The Journal of Seventeenth-Century Music

Volume 10 (2004) No. 1

The Dances in Lully’s Persée ***

Ken Pierce* and Jennifer Thorp**

Abstract

French dance theorists from the 1680s and later differentiated between the “ordinary” (formal) and the “imitative” aspects of theatrical dance. Their descriptions, evidence from scores and librettos, and commentaries on dancing in various operas provide insight into the context and function of dances in Lully’s opera Persée (1682). Dances may express general rejoicing, extend or suspend stage action, or accompany magical stage effects. Dances interact with solo songs and choruses. Extant choreographies by Pécour for the 1703 and 1710 revivals presumably reflect aspects of dance in Lully’s day, albeit with flexibility concerning the number of dancers and their spatial configuration.

1. Introduction

2. Writings on Theatrical Dancing during the Late Seventeenth and Early Eighteenth Centuries

3. The Evidence of Dance in the Scores and Librettos of Persée, 1682–1710

4. Extant Dances from Persée

5. Conclusion

References

Video Examples

Tables

1. Introduction

1.1 Dance played a significant part in all of Lully’s tragédies en musique, and Persée includes many instances of dance and stage movement performed by dancers. Premiered at the Palais-Royal theater in Paris in April 1682, Persée was repeated at Versailles in July; additional performances took place in Paris over the course of the next century.¹ Very little information on the dancing in the 1682 production survives beyond what appears in the score and libretto published that year.² The score identifies certain instrumental pieces within each act as dance pieces, either by naming a specific dance type (such as “Gavotte”) or with a more general title (“Entrée” or “Air”).³ In either case the title might also name characters who dance (for example, “Entrée des Fantosmes”). The libretto describes the setting for each act and the main action on stage in each scene, but gives only vague reference to dancing—for example by the standard phrase, “They demonstrate their joy by singing and dancing” (“ils témoignent leur joye en chantant et en dansant”).

1.2 Neither the score nor the libretto for the 1682 production names any dancers, although we know from later sources referring to that production that the two leading soloists were Mademoiselle Lafontaine in Act IV and Monsieur Pécour in Act V, and that Mademoiselle Desmatins made her debut in that production as a dancer and singer.⁴ We do not know, except in very general terms, who the choreographers were for any of the tragédies en musique performed in Lully’s lifetime, what form the dances took, nor how many dancers were on stage at any given moment. Librettos from later productions, however, provide more complete cast lists of dancers, act by act, and there are a number of extant choreographies by Guillaume-Louis Pécour (who had danced in many of the operas in Lully’s day) for the Paris revivals of Lully’s operas after 1700.⁵ In the case of Persée the best documented revivals are those of 1703 and 1710, and though we cannot know how or whether the extant choreographies, or the numbers and groupings suggested by the cast lists, reflect what went on in the original production, they do suggest that much of the dancing followed the concepts described in dance treatises and writings on theatrical dance of the late seventeenth and early eighteenth centuries. Thus, there seems good reason to look at all these sources, both from Lully’s day and from the first quarter of the eighteenth century, in an effort to understand the place of dance in late seventeenth-century Lullian opera. In this article we discuss those sources under the general headings of writings on theatrical dancing; the scores and librettos of Persée, 1682–1710; and extant choreographies from the 1703 and 1710 revivals of Persée.

2. Writings on Theatrical Dancing during the Late Seventeenth and Early Eighteenth Centuries

2.1 In 1682 Claude-François Menestrier identified three sorts of movement in theatrical dance: steps, figures, and expressions (“les ports du corps,” “les figures,” “les expressions”). His definitions for these were reiterated by the dance theorist Jacques Bonnet in 1724.⁶ Thus, it appears that these categories were in use throughout the period of Paris productions discussed in this article. The first, “les ports du corps,” comprises the steps of Baroque dance and the harmonious movements of which they are composed: turning, beating the legs, springing, rising, and so on. These steps and movements are familiar to dance historians from notation systems and treatises of the period.⁷ Dancing masters at the turn of the eighteenth century apparently thought in terms of a relatively small, fixed vocabulary of discrete steps that could be combined in various ways to create a dance.⁸

2.2 Menestrier and Bonnet’s second category, “les figures,” refers to the spatial patterns of the dances. For example, the dancers can be facing forward; back to back; in a circle, semi-circle, square, or cross; moving in straight or curved lines; chasing or fleeing one another; or interweaving; in fact, there can be as many different dances as there are geometric figures. All of these spatial patterns are found in extant theatrical dances, most of which are notated for one or two dancers. With only occasional exceptions, the patterns are symmetrical. Regrettably, there are only a few extant examples of notated theatrical dances for more than two dancers;⁹ in these, too, we find a high degree of spatial symmetry.

2.3 The third category, “les expressions,” is defined as actions that indicate (“qui marquent”), examples being the actions or characteristic movements of blacksmiths, boatmen, drunks, wrestlers, and so on. There is scope for different related actions to happen simultaneously, as, for instance, in a battle dance where one group of dancers might deliver blows with swords or clubs while another group parries them with shields.¹⁰

2.4 The abbé Dubos, writing in 1719, employed slightly different categories, dividing stage dance into “ordinary” and “imitative.”¹¹ Although writing at a generation’s removal from the events, and perhaps influenced by his deep respect for Lully, Dubos clearly had a strong and detailed knowledge of French opera and dance. His comments concerning Lully’s attitude toward dance therefore merit our attention. Dubos claimed that “sixty years ago fauns, shepherds, peasants, Cyclops, and Tritons danced almost uniformly, but nowadays [i.e., by 1719] dance is divided into different characters, even the female dancers adopting characterization as the men do.”¹² This change he attributes to Lully, who individualized the dances to reflect the characterization of his music, sometimes choreographing the dances himself. For example, Dubos recounts that Lully composed the steps and figures for the Chaconne in Cadmus et Hermione (a dance in which an African plants a palm tree in the middle of the stage, the tree is decorated with garlands, and four giants also appear) because he felt that Pierre Beauchamp, his usual choreographer, was not at all able to enter into the character of the music.¹³ Dubos also notes that six months before his death, Lully choreographed a dance for Cyclops in Acis et Galatée.¹⁴

2.5 In several of his operas, Lully provides occasion for “imitative” dance. Dubos refers to these dances as “half-choruses”—choruses without words—and explains that he means ballets with almost no dance steps, made up instead of gestures and imitative movements (“démonstrations”). He cites as examples the dance of the Old Men in Thésée; the funeral processions in Psyché and Alceste, both of which call for grief expressed by means of gestures; the “Songes funestes” (“Baneful Dreams”) from Atys; and the shivering scene in Isis. Dubos notes that these stage movements were choreographed by d’Olivet rather than by des Brosses or Beauchamp, whom Lully employed for “ordinary ballets.” He confirms that these “imitative” dances were successful in their day, even though executed by dancers unaccustomed, and not always receptive, to Lully’s demands.¹⁵

2.6 Following Lully’s death in 1687, there was another change in the approach to choreography at the Opéra in Paris. Beauchamp left and Pécour took over as resident choreographer. The Recueil général des opéra of 1703 notes that Pécour shared responsibility with Letang for the choreography of Thétis et Pelée (by Pascal Collasse, 1689), but that thereafter it was Pécour alone who created (or at least took the credit for) all the dances that were performed, either at court or in Paris, whether in new operas or in revivals of old ones, and in fêtes, ballets, and so on.¹⁶ In other words, there was a change from shared responsibility for choreography to individual responsibility. No longer might the dances in a given opera be the work of two or more choreographers, possibly including the composer himself, working in very different styles.

2.7 In examining the extent to which these components of theatrical dance—steps, figures, or imitative actions—might be used to describe dances in Persée, we have adopted Dubos’s term “ordinary” dance to refer to a dance with conventional steps and figures, and his term “imitative” to refer to movement incorporating gesture, mime, or other actions intended to represent a given character. Dubos seems to suggest that these two categories are distinct and mutually exclusive; the fact that Lully sometimes employed different choreographers for these two types reinforces this view. But it may be that in some cases an “ordinary” dance could include imitative moments, or an imitative dance might include ordinary steps.¹⁷

3. The Evidence of Dance in the Scores and Librettos of Persée, 1682–1710

3.1 Primary sources for the discussion that follows include the 1682 and 1710 scores and the librettos published for the Paris productions of 1682, 1703, and 1710.¹⁸

3.2 Appearances or possible appearances by dancers in Persée are shown in Table 1. The libretto explicitly calls for dance in every act except the third, and the movement called for in that act—flying, rising up into the air, climbing, running—would very likely have been performed by dancers as well.¹⁹ Comparison with the score shows that several of these instances of dance are quite extensive, with music for more than a single dance, or with the same dance music played more than once. Lully sometimes employs a palindrome structure, in which two or more pieces are played in order and then all but the last are repeated in reverse order. (See, for example, no. 3 in Table 1, a palindromic sequence followed by a chorus.)

3.3 To present danced passages in their larger contexts, Table 1 shows some vocal music that occurs between dances, or that immediately precedes or follows them. This is not to imply that there would of necessity have been dance during vocal music. There is no evidence that Lully would have wanted dance during vocal solos or duets. On the other hand, there is evidence to suggest that dance during choruses would sometimes have been called for. Several of the choruses in Persée have instrumental interludes, sometimes quite extensive, that may have been intended for dance. In certain instances, notably in a chorus at the very end of a divertissement, it may even have been appropriate for dance and song to happen simultaneously.²⁰

3.4 We see from Table 1 that when the libretto calls explicitly for dancing, the characters who dance are either humans who dance for reasons related to the story—celebration, propitiation, preparation for sacrifice—or idealized or otherworldly beings for whom dance is a normal and expected means of expression. When the libretto merely suggests movement that was likely performed by dancers, such as the flying mentioned above, or the combat between the followers of Persée and those of Phinée, the characters involved may be either mortal (Table 1, no. 11) or otherworldly (no. 8).

3.5 Whether a dance is “ordinary” or “imitative” depends both on the purpose of the dance in the context of the opera, and on the characters who dance it. In cases where the characters dance for reasons related to the story, as in the divertissement of Act IV (no. 9), we assume “ordinary” dances. Likewise for the dances of the Prologue, in which “ordinary” movement would surely suffice to amplify the characters of Virtue and Fortune, and to delineate the contrast between them. The “Air pour les Sacrificateurs” (no. 10) in Act V may have been processional or in other ways “ordinary,” or it may have incorporated elements of mime. The “Entrée des Fantosmes” (no. 8) in Act III and the combat between the followers of Phinée and Persée (no. 11) in Act V are likely to have incorporated mime or otherwise to have been heavily “imitative” in style.

3.6 For the first set of dances, for the Followers of Virtue (no. 1), the libretto instructs, “Innocence, the Innocent Pleasures, and all the Followers of Virtue demonstrate their joy with dancing and singing.” No dancers for the Followers of Virtue are named in the 1682 libretto; the 1703 and 1710 librettos each list six female dancers. The score shows a dance, a passepied, alternating with the two stanzas of a vocal setting of the same music. There might have been dancing throughout all of this, but it is more likely that dance and song alternated rather than competing for the audience’s attention. In any case, each instrumental segment would probably have merited a different choreography.²¹ Variety in figures and step sequences is an important component of Baroque choreography, and though some choreographies do employ repeated step sequences,²² there is no evidence of an entire dance being repeated unchanged at a repetition or reprise of the music. Moreover, there is evidence for two different choreographies to the same piece of music used in the same production.²³

3.7 Soon after the dances for the Followers of Virtue, the libretto indicates that the scene is transformed (no. 2), and “there is heard the resounding noise of a large number of instruments. Fortune approaches, Abundance and Magnificence accompany her, with richly adorned followers. All rejoice and all dance around Fortune” (no. 2). The music is a binary-form march for the Followers of Fortune. It could perhaps have been used solely as entrance music; more likely, it would have provided the dancers the opportunity to show off a contrasting style of movement to that of the previous dances. It may have been in a more masculine style—that is, involving relatively more jumps, turns, and ornamental beats—than the passepieds: the cast lists from 1703 and 1710 show only women in the Followers of Virtue, only men in the Followers of Fortune.

3.8 In the 1682 and 1710 scores the directions merely suggest a group entrance (“Marche pour les Suivants de la Fortune”), and this is confirmed in the 1710 libretto, which lists six male dancers. The 1682 libretto additionally indicates that a group dance takes place at this point: “all dance around Fortune,” after which la Vertu sings (“Me cherchez vous”). The 1703 libretto, however, lists seven names, setting Blondy apart as a soloist, and this raises the question of when Blondy danced in that scene. Was the March performed in 1703 as an entrance dance for a group, and then repeated as a solo dance for him,²⁴ or did the March include passages for both soloist and group, or did no solo dancing occur until the divertissement that ends the Prologue? We cannot tell from the sparse evidence of the score and libretto.

3.9 The Prologue ends (no. 3) with Virtue and Fortune united, as they purportedly were in the august person of Louis XIV, to whom the Prologue was addressed; and as they would be at the end of the opera, in the marriage of Andromède and Persée.²⁵ The libretto notes that “the Followers of Virtue and Fortune unite, and demonstrate their joy by their dances and songs.” Like the group of pieces for the Followers of Virtue (no. 1), this is another extended passage of dance and song, forming a divertissement with a palindrome structure, all in triple meter: an air en rondeau for strings; then an air for oboes, also en rondeau; then a vocal setting of the second air, for la Fortune and la Vertu (“Quel heureux jour pour nous!”); then the air for oboes again; and then the first air. Framing the divertissement, but standing outside the palindrome, is a pair of large-scale choruses that may have involved some dancing. (The first concludes the preceding conversation; the second belongs to this divertissement.)

3.10 It seems likely that all the dances in the Prologue were “ordinary” dances—that is, dances with steps and symmetrical patterns. Their purpose is straightforward: to elaborate the entrances of specific characters and exhibit shared joy.

3.11 The dances in Act I (no. 4) are used to represent the “Jeux junoniens,” competitions offered by Cassiope, Queen of the Ethiopians, to entertain and appease Juno, whom Cassiope had angered by her vanity. According to Bonnet, there may have been historical precedent for dance competitions:

Dance was so esteemed in the early days, that Lucien says that it was not admitted in the Olympic games because the Greeks did not feel they had prizes worthy of it; but in later times the inhabitants of Colchis added it to their public games: which [practice] passed into use among the Greeks, and the Romans, and almost in all the towns in the world.²⁶

According to the libretto, at the end of scene 4 Andromède says, “The Games are about to start, let us place ourselves to see them.” In the score a title follows her remark: “Games [in honor] of Juno, in which young persons compete in dance.” The Premier Air (no. 4a) ensues, ending the scene. The next direction in the libretto, “The Games begin, with competition for the prize in dance” (no. 4b), occurs during scene 5, after Cassiope and the chorus have made their entrance (during a ritournelle). Cassiope has invoked the goddess, and all have moved into place for the chorus “Laissez calmer.” It seems likely that the Premier Air ending scene 4 was not danced, but was used as processional music for the entrance of the young competitors,²⁷ for it is in the same key as the rest of scene 4. Scene 5 opens in a different key that remains throughout the sequence of the ritournelle, the solo invocation for Cassiope, and a palindromic sequence of dance and song: the chorus (“Laissez calmer”); the Second Air (in 3/8, the same meter as the chorus); the Troisième Air, apparently a bourrée; the Second Air again; and finally a repetition of the chorus. The competitors are small groups of young men and women.²⁸ The 1703 libretto names four men and four women; the 1710 libretto names six of each. Given their formal, competitive purpose, it almost goes without saying that these Act I dances are also “ordinary” dances, with steps rather than imitative actions. There are extant notations for dances to the Second Air and Troisième Air, which we discuss at greater length below.

3.12 What is meant to be the nature of these games? From Cassiope’s introduction to the Games, we learn that the dancers represent couples about to be married; thus, it seems likely that they would dance as couples, and that any contest would be among couples rather than individuals. Cassiope sings, “Each will show his skill to celebrate the games I’ve had prepared” (“Chacun va montrer son adresse / Pour celébrer les jeux que j’ai fait preparer”); but she does not indicate that this will happen sequentially, leaving open the possibility of a group dance, in which the “winners” are the couple who acquit themselves the best.

3.13 The instances of dance in Act II, scenes 8-10 (nos. 5–7 in Table 1), involve Cyclops, Amazons (Nymphes guerrières), and Gods of the Underworld (Divinités infernales), who, according to the scores and librettos, arrive at the behest of Mercury with gifts for Persée to aid him in slaying Méduse. In each case, the musical structure is the same: instrumental dance, song, reprise of instrumental dance, followed immediately by the opening dance for the next group of gift-givers. The actual bestowal of each gift can be dispensed with quickly, presumably as the solo singer describes the gift and its purpose. In the case of the Amazons’ gift, in Act II, scene 9, the libretto is explicit: one Amazon sings while giving the shield to Persée, and the others dance (no. 6). So the dances did not necessarily have a strong imitative (gift-giving) component.

3.14 The scene descriptions in the librettos indicate that the gift-givers’ entrances are cumulative, each group remaining on stage for the next. Moreover, the presence of a chorus at the end of scene 10 makes it clear that each entrance of gift-givers includes a group of singers as well as dancers. This chorus, in which the gift-givers call upon the underworld, the earth, and the heavens to bestow favor upon Persée,²⁹ ends both the scene and the act, and seems a plausible occasion for Cyclops, Amazons, and Gods of the Underworld to unite in dance as well, especially given that the chorus includes an instrumental interlude of over thirty-one measures.

3.15 Four male Cyclops, four female Amazons, and four male Gods of the Underworld are listed in the 1710 libretto. The numbers and genders are the same in 1703, except that the Gods of the Underworld also include a male soloist, Monsieur Balon. Nothing is known of who, or how many, danced in 1682.

3.16 The Cyclops were Vulcan’s blacksmiths (forgerons), and are referred to as such in the 1703 and 1710 cast lists for Persée. Bonnet suggests the character of dance for blacksmiths:

The more natural the types of movement, the more agreeable they are. The dance of the Winds should be light and quick; that of the Blacksmiths should have a beat, and intervals for striking the anvil.³⁰

He elaborates later, with reference to Le Ballet des Amours déguisez:

Act I represented at first the grotto of Vulcan, from which emerged eight Cupids so well disguised as Blacksmiths, that one could recognize them only by their application in making darts and arrows rather than other weapons.³¹

Thus it appears likely that the Cyclops’ dances in Persée (gavottes), were at least in part “imitative” dances, although, not being in Vulcan’s cave at the time but emerging from the underworld, it is uncertain whether they struck anvils in time to the music.³² The stage directions in the libretto simply read “Some Cyclops arrive, dancing, to give Persée, on behalf of Vulcan, a sword and winged sandals similar to those of Mercury” (Table 1, no. 5). However, the reference to their arrival dancing (“en dansant”) perhaps implies that even the singing Cyclops (Monsieur Labé in 1703, Monsieur Dun in 1710) also participated in this dance.

3.17 We can conjecture that the dances for the Amazons (no. 6) were “ordinary” dances, with steps, but we have found no information that might suggest details. No element of mime is implied, for according to the libretto, the singing Amazon does the gift-giving: “One of the Amazons presents to Persée, on behalf of Pallas, a shield of diamond; she sings as she makes him this gift, and the other Amazons dance.”

3.18 In considering the dances for the Gods of the Underworld, we note that these followers of Pluto were powerful but not evil; like the Cyclops and the Amazons, they were there to aid Persée, and the libretto explains that “Gods of the Underworld emerge from hell, carrying Pluto’s helmet, which they present to Persée. One of the Gods sings and the others dance” (Table 1, no. 7).³³ Dubos offers a fascinating passage about dances for underworld gods in Alceste (1674), in which (according to the libretto) “the followers of Pluto rejoice at the arrival of Alceste in the underworld, with a sort of festival” that includes a rather animated (assez-animé) dance in common time and a loure-like 6/4 air. He writes of “airs caracterisez,” in which the vocal music and the rhythm imitate the sort of music that is generally regarded as appropriate for certain types of people, even imaginary personages of antiquity.³⁴ This might suggest that the Persée dances for the Gods of the Underworld would have been “imitative” dances—imitative, that is, of commonly held ideas of how underworld gods move. But from the extant notation for this dance it appears that underworld gods were imagined, at least in 1710, to move very much like virtuoso French male dancers performing an “ordinary” dance, albeit with a careful choice of steps— sudden drops, turns, brushes, and so on—to evoke their subterranean and immortal origins.

3.19 The next likely appearance by dancers in Persée, the “Entrée des Phantosmes” in Act III, scene 4 (Table 1, no. 8), reminds us that dancers were often used in scenes requiring special effects, including transformation scenes and the use of machines and flying harness. Examples are documented from early seventeenth-century Italy,³⁵ and Parisian audiences in Lully’s day also expected to see impressive special effects, and rated the success of a staged work by them. Thus, among the cast for Alceste we find listed “Followers of Pluto, singing, dancing, and flying.”³⁶ In the Persée librettos, which called for four male dancers in 1703 (as the Followers of Méduse) and six in 1710, we find that “Chrysaor, Pégase, and several other monsters of bizarre and awe-inspiring appearance are formed from the blood of Méduse. Chrysaor and Pégase fly, some of the other monsters also raise themselves into the air, others climb, others run, and all seek Persée, who is hidden from their view because he is wearing Pluto’s helmet.” The degree of activity suggested by this description is corroborated by the music for this imitative “Entrée des Phantosmes,” which is in cut time and overflowing with sixteenth notes. Presumably this was intended to be a show-stopper.³⁷

3.20 The dances in Act IV, scene 7, for Ethiopians and Sailors (no. 9), represent the celebrations following Persée’s rescue of Andromède from the sea monster. The 1682 libretto indicates that “the Ethiopians climb down from the cliffs and show their joy in singing and dancing. Male and female sailors mingle in the public celebration. One of the Ethiopians sings in the midst of the sailors who dance.… Andromède, Cassiope and the Ethiopians repeat the verses that Céphée has sung, and the male and female sailors dance for joy at the rescue of Andromède.”³⁸ Once again we have an extended divertissement of song and dance, with a form described by Lois Rosow as “a complex and loosely symmetrical structure that is typical of the genre.”³⁹ A chorus, with interludes that may have been danced, is followed by a gigue; a danced minuet (repeated as a vocal solo); a second minuet, this one en rondeau, with instrumental, solo vocal, and choral segments alternating for two stanzas; a repetition of the gigue; and a sung air (repeated as a chorus, again with interludes that could have been danced), all in celebration of Andromède’s rescue.⁴⁰

3.21 The librettos suggest that the series of dances may have provided a showpiece for a female soloist, although it is not always certain what role she danced and whether a solo was included in all the early productions. For Act IV in the 1682 production, Parfaict names only one dancer (who thus presumably danced a solo at some point): Mademoiselle La Fontaine, as “une Ethiopienne.” ⁴¹ However, the libretto for the same production refers only to male Ethiopians, male and female sailors, and gives no hint of any solo dancing. The 1703 libretto makes no clear or implied reference to a soloist, but simply notes the dancers in Act IV as four male Ethiopians, four male and four female sailors (led by Mademoiselle Subligny, who is a plausible candidate for solo as well as group dances). The 1710 libretto names Mademoiselle Guyot as a female sailor, along with four male and four female sailors, which strongly suggests that she danced solo at some point in that scene. Whether for a soloist or groups, the dances in this divertissement were most likely “ordinary” dances, any solo passages perhaps including steps designed to display the dancer’s virtuosity.

3.22 In Act V, the first likely moment for dance is in scene 3, the “Air pour les Sacrificateurs” (no. 10). The 1703 and 1710 librettos each list four male dancers as Sacrificers, so it seems that dance of some sort took place. To what extent this was “ordinary” or “imitative” is much less clear.⁴² It may be worth noticing that in this scene the structure of the Act II gift-giving scenes is inverted, with a chorus and its reprise surrounding the dance, rather than the other way around.

3.23 “Imitative” dance—a combat of some sort—certainly seems called for in Act V, scene 5 and subsequent scenes, for which the librettos state that “Persée and Céphée and their followers pursue Phinée and his followers” (no. 11). The 1703 libretto names six male dancers as Phinée’s combatants; the 1710 libretto names five male dancers as Phinée’s combatants and six as Céphée’s. The music for this pursuit and combat, which ends when Persée reveals Méduse’s head and turns Phinée and his followers to stone, involves singers commenting on the action in a general way (“Concede … What horrors! … Let them not escape”), in alternation with instrumental passages that grow progressively shorter. It is tempting to imagine that choreographed pursuit and combat occurred during these instrumental interludes, and perhaps throughout the entire scene, but the scores give no instructions. One possible choreographic approach, not operatic in origin, is suggested by Lambranzi’s dance for battling statues: the statues jump from their pedestals and begin to grapple, and then hold their position for three measures before moving into another combative pose. They continue in this way through another four poses, at which point the dance ends.⁴³ A very different choreographic approach, from a later source, is found in the article on choreography in Diderot and d’Alembert’s Encyclopédie. The article discusses the opening five measures of the “Dance for two wrestlers, as danced by Messieurs Dupré and Javilliers in the opera Les Fêtes grecques et romaines.” The notation for the first ten measures of the dance accompanies the article, and the dance has many elements of an “ordinary” dance: simple steps, such as pas de bourrée or pas grave; complicated steps, including caprioles, assemblés battus en tournant; and symmetrical spatial patterns, beginning and ending in mirror symmetry. The notation shows the dancers taking hands at one point, and while this may seem unusual in a dance for two men, it is only from the accompanying article that we learn that the dancers should be “pretending to make the effort that two wrestlers make to overturn their adversary.”⁴⁴ Thus, though it appears from comments by Dubos and Bonnet that “imitative” dance was generally “almost without steps,” we must consider the possibility that this was not always the case, and that the notated steps of an “imitative” dance could sometimes be indistinguishable from those of an “ordinary” dance.

3.24 The Passacaille in Act V, scene 8 (Table 1, no. 12) occurs after the descent of an elaborate gloire depicting Vénus’s palace. (A “gloire” was a cloud machine suspended over the stage, representing heaven populated by divinities.)⁴⁵ In the score the sequence is as follows: Persée’s announcement of the arrival of Vénus and her entourage (“Cessons de redouter”), followed by the Passacaille, followed immediately by Vénus singing “Mortels, vivez en paix.” The libretto, however, ends scene 7 with “Cessons de redouter” followed by the descent of Vénus’s palace; it opens scene 8 with Vénus’s “Mortels, vivez en paix.” This raises an intriguing question concerning the exact place and purpose of the Passacaille. Unless the line “Le Palais de Vénus descend” in the libretto is printed in the wrong place (which is possible), the implication is that the Passacaille provides music for the descent of the gloire. But the evidence of Pécour’s notated choreography indicates that, at least in the 1703 production, the Passacaille was danced; and given the general nature of passacailles as dances, it seems likely that it was also danced in 1682. Moreover, there is more appropriate music than the Passacaille for the descent of Vénus’s palace: immediately after Persée petrifies Phinée and his followers comes a single strain (twenty-three measures) of instrumental music, which Persée then takes up in vocal form as “Cessons de redouter.”

3.25 The Passacaille from Persée is one of Lully’s earliest.⁴⁶ Geoffrey Burgess has argued, from the available evidence in scores and librettos, that chaconnes and passacailles on the stage “usually formed part of the celebrations of a company of peuples or courtisans and were danced by soloists and a troupe of dancers.”⁴⁷ It seems plausible that the Passacaille in Persée was danced by a group with passages for one or more soloists. There were certainly plenty of people on stage at the time: Vénus, l’Amour, l’Hymenée, les Graces, les Amours et les Jeux, all presumably in the cloud machine; Céphée, Cassiope, Persée, Andromède, awaiting their apotheosis; and troupes of Céphée’s courtiers and Ethiopians.

3.26 Regrettably, the librettos do not name the dancers of the Passacaille or the duple-meter Air later in the scene, and we cannot be certain how many dancers figured in each. For Act V in the 1682 production, the Parfaict brothers name only Pécour, as a courtier in the Followers of Céphée, although Mademoiselle Lafontaine would presumably have been on stage also, as an Ethiopian. For Act V in the 1703 production, neither the Parfaict brothers nor the libretto name any dancers beyond Sacrificers and Phinée’s warriors, although at least some of the male Ethiopians from Act IV presumably would have been on stage again.⁴⁸ However, the extant notation for the Passacaille, discussed below, indicates that Monsieur Balon and Mademoiselle Subligny (who were featured soloists elsewhere in this production) danced together in the Passacaille. The 1710 libretto lists fourteen dancers in the “Suite de la Peuple de Céphée,” with Monsieur Blondy and Mademoiselle Prévost indicated as soloists.⁴⁹ Possibly they too danced in the Passacaille as a couple.

3.27 The final instrumental dance, the duple-meter Air of Act V, scene 8 (Table 1, no. 13), follows Vénus’s solo “Mortels, vivez en paix” and the immortalization of Céphée, Cassiope, Persée and Andromède as heavenly constellations. The dance has the dotted rhythms and other musical characteristics of the first section of a French overture. In extant dance notations, this type of dance music is associated only with male dancers, whether in an “entrée” or an “entrée grave.”⁵⁰ Yet according to the Persée libretto, men and women dance at this point: “Céphée’s courtiers, and male and female Ethiopians, demonstrate their joy by their dances.” Perhaps in this case the women joined with the men in dancing the Air. Or perhaps the men danced the Air and the women joined in the dancing during the final choral reprise, bringing the opera to a close with both dance and song contributing to the general celebration.

4. Extant Dances from Persée

4.1 Although no choreographies from Lully’s day survive, we are fortunate that four of the dances created for revivals of Persée do still exist, in notations that were published within a few years of their likely performance. They are all the creation of Guillaume-Louis Pécour, who had danced in the 1682 premiere of Persée, and who, after working closely with Beauchamp for many years, succeeded him as compositeur des ballets at the Opéra.⁵¹ We do not know, however, the extent to which Pécour’s choreography corresponded to Lully’s ideas about the dancing in Persée.⁵² The notations of three of the dances—two from the Act I “Jeux junoniens” (Video 1) and the other the Act V Passacaille (Video 2)—indicate that they were danced in performances of the opera, very probably in the 1703 revival, by Monsieur Balon and Mademoiselle Subligny, the two leading theatrical dancers of their day.⁵³ A fourth dance by Pécour survives, probably from the revival of November 1710 (which was repeated in February 1711): one of the dances for two male Gods of the Underworld in Act II, scene 10, as danced by Messieurs Marcel and Gaudrau (Video 3 ). It was notated and published for Pécour by Michel Gaudrau around 1713,⁵⁴ and thus may be a rare first-hand record of a theatrical dance preserved by someone who had actually performed it.

4.2 An important but unresolved question concerns the extent to which these surviving notations of dances by Pécour represent what really happened on stage in 1703 and 1710. Rebecca Harris-Warrick has noted that while in each case the notation indicates that duos were danced, the scores and librettos indicate the presence on stage of a group of dancers.⁵⁵ There is no reason to doubt that the notations really do record steps performed by the two dancers named in each instance; nor is there any reason to doubt that the scores and librettos really do mean that whole troupes of dancers were on stage, and presumably dancing, during these scenes. So how might each notated duo have related to the larger group of dancers?

4.3 One possibility is that it did not—that the duo performed the dance as notated and that the larger group of dancers, along with the chorus, simply served as spectators and danced only elsewhere in the scene. That explanation is plausible for some of the scenes containing Pécour’s duos, but as Harris-Warrick points out, it is particularly unconvincing for the two entrées making up the Act I “Jeux junoniens”: a dance contest among affianced couples must surely have more than one couple as contestants.

4.4 Another possibility is that the whole group, or sections of the group in turn, danced to repetitions of the music that are not notated in the score. This might be a feasible option for the Act I “Jeux junoniens,” but it would make that scene so interminably long as to incur the wrath of the audience as well as Juno.

4.5 A third possibility is that the duo was cloned, so to speak, and the same steps performed simultaneously and in mirror image by other couples, to form a group dance. This is feasible for the eight named dancers in the 1703 version of the “Jeux junoniens,” and it could work equally well for the four dancing Gods of the Underworld named in the 1710 production. Yet another possibility is that the group danced at the same time as the duo, but behind them or in a semi-circular or square formation around them, sometimes static, sometimes dancing simpler steps or different patterns.⁵⁶

4.6 The discrepancy between the dance notations and the cast lists recorded in the librettos becomes acute when the Act V Passacaille from Persée is considered. The librettos of several French operas list the names of quite large groups of male and female dancers for scenes in which a passacaille was danced, so we must imagine that there were larger forces performing passacailles in some productions of Persée.⁵⁷ The 1710 libretto, for example, includes the star dancers Mesdemoiselles Chaillou and Prévost and les Sieurs Blondy and Marcel among its fourteen “peuples de la suite de Céphée et Persée” for the passacaille scene.⁵⁸

4.7 In addition to Pécour’s passacaille duo for Persée, only two other passacailles from French operas survive as notated choreographies. Both are notated as solos (and both, as it happens, represent dances performed by Mademoiselle Subligny). There is nothing in the notation to suggest that these dances were for larger numbers of dancers—no evidence of abridgement or rearrangement, no obvious spatial impossibilities. But we must allow the possibility that these dances, as well as other notated theatrical dances, represent group dances that were reduced to solos or duos, whether for legibility or for enhanced marketability.⁵⁹

4.8 In the case of Pécour’s version of the Passacaille from Persée as published in 1704, we are left with more questions than answers. As with many of the dances in Feuillet’s 1704 collection, Balon and Subligny are named as dancers. Perhaps their star status helped sell notations. The Passacaille (Video 2) undoubtedly functions choreographically as a duet, and gives every appearance of having been built that way; for example, the dancers move in axial symmetry—that is, around one another—for roughly half the dance, and only occasionally do they direct their attention toward the public. Contrast this with the much more presentational approach of the Act I “Jeux” dances (Video 1). Would this sense of privacy in the Passacaille, this lack of focus toward the public, work as well in a group dance? Perhaps. But if the passacaille notation does represent a duo that was danced in Persée by Balon and Subligny in 1703, then were the other dancers on stage for that scene simply to dance the final air that ends the opera? What about the evidence, suggested by the librettos of other works, that passacailles were usually performed by groups of dancers?

4.9 One further option presents itself, and we offer it here not as a definitive solution but as one possibility among several. Perhaps Balon and Subligny did dance the Passacaille as notated, but with others on stage also dancing during parts of it. The other dancers might have performed the same steps at the same time as the leading couple, or they might have joined in with similar or simpler steps at appropriate moments in the music, or at moments in the choreography in which the duo opens its focus toward the public, facing forward and moving in mirror symmetry. Table 2 shows a hypothetical distribution between duo and group dance in the Passacaille, with group dancers moving mainly during tutti sections in the music, during which mirror-symmetrical sections of the choreography always occur.⁶⁰

4.10 As a dance form on the Lullian stage, the passacaille was often associated with female characters, including Venus.⁶¹ Thus, perhaps at this particular point in Persée, the dance represented the influence of Venus not just on the harmony between man and woman, but also on the relationships between Fortune and Virtue, deities and humans, all of which are themes of this opera.⁶²

5. Conclusion

5.1 Theatrical dance in Lully’s day and for a generation thereafter seems to have followed the broad categories set out by writers from Menestrier to Bonnet as regards steps and figures (“ordinary” dances) and characterization (“imitative” dances). Persée apparently includes instances of both “ordinary” and “imitative” dances, some of them part of lengthy scenes involving vocal soloists and chorus as well as dancers. For the various productions from 1682 onwards, the librettos and the later record published by Parfaict identify some of the dancers concerned, but not consistently. The 1682 score indicates meter and musical form of dances, but gives no information about the number or arrangement of dancers; and the later scores give little if any new information about the later productions. Extant notations of four of the dances, created by Pécour for the 1703 and 1710 revivals, add more information about the nature and form of those dances, but draw attention to tantalizing ambiguities concerning the number of dancers involved and even the purpose of the notations themselves.

5.2 Although choreographic practice had changed since Lully’s death—in his day much of the work had been divided among several choreographers—Pécour’s contribution in and after 1703 is still of great significance for understanding dance in Lully’s lifetime. Pécour himself had danced in the 1682 premiere of Persée, had worked closely with the theatrical dancers and dancing-masters of that day, and had succeeded one of them as compositeur des ballets at the Opéra. Many of his choreographies, whether for ballroom or for theater, were set to music by Lully.⁶³

5.3 We have little way today of knowing what audience preferences were in Lully’s day for the balance of solo, duo, and group dances in opera, or the precise form of the dances assigned to whole troupes of dancers. However, a study of the theoretical writings available at the time, and their application to the extant dance notations of a choreographer who had been so closely linked with the productions of Lully’s own day, can suggest ways in which steps and figures might be combined, with appropriate variety of figures and choice of steps, to create dances that complement Lully’s music. A careful study of all the available sources can thus augment our appreciation of the possibilities for dance in Lully’s operas.

---