.A discussion of Meyerbeer's Robert le Diable. The first singers of Robert and the "Mario-aria" at the beginning of Act II

by Prof. Dr. Wolfgang Kühnhold

(University of Paderborn)

(revised May 15, 1998)

Now with Real Audio clips! (May 17, 1998)

Back to Meyerbeer Fan Club Home Page

When Giacomo Meyerbeer wrote his Robert le Diable, he knew that he could compose for the best singers of Paris, and that meant: of the entire world. In the beginning, this work was created not as a Grand Opéra, but as an Opéra comique. Guilbert de Pixérécourt of the Opéra comique made the initial contract. The librettists were Eugène Scribe and Germain Delavigne. The male singers were projected to be Louis-Antoine-Eléonor Ponchard (Robert) and August Huet (Bertram).

Huet was a well known tenor of the Opéra comique, not as a singer, but as an actor. It was at the end of his career that he got the roleof Bertram. Later on he was the director of the opera-house in Rouen. His voice range was from B flat to F sharp - only one and a half octave, which today is perhaps the range of a baritone. Meyerbeer wrote music for Bertram that fitted Huet's limited vocal expertise perfectly. We can see it in the drafts and those parts that remained in the score of the Grand Opéra. And the libretto shows us that in its original form, the main part of Bertram's role was spoken text.

Ponchard, on the other hand, was the most famous tenor of the Opéra comique. With the high musicality of his diction and the artificial variations of his means of expression he set a new standard for the French Opéra Comique. And he was a great teacher, too. Jean Baptiste Faure, Henri-Bernard Dabadie and Rosine Stoltz were among his students. Another was a young Italian vocalist we have to talk about later on: Giovanni Mario. It was for Ponchard that Meyerbeer wrote the first phrase in the introduction of his new opera: "Chevaliers, c'est à vous que je bois...". I found this first version in Meyerbeer's manuscript of the first act of Robert le Diable, which I discovered in Paris. The same version was found in the role-book of Nourrit, which is deposited in the library of the Paris Opéra. But there, it was sewn together and only the later definitive version could be seen. In the early version the tenor was to sing a trill on A over two measures, marching then over B flat to high C. And he was to do this in his 8th solo-measure, during the first minutes of his appearance! At least Meyerbeer changed these murderous measures in his definitive score. And this well known version is difficult enough!

But the first cast changed on December 29, 1829, when Meyerbeer signed the contract with the Académie de la musique and the three acts of Robert le Diable changed into a "Grand Opéra" with five acts and seven tableaux. The singer of Robert now was Adolphe Nourrit and for Bertram at first it was Dabadie. Lubbert - and not Verron, who told in his mémoires, that it was his idea - proposed that the great Nicolas Prosper Levasseur should sing the role of the fallen angel. Lubbert was the director of the Opéra at the beginning of 1831. Meyerbeer had to revise his opera. For the new tenor only small alterations necessary: Nourrit was one of the best singers of the time. Both Nourrit and Ponchard could sing the high D, which Meyerbeer wrote in the duo between Robert and Bertram in the third act. Only the role of Bertram had to be written into bass-clef. Although Meyerbeer knew Levasseur very well, he did not work easily on this transposition, and there much for Meyerbeer to do in early 1831. In this essay I take into consideration only the casting of the male-roles of Robert and Bertram. The development of the female-roles are much more complicated. And during the first months the whole third act has been changed, the scene of the nuns was created and the new director Verron wished to have a new ballet-scene in the second act too.

We all know that the first performance of Robert le Diable was an immense success. Overnight, Meyerbeer became the most important composer of the entire world. He knew that he owed a great part of this result to his actors. Nouritt was one of the greatest and most intelligent musicians of the time. A great singer, a composer, a writer - he wrote scenarios for ballets of Paris Opera - Giselle for example - created by Marie Taglioni and Fanny Elssler. His voice was of great brilliancy and wealth. People admired his high musicality and the intelligence of his execution. For fifteen years he was in the center of Paris music life. Rossini admired him and he wrote for him the role of Neocles in Le Siège de Corinthe. And he was the first Arnold in Guillaume Tell. The list of his roles is immense: Massaniello (La Muette de Portici),Eleazar (La Juive), Raoul (Les Huguenots), Gustav III (La Bal Masque) and so on. He sang the role of Robert from 1831 until 1837. Only in a few performances Lafont, who normally sang in the role of Raimbaud, alternated with him.

The end of his career was tragic. In 1837 Gilbert Duprez became the first tenor of the Opéra. Nourrit felt cast off. He went to Italy and there he died: In a paroxysm of melancholia he threw himself from the window. But Gilbert Duprez made radical changes in the vocal technique. In the role of Arnold he sang the high C in chest voice for the first time. Rossini, Meyerbeer and the contemporary composers and, especially, the singing-teachers were shocked. But it was a turning-point of the art of singing. Since then the tenor lost more and more the ability to use the head voice and "voix mixte" in the right way. And they became unable to sing more than the high C. Today high D or others are in a range only for specialists. Duprez changed in that way the French art of singing, but it was Italy that adapted his placing of voice in new techniques. From this the "heroic tenor" arose.

After the premiere of Les Huguenots on February 29, 1836, which once again moved the public to enthusiasm, Robert le diable was performed 16 more times, although only 5 times in 1837 in its full length, and on 27 March the fifth act alone. In letters to his wife Minna, Meyerbeer again and again expressed his annoyance with this practice. However, mention must be made of one particular performance. As the new director of the Opera, Duponchel had arranged a concert on the occasion of the marriage of the Duc d'Orleans on 10 June 1837 in Fontainebleau: the third and fifth acts of "Robert le diable" were performed, with Guilbert Duprez in the title role for the first time. Although the newspapers were full of praise, Meyerbeer did not like him; without question, he preferred Nourrit. In any event, Duponchel stated his intention after the concert of preparing a revival of "Robert le diable". But Duprez did not want to sing the title role. Whether he had heard Meyerbeer's negative criticism, or whether he himself had second thoughts -- both are possible. In any case, Duprez then sang the entire part for the first time on 4 November 1840. After this concert, "Robert le diable" was presented in 1837 only one more time, on 20 November. On that occasion, Lafont again sang the title role. Meyerbeer was not at all satisfied with him in this role. As early as 23 September, he discussed Lafont with his secretary Gouin: in no case should he participate in the revival of "Robert". At the same time, he inquired about when the revival would finally take place. Since Duprez did not want to sing the role now, either, he suspected intrigue. On 2 November, however, he inquired about a new tenor who had come from Italy. His publisher, Schlesinger, informed him on 8 November that the tenor had received 200 Francs a month from Duponchel in order to study for six months. That must have aroused Meyerbeer's interest. He tried to establish contact with this Italian, but only on 29 December did he succeed in hearing him. He wrote immediately to his wife Minna: "A half an hour ago, I finally heard de Candia, after his leading me around until now. He confessed to me that he has been suffering from a venereal disease for the last 8 months, which has weakened his chest and voice very much, the latter being particularly very rough (a nice prerequisite for making one's debut in a role like 'Robert'). His voice is by nature very nice, he seems to have natural abilities, but has never sung in French, and has never entered a theater, though he is musical. -- In a few days, he will begin to study 'Robert', and in two weeks I hope that I will then be able to tell whether or not it will be possible for him to make his debut in 'Robert'."

Then, on January 8th of the new year, he decided to proceed, although Mario on the stage "cannot walk, cannot stand, cannot sing, cannot speak French." His first voice teachers were the well-known Ponchard and Allari, then later Giulio Marco Bordogni. Mario's debut was planned for May. Meyerbeer even promised him a new aria and a duet. Thereupon Mario expressed his thanks to him in a delightful letter, in which he provides evidence that his French is indeed still very unusual. On 17 January, then, he saw "Candia" again, and yet again on 22, 26, and 29 January. -- This suggests that Meyerbeer was working with the young man on his role. Evidently he had confidence in de Candia and wanted to help him improve. He was then receiving instruction in drama from Pierre Marie Nicolas Michelot.

Again and again, Meyerbeer wrote to Minna about problems with "Robert" and Mario -- to be sure, another factor seems to play a role, too, namely, that he did not want to return to Berlin, which he detested, and therefore was looking for a reason to put off his departure. He finally left on 24 February, but not before he rehearsed with Mario several more times, as his pocket calendar indicates. Traveling via Dresden, where the performance of the Les Huguenots was being prepared, he finally arrived in Berlin on 17 March. Since Meyerbeer was very dissatisfied with the revival of Robert le diable on 20 November 1837, mainly, it seems, with Lafont in the title role, the opera was performed only four more times through the summer of 1838: on 29 April, 1 and 25 June, and on 25 July. On 19 May, Meyerbeer was back in Paris, and no later than 22 May he again visited deCandia. He was still not satisfied with him; he had not made any real progress. However, he played his new aria for him and his teacher Bordogni; both were quite enthusiastic. To be sure, he himself kept postponing the debut. In the following months, we find the name "de Candia" again and again in his calendar -- proof that he was continuing to work with him. Then, on 30 November, the day finally arrived. The long awaited revival of Robert le diable, with the female roles sung by Nau (Isabelle) and Dorus-Gras (Alice) and the male roles by Derivis (Bertram) and Wartel (Raimbaut), again drew the Paris audience into the Opera. And with them, a new "Robert" appeared on the stage, a young Italian nobleman: Giovanni Matteo Cavaliere di Candia, known as "Mario." His debut added great success in the numbers of the performances of Robert le diable. When he sang the opening phrase of the role of Robert, the audience was highly enthusiastic. But in the second act the audience was treated to something new. After the well known Entr'acte the audience did not hear the air of Isabelle. Instead, they heard a previously unknown recitative and a new air of Robert, which filled the audience with enthusiasm. After this a star was born!

Cavaliere di Candia was born on October 17, 1810 in Cagliari. He should have become an officer, but political reasons made him emigrate to France -- and there was a love-story with a ballerina. His family renounced him, so he came to Paris. It is said, that his voice was one of the most beautiful tenor-voices of all. The elegance of his diction and the high musicality of his stylistic sense have been praised. He possessed a brilliant appearance on stage and was a great talent as an actor. He also created the role of Ernesto in the first night (January 3, 1843) of Donzetti's Don Pasquale together with Luigi Lablache (Pasquale), Antonio Tamburini (Malatesta) and Giulia Grisi (Norina), whom he married in 1844. Since 1847 he became the first tenor at Covent Garden. He sang there until 1871. His daughter, Mrs. Godfrey Pearce, wrote his biography, The Romance of a Great Singer (London, 1910). His career is, if nothing else, a demonstration of Meyerbeer's infallible instinct for the quality of singers' voices. Despite the many mishaps, despite de Candia's embarrassing illness, despite his deficient French, despite his lack of stage experience, Meyerbeer stood by this young singer steadfastly. The fact that he never sang the role of Robert later outside of Paris is something one can hardly understand. Perhaps it was because of the special situation of the Paris opera.

But what about this aria? It is a strange fact, that there isn't any solo aria for the title hero in Robert le diable. You often hear that the main role is that of Bertram. I would say that is right, but the opera is not called "Bertram le Diable". Even though Robert sings the wonderful "Sicilienne" and the "Grand Duos" with Isabelle or Bertram, it still seems that Robert is given a disproportionately small role. Perhaps Meyerbeer corrected this disproportion for the first performance in London which took place on June 11, 1832. The whole Paris cast was there, but they had to sing in Italian. For this occasion Meyerbeer wrote a new aria for Nourrit. But this piece of music existed only in a manuscript of the author and in a copy, both part of Meyerbeer's library in Berlin. Both are missing since the Second World War. Perhaps it was the same music we can find today in the aria for Mario. We don't know it exactly, because the autograph copy of the "Air de Mario" was in Berlin, too. We cannot say anything about this piece of music, whether it is a new composition or an altered edition of the "Nourrit-aria". Meyerbeer's library was brought out of Berlin in the last weeks of the war. Since these days it is missing.

It seems probable, however, that Meyerbeer had worked out this aria between 9 January and 21 June 1838 (cf. the dates above); whether it was a completely new composition or a revision-- his diaries and pocket calendars say nothing about it. As is the case so often!

And like the whereabouts of these original scores, so it is in the case of the text of this aria. The author is said to be Emile Deschamps -- or so Théophil Gautier tells us in his Histoire de L'art Dramatique. But except for this short notice, there is no reference in any existing source. After a stylistic comparison I suppose it was Delavigne, but I don't have any other documentation.

In 1988 two great singers created these roles again: Chris Merritt and Samuel Ramey in Carnegie Hall concert with Eve Queler as conductor. In this concert Chris Merritt sang the "Mario-aria" and it was a miracle. Not only the wonderful musicality of the singer and his admirable high C at the end, but also the fact that he sang this piece in its full length.

In the pressed editions of the opera (1832 and 1837) one does not find this aria. Only in the supplement of the piano scores there is a version of this aria: Recitative in G, aria in D (the original version, which Chris Merritt sang, is in E flat). But there appears only the first part of the aria, "Priere". We don't know of the existence of any pressed score or orchestra-material of this piece, although there must have been something like that to perform it. For example: In the German libretto you can find this aria and it must has been heard in the 19th century. So it was very interesting to learn how Chris Merritt and Eve Queler gotthe score. Chris Merritt discussed his research with Clarissa Lablache.[Click here for a full text of the interview with Chris Merritt]. And it is indeed a wonderful feeling finding an unknown piece of music in archives or libraries. We also got this feeling, when we started the new edition of Robert le diable in the "Meyerbeer Werkausgabe", which was established at Paderborn University in collaboration with Ricordi publishers and the Meyerbeer Institute. Of course we knew the recitativeand aria "Ou me caché / O ma mère", but we thought to be the first ones who knew about the existence of the material and the second part, which was unknown since 1988. But we were not correct. Chris Merritt knows it, too. The aria is transmitted in a copy of the first score of Robert le diable. The original "director's score" is missing. Remember the number of those performances! It was from the first night November 21, 1831 until 1841 exactly 223 performances. So this old score became unreadable after about ten years. As an important source it has been copied for the archive of the Opera and then became waste-paper. This copy, the so called "Repetitionspartitur" is a very important source, especially for our new edition. You can find many measures there which do not appear in the pressed scores. Some of them are in Meyerbeer's own handwriting. I think this must have been the manuscript Chris Merritt spoke about.

Indeed the recitative and the first part of the aria were well known out of the piano score. It was performed in all countries and it was published. On December 22, 1839 Meyerbeer noted in his pocket calendar that he was involved with the third revision of de Candia's aria. It was supposed to appear in a collection of Lieder that Maurice Schlesinger subsequently published at the beginning of March 1839 as a "Collection des melodies de Giacomo Meyerbeer": 22 titles, of which the last was the "Scene et prière, composees pour Mario dans Robert-le-Diable". But the aria has two parts: "Prière et marche". In this connection it is of importance that there are different sources of this piece in the Paris "Bibliothèque de l'Opera" but they are not easy to find. Especially the "role book" of "de Candia" is of great importance. There you can find a curiosity. The second part -- Chris Merritt called it "Cabaletta" -- has been sewn together. Of course we had to open these pages and saw the same notes which we knew out of the score. Together with this and the material for the orchestra we can reconstruct the fact that this second part was rehearsed. But probably it has never been performed. The critic for Théophil Gautier, which wrote for the debut of Mario, told us only about the "prière"! If Mario sang the "marche" too, I think we would know it from the critics of the day. To be sure, there is a reference in the periodical "France musicale" of December 9, 1838 (No. 50) to a second part. The reviewer, after he had complained as late as December 2nd that Mario could not make one forget Nourrit, now said that Meyerbeer had been wise to shorten the new aria in the second act. The second part was superfluous, he continued, the andante alone sufficed to assure Mario of his applause. Perhaps the second part had already been omitted in the performances on December 3rd and 5th, perhaps the reviewer had learned of its existence from various other sources -- we cannot say more precisely. So it may be that Chris Merritt is the first tenor who sang the whole aria! At least he shows that both parts definitely belong together! To be sure, the applause of the audience sets in after the first part, but the escalation of the march is needed, with the possibility of reaching high C, in order to bring this aria to its triumphant conclusion. The achievement of Chris Merritt is a very important act that we must admire! Now it is our part in Paderborn to edit the full score of Robert le diable. There are still more previously unknown and wonderful pieces in this opera. The version by Eve Queler has many cuts, even in the "Mario-aria;" eight measures are lacking and a few, but little mistakes are heard. The new score with a new piano score and the whole orchestra part will be published in 1999. And we all hope for a wonderful performance of Robert le diable at "Berliner Staatsoper" in 2000. Perhaps the new century will become Meyerbeer's time.

Prof. Dr. Wolfgang Kühnhold

"Mario aria" - prière clip

"Mario-aria" - Marche clip

Meyerbeer Fan Club Home Page

The story of Robert le Diable

A Few Words About Robert le Diable, by Tom Kaufman

Interview with Chris Merritt, with Clarissa Lablache

Review of Robert le Diable by Tom Kaufman

"Mario-aria" libretto

Join the Meyerbeer Fan Club! Feedback and Membership MFC Members may obtain a copy of the 1988 Carnegie Hall performance of Robert le Diable! Please inquire!


Send us your comments! mailto:meyerb@meyerbeer.com