(Fr.).
The most important species of French opera in the period from Lully to Rameau (1673–1764). In its organization and musical character it shares many features with lesser genres such as the pastorale-héroïque. It is nevertheless distinguished by its five-act structure (others usually had fewer acts) and by a greater dramatic intensity and seriousness of tone. In the later 18th century the tragédie en musique profoundly influenced the operatic reforms of Jommelli, Traetta and, above all, Gluck. In their turn the French tragedies of Gluck, Piccinni, Sacchini and Salieri, produced in the 1770s and 80s, represent a transformation and final flowering of the genre.
3. Dramatic and musical characteristics, 1673–1773.
GRAHAM SADLER
During the greater part of the Lully–Rameau period, the term ‘tragédie lyrique’ is rarely encountered. For the first 85 years of the genre’s existence, this expression was scarcely ever employed by librettists or composers. Almost without exception, librettos printed before 1760 use the terms ‘tragédie’ or ‘tragédie en musique’. On printed scores the same terms are found with similar consistency, together with such variants as ‘tragédie mise en musique’ or (after the mid-18th century, when old librettos began to be recycled) ‘remise en musique’ (e.g. Dauvergne’s Enée et Lavinie, 1758). Le Cerf de la Viéville is typical of his time in referring to ‘nos Opéra que nous appelons des Tragédies en Musique’ (Comparaison de la musique italienne et de la musique françoise, 1704–6, iii, 3). During the 18th century, as other genres grew in stature and popularity, writers found the need to invent additional makeshift terms (e.g. ‘tragédie-opéra’, ‘opéra tragédie’).
It was in literary circles, and not until the mid-18th century, that the expression ‘tragédie lyrique’ first became fashionable. Rémond de Saint-Mard (1741), for example, uses ‘tragédie lyrique’ and ‘tragédie en musique’ interchangeably; Cahusac (1751, 1754) is more consistent in employing the former to distinguish lyric from spoken tragedy. During the 1760s the new term is found sporadically on title-pages (e.g. Dauvergne’s Hercule mourant, 1761). But only in the following decade, particularly after the arrival in Paris of Gluck’s operas, does it appear at all frequently in this context.
For much of the 20th century ‘tragédie lyrique’ was the preferred term. Yet present-day scholars are increasingly following Girdlestone’s lead in returning to the terms used by librettists and composers, at least for the Lully–Rameau period.
A year after seizing control of the Académie Royale de Musique and hence acquiring a monopoly of French opera, Lully produced his first tragédie, Cadmus et Hermione (1673). Thus was born the genre that was to dominate the composer’s output and bear his imprint for many years to come: between them, Lully’s 13 completed tragédies, all but two to librettos by Philippe Quinault, set the dramatic and in some respects the musical tone of the genre for almost a century.
The tragédie en musique emerged more or less fully fledged. Although various developments may be traced through Lully’s subsequent works, Cadmus already includes most features characteristic of the genre. The Lullian tragédie is, in fact, an extraordinary amalgam of pre-existing elements. From the tragedies of Corneille and Racine come the five-act structure and the use, if no longer exclusive, of alexandrines, while the ‘déclamation enflée et chantante’ of the tragédienne C.C. La Champmeslé influenced Lully’s conception of recitative. From the ballet de cour and elsewhere come the panegyrical prologues, stage spectacle, dances, symphonies and choruses. From the tragédie à machines, the pastorale and pre-Cornelian tragedy, the comédie- and tragédie-ballet come other elements, both musical and dramatic. Lully’s supreme achievement is to have synthesized all these into an art-form in which everything – music, drama, dance, staging – was subservient to an overriding dramatic unity: a true Gesamtkunstwerk ‘avant la lettre’.
This eclectic mixture proved astonishingly popular at all levels of society, from Louis XIV down. Having created a taste for the tragédie en musique, Lully used his monopoly to ensure that only his operas could satisfy that taste. In any year after 1673, audiences could see a new Lully tragédie (they appeared at the rate of about one a year) and at least one revival of his existing operas, but no others. Given that devotees could buy season tickets and that the season lasted most of the year, it is hardly surprising that audiences got to know these works by heart; there are, indeed, reports of their joining in with the actors.
Lully’s unexpected death in 1687 left something of a vacuum. Because of his monopoly, few French composers had first-hand experience of composing opera. (M.-A. Charpentier was a rare exception, having written music for various private productions, including biblical operas for the Jesuits.) To fill the vacuum, the practice of reviving Lully’s tragédies continued, reinforcing their dominant position. Many went on to be revived for 60, 70 or 80 years after their creation; Thésée lasted an astonishing 104 years, until 1779.
Such new works as appeared in the remaining years of the 17th century were almost all tragédies in the Lullian mould. Among the composers were Lully’s son, Louis, and former colleagues or pupils, Collasse, Marais and Desmarets. The works of these years are not without originality or musical interest. Indeed, Charpentier’s Médée (1693) is one of the finest in the repertory. But almost all were judged unsatisfactory by comparison with the works of the revered Lully – Charpentier’s because of its too-adventurous harmonic idiom, others because they seemed pale imitations. Only Collasse’s Thétis et Pélée (1689) enjoyed a success comparable with the master’s.
Almost half a century separates the last of Lully’s completed tragédies (1686) and the first of Rameau’s (1733). During that time several librettists emerged, among them Danchet, Lamotte and Roy, who came near to equalling the great Quinault without deviating much from his style. Yet even composers of the calibre of Campra, Destouches, Marais and Montéclair seldom managed the musical ingredients in a manner that carried dramatic conviction. Such men often seemed more comfortable in the lighter genres. The gaps between successful tragédies became steadily wider: Destouches’ Omphale (1701), Campra’s Tancrède (1702), Desmarets and Campra’s Iphigénie en Tauride (1704), Marais’ Alcyone (1706), Gervais’ Hypermnestre (1716), Montéclair’s Jephté (1732). By the late 1720s, indeed, the Opéra had become ‘a veritable graveyard of the tragédie lyrique’ (Anthony).
Yet for Rameau the tragédie remained the most esteemed genre. In all, he devoted to it about a third of his operatic output. Few others now showed the same commitment. In the 20-year period 1739–58, the Académie Royale de Musique presented only six new tragédies. (A similar period between 1718 and 1737 included three times as many.) If Rameau’s first operas initially appeared controversial, that was not through any radical break with tradition: most elements of his style are at least hinted at in ‘pré-ramiste’ operas of the 1720s and early 30s. The Lulliste–Ramiste dispute had more to do with the complexity and intensity of Rameau’s idiom. Once audiences had accepted this, they readily placed Castor and Dardanus on a level equal to, or even higher than, Lully’s masterpieces.
From its beginnings, the tragédie en musique had been inseparably linked with the court and the Académie Royale de Musique (whose entrepreneurs inherited Lully’s monopoly), and thus identified with the political and social hierarchy. Inevitably, therefore, the genre found itself attacked in the 1750s by the anti-establishment intellectuals of the new Enlightenment, prominent among them F.M. Grimm and J.-J. Rousseau. The resulting Querelle des Bouffons, as much political as artistic in its motivation, succeeded in undermining the whole aesthetic basis of serious French opera and hastened the search for alternatives. There were those who clung to the tragédie: if anything, the Querelle led to an upsurge of activity. Between 1758 and 1773 a dozen new tragédies appeared, including such anticipations of Gluck as F.-A.D. Philidor’s Ernelinde (1767) and Gossec’s Sabinus (1773). Dauvergne’s Enée et Lavinie (1758) is one of several to take the hitherto unthinkable step of re-using librettos from the golden age; Mondonville (1765) even had the temerity to reset Quinault’s Thésée. Yet few were successful enough to merit a revival, and it was gradually conceded that the genre was moribund. Moribund but not dead, for the arrival of Gluck’s operas in Paris in 1774 was to provide a massive and invigorating transfusion. Gluck’s work was ‘at once a blow to French opera and a renewal of it’ (Einstein, Gluck, 1936, p.138).
Opera in France emerged at a time when the native spoken theatre was at its peak. This fact, and the fact that the ‘classic’ French theatre of Corneille and Racine was governed by a strict code of rules explains many characteristics of the tragédie en musique. Yet although the new genre shares features with spoken tragedy (most obviously the five-act structure), it was never conceived of as merely an adaptation of the existing one. Spoken tragedy, it was widely believed, had reached a state of perfection: to graft music on to it would be inappropriate. Far better that lyric tragedy should aim at a perfection equal to that of Corneille or Racine without directly imitating it. Indeed, lyric tragedy came to be seen not only as the counterpart but also the inverse of spoken tragedy.
The inversion applies first to the subject matter. Whereas spoken tragedy preferred to treat historical figures, such subjects were considered inappropriate for opera. What rational mind could tolerate the idea of a Brutus or a Pompey expressing himself solely through singing? But if opera restricted itself to a world inhabited by gods and allegorical or legendary figures, the problem disappeared. Hence the preponderance of tragédies based on classical myth or legend (Lully, Alceste; Rameau, Castor et Pollux), Italian epic (Lully, Armide; Campra, Alcine) or Spanish romance (Lully, Amadis; Destouches, Amadis de Grèce). For the same reasons, librettists turned very occasionally to the Old Testament (Montéclair, Jephté) or other ancient religious sources (Rameau, Zoroastre).
Almost by definition, then, opera made extensive use of supernatural elements, collectively known as le merveilleux. In spoken tragedy this had been marginalized; certainly it rarely played a part in the dénouement. In the tragédie en musique, conversely, the supernatural was considered not only desirable but essential. Indeed, for many writers le merveilleux was the key feature – ‘la pierre fondamentale de l’édifice’ (Cahusac, 1751) – that distinguished lyric from spoken tragedy. If it was pointless to imitate what spoken tragedy did perfectly, Cahusac argued, lyric tragedy was bound to explore what the theatre was forbidden from doing. In this way tragédie en musique became the antithesis of its spoken counterpart. Le merveilleux provided justification for this inversion. Where spoken tragedy eschewed stage machinery, the supernatural happenings in opera cried out for elaborate effects. Where violent events in spoken tragedy were perforce enacted offstage, they happened in full view in lyric tragedy (compare the treatment of the sea monster in Racine’s Phèdre and in Rameau’s Hippolyte). Where dance was seldom required in spoken tragedy, it was justified within the supernatural milieu of opera in the same way that music was. Even the balance of elements in the plot was inverted: whereas love interest in the theatre was usually secondary to themes of a political nature, overtly political themes were generally thought ill-suited to opera. Rather, the typical tragédie en musique adopted one of a number of formulae involving a pair of lovers and one or more rivals.
Similar inversions may be seen to apply to the Cornelian unities. In lyric tragedy, unity of place was routinely ignored: it was usual for each act to be set in a different location, and individual acts might include further scene changes. Such infringements were positively encouraged by La Bruyère (‘il faut des changements’) though discouraged by the Abbé de Mably. Unity of time was likewise generally ignored. As for unity of plot, the episodic nature of a genre so devoted to divertissement and spectacle made this difficult to respect, although the comic subplots in Lully’s early operas are avoided in all later ones.
The object of both lyric and spoken tragedy was, in Aristotle’s phrase, to excite pity and terror. Yet this was never the sole or main aim of the tragédie en musique. The prevailing tone, at least until the time of Rameau, was more elegiac than tragic. Any pity and terror engendered was quickly dissipated by scenes of calm and joy.
For La Bruyère, indeed, the essential characteristic of the genre was not so much to plumb the depths of human experience as ‘to hold the spirit, the eyes and the ears in an equal enchantment’. This was a view echoed many times in the 18th century to justify the mixture of drama, spectacle, singing and dancing. In order to accommodate the competing demands of these elements, Lully and Quinault established a convention whereby the action would give way at some point in each act to a fête or divertissement. Whereas the dramatic scènes would be declaimed predominantly in recitative, the fêtes would involve airs, choruses, ballet and spectacle in continuous sequences that between them might make up a third of the whole work or more.
Fêtes could nevertheless help further the action: the goddess Cybèle uses a dream-sequence to apprise Atys of her love and warn him of the consequences if he spurns her (Atys, Act 3). Moreover, fêtes were often effective in preparing for what Aristotle calls peripeteia – unexpected reversals of fortune. It is during pre-nuptial rejoicings that Roland learns with horror of the betrothal of Angélique to his rival Médor (Roland, Act 3). Such fêtes may even generate a degree of dramatic irony, where the audience knows that the celebrations are premature (e.g. Hippolyte et Aricie, Act 3 and 4). Significantly, the contrast between action and fête is often reflected in a work’s tonality, which will be in a state of flux during the former but static during the latter.
In Lully’s day the scènes consisted largely of declamatory, French-style recitative, with painstakingly notated rhythms, fluctuating metres and active bass lines. This was periodically relieved by binary or rondo petits airs usually expressing some sententious maxim, or – less often, and only in passages of high emotion – by accompanied recitative, elegiac monologues or laments. During the 18th century the recitative changed remarkably little: Rameau could genuinely claim to have taken as guide ‘la belle déclamation & [le] beau tour du Chant qui regnent dans le Récitatif du Grand Lully’. Steadily, however, the proportion of recitative to set piece decreased, while that of accompanied to unaccompanied recitative grew. Likewise there was greater musical involvement in the scènes: accompaniments of petits airs became more elaborate, while the monologue, in the hands of Destouches, Campra and, above all, Rameau, could generate an intensity undreamt of by Lully. Nevertheless, the primacy of the action outside the fêtes was seldom forgotten: however intricate the accompaniments, the vocal line remained predominantly syllabic and the transition from recitative to set piece flexible. Monologues often lack a final ritornello, so that the ensuing recitative immediately follows the singer’s last words; they are usually placed at the beginnings of acts so as not to interrupt the action.
Indeed, the principal differences between Lullian and Ramellian tragédie have little to do with dramatic structure and much more with musical expansion. Campra may have thought Rameau’s Hippolyte contained enough music to make ten operas; but Campra himself, with Destouches, Montéclair and others, had contributed much to this expansion. It affected every aspect, including choruses, dramatic symphonies, dances and ariettes. These last, unknown to Lully, were fully developed da capo arias, the only movements to allow extensive solo vocal display. Characteristically, though, their role was purely decorative. As such, they were confined to the fêtes and, in many tragédies, limited to a single example. The ariette was French Baroque opera’s most obvious debt to Italian opera; yet in the course of the 18th century the tragédie gradually succumbed to the general influence of italianate music even if most composers contrived to disguise it with a French accent.
For the first 75 years of its existence, the tragédie en musique was prefaced by a prologue, roughly the length of a single act. In Lully’s day the prologue had a political function – to focus attention on the monarch: Louis XIV’s achievements, real or imagined, were allegorized in a brief action usually independent of the ensuing tragédie. During the 18th century, and especially after Louis’ death in 1715 when such panegyric became unfashionable, the subject matter of prologues varied widely: some (for example that in Campra’s Idomenée) act as general introductions to the ensuing drama; others (Hippolyte et Aricie) justify the treatment of the source material; one (Campra’s Achille et Déidamie) even makes a polemical point in the Lulliste–Ramiste dispute. With Rameau’s Zoroastre the prologue was, with rare exceptions, abandoned. In his preface, the librettist Cahusac argued that the time devoted to the prologue would be better spent on the main plot; instead, ‘the overture serves as prologue’. Indeed, for some time (and several decades before Gluck) Rameau had been experimenting with ways of linking overture and drama.
When Gluck published his preface to Alceste (1769), many of the reforms he advocated for Italian opera seria had long been established practice in France – not just the integration of the overture but the avoidance of excessive vocal display, irrelevant ritornellos, da capos, superfluous ornament, too sharp a contrast between recitative and aria. In fact, Gluck’s librettist Calzabigi (possible author of the Alceste preface) had long appreciated the potentiality of French opera. Meanwhile, Diderot, D’Alembert and other Encyclopedists advocated a simplification and humanizing of the tragédie and the adoption of an international musical language. Thus Gluck, disappointed with his lack of success in Vienna, was well aware of the advantages of moving to Paris. His six French tragédies (Iphigénie en Aulide, Orphée et Euridice, Alceste, Armide, Iphigénie en Tauride and Echo et Narcisse) exactly capture the ideals of all those reformers and others, among them Algarotti and Noverre. In these works he preserves the best of French practice – the structural flexibility and the subordination of aria, ballet, chorus, symphonie to the requirements of the drama – but subjects it to a rigorous simplification. Although le merveilleux is not eliminated, it is never allowed to detract from the sincere expression of human feelings. The traditional French musical idiom was now decisively abandoned; Gluck’s claim that ‘I have found a musical language fit for all nations, and hope to abolish the ridiculous distinctions between national styles of music’ (Mercure de France, February 1773) was to some extent justified.
There were many who nevertheless found such operas too severe. As a rival to Gluck, the Italian composer Piccinni was brought to Paris and championed by the librettist Marmontel. Piccinni’s ‘tragédies lyriques’ have the structural flexibility and even something of the musical style of Gluck’s French operas; insipid as they now seem, they integrate the ballet, chorus and spectacle in a way that would satisfy the Encyclopedists. Piccinni’s scoring is fuller than Gluck’s, and he no longer confines long arias to soliloquy. Works like Roland (1778) and Atys (1780) are based on Quinault librettos, albeit ‘Marmontelized’ (Grimm’s term) into three acts; they thus make extensive use of the supernatural. In the subsequent development of tragédie lyrique by Lemoyne, Gossec, Sacchini, Salieri, J.C. Bach and others, it was usually Piccinni rather than Gluck who was taken as a model. As well as further resettings of Quinault and other venerable old librettos, operas of the 1780s include adaptations of Corneille (e.g. Sacchini, Chimène, 1783), Greek tragedy (Lemoyne, Electre, 1782; Salieri, Les Danaïdes, 1784) and historical subjects (Sacchini, Arvire et Evelina, 1788).
J.N. de Francini and others, eds.: Recueil général des opéra représentés par l’Académie royale de musique depuis son établissement (Paris, 1703–45)
F. and C. Parfaict: Histoire de l’Académie royale de musique depuis son établissement jusqu’à présent (MS, 1741, F-Pn)
Mémoires pour servir à l’histoire de l’Académie royale de musique vulgairement l’Opéra depuis son établissement en 1669 jusqu'en l’année 1758 (MS Amelot, F- Po)
L.-F. Beffara: Dictionnaire de l’Académie royale de musique (autograph MS, 1783–4, F-Po)
C. Perrault: Critique de l’opéra, ou Examen de la tragédie intitulée ‘Alceste ou Le triomphe d’Alcide’ (Paris, 1674/R1987 in Textes sur Lully et l’opéra français [with introduction by F. Lesure])
J. de La Bruyère: Les caractères, ou Les moeurs de ce siècle (Paris, 10/1699)
A.L. Le Brun: Théâtre lyrique; avec une préface où l’on traite du poëme de l’opéra, et la réponse à une épître satyrique contre ce spectacle (Paris, 1712/R1987 in Textes sur Lully et l’opéra français [with introduction by F. Lesure])
J.-B. Dubos: Réflexions critiques sur la poésie et sur la peinture (Paris, 1719, 7/1770/R; Eng. trans. of 5th edn, 1748/R)
G. Bonnet de Mably: Lettres à Madame la marquise de P*** sur l’opéra (Paris, 1741/R)
T. Rémond de Saint-Mard: Réflexions sur l’opéra (The Hague, 1741/R)
C. Batteux: Les beaux-arts réduit à un même principe (Paris, 1746, 3/1773/R)
P.C. Roy: ‘Lettre sur l’opéra’, in E.C. Fréron: Lettres sur quelques écrits de ce temps, ii (Geneva, 1749), 7–22
L. de Cahusac: ‘Enchantement’, Encyclopédie, ou Dictionnaire raisonné des sciences, des arts et des métiers, ed. D. Diderot and others (Paris, 1751–80)
L. de Cahusac: La danse ancienne et moderne, ou Traité historique de la danse (The Hague, 1754/R)
F. Algarotti: Saggio sopra l’opera in musica (Bologna, 1755/R, 2/1763/R; Eng. trans., 1767)
R. de Calzabigi: Dissertazione … su le poesie drammatiche del sig. Abate Pietro Metastasio (Paris, 1755)
J.-F. Marmontel: Eléments de littérature, Oeuvres complètes, xii–xv (Paris, 1787, 2/1818–19/R)
D. Launay, ed.: La querelle des bouffons (Geneva, 1973) [facs. of 61 pamphlets pubd 1752–4]
AnthonyFB
J. Ecorcheville: De Lulli à Rameau, 1690–1730: l’esthétique musicale (Paris, 1906/R)
E. Gros: Philippe Quinault (Paris, 1926/R)
P.-M. Masson: L’opéra de Rameau (Paris, 1930/R)
C. Girdlestone: Jean-Philippe Rameau: his Life and Work (London, 1957, 2/1969)
P. Howard: Gluck and the Birth of Modern Opera (London, 1963)
D. Heartz: ‘From Garrick to Gluck: the Reform of Theatre and Opera in the Mid-Eighteenth Century’, PRMA, xciv (1967–8), 111–27
J.G. Rushton: Music and Drama at the Académie Royale de Musique (Paris) 1774–1789 (diss., U. of Oxford, 1969)
C. Girdlestone: La tragédie en musique (1673–1750) considérée comme genre littéraire (Geneva, 1972)
R.M. Isherwood: Music in the Service of the King (Ithaca, NY, 1973)
G. Cowart: The Origins of Modern Musical Criticism: French and Italian Music 1600–1750 (Ann Arbor, 1981)
L. Rosow: Lully’s ‘Armide’ at the Paris Opéra: a Performance History, 1686–1766 (diss., Brandeis U., 1981)
H. Schneider: ‘Tragédie et tragédie en musique: querelles autour de l’autonomie d’un nouveau genre’, Literatur und die anderen Künste (Bayreuth, 1982), 43–58
C. Kintzler: Jean-Philippe Rameau: splendeur et naufrage de l’esthétique du plaisir à l’âge classique (Paris, 1983, 2/1988)
L.E. Brown: ‘Departure from the Lullian Convention in the Tragédie lyrique of the préramiste Era’, RMFC, xxii (1984), 59–78
R. Fajon: L’Opéra à Paris du Roi Soleil à Louis le Bien-Aimé (Geneva, 1984)
W. Weber: ‘La musique ancienne in the Waning of the Ancien Régime’, Journal of Modern History, lvi (1984), 58–88
L.E. Auld: The ‘Lyric Art’ of Pierre Perrin, Founder of French Opera (Henryville, PA, 1986)
P. Howard: ‘The Influence of the Précieuses on Content and Structure in Quinault’s and Lully’s Tragédies lyriques’, AcM, lxiii (1991), 57–72
C. Kintzler: Poétique de l’opéra français de Corneille à Rousseau (Paris, 1991)
C. Wood: Music and Drama in the ‘Tragédie en musique’, 1673–1715: Jean-Baptiste Lully and his Successors (New York, 1996)
C. Dill: Monstrous Opera: Rameau and the Tragic Tradition (Princeton, NJ, 1998)