to the scene.