A Forgotten Composer and Work Revealed

The name of Félicien David was nearly forgotten to music history for about a century. In recent decades, though, an important life-and-works study was published by Dorothy V. Hagan (1985), and Richard Taruskin devoted several pages of his pathbreaking Oxford History of Music (2005) to David's influential 1844 concert work Le désert. (The latter is for tenor, men's chorus, orchestra and narrator; because of the mixture of spoken verses and orchestral passages, David dubbed the work an ode-symphonie.)Footnote ¹ Le désert is now available in two excellent recorded performances; similarly fine recordings also now exist of Christophe Colomb (his second and last ode-symphonie), his three piano trios, the four string quartets, dozens of his remarkable songs and the most long-lived of his four comic operas, Lalla-Roukh (1862), an enchanting work based on a famous book by the Irish writer Thomas Moore. Indeed, by 2014, one French journalist was proudly calling this once-obscure French composer ‘the revelation of the moment’.Footnote ²

The enthusiasm for David has spread beyond France, not least thanks to two imaginative and very different stagings of Lalla-Roukh: one, by Opera Lafayette, in Washington DC and New York City in 2013 (which resulted in the widely praised recording just mentioned); the other at Wexford Festival Opera in October 2022 (in a production received with delight by the audience and by prominent opera critics).Footnote ³

Even so, there was special anticipation for the recording – the first ever – of David's sole grand opéra (released in 2015).Footnote ⁴ The cast was headed by three major stars of the opera world: Véronique Gens, Karine Deshayes and Edgaras Montvidas, and the alert conductor was Hervé Niquet. Herculanum was composed for the Paris Opéra and was performed there no fewer than 74 times during the years 1859–68, to full houses and sustained enthusiasm. It was also staged in Brussels, Venice and St. Petersburg. But then it disappeared from sight, except for one aria: Lilia's stirring, hymnlike Credo (from Act III).Footnote ⁵

Today's opera lovers had little idea what to expect. The libretto is set in the Roman-controlled town of Herculaneum (French: Herculanum; Italian: Ercolano) in 79 CE, and one of the five major characters is Satan, yet the work was written by a composer most often described as having a gift for melodic charm, piquant orchestration and the evocation of places and peoples distant from Europe, such as Arabia and colonial Brazil. Fortunately, two scholars who have written about the work provided encouraging descriptions. Dorothy Hagan, while admitting that the work is full of ‘the stereotypes without which such an opera was unacceptable to its Parisian public’, praises David's ability to create ‘romantic, lyric scene[s]’ that powerfully combine ‘Italianate expression’ with, as she put it, ‘Saint-Simonian ecstasy’. (Hagan is referring to David's involvement, in his early years, with Saint-Simonism, a French early-socialist movement that emphasized the importance of solidarity across the borders of nation and social class.)Footnote ⁶ Alexandre Dratwicki, emphasizing other aspects, describes the work as ‘an example of the “second wave” of a genre that had been developed in the 1830s and that, beginning in 1850, had begun to come under the influence of the new post-Donizettian Italian opera style, more particularly that of Verdi’. Dratwicki draws particular attention to the vocal writing for Olympia and her brother Nicanor, which is forthright and emphatic yet still florid in ways that relate it to Rossinian norms from several decades earlier. And, he adds, David's grand opera carries out this complex task of stylistic and formal coordination ‘in a refined and easily comprehensible manner’.Footnote ⁷

Multiple reviews of the CD recording, often favourable, were published in France and many other countries. Some of the reviewers raised questions about the opera's convoluted and implausible plot. (Tastes have greatly changed over the past 150 years!) But they readily dubbed David's music ‘charming and tuneful’, or declared that a number was ‘lovely in its delicate melancholy’, or welcomed a vocal line's ‘coloratura escapades’. To be sure, several felt that David's music in Herculanum lacks dramatic intensity and psychological depth, a criticism that has long been applied equally to French grand operas by, say, Meyerbeer, Donizetti, Halévy and Gounod. (Though not to two other works in that genre: Verdi's Don Carlos and Berlioz's Les Troyens.) One reviewer complained that the volcanic eruption that brings the opera to a devastating close is accompanied by music that is too brief and insufficiently vivid – at least for home listening, in the absence of all requisite stage effects.Footnote ⁸

Alas, as with Meyerbeer's remarkable four French grand operas (Robert le diable, Les Huguenots, Le prophète and L'Africaine), many of these same critics tended to take for granted, instead of pointing out appreciatively, how imaginative the orchestral writing is, and how carefully suited the vocal parts are for each onstage character and for the two ‘peoples’ played by the chorus: early Christians (in Act II, scene 1) and pagan Romans and their vassals from other lands (in the rest of the work).Footnote ⁹

In 1859, the music and theatre critics of Paris likewise made their own range of pronouncements about the work. Ernest Reyer (1823–1909) and Hector Berlioz wrote at length and, for the most part, supportively.Footnote ¹⁰ The other critics tended to praise certain tuneful and touching solo moments, whether in an aria or a duet, but also often found faults. Scoffers objected that the chorus's ridicule of the priest Magnus at the end of Act I seemed like something out of a comic opera, or they found certain passages in the libretto or score too reminiscent of specific well-known works by previous composers. Indeed, the scene in which Lilia and Nicanor stand beside a large cross does clearly echo one in the last act of Meyerbeer's Robert le diable (1831) between Alice (a likewise innocent yet determined soprano) and Bertram (a likewise devious, diabolical baritone). Unfortunately, most French opera critics in the middle of the nineteenth century were literary figures with almost no ability to discuss basic musical matters. It is not by chance that Reyer and Berlioz, the work's most emphatic supporters, were themselves highly accomplished composers. And their support for the work seems to have been sincere. Neither of them was, so far as we know, friendly with David, who tended to keep to himself.Footnote ¹¹

The small-format book that comes with the CD recording features an important essay (by Gunther Braam) that offers substantial excerpts from many of the opening-night reviews. The book also reprints, as a separate chapter, large excerpts from Berlioz's thoughtful and laudatory report.Footnote ¹² But Reyer's, which was equally positive, seems to have escaped notice until now. It focuses, among other things, on David's attractive and memorable melodic lines.

Berlioz and Reyer were right. The work is a strong one, though of course also marked by the conventions of its age, and it has, rather suddenly, made a place for itself on the ‘mattering map’ of a new generation of performers, listeners and critics, not least through two equally skilful productions since the CD recording came out: namely, a staged version at Wexford Festival Opera (2016) and a concert rendering in Budapest (2021).Footnote ¹⁴ And the work deserves examination, in the context of the present special issue of this journal, as an intriguing instance of how a talented creative team renewed the longstanding tradition of ‘ancient world’ operas by combining that tradition with an element that was relatively new to the Parisian stage: early Christian history, seen as an instance in the long history of religious persecution.

Opera and the Ancient World, Opera and Religion

By 1859, ancient Rome was hardly a novel setting for an opera. From its very beginnings in the early 1600s, opera had often placed its plots in the worlds of ancient Greece or Rome. This brought a wide range of advantages. Those two civilizations, taken separately and together, were widely understood in Europe at the time (and eventually in the Americas) as being the fount of modern life and values. Greek democracy, Greek architecture, Greek philosophy and Greek tragic drama were regularly hailed as high achievements deserving of emulation, as were, for example, Roman engineering and Roman military organization.

In addition, ancient Greece and Rome seemed peculiarly appropriate antecedents to mid-nineteenth-century Europe because each of those two societies had been an expansive imperial power. Having an opera deal with, say, Alexander the Great's conquest of Persia or India or with the Romans’ occupying army in North Africa could cause the work to resonate with a variety of more current-day (though unnamed) European developments and ventures. Those current developments and ventures could include political tensions, or even warfare, within the home country (such as between the contentious religio-political factions in British politics during Handel's era); hostilities between the home country and other European lands (such as between France and the Triple Alliance – England, Sweden and the Dutch Republic – in 1668, relevant to certain Lully operas); tensions – or, again, military engagements – with more distant lands, such as between the major European powers and the expansive Ottoman Turkish empire (echoed in Mozart's Die Entführung aus dem Serail); or colonialist and imperialist efforts in the eighteenth and nineteenth centuries, such as those of Britain in the Caribbean and in India (in numerous operas by, again, Handel or – much later – in Delibes's Lakmé) or those of Spain or Portugal in the Americas (in, say, Offenbach's La Périchole).

Religion, by contrast, was a topic that, at least until the late 1700s, the makers of opera generally steered clear of mentioning, no doubt for fear of censorship by church and state and because opera often relied on direct patronage, whether by the government or powerful aristocratic families. True, some operas did bring up the question of non-Christian religious customs, such as the Hindu practice of satih. For example, in the last act of Metastasio's libretto Alessandro nell'Indie, a funeral pyre is set up for the purpose of immolating an Indian leader's surviving wife, but the Western leader – Alexander the Great – is horrified and forbids the custom, thereby saving the life of the main female character, Queen Cleofide. The libretto became the basis of operas by over three dozen composers, including Handel and Hasse.Footnote ¹⁵ By contrast, allusions to Christian worship and belief were rare or non-existent in most operatic traditions before 1800. Indeed, numerous librettists, during opera's first two hundred years in various lands, made sure that characters, when under the stress of a powerful emotion, would cry out to ‘the gods’, in the plural (for example, ‘grands dieux!’, ‘giusti dei!’, or ‘o Numi!’), rather than to ‘God’ in the singular and capitalized (‘Dieu’, ‘Dio’, and so forth), even if the plot were set in current- or recent-day Europe. Furthermore, plots drawn from the Bible (whether the Old or New Testament) or later religious tales (such as the lives of saints) were often banned entirely from the operatic stage, forcing them to find a different kind of home in the unstaged – or minimally staged – genres of the ‘sacred drama’ (a spoken work, usually with little or no music) and oratorio.Footnote ¹⁶

The near-ban on representations of religion – and of religious persecution – in opera began to crumble during the decades around the year 1800 and increasingly throughout the nineteenth century – and quite notably (if not uniquely) in France. The French Revolution's severing of the longstanding ties between church and state led to an increasing tolerance of religious difference, largely ending the longstanding tensions between French Catholics and French Protestants and allowing Jews and other ethno-religious minorities full access to citizenship. Public discussions of religion now regularly included warnings against such things as parochial attitudes (that is: viewing one's own version of religion as uniquely correct), prejudice against minority religious and ethnic groups, and the misuses of religion by individuals in positions of power. These warnings were dramatized in French operas of the day, works that were often set back a few centuries but that clearly carried a message about state- or nation-based religion and its often high-handed representatives in the current day. Three such works were among the most successful (both financially and artistically) of that time: Halévy's La juive (1835, set in Konstanz, in southwestern Germany, near the eastern border of Switzerland, in 1414), Meyerbeer's Les Huguenots (1836, set in France in the year 1572) and that same composer's Le prophète (1849, set in sixteenth-century Dordrecht and Münster).Footnote ¹⁷

Herculanum is notable for being a comparably successful work that combines the two trends that I have sketched above: that is, it engages explicitly with Christian history and doctrine and even portrays Christian worship onstage; and it engages with a truly distant past, specifically the Roman Empire in the year 79 CE.Footnote ¹⁸ The advantages were double. Bringing early Christianity onto the stage helped refresh the centuries-long practice of setting operas in ancient Greece or Rome. And placing a tale of Christians nearly two millennia in the past helped the opera evade the kinds of harsh criticism and official censorship that often awaited operas in which the events shown onstage were closer to the current day.

Méry and David's opera pits the Roman authorities – including the officials controlling its eastern provinces (in what are today Turkey and Syria: that is, the northwesterly regions of former Babylonia) – against the early supporters of Christianity. With a well-informed and artfully phrased libretto written primarily by the poet Joseph Méry (plus some late tweaking by Térence Hadot, a friend of the composer), the work demonstrates the resonance that a savvy librettist at that time could generate by mixing elements from Roman history and Classical mythology (for example, references to Venus) with repeated evocations of the spiritual power and political powerlessness of the nascent Christian movement. Among other things, the prominent emphasis on Christian belief and practice in Herculanum comprises a clear attempt at – as the editors of this special issue put it – reinvigorating an ancient Roman setting at a time when interest in such topics was waning.

The three recent revivals of Herculanum have helped reveal the imaginative ways in which this important work plays with the mixed legacies of ancient Rome and early Christianity. The result, we can now appreciate, mingles the grandiose and the intimate, the statuesque and the sensuously seductive, and it can delight and fascinate an opera-house audience or at-home listeners even a century and a half later. Herculanum may not be psychologically sophisticated (in the manner of, say, La traviata, Lohengrin, Carmen, Madama Butterfly, Pelléas et Mélisande, or Wozzeck), but it is one of the most appealing and best-crafted serious French operas of its era and fully rewards the close examination that, until now, it has largely been denied.

Méry's Evocative Preface

The preface to the libretto plunges us directly into the tensions between Rome and early Christianity. This, like the libretto itself, was written by Joseph Méry (1797–1866), a poet and short-story writer who had contributed librettos (alone or in collaboration) to three 1855 stage works by Offenbach and who would attain a place in the annals of opera for co-writing, with Camille du Locle, the text for Verdi's Don Carlos (first performed in 1867, a year after Méry's death).Footnote ¹⁹ More crucially in the present context, his mystère entitled L’Éden (1844) had been set as an oratorio by David in 1848, and he had provided most of the text for the second of David's two odes-symphonies, entitled Christophe Colomb (1847).Footnote ²⁰

Berlioz quoted about half of Méry's preface in his largely admiring review (in the Journal des débats) of the opera's first performance, but the preface has never been reprinted or translated in full. I therefore give it here, in the appendix, with basic explanatory annotations. One thing that Méry takes for granted is that the reader knows about the rediscovery of the ruins of Herculaneum; this rediscovery began in 1709 (in a town named Resina, built on melted volcanic ash that had hardened into rock) and continued through much of the eighteenth century, causing the site to be included in many a Grand Tour. But, as he neglects to point out, the ruins were more deeply buried than those of Pompeii and thus remained (and remain today) largely hidden from view. The decision to set an opera in Herculaneum thus provided very limited data for a librettist and the set-design team, but also left them freer to include elements having nothing to do with the locale: in this case, two central characters from the Euphrates valley (part of what the French broadly called ‘l'Orient’ – ‘the East’), as well as supernatural elements (Satan, and the miraculous vision that he summons up in Act II in hopes of driving Lilia into his arms and putting her under his power).

Méry's preface is given in its entirety in the appendix to the present article. In it, Méry unleashes a lava-flow of erudite allusions to names from ancient Rome. But the preface must have been a powerful stimulus to audience members who read even just bits of it when perusing the printed libretto. (Librettos were available for sale before opera performances, and people often brought a copy with them into the theatre and could read it there since theatre lights were in those days not yet dimmed.) Berlioz made a point of calling Méry, in his laudatory review of the libretto and the score, ‘one of the best Latinists among today's writers, and one of those most deeply versed in a knowledge of Roman antiquity’.Footnote ²¹ This is high praise from a man who knew by heart most of Virgil's Aeneid in Latin and who had recently finished the main compositional work on Les Troyens, a massive five-act opera based on that epic poem. (Les Troyens would not reach the stage until 1863, in a version that omitted its first two acts.)

Indeed, Méry's images of grandeur and imperial sway marked the sets and costumes of Herculanum, which reviewers assured their readers were among the most elaborate and (at least seemingly) authentic that had ever been seen on the stage of the Opéra. The net result was crucial to the impact made by the various characters. The Roman ruling class and its vassal princes and princesses from abroad – notably Nicanor and his sister Olympia – matched the scenery in the ostentatious apparel that they wore and (to judge by the libretto and the music) in the confidence and spontaneity with which they moved and expressed themselves. The oppressed Christians, in stark contrast, wore simple linen gowns and were obliged to comport themselves, again and again, in postures of modesty, piety, supplication and fear. More precisely: the chorus of Christians is present only in the opening of Act II. In the bulk of the opera, Lilia and Hélios are thus the only Christians, vulnerable in a sea of pagans. Queen Olympia (in Act III) ridicules Hélios's people as ascetic desert-dwellers who eat dry bread under palm trees, but this image backfires on her, since it can evoke the Jews’ forty years of wandering or, equally well, David's own Le désert, in which brave members of a caravan face sandstorms and critique the soft life of city-dwellers who enjoy the safety of solid, tomb-like houses.Footnote ²²

One should also notice the historical details that Méry omits from his preface. He does not mention the slave revolt led by Spartacus and others (73–71 BCE), even though it occurred in this very region (starting in Capua, north of what is today Naples, and continuing in an encampment near Vesuvius itself) and is explicitly evoked by Nicanor in Act IV, scene (tableau) 1, when he urges the slaves to rise up against their masters. Even more basically, Méry does not mention a single individual Christian. Perhaps this was his way of admitting that the Christian characters that we are about to meet in his and David's opera are invented by him. Indeed, he describes the entire work as légendaire and at most implies that Olympia is somehow analogous to Zenobia, the renowned queen of Syria who, with Roman support, extended her realm westward to include Egypt; her dates were approximately 240–274 CE.Footnote ²³ As for Olympia's brother Nicanor, Méry most likely borrowed that name from 2 Maccabees; there, Nicanor is a Syrian-Seleucid Greek general sent by king Antiochus Epiphanes to subdue the Jewish population of Judea, and he is described as ‘thrice accursed’, a phrase that the opera's baritone role fully deserves.Footnote ²⁴ Olympia displays many traits typical of the eastern queens and sorceresses from early generations of opera. (For example, she has access to magic potions.) But Nicanor is an even more distasteful type of distant foreigner: he becomes positively smarmy in the end of Act IV when he comments that, if Hélios leaves Olympia (who, we recall, is Nicanor's sister) and returns to the Christian virgin Lilia, he is ‘walk[ing] away from pleasures barely tasted’ (‘abandonne / des plaisirs à peine goûtés!’). More generally, then, Olympia and Nicanor, though endowed with power by Rome, are the latest in a long series of portrayals of exotic and corrupt members of the ruling class, and carry resonances of, for example, Belshazzar (in the Book of Daniel) and the ‘whore of Babylon’ (in the Book of Revelation, a passage quoted by the Christian preacher Magnus in Act I).

That Olympia and Nicanor come from a distant land – one ruled earlier by, indeed, Babylonia – was perhaps a nod to a widely read English novel, The Last Days of Pompeii (1834), by Edward Bulwer-Lytton, which features characters from Rome, Greece and Egypt.Footnote ²⁵ By contrast, the characters in, for example, Andrea Leone Tottola's libretto for Giovanni Pacini's L'ultimo giorno di Pompeii (1825), an opera about the same basic event – the eruption of Vesuvius in 79 CE – consist entirely of Romans. Méry seems to have been one of the first to realize what advantages could be had by merging Roman decadence (for example, the empty display of wealth, built upon slave labour) with, as it was understood, eastern voluptuousness and manipulativeness.

The librettist, and the directors of the Opéra, may have also felt that bringing exotic ethnicity into the heart of the Roman Empire would give the composer Félicien David a chance to strut his stuff, given the latter's success up to that point with works set in distant lands, especially in Le désert, Christophe Colomb (1847), La perle du Brésil (1851), and certain songs and descriptive piano pieces (including his remarkable song Tristesse de l'odalisque, to a heartbreaking text by Théophile Gautier).Footnote ²⁶ The resulting opera libretto worked out well, freshening the Roman Empire setting with new colours and qualities (such as invocations of the supernatural), and giving ample scope to David's abilities and predilections.

The Opera's Plot: A Christian Minority Oppressed by Roman Rulers from the East

The opera's plot, adumbrated in Méry's unusually learned if chronologically confused preface, deserves a more extensive summary phrased in terms of the opera's central and enlivening tension between ancient Rome and early Christianity.Footnote ²⁷ In the following, I also offer occasional hints about David's musical treatment, to be explored more fully later.Footnote ²⁸

When the Christian maiden returns to consciousness, Satan shows her, magically, a vision of Hélios (Figure 1), smitten, at the feet of Olympia and repeating his love song from Act I, now cushioned by slow chords from an unseen chorus (SSTTBB), humming dreamily.Footnote ³³ Lilia is not so much jealous as horrified. Satan snatches the dead proconsul's cape and declares his intent henceforth to impersonate Nicanor. His precise words are ‘Et maintenant le proconsul c'est moi’, likely evoking the tyrannical motto long associated with Louis XIV: ‘L’état c'est moi’.

Fig. 1 The moment in Act II in which Satan shows Lilia, by means of a ‘magical image’, her beloved Hélios singing a love song to Olympia, whom the Romans have installed as queen over her native lands near the Euphrates River. Front-page illustration in L‘Univers illustré, no. 45, 26 March 1859; Gallica.bnf.fr, Bibliothèque nationale de France.

The opera thus ends very quickly, with the death of everyone on stage (except perhaps Satan, though the stage directions at the end of the libretto do not mention him). This drastic foreshortening did not upset critics at the time. They were presumably accustomed to it from other operas of the period. For example, Halévy's La juive and Verdi's Il trovatore both end with a horrifying death, moments before the curtain falls. This was not Méry and David's original plan: they intended a subsequent scene portraying the Last Judgement, in which the population of Herculanum (or of the entire world, it seems) would be divided into saints and sinners and go off to enjoy or bewail their respective eternal fate. The scene was excised before the first performance.Footnote ³⁶

Characterization and Musical Style: Ancient or Exotic?

The elaborate set-up that Méry provided in his preface might lead us today to expect a work rich in what during the nineteenth century was called couleur locale, a phrase that could refer not just to place – that is: ethnic and cultural traits of a region – but also to time period. Yet there is nothing in David's music to suggest that the Roman characters and their vassals from the Middle East (Olympia and Nicanor) lived in the year 79 CE. Quite the contrary, the processional marches and the choruses of praise for the rulers, near the openings of Acts I and III, sound very much like other mid-century operatic numbers that represent the grandeur and might of some much more recent European ruler. This in itself might be seen as an invitation to the audience to consider the nefarious Olympia and Nicanor as symbolizing not so much the ancient ‘Orient’ as individuals in positions of power ‘here’ and ‘now’.

Indeed, quite notably, those two characters are not exoticized in the music that they themselves sing. Olympia's Act I drinking song is in polonaise rhythm, to emphasize the grand, festive mood. (Berlioz found the number strikingly ‘original’, though he didn't say in what way.)Footnote ³⁷ As for the Greek prince Hélios, one prominent drama critic (Paul de Saint-Victor) exercised his creativity in finding something ‘Aeolian’ or, continuing even more dubiously, a mixture of ‘Greek smoothness [suavité]’ and ‘Creole nonchalance’ in Hélios's ecstatic outburst of love for Olympia after the love potion takes effect.Footnote ³⁸

Similarly, most of the Act III ballet is not at all exotic in style. The main danced divertissement, which follows Olympia's ‘Viens blonde déesse’, is a six-movement pas de deux for ‘the Faun and the Bacchante’ [a female follower of Bacchus, god of wine], evoking characters from Roman mythology but in no way sounding any different from hundreds of ballet tunes of the period. (Berlioz admired one movement, featuring piccolos, pizzicato violins and a lyrical counter-melody in the cellos.)Footnote ³⁹ Olympia's aria is preceded by a shorter ‘Pas des muses’ (Dance of the Muses), which does convey a distinctive sense of Otherness, sounding like a mazurka or perhaps a highly countrified Ländler or Bohemian dance: it is in 6/8 in A minor, with recurring syncopations and – in the contrasting episode in A major – with a recurring raised fourth degree.Footnote ⁴⁰ Why the Greek Muses should sound Germano-Slavic is, today, anybody's guess. Perhaps David meant the music to sound somewhat folklike. (French hurdy-gurdies in the countryside, like certain Polish mazurkas, sometimes use a raised fourth.) Or was he thinking of the (medieval, actually – not ancient) Lydian mode? Whatever the intent, the result is one of the best and most intriguing dance movements by a French composer of the time.

By contrast to the ballet (aside from this Lydian moment), the chorus of early Christians is pointedly given music that sounds quite unlike what was currently normative in the theatres or the salons. At the beginning of Act II, we encounter the Christian community for the first time. They are praying at a cross that they have erected in the wilderness as a place where they might recall ‘the tombs of their ancestors’ (perhaps located very nearby). They sing two choral numbers in immediate succession. The first, in pure chordal homophony, reflects the fact that they are entering the spot in a slow procession, taking care not to be heard. Méry built the sense of quiet footsteps into the verses: every line (except the first) is in three syllables, presumably suggesting that the chorus move one small step forward at the end of each line, exactly as David sets it. (I italicize each syllable that coincides with an elongated beat and apparently coincides with a footstep.)

The music, in G minor, is marked by a recurrent dactylic rhythm (essentially: crotchet quaver quaver) reminiscent of the second movement of Beethoven's Seventh Symphony (the famed Allegretto).Footnote ⁴² The singing begins pianissimo, rises briefly at one moment of fear, and then subsides. These same Christians then, apparently having reached the large cross, join in an unaccompanied canticle in F major that Berlioz rightly praised for its ‘felicitous’ contrast of phrases for the women and brief chordal statements (pianissimo) by the men confirming cadential arrivals in I, iii, V and finally I again. The aural separation of women and men in this second choral movement surely was meant to suggest a certain emphasis on celibacy among the Christian adepts, quite different from the possessive yearnings of Olympia for the handsome Hélios and – as the audience is about to witness – Nicanor's manipulative praise of Lilia's physical beauty (and his then grabbing her by the arm).

But, as I have demonstrated in various of my previous publications, exoticism in an opera need not rely upon exotic musical features.Footnote ⁴³ If a character or group/people is represented – through the sung words, onstage action, and sets and costumes – as being different from Us (that is: different, in some basic way, from the work's original audience), then the composer can feel free to engage all variety of normal musical materials to signal that difference, that Otherness, to his or her listeners. (For example, the role of an exotic temptress may be written for a deep-toned mezzo-soprano voice, historically associated with female sensuality; and an exotic male ruler may be portrayed as intemperate and punitive through stentorian pronouncements typical of authority figures generally and/or through devices regularly used for indicating anger and threats, such as stormy rushing scales and emphatic diminished-seventh chords.) This leads us, finally, to discuss the opera's main characters and some specific aspects of their words, actions, arias, duets and larger ensembles.

A Roman Opera without Romans

One of the most surprising features of Herculanum is that, though set entirely in the south of Italy – a region formerly controlled by Greeks but, in the year 79, part of the Roman empire – the opera has no characters who are identifiably Roman (apart from the chorus, when it represents the collective inhabitants of the town's fancy palazzos). Rather, it pits two innocent young Christian adepts from Europe against two rather worldly and selfish individuals who have come from some (former) Babylonian-controlled land and who have been raised to unaccustomed power by the Roman authorities. This naturally invites a double question: which of the characters were meant to be perceived as exotic, and in what way. The question can best be answered by drawing attention to the traits that are sketched for the various characters in Méry's libretto and further emphasized (or concretized) by the musical features that David chose to apply. It pays to remember that, in the late 1850s, the composer was at the height of his powers (which he would retain at least as far as the aforementioned Lalla-Roukh, 1862).

Olympia is a coloratura mezzo, a category that, though familiar from Rossini operas, took on increasingly exotic and femme fatale qualities in French operas, notably in Samson et Dalila, an opera that Saint-Saëns started composing in the late 1850s. (Because of the persistent institutional nervousness about putting biblical characters and stories on stage, Samson would not reach performance anywhere until 1877 in Weimar, translated into German; it finally reached Paris in 1890.) Olympia even emits a remark of triumph, toward the end of Act I, that could have been put in Dalila's mouth: ‘Le désir enfin brûle son coeur … / Il m'appartient, il est à nous!’ (‘Desire [for me] is finally burning in his heart … / He belongs to me, he is ours!’).

Olympia is a conniving sensualist. She is acting on behalf of the Roman authorities to subvert the Christians any way that she can: by her physical beauty, by her singing voice (her big Act I and Act III arias are both explicit ‘diegetic’ numbers, namely a call to drink and a hymn to Venus), and by the wonders of her realm, which is filled with lovely vegetation, symbolic of her own physical ripeness.Footnote ⁴⁴ Like the sorceresses – often associated with ‘the East’ – in many early operas set in ancient days (for example, Armida, queen of Damascus; or Medea, from Colchis, which is now a part of Georgia), she has supernatural powers, namely access to love potions. And David rose to the occasion, as Berlioz observed: ‘The chromatic figure in the muted violins, as Hélios feels the love potion taking effect, is vividly appropriate’.Footnote ⁴⁵

Olympia's self-possession and willingness to display her beauty are understood as being marks of her political power and also of her eastern-ness. Presumably she was highborn in her homeland, and accustomed to being feted and adored. The beauty that makes her so desirable is represented as much by her powers of vocal display as by her bejewelled finery. This is evident in her Act I drinking aria, in which she invites Hélios to imbibe with her: this is perhaps the most extroverted number in the whole work, and one of the flashiest. Berlioz found it ‘exceptionally well written for the voice of Madame [Adelaide] Borghi-Mamo’, ‘enormously effective’, and, from first note to last, marked by great ‘originality’ (high praise from one of the century's most original composers).Footnote ⁴⁶ The coloratura is even more luxuriant in Olympia's much-praised ‘Hymne à Vénus’ in Act III: ‘Ô, viens blonde déesse’. The aria ends with four successive ‘bounces’ from a low D to a trilled middle D, and then one last scalar ascent to a high A (the approximate equivalent, for a mezzo, of a soprano's high C). Audience members may well have thought of a previous exotic female, the native Brazilian woman Zora in David's La perle du Brésil, when hearing similar bounces (a crotchet on the downbeat, followed by a minim, in 3/4 metre) at the end of the middle section of Olympia's aforementioned first-act drinking aria: there they occur on the second syllable of her emblematic word ‘plaisir’; and in Zora's ‘Charmant oiseau’ they had occurred on the emotionally rather similar word ‘joyeux’.

Nicanor, Olympia's scheming brother, is a sometimes stentorian, sometimes wheedling bass-baritone, close kin to Robert's diabolical father Bertram in Meyerbeer's Robert le diable. His role, like that of Olympia, is studded with coloratura, most extensively in the first tableau of Act IV, when he – or, more accurately, Satan posing as him – urges the slave population of Herculanum (male chorus) to rebel against their masters and mistresses.Footnote ⁴⁷ Consistent with the distorted ethics that drive Satan, the coloratura is somewhat bizarre, including four bars entirely in dotted quavers and semiquavers (on the second syllable of ‘[mar-]chons!’). His part also includes some striking downward leaps to unusually low notes, including a creepy low G (over a diminished-seventh harmony in that same scene with the chorus of slaves).Footnote ⁴⁸

As for the galop-like music in the stretta (to use Italian terminology) with which the onstage Romans and easterners end Act I, and perhaps especially its tarantella-like coda, some critics at the time found it insufficiently dignified for the Opéra, perhaps especially the shift, on the part of the chorus, from fear to laughter (at the warnings from the one onstage Christian: the prophet Magnus).Footnote ⁴⁹ But, in fact, the passage quite appropriately skewers Olympia, Nicanor and their followers, showing that they prefer a life of sensual pleasure and that they disdain the warnings being announced to them by a man who takes seriously the precepts that God has laid down for living a good and moral life. Berlioz felt that the stretta grows and grows appropriately – from the moment when the people break out in (rhythmicized) laughter – and ends with real punch.Footnote ⁵⁰

By contrast to the demonstrative coloratura of Olympia and Nicanor, Lilia's and Hélios's roles are written in an almost purely syllabic manner, thus etching in note values those characters’ basic simplicity and honesty – traits that help us accept that Hélios will eventually return to his rightful Christian beloved, once he is no longer under the influence of the love potion. (The events explicitly take place over the course of a single day.) The role of Hélios is similar in manner and technique to two – namely the title character in Robert le diable and Raoul in Les Huguenots – that Meyerbeer wrote for Adolphe Nourrit, a singer renowned for his confident vocalizing and excellent skills as an actor. Lilia is, by contrast to the coloratura-launching Olympia, a pure lyric soprano and is perfectly paired in that regard with Hélios. (Their names are similar in sound, hence tend to get set to similar identical rhythms, and have the added advantage of referring to elements in nature: the lily and the sun.)Footnote ⁵¹ Thus, oddly, the two Christians, though representing a religion that arose in the eastern Mediterranean, are more normative – more European – than the two representatives of Roman power, who hail from the distant land that had once been known as Babylonia until its conquest by Greece and then by Rome.

Magnus, the Christian priest whose sole assignment in the opera is to denounce the aristocrats in the ends of Acts I and IV, is ethnically unmarked. He was apparently born nearby, as Olympia dubs him ‘this Italian clown’ (‘ce bouffon d'Italie’), a phrase that also reminds us that her own realm, by contrast, lies far eastward. Magnus's denunciation of Olympia and her court in Act I ends with sung recitation of phrases from the Book of Revelation.Footnote ⁵² This music (setting dozens of words in a row to a single pitch and then concluding each statement with a short cadential figure) is closely allied to chant formulas that have been used across the centuries in Catholic liturgy (for example, for psalms). Thus, it most likely sounded very familiar, not at all foreign, to most audience members.

A Christian Opera Focusing Heavily on Satan

The role that Satan plays in this work is crucial to the work's treatment of the pagan vs. Christian binarism. An opera about ancient Rome in the older, eighteenth-century style would have had no Satan in it. The very existence of Satan in the plot – and indeed on the stage – further emphasizes that the opera is viewing Rome through a Christian lens (Satan, manipulatively, assures the slaves in Act IV that the Roman gods – ‘les dieux’ – will support them if they pillage the nearby palaces).Footnote ⁵³ Satan thus forms the evil counterpart of Magnus, the upstanding leader of the Christians. The two roles are, significantly, both written for a bass-baritone, thereby stressing their polar opposition, and the two cast members return at the same moment toward the end of Act IV as a kind of double-nemesis. Also, each is, at times, accompanied by a brass ensemble, evocative of power and authority.Footnote ⁵⁴

Satan is not just a representative of evil: he is, as in the Bible, a superhuman, semi-divine being. He first appears on stage toward the end of Act II, just after Nicanor, having denied the existence of God, is struck dead by lightning. It seems that somehow God's act of striking Nicanor dead has opened a fissure in the universe, or in the surface of the earth, allowing Satan to crawl out of his underground abode and wreak his wonted mischief. Or perhaps this was a broader act of divine retribution: after all, Magnus declared in Act I that, as a result of the debauchery of the ‘harlot of Babylon’ and the ‘kings of the earth […] an angel […] came down from Heaven holding in his hand the key to the abyss, and, having opened it, he unleashed the devil, whom the Son of God had held in bonds’.Footnote ⁵⁵

When Act II ends, Satan, as noted earlier, dons the cape of the dead Nicanor, after which he pretends (successfully) to be Nicanor until shortly before the end of the opera. This switch of identities raises an interesting question that, I believe, has never been raised in the (admittedly few) writings about this opera: who is to sing the role of Nicanor disguised as Satan in Acts III and IV? This is of course a crucial question, given that the two are figures of power in, respectively, Roman government and Christian theology. The original printed libretto and the early reviews state that Louis-Henri Obin sang the role of Nicanor and that of Satan-when-disguised-as-Nicanor, and that Théodore Coulon sang Satan (presumably when not disguised – that is, only in the end of Act II or possibly also at the very end of the opera, when Satan re-enters and reveals himself).Footnote ⁵⁶ Later productions seem to have simplified the matter, giving the role to a single singer.Footnote ⁵⁷

The most likely explanation is that, in the first run of performances, Coulon played Satan from the latter's first appearance in the middle of Act II through to the end of the act (when he dons Nicanor's cape and the curtain falls), and that this was considered necessary since Nicanor's dead body was still visible on stage. In Acts III and IV, the role was assumed by the singer of the role of Nicanor, since Satan now bears his traits (or at least wears his cloak, perhaps hooded to make his face harder to see). One review of the first production specifically praised Coulon for generously appearing in a role that was ‘épisodique’: that is, Coulon appeared in just one scene.Footnote ⁵⁸ But Nicanor and Satan are never alive on stage at the same time. Thus, for later performances, stage directors must have come up with ways to enable a single singer to encompass both roles in their entirety, such as by having Nicanor's dead body be replaced (under cover of darkness) by a mannikin. Or perhaps replaced by a supernumerary dressed as Nicanor, while the singer of the role of Nicanor quickly changed into the costume for Satan. A third possibility: maybe the singer was already wearing the Satan costume, under his cape as Nicanor.Footnote ⁵⁹

Bringing the Ingredients, and the Characters, Together

There is a danger, when talking about a forgotten opera, that one will separate out its various strands and judge each of them on its own merits. This is certainly a trap into which some already ill-intentioned critics fell when reporting on the work's opening night in 1859. But, with the advantage of hindsight, and the published full score and good recordings to consult, we can now draw attention to moments in which the various elements come together with a powerfully engaging effect. (True, we have to imagine the costumes, sets and staging, but the book that comes with the Palazzetto Bru Zane recording reproduces a number of illustrations from the period.)Footnote ⁶⁰

The orchestral preludes and postludes in the work are particularly varied and distinctive in character. Often they do not simply anticipate or repeat material from one or another musical number; rather, they alter or develop it, juxtapose one such excerpt with another, or even superimpose one upon another. This is true, for example, of the Introductions to the four acts and the aforementioned orchestral music underlying Hélios's drinking of the love potion (middle of Act I). One particularly imaginative orchestral response to a moment in the text comes near the beginning of Act IV: Hélios's yearning cries for ‘Lilia!’ – on a rising fifth, followed by a second fifth a half-step higher – are immediately countered by chromatic sobs in the English horn, clarinets and horns, indicating both the terrible silence that is meeting him and his fear that his beloved has perished. Important melodies also recur in fresh dramatic situations and are adapted accordingly, as when (in Act II) Satan allows Lilia to view, magically, Hélios extolling Olympia in song, or when (in Act III) Lilia and the English horn remind Hélios of the love he once avowed for Lilia and for God.

The scene between Satan and the slaves (the ‘sons of Spartacus’), not least some of its orchestral commentary, may remind listeners of diabolical moments in Weber's Der Freischütz (the ‘night birds’ squawking in the Wolf's Glen Scene) or Berlioz's La damnation de Faust (specifically, the several apparitions that threaten Faust during his horseback ride to Hell). Audiences at the time might have likened it more to the scene for Bertram and the spirits of hell in Act II of Meyerbeer's aforementioned Robert le diable, a work that was much better known in France than either Der Freischütz or La damnation.

The Christian religious element is well caught by the composer in Lilia's brave aria at the court of Olympia, ‘Je crois au Dieu que tout le Ciel révère’. The tune's solid squareness likens it to a hymn and helps us feel Lilia's bravery and commitment. One critic at the time, Léon Escudier (better known as a music publisher in Paris, and co-owner of the journal La France musicale), claimed to find a resemblance between the melody and that of Adolphe Adam's ‘O Holy Night’ (‘Minuit, chrétiens’, composed 12 years earlier), implying – unfairly – that David had quasi-plagiarized.Footnote ⁶¹ The resemblance is actually fleeting. More to the point is that Lilia's hymn, after its first strophe, goes on to be treated in a powerfully operatic manner, as the crowd of pagans curses this isolated, defenceless woman in almost machine-gun-like quick repeated notes while she adamantly continues her profession of faith.Footnote ⁶²

The moments that most effectively embody the work's tension between Christian morality and the multiple dangers of Eastern-pagan sensuality, Roman power and satanic scheming (themes I have discussed above) are not the solos, as effective as they are and as highly praised (and indeed often encored) as they were, but the duet for Lilia and Nicanor (in the middle of Act II) and the duet for Lilia and Hélios (near the beginning of Act IV), and the broadly laid-out ensemble finales of Acts I and III. Here David had plentiful models of musico-dramatic structure that he could adapt, not least from operas of Donizetti and Verdi.

Each of those two duets is laid out in multiple movements (an opening one, a more lyrical, self-enclosed one, a transitional one and a concluding up-tempo stretta), a pattern that Rossini had codified in serious operas such as Tancredi (1813), Semiramide (1823) and Le siège de Corinthe (1826). But the vocal lines are heartier and more thrusting than Rossini's tend to be, relying on full-volume singing in the upper register, such as one encounters in Edgardo's outburst in the second-act finale of Donizetti's Lucia di Lammermoor (1838), ‘Maledetto sia l'istante’.Footnote ⁶³

In spite of this heavy reliance on Italian models, the work does not come across as a pale copy, in large part because of David's imaginative and varied orchestration and also his often-subtle manipulation of melodic material and phrase structure. Though Reyer said in his review that he carried entire melodies in his head from having heard them on opening night, he may have been referring mainly to their opening phrases.Footnote ⁶⁴ There is hardly a single melody in the work that returns to its opening phrase at the end, in the manner of so many AABA (or, using prime marks, AA′BA″) tunes of the Italians. David's most self-enclosed melodies often move from one kind of material to another and close with a third, in a way that is meaningless to reduce to capital letters (for example, AA′BC) because the materials actually have linkages and contrasts that it would take a separate study to demonstrate (something more like AA′A″A′″), creating a sense of forward movement and development that matches the complex emotions that a given character is feeling at the time and also reflecting the sonic and emotional weight of specific words and syllables.

Harmony is an effective tool in this process. In the first movement of the Act II duet with Nicanor (the movement that Italians at the time sometimes called the tempo d'attacco), Lilia's opening melody begins with a closed eight-bar period (two four-bar phrases) in A-flat major, each phrase ending placidly in the tonic, reflecting her pious statements about praying on behalf of Olympia, Nicanor and her beloved Hélios. But it continues with two further phrases that deepen the mood: first modulating to the relative minor, on words in which Lilia describes herself as weak and living in isolation from the world (‘Faible et dans l'ombre retirée, / que pouvez-vous craindre de moi?’), and then returning to the tonic major as she asserts her desire to live unnoticed, devoted to her love and her faith (‘Ah! laissez-moi vivre ignorée, / avec mon amour et ma foi’).Footnote ⁶⁵ But the third phrase (the one in F minor) incorporates a five-note descending motive (with a dotted rhythm in it) that had been heard in the second A section, not in the first, and that will then dominate the fourth and concluding phrase. The process feels almost inevitable once it is over, yet it was in no way predictable.

Even more remarkable is the passage in Lilia's Act IV duet with Hélios in which she agrees to forgive him: an 11-bar phrase that is utterly coherent and connected, each phrase growing out of and responding to the previous one in ways that defy simple explanation. The key is the extreme one of G-flat major, often used at the time in situations invoking passionate love; and the orchestration features sextuplets in two harps, suiting Lilia's prayer for Hélios to receive heavenly pardon.Footnote ⁶⁶

Other cases of melodies that continue in unpredictable but, finally, gratifying ways include Olympia's ‘Suis-la, suis-la donc’ in the Act III finale, built entirely of four-bar phrases until the final phrase, which expands into five bars to allow the singer to emphasize the queen's sarcastic taunt to Hélios (while rising to a C in her powerful high-middle register), namely that, if he rejects her, he can perfectly well enjoy a grim life, eating air and praying under the palm trees (‘Et vivre d'air et de prière / Sous le dôme vert des palmiers!’).Footnote ⁶⁷ And, in regard to that same Act III finale, Berlioz – so fond of rhythmic complexities – expressed particular delight at the offbeat accents in the coda, sung out emphatically by everybody on stage.Footnote ⁶⁸

These kinds of compositional subtleties testify to David's familiarity with the works of the Austrian and German composers he so admired (he particularly mentioned Beethoven and Weber in early letters) and with the current reigning masterpieces of French opera, notably those by Meyerbeer.Footnote ⁶⁹ We might recall that he had, by this point, composed symphonies, quartets and piano trios that were influenced by Beethoven, Mendelssohn, (arguably) Schubert and other composers from across the Rhine, and had even gotten most of these performed and published. To tell a story of largely distinct national styles in mid-nineteenth-century European music (French vs. German vs. Italian) is to deny the basic reality of cross-cultural admiration and mutual influence: ‘cultural transfer’, as it is called by historians today.

The layout and flow of the Lilia–Nicanor duet (Act II) likewise demonstrates David's (and Méry's) having absorbed the effective methods of scene-construction that had been more or less codified by Rossini and then further developed in recent years by, most notably, Donizetti and Verdi.Footnote ⁷⁰ It begins with what is, effectively, a typical tempo d'attacco, done as a ‘similar’ movement, with Nicanor repeating, for the most part, Lilia's melody (discussed above) but in the dominant, and with two energetic leaps to a high C in the final phrase. After he promises her the ‘treasures’ of ‘the Orient’, to a tune that Lilia bluntly ignores, she launches what is, structurally, the cantabile (second main movement) of the duet, but in an emphatic, Donizettian manner full of wide leaps up and down. Nicanor attempts to reply, in the parallel minor, with his most outrageous lie (that, to prove his love for her, he has become a Christian), and then they battle out the matter in simultaneous singing, ending in Lilia's major mode. After what scholars today would call a short tempo di mezzo, Nicanor now is the one to launch the cabaletta, but it is the ‘dissimilar’ type pioneered by Donizetti and used extensively by Verdi: Nicanor here uses again the emphatic style of the second movement, made menacing by being in the minor (F), but Lilia responds in the parallel major and with a much more conjunct melodic line, to words begging God for a miracle or else for death. And the duet comes to a highly theatrical conclusion as Nicanor ‘seizes Lilia’ (according to the printed libretto and the piano-vocal score), then cries out ‘Your God does not exist’.Footnote ⁷¹ The ground shakes and the ray of light on the cross turns red (according to a stage direction in the orchestral score). Lilia comments, in a tone of ‘exaltation’: ‘God replies to your blasphemy with a wondrous sign’ (‘A ton blasphème Dieu réponds par un prodige’). The Lilia-Hélios duet in Act IV is, if anything, even more consistent in musical quality and dramatic force.Footnote ⁷²

As for the finales to Acts I and III, they testify to David's growth as a composer and his increasing ability to match music to single characters or groups of characters in a way analogous to, yet distinct from, what Verdi achieved in, for example, the famous quartet from the last act of Rigoletto. We have already seen this in the concluding stretta of the Act I finale that Berlioz so praised.

Even more impressive is the Act III finale. In its remarkable second movement, Olympia, Nicanor and Lilia struggle over Hélios, each urgently informing him what he should now do (‘Parle, réponds, décide!’). A propulsive figure, grimly orchestrated for bassoons and strings in unison, keeps modulating from key to key, leading the beleaguered man to cry out: ‘Quelle nuit que ce jour!’ (‘How fearsomely dark is this day!’) and, finally to implore them all: ‘Par pitié! laissez-moi!’ (‘For pity's sake, let me be!’).Footnote ⁷³ This brings Lilia to lament the situation that Hélios has created, in broken phrases while the English horn – frequently associated in the nineteenth century with the exotic – repeats the melody of the declaration that Hélios and then Lilia had made to Olympia and Nicanor in Act I.Footnote ⁷⁴ The tension between Christian virtue and pagan (and Satanic) manipulation – or, speaking more generally, the struggle between two irreconcilable forces, each seeking to enlist the support of one valued, beleaguered individual – has never been better expressed than in this four-minute stretch of music. Or six minutes, if one starts with Olympia's powerful, sardonic ‘Suis-la, suis-la donc, si tu l'aimes’ (‘Go, follow that woman, if you love her’), discussed earlier.

The Act III finale concludes with a powerful ensemble movement in which Hélios declares his love for Olympia (in order to save the life of Lilia); Satan sings, gloatingly, ‘Gloire à moi’; and everyone on stage and in the pit joins in the highly syncopated coda that (as mentioned) so delighted Berlioz. Hearing this whole act-finale, after its having gone unperformed for a century and a half helps listeners today understand why in 1867 David was granted a massive financial prize (20,000 francs) for this opera and his subsequent one (Lalla-Roukh) by the Académie des Beaux-arts and, two years later, was elected unanimously to replace Berlioz as a member of that same Académie, a ‘section’ of the distinguished Institut de France.

* * *

Herculanum, which relies on fire and heavy ash for its dramatic final minutes, itself became a victim of fire in 1873 when the imposing sets – and presumably the lavish costumes – were destroyed by the massive conflagration that consumed the Opéra's theatre (the Salle Le Peletier). As a result, the work would thereafter have been too expensive to revive, a fate it shared with some other highly worthy operas in the company's repertory.

But Herculanum has been brought back to life again in our own day (indeed three times, with entirely different casts, orchestras and conductors). The unexpected return of this remarkable work helps remind us of the many vital creations that once held the musical and theatre-going world enthralled in Paris and in other major European cities, and shows how a savvy creative team was able to knit together in new ways and with great dramatic and musical vitality historical and legendary accounts about the widespread Roman Empire and its struggle with early Christianity.