In 2013 Arthaus Musik assembled three previously released ballets – Coppélia (2001), Paquita (2003) and Giselle (2006), each filmed in front of an audience at the Opéra National de Paris – and released them as a box set entitled ‘The Ballet Classics’. Such a title raises a question that historians and producers of ballet are beginning to examine: What exactly is a ‘ballet classic’? The term, steeped in an ideology of canonicity, seems to refer to a quality of performance that resembles the original production as closely as possible or at least is somehow consistent with past moments of balletic excellence. ‘Classic’ in this context contributes a sense of historical continuity to a varied, international, unevenly preserved tradition: ‘ballet classic’ as a term invites participants and onlookers to smooth the seams between the many historical antecedents and resonances associated with these works.Footnote ¹

The repertory developed at the Imperial Russian Ballet from about 1890 to the October Revolution has played a central role in the process of canon creation. Beth Genné has argued that Dame Ninette de Valois, the founder and artistic director of the company now known as the Royal Ballet, created the idea of the ‘ballet classic’ for England when she made a standard practice of reviving old works while also fostering the creation of a modern repertory. In doing so, she built on her experience performing with the Ballets Russes during the 1920s, employing the regisseur Nicolai Sergeev and the historical documents in his possession as a basis for the works she revived.Footnote ² The canon Genné discusses includes five works, three Russian (The Nutcracker, Swan Lake, Sleeping Beauty) and two French (Giselle and Coppélia). By the 1930s, when de Valois revived these works, few remnants of their original productions survived. One might think that documents related to historical premieres would eliminate uncertainty, but documentation does not always exist, and when it does it may be inaccurate or unclear, representing but one moment in a diverse performance history. Sergeev, the regisseur whom de Valois hired to reconstruct the dances, possessed documentation but lacked musicality, imagination and communication skills to realize the choreography, which made working with him a challenge. But constructing an audience was more important than staging old works authentically, and that may in the end be de Valois's more important contribution to ballet in the twentieth century.

These DVDs continue the process of audience development, as they provide a fuzzy Franco-Russian frame for understanding these works and present them with a French television audience in mind. Shortly after their nineteenth-century premieres, each production included here fell out of the repertory at the Paris Opéra, the theatre featured in these DVDs. To ‘reconstruct’ these productions, elements were pulled together across a discontinuous international performance history. These three productions in particular call attention to the stylistic tension between historical ballet traditions in Paris and St Petersburg. Of the three ballets in this box set, Paquita's performance history is the most fractured and incomplete. It was performed during most of the twentieth century in excerpted divertissements: the grand pas and the pas de trois from the second act.Footnote ³ But here it is made ‘classic’, despite its absence from the stage as a complete work for most of the twentieth century.

Since his reconstruction of La sylphide in the early 1970s, Pierre Lacotte's reputation as a choreographer has been built in part on his exhumation of spectacular yet forgotten story ballets. As performed on the DVD in this set, Paquita represents a work of new choreography informed by Lacotte's knowledge of Marius Petipa's style. In the interview with him found in the DVD's bonus material, he discusses how, when documentation is available, he examines it; but he depends more on his intuition as a dancer steeped in international choreographic traditions.Footnote ⁴ Such an approach is necessary for Paquita, because, despite the fame of the grand pas from the second act, few documents are known to survive that capture the full dimensions of the ballet – a situation that is true of most if not all nineteenth-century ballets.

Historical evidence supports Lacotte's most impressive feat in Paquita's production: the reconstruction of the long mime scene of the first act, second tableau.Footnote ⁵ The scene depicts Paquita overhearing a plot to murder Lucien d'Hervilly and then responding, with the support of Lucien, to subvert the plan. Lacotte based the reconstruction on documents he found in ‘a privately held collection in Germany’, according to his interview in the DVD's bonus material. The resulting performance effectively responds to what we presume to be Edouard Delvedez's score, with histrionic acting and gesture precisely timed to the music. This scene not only enhances the narrative flow of the story ballet, but also exposes audiences to some of the nineteenth-century French antecedents to more familiar Russian versions of these ballets. Indeed, the recovered mime makes the old abstract ballet complete, ‘classic’ – but also strange and new. Lacotte's work, then, contributes to a hybrid production of Paquita, with historical and contemporary material and techniques – presumably from both Paris and St. Petersburg – all blended to restore a ‘classic’ with new levels of completeness. (David Coleman is credited with reconstructing the score, presumably from sources available at the Bibliothèque national de l'Opéra in coordination with Lacotte's choreography.)

Coppélia presents different historical problems. In the original Parisian ballet, the male character of Frantz was played by a female dancer; but in the Russian version, a male dancer has traditionally assumed the role.Footnote ⁶ Eugénie Fiocre, the woman who played Frantz in 1870, was known for her mime, exemplifying a type of dancer later excised from the balletic tradition. Fiocre was a beautiful, shapely woman, who first made a sensation in 1864 as Cupid, a traditionally cross-dressed character, in the ballet-pantomime Neméa. After this success she played a series of mimed roles, including Fenella in the 1865 revival of La muette de Portici. Despite her meagre skills as a dancer, her popularity was such that in 1866 she was given the dancing role of Noureda in La source, but her undisciplined performance received mixed reviews, one author suggesting that she had almost learned to dance, and another disdaining the ‘cancan’ quality of her oriental pas.Footnote ⁷ Despite her low level of skill, she drew a large audience including wealthy admirers, a following the administration of the Opéra could not ignore. They further cultivated Fiocre's allure by placing her in mime roles in which her attractiveness, rather than her dancing skills, could be showcased: in the role of Frantz, for example, the cross-dressed dancer exaggerated the post-pubescent desires of the young man. Fiocre's sensuality was further emphasized in contrast to the highly disciplined Giuseppina Bozzacchi, who at seventeen created the virtuosic, Taglioni-like role of Swanilda.

The melody that accompanies Frantz's entrance raises issues for how his role as the male love interest in the ballet can be understood. One analysis is preserved in a British television program from 1970 called ‘How Dance Was Saved’, which describes the reintegration of the male dancer into ballet through the Russian school.Footnote ⁸ The narrator identifies the soft and unaccented melody accompanying Frantz's first entrance as ‘too weak for a man’, an interpretation further supported by the revelation that it originally accompanied a cross-dressed woman. Despite this resonance with femininity, the melody forms a tolerable, even appropriate and enjoyable, accompaniment to the male dancer as Frantz; when the Russian ‘correction’ of Coppélia's choreography was made, the tune remained unaltered. Its feminine quality, presumably, was not problematic.

One reason for the toleration of the tune may lie in its lack of character specificity, despite its clear connection in most cases with Frantz's entrances. In addition to suggesting Frantz's femininity, the tune and its transformations articulate the instability of his romantic relationship with Swanilda: it delineates Frantz's shift of affection between his fiancée Swanilda and the inanimate Coppélia, and musically celebrates the quixotic and romantic nature associated with cross-dressed characters on the Opéra's stage, be they cupids, pages or a young lover. Frantz's tune may also be interpreted as a musical ‘visiting card’, its function resembling that of the recurring melody associated with Verdi's Aida: both adapt to new tonal and dramatic contexts to associate with the topics of romance and even the gaze.Footnote ⁹

Of the three performances on these DVDs, Coppélia may best represent the mid-nineteenth-century French narrative impulse, with its integration of mime and character dance accompanied by Delibes's inventive and tuneful orchestral score, and performed by the highly skilled but noticeably youthful students of the Paris National Opera Ballet School of Dance.Footnote ¹⁰ The third act divertissement (the consecration of the bells) is omitted, further emphasizing the story ballet's narrative by removing a significant amount of abstract dance. Again, Pierre Lacotte is involved in the production, directing choreography attributed to Albert Aveline.

The version of Giselle presented here is the most dated of the three performances, in that in recent years documents have been uncovered that provide new insights into the ballet and its performance history; these resources were not available in 1991, when this production premiered. It lacks the freshness of the newly constructed Paquita and the youth-infused energy of Coppélia, and it feels old and tired for the interpolations of abstract Russian pas within the French frame of pantomime and character dance. New research on Giselle suggests the potential for alternative interpretations that would approach more closely the ballet's historical antecedents. Musicologist and ballet historian Marian Smith has published numerous descriptions of documents associated with Giselle that date from before its productions in St Petersburg. These documents provide additional insight into the mime that was central to the expression of the ballet-pantomime's narrative. These documents include a Parisian rehearsal score from the 1840s that is now housed in St Petersburg, and a document called the ‘Justamant manuscript’ at the Deutsches Tanzarchiv Köln.

Mime has been drastically reduced in twentieth-century productions of ballet-pantomimes, but so have character dances. Smith's monograph on the ballet, Ballet and Opera in the Age of Giselle, suggests ways that choreographic variety could be introduced to the second act by including different national styles in the dances of the Wilis. The score suggests the possibility of Wilis dancing steps stereotypical of the Bayadère and the Odalisque as well as French, German and Spanish styles.Footnote ¹¹ Emphasizing this variety would make for an altogether unfamiliar work for modern audiences, but it would also recover an aspect of Giselle associated with its earliest incarnation in Paris. The Pacific Northwest Ballet (PNB) in Seattle has recently taken advantage of these newly discovered manuscripts in a new production of Giselle, informed by Doug Fullington's knowledge of Stepanov notation and Marian Smith's understanding of mime and its relationship to the musical score. The result brings a new energy to the choreography. The precision with which the physical gestures of the dancers connect with the musical gesture of the orchestra returns precision to the narrative.Footnote ¹²

Adherence to tradition and classicism were arguably less of a concern in the Paris of the 1840s than the St Petersburg of the 1880s. Performances at the Paris Opéra, then on the Rue Le Pelletier, were more about innovation than tradition, more about visual stage effects and less about dancers trained to achieve stunning effects themselves. The artistic reputation of ballets suffered in part because of the theatre's dependence on spectacle. In addition, these productions were designed to appeal to the bourgeois man's eye. In 1841 for Giselle's premiere, and then again in 1846 for Paquita's, the wondrous ballerina Carlotta Grisi danced alongside less-skilled women whose presence onstage and backstage contributed to the Opéra's scandalous reputation, and ballet's degradation as a genre.Footnote ¹³ The internationally known virtuosic women dancers of this time rarely trained in Paris. Grisi, for example, developed her skills in Italy.

These three ballets experienced a rebirth of sorts in St Petersburg that allowed them separation from their Parisian origins, but also reintegration with a Parisian culture of dance at the present moment. Each ballet combines a mid-nineteenth-century French narrative realized through pantomime and character dances alongside expanded or newly introduced pas of Russian origin, performed by the most rigorously trained team of artists in twenty-first-century Paris. The recordings thus promulgate the balletic myth of a timeless art, yet also acknowledge that this myth evolved as part of artistic exchange between Paris and St Petersburg.