With Debussy’s Pelléas et Mélisande and Dukas’s Ariane et Barbe-bleue , French opera moved decisively beyond broadly imitative ‘wagnérisme’ to a more individual expressive language, and a far stronger sense of synthesizing Wagner’s achievements rather than producing rather pale copies. But despite the innovations of Debussy and Dukas, the shadow of Wagner was to hang over a good deal of French opera for the first few decades of the twentieth century.
Gabriel Fauré – like Dukas – was among the many French pilgrims to Bayreuth in his youth, but his only completed opera came towards the end of his career. Already a successful composer of incidental music for plays – notably for Edmond Haraucourt’s Shakespeare adaptation Shylock (1889) and Maeterlinck’s Pelléas et Mélisande (1898) – Fauré began work in about 1907 on Pénélope , a ‘drame lyrique’ in three acts to a libretto by René Fauchois. The opera was complete in piano-vocal score by 1912 (when it was published by Heugel; a revised edition appeared in 1913). The orchestration, mostly by Fauré but partly by Fernand Pécoud, was completed early the next year in time for the premieres at Monte Carlo on 4 March 1913 and at the brand-new Théâtre des Champs-Elysées in Paris on 10 May 1913. Pénélope was a work which took Fauré considerably longer than any other and its attraction for him seems to have been as an affirmation of conjugal love (something Fauré himself only experienced intermittently); his son Philippe Fauré-Fremiet described it as ‘a new Bonne Chanson on a mythic scale, sung by characters who are larger than life’ (Fauré-Fremiet 1945 , 17). This is a perceptive view of a work which, though organized on ostensibly Wagnerian principles, lacks something in dramatic coherence and, crucially, pace. But there is much to enjoy in Fauré’s score, especially the many moments of intense lyricism; and while the opera avoids conventional arias, the music often blossoms into song-like writing which shows Fauré in his natural habitat .
The opening season of the Théâtre des Champs-Elysées in April, May and June 1913 – billed by Gabriel Astruc as the ‘Grande Saison de Paris’ – provides as good a place as any to see the cross-currents which were coursing through Parisian cultural life immediately before the outbreak of the First World War. Though this season has long acquired legendary status on account of the new ballets presented by Diaghilev’s Ballets Russes ( Debussy’s Jeux and Stravinsky’s Rite of Spring within a fortnight of each other), opera also had a major part to play, not least with the first Paris performance of Pénélope . After an inaugural concert on 2 April (in which Saint-Saëns, Fauré, d’Indy, Debussy and Dukas all conducted their own works), the theatre began its life on 3 April with Berlioz’s Benvenuto Cellini , conducted by Désiré-Emile Inghelbrecht; it was a long evening, since the opera was followed by a ‘spectacle de danse’ starring Anna Pavlova. Benvenuto Cellini was more of a novelty than might be imagined, as this was the first time it had been revived in Paris since its disastrous premiere at the Opéra in 1838. Also on the bill were two of Musorgsky’s masterpieces, Boris Godunov and Khovanschina .
The influences of Russian music – especially Musorgsky – and of composers from Spain (several of them resident in Paris) were to have an important impact on French opera, not least as a counterpoise to Wagner; and in France, as elsewhere, composers were starting to experiment with the genre: in particular there was a move towards one-act operas. These were nothing new in themselves, but had been something of a Russian speciality in the later nineteenth century.
Ravel completed the piano-vocal score of his one-act opera L’Heure espagnole in October 1907, and it was published by Durand the following year. In 1909 Ravel finished the orchestration, but it was not until two years later that the work finally reached the stage, at the Opéra-Comique on 19 May 1911, partly because of some rather improbable scruples by the house’s director, Albert Carré, over the delightfully saucy text. During 1910 extracts were given in concert and at least one unidentified critic asked the question which had been bothering Ravel for years: ‘When will L’heure espagnole be performed at the Opéra-Comique?’ (Orenstein 1975 , 55). Two days before the long-delayed premiere, Ravel wrote a letter to Le Figaro explaining his aims in the opera:
What have I attempted to do in writing L’heure espagnole ? It is rather ambitious: to regenerate the Italian opera buffa – the principle only. This work is not conceived in traditional form. Like its ancestor, its only direct ancestor, Musorgsky’s Marriage , which is a faithful interpretation of Gogol’s play, L’heure espagnole is a musical comedy . Apart from a few cuts, I have not altered anything in Franc-Nohain’s text. Only the concluding quintet, by its general layout, its vocalises and vocal effects, might recall the usual repertory ensembles. Except for this quintet, one finds mostly ordinary declamation rather than singing. The French language, like any other, has its own accents and musical inflections, and I do not see why one should not take advantage of these qualities in order to arrive at correct prosody. The spirit of the work is frankly humoristic . . . I was thinking of a humorous musical work for some time, and the modern orchestra seemed perfectly adapted to underline and exaggerate comic effects.
(55–6)
The ‘ancestry’ of Musorgsky’s The Marriage is an intriguing reference, since Ravel could not apparently have known this work until its first publication in 1908: even Michel-Dimitri Calvocoressi – who introduced Ravel to much Russian music, was a fellow member of ‘Les Apaches’ and also a great authority on Musorgsky – had not come across it until then . The critical reaction to L’Heure espagnole was decidedly mixed: Gaston Carraud called the libretto a ‘mildly pornographic vaudeville’, while Emile Vuillermoz found the music immensely accomplished, but too clever by half and rather calculated: ‘In the name of logic, Ravel removes from the musical language not only its internationalism and its universality, but its simple humanity’ (57). One of the critics to praise the work was Ravel’s former teacher Fauré, in his review for Le Figaro (20 May 1911). Fauré was impressed by Ravel’s fidelity to Franc-Nohain and the care with which every nuance of the text seemed to be reflected in the score. He was also delighted by Ravel’s brilliantly inventive sound-world: ‘What harmonic and orchestral discoveries, what originality, what subtle ingenuity, and what gaiety of spirit!’ (Fauré 1930 , 116–7).
Louis Laloy was another enthusiast for the work, and in La Grande Revue (25 April 1911) he echoed some of Fauré’s thoughts:
Monsieur Ravel is such a pure musician that he never exceeds the limits of beauty; in his style, even when he disguises it intentionally, harmony is innate; and if we smile on hearing it, we are also moved by a tenderness whose object is none other than the music itself. Already the charm of the Histoires naturelles arose from this sweet soul which could be guessed at through the external irony. It is a much more significant work this time: never has the author shown himself to be more inventive, nor more the master of this genre which he has made his own.
(Priest 1999 , 259–60)
Ravel’s only other completed opera, L’Enfant et les sortilèges (discussed below), was to be a work of even greater inventiveness, and one of astonishing, enchanting boldness .
The French had always been fascinated by Spain, and many French operas were based on Spanish themes. But what about a Spanish opera by a Spanish composer? In 1905, Falla’s La vida breve won a competition organized by the Academia de Bellas Artes in Madrid for the best new one-act opera. Though a production was supposed to follow the award of the prize, nothing came of this. Falla moved to Paris in the summer of 1907, and that autumn he played the opera – which he considered his first fully mature work – to Dukas and Albeniz; both admired it greatly (Pahissa 1954 , 42–3). The work was first published in Paris, by the firm of Max Eschig, and the premiere eventually took place at the Municipal Casino in Nice on 1 April 1913, quickly followed by a first Paris performance at the Opéra-Comique on 30 December 1913. Pierre Lalo, often a hostile critic of Debussy and Ravel, reviewed the opera in Le Temps . It was the romantic love music that Lalo enjoyed least, finding it too reminiscent of the ‘Italian influence which dominated Spanish music for so long’; but the picturesque qualities of the opera had, he thought,
a particularly intense charm – no excess of colour, no deliberate searching for effect, but a subtle restraint, delicate and precise shading, discrimination and good taste. The most felicitous passage is that at the end of the first scene which describes twilight in Granada – a page of penetrating poetry which preserves in its sensitivity and melancholy accents, something intimate and concentrated.
(Quoted in Pahissa 1954 , 64)
While Lalo’s comments are perceptive, it is amusing to note that at the time Falla wrote this music, he had never been to Granada, though he was to settle there permanently in 1920.
Before writing La vida breve , Falla had tried his hand at half a dozen zarzuelas, none of which had any success . But his enthusiastic interest in this popular theatrical form was to influence some of his subsequent stage works; so, too, was Falla’s passion for the traditional music of his country, and Spanish music from the Renaissance and earlier. It was to be one of the great Spanish classics that provided the literary source for his most original operatic project. El retablo de Maese Pedro (‘Master Peter’s Puppet Show’) was composed in 1919–22 as a ‘puppet opera’ in one act, with a libretto by Falla after Cervantes’ Don Quixote . The work was a commission from the Princesse de Polignac, for a performance by puppets in her Paris home. In fact the first performance took place in Seville on 23 March 1923, and the Paris premiere was given to an invited audience in the Princess’s salon at the Avenue Henri-Martin on the following 25 June. Wanda Landowska played the harpsichord part, while the puppet of Don Quixote was worked by Ricardo Viñes, assisted by his pupil Francis Poulenc. The more refined musical language of this work, drawing on many aspects of Spanish music as well as on trends in contemporary European musical thought, produced what Sylvia Kahan has described as ‘something wholly original in all Spanish music’ ( 2003 , 236).
Diaghilev’s Ballets Russes seasons in Paris were of an artistic significance that is now the stuff of legend. It should also be remembered that Diaghilev introduced Parisian audiences to Russian opera, including Musorgsky’s Boris Godunov and Khovanschina , Borodin’s Prince Igor and at least one significant novelty, Stravinsky’s Le Rossignol , a ‘conte lyrique en trois actes’ with a libretto by the composer and Stepan Mitusov based on the fairy tale by Hans Christian Andersen. Stravinsky began the work as early as 1908 – the ageing Rimsky-Korsakov reacted with approval when he was shown the sketches for Act I – but it was not completed until 1914 (in the interim Stravinsky was busy writing The Firebird , Petrushka and The Rite of Spring ). The first performance of Le Rossignol was given at the Paris Opéra on 26 May 1914 and it revealed a work – lasting less than an hour – of slightly bewildering stylistic diversity (hardly surprising given what Stravinsky composed in between the first and second acts), which drew on both Russian and French predecessors, specifically Musorgsky and Debussy. A year earlier, Stravinsky talked of his unease with opera as a genre: in an interview with the Daily Mail (13 February 1913) he declared: ‘I dislike opera. Music can be married to gesture or to words – not to both without bigamy. That is why the artistic basis of opera is wrong and why Wagner sounds best in the concert-room. In any case opera is a backwater. What operas have been written since Parsifal ? Only two that count – [Strauss’s] Elektra and Debussy’s Pelléas ’ (quoted in White 1979 , 225). So what of his own first effort in the genre? Though he quite enjoyed the hints of Boris Godunov in the Emperor’s death-bed scene of Le Rossignol , Stravinsky’s later judgement was chilly: ‘Perhaps The Nightingale proves that I was right to compose ballets since I was not yet ready for an opera’ (Stravinsky and Craft 1962b , 62 n. 2). The work was, however, a veritable feast for the eyes, and Stravinsky was generous with his praise of Benois: ‘scenically, thanks to Alexandre Benois who designed the costumes and sets, it was the most beautiful of all my early Diaghilev works’ (Stravinsky and Craft 1960 , 132).
The influences on Le Rossignol , especially the earlier scenes, were almost as much French as they were Russian, and Stravinsky recalled this in describing the reaction to the first performance:
The premiere was unsuccessful only in the sense that it failed to create a scandal . . . As to its reception, the ‘advanced’ musicians were genuinely enthusiastic – or so I thought. That Ravel liked it, I am certain, but I am almost as convinced that Debussy did not, for I heard nothing from him about it. I remember this well, for I expected him to question me about the great difference between the music of Act I and the later acts, and though I knew he would have liked the Mussorgsky–Debussy beginning, he probably would have said about that, too, ‘Young man, I do it better.’ On my last trip to Russia I remember reading a remark in my diary – I kept a diary from 1906 to 1910 – written when I was composing the first act of The Nightingale : ‘Why should I be following Debussy so closely, when the real originator of this operatic style was Mussorgsky?’ But, in justice to Debussy, I must own that I saw him only very infrequently in the weeks after The Nightingale , and perhaps he simply had no opportunity to tell me his true impressions .
(132–3)
Laloy, writing in Comoedia , gave the work a heroic welcome: ‘A masterpiece, as has been declared here right from the first. A pure masterpiece. Superhuman music . . . Supernatural music . . . The revelation of Le rossignol takes possession of our soul and renews it: only the revelations of Parsifal , Boris Godunov , Pelléas et Mélisande and [Debussy’s Le martyre de ] Saint Sébastien are comparable. What a fortunate time we live in, with so many unexplored perspectives being discovered one after the other!’ (Priest 1999 , 285–9).
On 15 May 1914, just over a week before Le Rossignol was first performed, Henri Rabaud’s Mârouf – one of the biggest popular successes of prewar French opera and now largely forgotten – had its premiere at the Opéra-Comique. Rabaud came from a family with strong operatic connections: his great-aunt was the singer Julie Dorus-Gras, who had created major roles in operas such as Robert le Diable , La Juive , Les Huguenots and Benvenuto Cellini . After studies with Massenet, Rabaud won the Prix de Rome in 1894 and during his time in Italy became an enthusiastic admirer of Verdi and Puccini. But his Wagnerian roots went deep and Mârouf is a diverting amalgam of oriental story-telling, perfumed and exotic orchestral colours, and structures which owe much to Wagner. In 1922 Rabaud succeeded Fauré as Director of the Paris Conservatoire and his L’Appel de la mer , based on J. M. Synge’s Riders to the Sea (also the basis for Vaughan Williams’s opera) was first performed at the Opéra-Comique in 1924 .
Albert Roussel was another composer for whom the Orient was an irresistible lure. He wrote the ‘opéra-ballet’ Padmâvatî during the First World War, to a libretto by Laloy – a man of prodigious talent whose importance as a critic has tended to overshadow his activities as a scholar and writer. Roussel and Laloy had known each other for almost twenty years, having met in the composition classes at the Schola Cantorum. Both were fascinated by the Orient: Roussel had spent several months during the autumn of 1909 on a voyage to India and Cambodia with his wife Blanche; Laloy had published a good deal on the music of the Far East, including an article on music and dance in Cambodia ( 1906 ) and one of the first books in French on Chinese music ( 1912 ). In 1914, following the success of Roussel’s ballet Le Festin de l’araignée , Jacques Rouché, recently appointed as Director of the Paris Opéra, commissioned Roussel to write a new stage work. Inevitably, the war disrupted creative activity and progress was sporadic, since the composer was serving as a transport officer on the Somme and elsewhere in 1915–18. And while Padmâvatî was finished in 1918, it was to be another five years before the piece received its first performance, at the Opéra on 1 June 1923. An ‘opéra-ballet’ in two acts, it clearly reflects Roussel’s love of India, not least in terms of the influence on his musical language. Nicole Labelle has summarized the work’s significance:
Padmâvatî represents the culmination of Roussel’s fascination with India, in its subject matter – the legend of the Queen of Chitor – and in its masterful integration of an Indian modal language into the composer’s harmonic style. Dark, brooding orchestral colours, emotionally effective choruses and danced numbers, and poignant solo writing all evoke the majesty of Hindu temples and the tragic destiny of the characters.
(2001, 808)
Critical reaction was enthusiastic. Dukas wrote: ‘I believe sincerely that of the new generation of musicians, M. Albert Roussel is one of those who makes the strongest impression, through the combination of traditional skills and the most daring harmonic experimentation’ (Le Quotidien , 7 June 1923). André Messager in Le Figaro was equally impressed, drawing particular attention to the primordial power of Roussel’s rhythm, and other composers who wrote admiringly of the work included Florent Schmitt and Darius Milhaud (see Hoérée 1938 , 57–8) .
Ravel’s collaboration with Colette on L’Enfant et les sortilèges produced a one-act opera which is extraordinarily touching, funny and filled with a sense of the marvellous. First performed at Monte Carlo Opera on 21 March 1925, it has been described by Richard Langham Smith as ‘high on the list of works which at one level deal with the child within the adult’ ( 2000 , 200). Parodistic humour is an important feature of the score, but the work is far from lightweight. Ravel was characteristically self-effacing about it, but Colette certainly saw beyond the fun to the opera’s deeper realms:
How can I describe my emotion when, for the first time, I heard the little drum accompanying the shepherd’s procession? The moonlight in the garden, the flight of the dragonflies and bats . . . ‘Isn’t it fun?’ Ravel would say. But I could feel a knot of tears tightening in my throat .
(Quoted in Nichols 1987 , 58)
Ravel’s handling of the orchestra in L’Enfant is breathtaking – a mixture of the dazzling and the delicate which results in a highly original and astonishingly beautiful sound-world quite unlike that of any other opera. His use of eccentric instrumentation, notably the Swanee whistle in the scene in the garden, was another inspired touch. The vocal writing is far more overtly melodic than in L’Heure espagnole , though Ravel’s characteristic care over word-setting is apparent throughout. The close of L’Enfant – as The Child sings ‘Maman’ – is remarkable for its simplicity and its lack of obvious finality despite being a mildly coloured perfect cadence. This gesture is typical of the whole work: ostensibly simple means deployed to ends that seem effortlessly to fuse the childlike and the profound .
Darius Milhaud composed almost forty operas and ballets, and in several of these dramatic works he is at his most original and inventive. His first operatic project was La Brebis égarée , composed in 1910–14 during his years as a student at the Paris Conservatoire, to a libretto by Francis Jammes. It was first performed at the Opéra-Comique on 8 December 1923, almost a decade after its completion. The audience was hostile, as was much of the subsequent press reaction, but an exception was the typically thoughtful review by Dukas. While recognizing that Milhaud had come a long way in the ten years since writing the opera, he found much to admire:
The basic material of the work is as simple, banal and barren as it could possibly be, and most likely it is this way on purpose. Apparently the reason is to make vivid the utter contrast between the radiant souls of the protagonists and the miserable platitude of their lives . . . Nevertheless, the two elements are expressively unified at those moments where the poetry intensifies and the dramatic situation reaches a climax, as in the church scene where Pierre is at prayer, and in the one in which Françoise, on her hospital bed, reads Paul’s long letter of forgiveness. These two scenes mark the culminating points of the score as well as of the play. They are also the ones which made the most striking and profound impression on the audience. They reveal in M. Darius Milhaud a born musician of the theatre .
(1948, 663–7)
Les Malheurs d’Orphée was written in the autumn of 1924. Its subject is a contemporary reworking of Orpheus’s desolation following Euridice’s death; despite being designated an opera in three acts, the work lasts no more than forty minutes. The music which Milhaud wrote to Armand Lunel’s libretto is generally terse, austere and economical. Paul Collaer described it as follows:
Each section is brief, concentrated, stripped to the bone, completely devoid of development. It is as though ‘mere music-making’ is superfluous in the face of such sorrow. Measure after measure represents a cry, a sigh, or a shiver, as though the heart were being torn out piece by piece. Each note must sound true, necessary, beautiful. This is a work of quality rather than quantity, a concentration rather than a diffusion of sentiment. Moreover, the music must be performed as a kind of offering, a ritual prayer to console and soothe the wounded spirit.
(1988, 81)
Another admirer was Ravel, who discussed the work in an interview with Roland-Manuel first published in Les Nouvelles littéraires (2 April 1927). Ravel praised Milhaud at the same time as taking a swipe at an old enemy of his own, the critic Pierre Lalo:
It is with respect to Darius Milhaud and his Malheurs d’Orphée that M. Lalo attains the height of impertinence. Here is a moving, magnificent work, Milhaud’s best, and one of the finest achievements that our young school has produced in a long time. M. Lalo seeks in vain for ‘something vibrant and expressive.’ He complains that ‘the progression is almost always slow,’ while at every moment I find rapid progressions which indicate extraordinary rhythmic inventiveness. The orchestration of Les Malheurs d’Orphée is always very skilfully balanced. M. Lalo declares it to be abominable.
(Orenstein 1990 , 446)
Three years after finishing Les Malheurs d’Orphée , Milhaud began another series of compact little operas, the three ‘opéra-minutes’, none lasting more than ten minutes. The first was L’Enlèvement d’Europe , written at the request of Paul Hindemith and first performed at Baden-Baden in July 1927, on the same programme as Hindemith’s own Hin und Zurück . L’Abandon d’Ariane and La Délivrance de Thésée complete this trio, all of which were written in 1927.
Esther de Carpentras was a project particularly dear to Milhaud’s heart, concerning as it does a Provençal tradition to celebrate one of the great Jewish Festivals. Based on a libretto by Armand Lunel, the work is a comic opera in two acts, composed in 1925–7. But as well as moments of carnival atmosphere for the staging of the story of Queen Esther and a rich seam of melodic invention, the opera also explores the tensions between the Jews of the Carpentras ghetto and the Catholic cardinal from whom permission to celebrate the festival had to be obtained. The end of the opera, where a joyous song of praise from the Jews mingles with an anthem sung by the departing Catholics, is a remarkable moment, before the curiously subdued close where the character Cacan announces that ‘the masquerade has ended up as a sermon’ (‘La mascarade s’achève en sermon’). The first performance was planned for the opera in Monte Carlo, but this fell through and the work had to wait ten years for its premiere, at the start of February 1938 at the Opéra-Comique (having already been broadcast from Rennes in the previous year). One pleasing juxtaposition was the review by a great Catholic composer of a work by his older Jewish colleague: Olivier Messiaen wrote about the opera for the Brussels magazine Syrinx . He was enchanted by the ‘exuberant gaiety’ of the score, with its ‘orchestration which is comical, powerful, and joyously unbuttoned’, and he attempted to summarize its particular quality: ‘The spirit is not at all Rabelaisian, but Provençal, with all that word suggests in terms of light and of good humour. It is this quality with which Milhaud has infused so many places in his score’ (Messiaen 1938 , 25–6).
In 1928, Milhaud completed the first of his epic operas based on historical characters in the Americas. Christophe Colomb was first performed in Berlin (5 May 1930) under Erich Kleiber and it has come to be viewed by many, including Milhaud’s friend Collaer, as one of his greatest achievements. (For a detailed discussion of the work, see Collaer 1988 , 128–37.) It was followed by Maximilien (composed in 1930, first performed in 1932) and Bolivar (composed in 1943, first performed in 1950). Médée (1938) received its premiere in Antwerp (7 October 1939), and the first performance in Paris (on 8 May 1940) was a poignant occasion: it was the last new work to be performed at the Opéra before the German Occupation.
Milhaud continued to produce a seemingly unstoppable flow of music after the Second World War and this included several dramatic works. The grandest of these was the five-act David , a commission from the Koussevitzky Foundation for a work to celebrate ‘the 3,000th birthday of King David and of the foundation of Jerusalem’. Dedicated to ‘the people of Israel’, this epic work was first performed in concert in Jerusalem on 1 June 1954, and the stage premiere took place a few months later, on 2 February 1955, at La Scala, Milan. Milhaud’s next operatic venture could hardly have been more different: Fiesta was, on Poulenc’s suggestion, a collaboration with Boris Vian – poet, songwriter, jazzman and one of the iconic figures of St-Germain-des-Prés in the 1950s. The short (twenty-minute) one-act opera resulting from this rather unlikely partnership was commissioned by Hermann Scherchen for the 1958 Berlin Festival and first performed there on 3 October 1958. In 1966 Milhaud returned to the world of Le nozze di Figaro . The trilogy of ‘Figaro’ plays by Beaumarchais ended with La Mère coupable , widely considered to be a dud, but clearly Milhaud thought it had potential. Madeleine Milhaud’s libretto cleverly tightened up Beaumarchais’s text, but the story never comes alive, and Collaer was forced to conclude that the score – Milhaud’s last conventional opera – was ‘tiresomely monotonous’ ( 1988 , 156).
The theatre was one of Arthur Honegger’s abiding passions, but his operas are not widely performed. A prolific composer of incidental music for plays and of pioneering film scores, as well as of ‘dramatic oratorios’ which lent themselves to staged presentation (such as Le Roi David and Jeanne d’Arc au bûcher ), his first opera was Antigone , composed in 1924–7 to a libretto by Jean Cocteau (after Sophocles). This was first performed at the Théâtre de la Monnaie, Brussels, on 28 December 1927. Greek tragedy was very much à la mode at the time, with Satie’s Socrate and Stravinsky’s Oedipus Rex near contemporaries. Honegger himself wrote that he chose the story ‘because it is not the standard anecdote of love which is the base of nearly all lyric theatre’ (Spratt 1987 , 94). The composer’s declared musical intention, set out in the preface to the score, was ‘to envelop the drama with a tight symphonic construction without the movement seeming heavy.’ Equally significant was the attention Honegger gave to word-setting, in particular the accentuation of particular syllables in order to achieve a natural and dramatically compelling prosody. Much later Honegger discussed the challenge which this presented: ‘What I had to work at at all costs was the means whereby I might make others understand the lyric text: that, in my opinion, was the rule of the game in the realm of the lyric. French dramatic musicians show an exclusive concern for the melodic design and a quite subordinate care for the conformity of text and music’ (Honegger 1966 , 96). He goes on to discuss specific details which concerned him in Antigone , for instance:
What is important in the word is not the vowel, but the consonant: it really plays the role of a locomotive, dragging the whole word behind it . . . In our time, and for a dramatic delivery, the consonants project the word into the hall, they make it resound. Each word contains its potential, its melodic line. The addition of a melodic line in opposition to its own paralyses its flight, and the word collapses on the floor of the stage. My personal rule is to respect the word’s plasticity as a means of giving it its full power.
Contemporaneous reaction to Honegger’s word-setting in Antigone was generally positive, not least because the audience was able to hear all the words clearly. Some critics felt that Honegger’s theories on prosody had restricted the melodic invention in the opera. Certainly the musical language of Antigone is notable for its austerity – severity, even – and for its avoidance of lyricism in any conventional sense. But it is precisely this lack of obvious allure, and the nature of Honegger’s text-setting, which mark it out as a work of decisive importance in the history of French opera after Pelléas . Its failure with the public was a blow to Honegger and his next stage projects were very different .
Amphion (1929) was an unusual melodrama written for Ida Rubinstein and first performed by her and Charles Panzéra at the Opéra on 23 June 1931. Also in 1929 Honegger started work on Les Aventures du Roi Pausole , an operetta completed the following year and first performed on 12 December 1930 at the Théâtre des Bouffes-Parisiens – the celebrated birthplace of many of Offenbach’s operettas. Honegger’s operetta was based on a story by Pierre Louÿs which was originally intended for Debussy, who had asked his friend for an operetta text in 1916. Honegger’s deft and utterly delightful score pokes amiable fun at operettas by Offenbach, Messager, Chabrier and others, but does so with a complete understanding of the genre as well as a real affection for it. The solemn and earnest composer of Antigone is replaced by one whose lightness of touch here resulted in a major box-office success: the original production of Les Aventures ran for 500 performances at the Bouffes-Parisiens . His later operettas were collaborations with Jacques Ibert: L’Aiglon (1936–7) and Les Petites Cardinal (1937) . The remainder of Honegger’s music for the lyric stage is not strictly operatic at all: the ‘dramatic oratorios’, of which Jeanne d’Arc au bûcher (1933–5, with a prologue added in 1944) is much the most spectacular, also included Nicholas de Flue , a kind of pageant written for the Swiss National Exhibition in 1939 .
Among the oddities of the French repertoire in the 1930s, none is perhaps odder than Vercingétorix , the most lavish operatic venture of Joseph Canteloube. Over a fifteen-year period, from 1910 until 1925, he worked on his first opera, Le Mas , which won the Heugel Prize in January 1926, but then had to wait another three years before receiving its premiere at the Paris Opéra (on 3 April 1929). After the modest theatrical success of Le Mas , Canteloube turned to Vercingétorix , an ‘épopée lyrique’ (lyric epic) in four acts. This was composed in 1930–32 and first performed at the Paris Opéra on 22 June 1933. The large orchestra includes real novelty with its parts for four Ondes Martenot to evoke, according to the composer, ‘the mystic moments in the score’. Vercingétorix is a tale of Celtic heroism – the eponymous character was the leader of the Gauls who freed France from Roman occupation – and its visionary world was described by Emile Vuillermoz in his review for Excelsior published on 26 June 1933:
The libretto of Vercingétorix is exactly what Richard Wagner, the author of Siegfried , would have written if he had been a Frenchman. The same preoccupations with deeply rooted ethnicity, the same philosophical and historical viewpoints . . . the same moral and religious mysticism, the same theory of renunciation and of atonement through sacrifice, the same conception of heroism, the same suspicion of human love, and the same exaltation of Parsifal-like chastity.
Wagner’s influence on the French lyric theatre was still at work in the 1930s, especially in an opera like Vercingétorix : Canteloube’s nationalist epic is one of the most conspicuous (and grandiose) examples of this enduring legacy. The score is ripely post-Romantic, full of aspirational leitmotives, one of which (beginning with a rising fourth) evoked for one early critic not only the spirit but also the tune of La Marseillaise . While it is easy to smile at a cast-list which reads like the characters from the Astérix stories, and while the setting of the opening – on the summit of the Puy de Dôme, the ‘montagne sacrée’ – suggests a decidedly Wagnerian approach to stage mysticism, there is a certain nobility in Canteloube’s music (quite unlike his more familiar folk-song arrangements) which is genuinely impressive. This vast score could be an interesting candidate for revival .
A glance at the list of operas first performed in Paris during the 1930s reveals a number of other works which have lapsed from the repertoire. 1930 saw Ibert’s Le Roi d’Yvetot , Marcel Delannoy’s Fou de la dame and Manuel Rosenthal’s Rayon de Soieries at the Opéra-Comique, and Raoul Brunel’s La Tentation de Saint-Antoine at the Opéra. The following year, the Opéra staged Virginie by Alfred Bruneau, the last in a long line of operas by a composer who had made his name in the 1890s thanks to his collaborations with Zola; 1931 also saw the posthumous premiere at the Opéra of Guercoeur by Albéric Magnard (composed in 1897–1901). In 1932, Milhaud’s Maximilien and Alfred Bachelet’s Un Jardin sur l’Oronte were the novelties at the Opéra, while 1933 saw the premiere of Rosenthal’s operetta Bootleggers (libretto by Nino) at the Art-Déco Théâtre Pigalle. At the Opéra, Rabaud’s Rolande et le mauvais garçon was first performed in 1934, followed by Reynaldo Hahn’s Le Marchand de Venise (after Shakespeare) in 1935. The arrival of Enescu’s Oedipe at the Opéra in 1936 meant that, after several years of near-misses, the company had a major new work on its hands. It was not until 1939 that the Palais Garnier put on another novelty of comparable scale: Henri Sauguet’s output was dominated by several successful ballets for Diaghilev, Roland Petit and others, such as La Chatte and Les Forains ; but his largest stage work was the opera La Chartreuse de Parme , composed over a period of ten years (1927–36), dedicated to Milhaud and first performed at the Paris Opéra on 6 March 1939. The libretto is by Milhaud’s friend Armand Lunel (after Stendhal) and, though the opera is in most respects traditional, the music has a melodic grace which raises it well above the level of some of the works composed in the years immediately before the outbreak of the Second World War .
Through harmonic writing nourished by the old Venetian contrapuntists, through instrumentation meticulously crafted and modern, through the original development of ideas . . . the musical quality of the operas of Malipiero is superior to anything which has been seen in Italian theatres since Verdi.
This ringing endorsement of the operatic output of Gian Francesco Malipiero, proclaimed by Massimo Mila ( 1947–8 , 109), is ample encouragement to re-evaluate these largely neglected works, widely recognized at the time as an important new departure for Italian opera, making as they did a definitive and highly imaginative break with the verismo tradition. Malipiero’s triptych with the collective title L’orfeide was composed in 1918–22 and comprises three works: La morte delle maschere , Sette canzoni and Orfeo, ovvero L’ottava canzone . Of these, perhaps the most striking is the first to be composed, Sette canzoni , consisting of seven miniature operas unconnected by plot and ‘threaded together like beads on a string’ (Waterhouse 2001b , 699). It was first performed at the Paris Opéra, conducted by Gabriel Grovlez, on 10 July 1920, but this was an unhappy occasion, as René Dumesnil reported ten years after the event:
Francesco Malipiero has taken up the battle against Mascagni, Leoncavallo and Puccini . . . He is also a friend of d’Annunzio, and his works, the cause of violent controversy, have been the most courageous manifestations against verismo . It would be reasonable to think that the Sette canzoni contained something really subversive, since the audience at the Paris Opéra heard them against an uproar and it was necessary to wait for Mme Bériza to revive them at her theatre in order to be able to listen to them. In truth, because it is both sincere and original, this music stays in the memory. It contains passages . . . which made a lasting impression, even at the first performance, on anyone who didn’t systematically refuse to understand it.
Despite these unpromising beginnings, the opera was soon recognized as a work of considerable significance and as early as 1929 it was the subject of a collection of essays entitled Malipiero e le sue Sette canzoni (Ciarlantini 1929 ), written by some distinguished Italian contemporaries including Franco Alfano, Alfredo Casella and Mario Castelnuovo-Tedesco. Malipiero’s fondness for operatic triptychs is reflected in other works from the 1920s and 1930s: 3 commedie goldoniane (1920–25), Il mistero di Venezia (1925–8) and I trionfi d’amore (1930–31). Usually the author of his own librettos, Malipiero did collaborate in 1932–3 with a leading Italian literary figure, the great Luigi Pirandello, on La favola del figlio cambiato . Malipiero’s interest in earlier music, especially Monteverdi, is reflected in San Franceso d’Assisi (1920–21) which received its first (concert) performance in Carnegie Hall, New York, on 29 March 1922. Merlino, mastro d’organi (1926–7) is rather more of a curiosity given its bizarre and tortuous plot: credulity is strained by a story-line in which a vast magic organ kills all who hear it; its evil-doings are only brought to an end when a deaf mute murders the organist and then turns out to be a reincarnation of his own victim. A good deal more plausible is the brooding Torneo notturno (1929), considered by Mila to be ‘perhaps Malipiero’s operatic masterpiece, which deploys a number of expressive possibilities: Goldonian comedy, mixed with a bitter dose of sarcasm, religious fervour and mystical elevation, and above all a love of the fantastic and of the artificial which is a constant theme, the motor, so to speak, of his artistic creativity’ (1947–8, 109–10). John Waterhouse has characterized the music of this opera as ‘hauntingly enigmatic . . . another of Malipiero’s supreme achievements, in which the obsessively recurring “canzone del tempo” evokes the inexorable destructiveness of time’ ( 2001b , 699).
Antoine de Saint-Exupéry’s largely autobiographical novel Vol de nuit was first published in 1931 and its influence at the time was considerable, even inspiring such ephemeral delights as the ‘Vol de nuit’ perfume by Guerlin, as well as the film Night Flight (1933) starring Clark Gable. One of the pioneers of commercial flying, Saint-Exupéry worked during the 1920s in Africa, then South America where he was a director for Aéropostale in Argentina. Vol de nuit was also to serve as the inspiration for a remarkable one-act opera composed by Luigi Dallapiccola in 1937–9 and first performed at the Teatro della Pergola in Florence on 18 May 1940 (conducted by Fernando Previtali). Dallapiccola had discovered the music of Debussy – especially Pelléas – in his teens, and its impact was such that he stopped composing altogether for three years (1921–4). His enthusiasm for French music was thus an important early influence, and his encounter with Berg in 1934 was to prove decisive. Dallapiccola’s older compatriot Casella wrote admiringly of Volo di notte in his memoirs (first published in Italian in 1941), praising it as among ‘the fruits of one of the richest imaginations in music today, not only in our country but even in the whole of Europe . . . Dallapiccola . . . represents one of the greatest energies to which our musical future can be confided’ ( 1955 , 200). This was the composer’s first stage work and its visionary qualities are perhaps its most enduring feature. The opera
re-uses material from [Dallapiccola’s] Tre laudi . This transference of music originally associated with medieval religious texts to an opera about night flying in the Andes is less incongruous than it may seem, for Dallapiccola’s libretto contains a strong element of religious symbolism. When, at the climax, the pilot Fabien rises above the storm and, just before death, glimpses the infinite, eternal beauty of the stars, his experience has mystical connotations: for Dallapiccola the stars were a symbol of God.
(Waterhouse 2001a , 855–6)
In terms of sheer novelty, the most startling feature of the opera was the introduction to the lyric stage of modern technology, but this was complemented by music of genuine modernity. Mila’s article (referred to above) in the progressive French journal Polyphonie described this successful combination: ‘Volo di notte , drawn from the masterpiece by Saint-Exupéry, is one of the most compelling of modern operas. It goes without saying that its modernism is not only evident in the appearance of aeroplanes and radio transmitters on the stage, but it extends to the score, which it penetrates deeply’ ( 1947–8 , 112).
Paris was under German Occupation from June 1940 until August 1944, but the Opéra continued to flourish, albeit under the ultimate control of the occupying authorities. Visits by German companies attracted a good deal of notice (including the celebrated Bayreuth production of Tristan und Isolde which was brought to Paris in May 1941, with Max Lorenz and Germaine Lubin in the title roles, conducted by Herbert von Karajan), and German repertoire was to dominate the programme, including modern works like Pfitzner’s Palestrina (1917) and Werner Egk’s Peer Gynt (1938) as well as generous helpings of Wagner. No new French opera was put on at the Opéra under the Occupation – a striking contrast with the ballet repertoire which included important new works by Poulenc ( Les Animaux modèles ) and Jolivet ( Guignol et Pandore ) – though there were new productions of Fauré’s Pénélope and Honegger’s Antigone in the spring of 1943.
Elsewhere in Paris things were a good deal livelier – indeed a production from the Opéra-Comique formed the basis of a heroic affirmation of French culture: the first major recording project in France during the period was the famous set of Pelléas et Mélisande featuring Irène Joachim, Jacques Jansen and Henri Etcheverry, conducted by Roger Désormière. It was made by Pathé-Marconi between 26 April and 26 May 1941 and originally issued on twenty 78rpm discs. Music by Offenbach, who was Jewish, was banned during the Occupation, but at the Opéra-Comique Désormière chose another enchanting example of French light opera, Chabrier’s L’Etoile , to coincide with the composer’s centenary in 1941. Extracts from L’Etoile were recorded by Opéra-Comique forces in 1943. The company also put on a new work which enjoyed considerable success at the time. Commissioned by the state in 1938, Delannoy’s Ginevra was finished in 1942 and first performed at the Opéra-Comique the same year (25 July). For some of the musical material, this work draws on a source which would have had nostalgic resonances for the audience of the time: French Renaissance chansons .
Poulenc’s first venture into opera came after more than two decades of working as a composer of ballets and of incidental music for the theatre. His setting of Les Mamelles de Tirésias – by turns uproarious and radiant – is based on a text by Guillaume Apollinaire, who had been such a decisive influence on Poulenc’s work. Poulenc had attended the first performance of Apollinaire’s play in June 1917, when he was in good company, as the audience also included Matisse, Picasso, Braque, Dufy, Cocteau, Eluard, Satie, Diaghilev and Breton. Composed during the war years, Poulenc’s ‘opéra-bouffe’ on Les Mamelles was first performed on 3 June 1947, though Poulenc had performed it privately autour du piano as early as November 1944. His own assessment of the work (for which he declared ‘a passionate fondness’) was that it was ‘one of the few things I have done where I wouldn’t change a single note’ (Schmidt 1995 , 354). Perhaps more than most of his extended pieces, it combines brilliantly the different aspects of Poulenc’s musical personality: farcical humour and rapturous tenderness happily co-exist in a score which is both exuberant and touching .
The climax of Poulenc’s output of religious music (from 1936 onwards), as well as his zenith as an opera composer, is Dialogues des Carmélites , based on an unused film script by Georges Bernanos and first performed at La Scala, Milan, on 26 January 1957 (when it was sung in Italian). The work has been criticized for its episodic construction and its short-breathed musical phrases, but Poulenc’s score burns with that most highly prized of qualities among French composers – ‘sincérité’ – and at its best, the music is inspired. This is Poulenc’s longest work and it is thus no surprise to find it full of the self-borrowings (from a bewildering diversity of earlier works) which make pinning down his elusive expressive and aesthetic intentions so problematic. The religious nature of the work is underlined by the inclusion of sections setting Latin liturgical texts. The most famous – to some, notorious – of these is the remorseless ‘Salve Regina’ sung by the nuns as they make their way to the guillotine (immediately preceded by a march which has its origins in Deux marches et un intermède , a short orchestral work written for a dinner party at the 1937 Paris Exposition); elsewhere a kind of rapt ecstasy can be found in these moments of semi-ritual. The Priest’s farewell Mass, where he and the Sisters sing the ‘Ave verum corpus’, is a memorably beautiful case in point .
Poulenc’s last operatic venture is also his most unusual. Based on a libretto by his old friend Cocteau, La Voix humaine (first performed on 6 February 1959) is a monologue for one singer – Poulenc wrote it for the soprano Denise Duval – with a set comprising a couch and, crucially, a telephone. The result is a gripping one-act drama which has a sustained, claustrophobic intensity rare for Poulenc .
The Paris Opéra was plunged into administrative chaos after the Liberation of the city in September 1944 and this inevitably led to a rather cautious attitude by the management. While new ballets continued to appear regularly, new operas were few and far between. The company itself was constantly dogged by strikes: the theatre closed for a month in 1945 due to industrial action; the musicians were on strike for several months in the first half of 1946; the stage technicians followed suit in 1947 to protest at Serge Lifar’s return to the company and a new dispute closed the theatre for a month in 1948; the musicians were on strike from 28 November 1949 until 10 January 1950; and six weeks later the technicians withdrew their labour for over a month. The fire which broke out on Christmas Day 1950 was just one more reason to close the theatre, this time until April 1951. In almost every season for the next two decades, there were more disputes, including a three-week lock-out in October–November 1953. Finally, in 1969, the theatre closed, ostensibly for repairs, but primarily in an attempt to sort out working practices, contracts and artistic policy. The following year there was no attempt to disguise the reasons: on 30 July 1970 the theatre closed and it was not until 14 months later, on 30 September 1971, that performances resumed in the Palais Garnier. Given the precarious state of the Opéra, its appalling industrial relations and its ageing infrastructure, it is perhaps unsurprising that few important new works saw the light of day there during these turbulent years. The announcement that Rolf Liebermann was to become administrator of the Opéra in 1973 led many to hope that there would be a more innovative artistic policy, and this turned out to be the case .
It was a commission from Liebermann which produced one of the grandest of French operas composed since 1945: Messiaen’s Saint-François d’Assise , first performed at the Palais Garnier on 28 November 1983, after almost a decade of work on the composition and orchestration. Messiaen’s position on opera was ambivalent: while Mozart, Wagner, Gluck and Debussy’s Pelléas had been among his earliest inspirations (and recurred frequently in his own teaching) he declared more than once that he would never write an opera, but admitted elsewhere (even as early as 1948) that it was something he wanted very much to attempt. His mammoth Saint-François d’Assise was conceived on a Parsifal -like scale, and while the music has been greatly (and rightly) admired, the work has been criticized for being too static, too monumental. However, while Messiaen’s intention was never to write a fast-paced drama, to claim that Saint-François is a kind of glorified oratorio is to miss the point: the human drama of the story, the passionate engagement with nature by the principal character and the composer, and the blazing fervour (and incredible beauty) of the music suggest that this is very much an opera, albeit an opera as only Messiaen could (or would) compose – and the work, despite its huge instrumental and vocal demands, has started to make its way into the international repertoire .
Other French operas from the second half of the century include several by Marcel Landowski. A member of the Académie des Beaux-Arts from 1975 (and its Secrétaire Perpétuel from 1986), his works for the lyric stage included Le Fou (1948–55), the one-act Le Ventriloque (1954–5), Les Adieux (given its first staged performance at the Opéra-Comique on 8 October 1960), the children’s opera La Sorcière du placard aux balais (‘The Witch of the Broom Cupboard’, completed in 1983), and three full-length operas. The first of these, Montségur , received its premiere in Toulouse on 1 February 1985 and was described by Landowski as ‘born of the meeting between the passionate story of an impossible love and the equally stirring one of the bloody conflict between two absolute and rival religious faiths’ (Landowski (catalogue of works), 1996). La Vieille maison (1987), to a libretto by the composer himself, was first performed at Nantes on 25 February 1988. Galina (1995), jointly commissioned by the French Government and the Opéra de Lyon and first performed on 17 March 1996, is a most unusual work: Landowski’s own libretto is based on the autobiography of the Russian singer Galina Vishnevskaya, and Vishnevskaya herself was present in the audience at the opera’s premiere. Despite quite a prolific operatic output, Landowski – whose music is rooted in tradition but is sometimes rather anonymous – has failed to make any real impact beyond France. His importance perhaps lies as much in his work as an administrator: he appointed Liebermann to the Paris Opéra and his reforms of musical education, and policy of decentralization, have made a lasting impact on French cultural life .
The Romanian-born Marcel Mihalovici wrote several operas. Apart from the early L’Intransigeant Pluton (1928), these include Phèdre (completed in 1949) and Le Retour (1954), based on a story by Guy de Maupassant. In 1959, Mihalovici composed what is perhaps his most intriguing opera, Krapp, ou la dernière bande , a setting of Samuel Beckett’s Krapp’s Last Tape , lasting about one hour. Described as an ‘opera’, it has a cast of one (like Poulenc’s La Voix humaine ), in this case a single baritone. The role features quite extensive use of Sprechgesang , and the otherwise small orchestra includes a large percussion section, notably a part for vibraphone requiring four players. This work is noteworthy not least because operatic settings of Beckett are so few and far between. It was first performed in a concert version given by Radio France forces under Serge Baudo on 13 February 1961, and the first stage performance took place 12 days later at the Bielefeld Opera .
Claude Prey studied with Messiaen and Milhaud at the Paris Conservatoire and the vast majority of his works were written for small experimental groups specializing in music theatre. The composer of at least 30 such pieces, he had a great deal of fun inventing new operatic sub-genres. His stage works (all to his own texts) include an ‘opéra-cruciverbal’, an ‘opéra-d’appartement’, a ‘mono-mimo-mélodrame’, an ‘opéra-test’, an ‘opéra-epistolaire’, an ‘opéra-kit’, an ‘opéra opus Proust’ and – with a kind of lunatic inevitability – an ‘opéra-opéra’. Prey’s brilliant games with music (parodies and allusions to Wagner, Beethoven, Fauré and others) and with written and spoken language (for instance using only the 12 letters in the title of L’Escalier de Chambord ) can be seen at their most elaborate in O comme eau ou L’ora dopo (1984), subtitled an ‘ode homophone’. Set in an underwater world after the drowning of Venice, this piece employs the vowel ‘o’ throughout, since it is the only sound which the inhabitants can remember. Prey’s works in some ways look back to much earlier innovations like Milhaud’s ‘opéras-minutes’, but their wit, cleverness and their modest scale also ensured that – unlike most of the more grandiose ventures of the 1980s and 1990s – they were quite widely performed. The Prix de Rome for music – intended above all as a nursery for future opera composers – was awarded for the last time in 1968. It had become an anachronism and the student uprisings, the ‘évènements’, of that momentous year brought about its abolition. The work of Prey and others offer the possibility of some intriguing new directions a century on from Debussy’s Pelléas et Mélisande .