Backward glances: scene and narrative

The stage was festooned with long dresses. Some were elaborate, bell-shaped creations, worn by characters with fancy coiffures and fans; others were made from muslin or gingham, accessorized with pinafores and bonnets; still more were in the making, half-stitched cuttings of satin, silk, gauze and tarlatan lying on a table, pored over by eager assemblers.

This was the setting of Thérèse, Act One: a fashion designer’s studio, with seamstresses attending to their clients’ demands, offering garments for try-on sessions and rhapsodizing on the tried-on. The dresses of course are a clue to the studio scene. The animate ones, filled with performing bodies, help differentiate between character and class, customer and employee; the inanimate suggest industry or work, as well as the article of trade. But both types, animate and inanimate, betray something beyond a local dramatic scene; for these were not dresses of a style or taste to suit 1910s Paris. Instead, they were vestiges from a bygone age, material residue of a former fashion that, along with other scenic elements, signalled a time other than the present day. In the words of critic F. Mobisson:

Louis-Philippe had reigned some seventy years earlier, in a period characterized by ostensible liberalism and reform, along with a marked historical consciousness: the king famously converted the château at Versailles into a museum dedicated ‘à toutes les gloires de la France’.²² Indeed, there is something of the museum about Thérèse’s Act One scene. Besides the quasi-authentic props (vases, dressing tables, curtains) and costumes, the stage was peppered with real-life remains. At Mendès’s request, portraits of some of the most famous French ladies of the day hung on the studio walls, reflected by mirrors and lit with a vibrant gold light. The writer Émile de Girardin, the duchesse d’Abrantès, the dancer Carlotta Grisi, the actresses Virginie Déjazet, Marie Dorval and ‘Rachel’ (Élisabeth Rachel Félix): these icons of the July Monarchy came to life in full material finery. They were all-consuming, inescapable; the studio was a sort of iridescent shrine.

So much was noted by the press. Journalistic nods to the scene – ‘exquis’, ‘charmant’, ‘ingénieux’ and, most important, ‘très français’ – were rife. Thérèse, on a visual surface at least, recalled French history, perhaps even in a manner similar to Louis-Philippe’s recycled chateau. Indeed, just as the ballet’s visuals appealed to a national heritage, its plot evoked a specifically balletic one.

Here is a short synopsis:

What is traditional about this? According to Opéra director André Messager, not a lot: ‘Ce qui nous séduisit, c’est que La Fête chez Thérèse est un ballet d’un genre tout nouveau à l’Opéra, c’est une chose intime, évoquant la vie et les costumes de 1840’ (What attracted us is that La Fête chez Thérèse represents a new type of ballet for the Opéra; it has something intimate about it, evoking the life and costumes of 1840).²³ ‘Intimate’, certainly: Thérèse tells of a personal affair of the heart, a domestic intrigue set against a picturesque backdrop of (in Act One) an interior workplace and (in Act Two) a private party. But ‘a brand new genre’? Thérèse betrays little of the Russian-inspired exoticism or abstract poeticism that infiltrates contemporary works; and its local, anecdotal history seems out of place amongst the mythological settings and themes favoured by the Opéra. (Eight of the eleven ballets commissioned by the Opéra between 1900 and 1914 were based on legend.²⁴) Yet the ballet was not wholly anomalous. The historical setting, historically authentic backdrop, opulent staging, elaborate plot, domestic intrigue and plot formulae (suspense, social contrast, cause-and-effect incidents): all these features invite comparison with the ballet-pantomime – the self-standing dramatic genre, combining dance and mime, that featured heavily at the Opéra during the July Monarchy, the period depicted in Thérèse.

Much is known of this double-barrelled species, thanks in large part to the work of Marian Smith.²⁵ A few parallels with Thérèse are particularly worthy of note. Perhaps the most vivid concerns the centrality of a love story and certain amorous plot situations: a man’s love for and pursuit of a woman of higher social standing (in Thérèse, Théodore’s desire for the duchess); a love triangle or scene of romantic crisis at the end of the first act (Mimi’s desperation over her beloved’s newfound love interest); a woman’s attraction to a man, disguised, whose name and identity she does not know (Thérèse’s fondness for the masked Théodore).²⁶ This last situation leads to an additional parallel, in that Thérèse also follows the standard ballet-pantomime in its staging of a masked ball. (The popularity of ball scenes in ballet-pantomimes of the July Monarchy is thought to result from their scenographic potential and narrative verisimilitude: the scenes provided a rationale for stage dancing and thus alleviated librettists’ concern for dramatic realism.²⁷) There is also the matter of dramatic mood. The ballet-pantomime, Smith argues, was light-hearted and lightweight entertainment; ballet itself was considered inappropriate as a medium for conveying political and religious content, and more suited to romantic themes, happy endings and ‘soothing’, pastoral backdrops.²⁸ Thérèse offers a similar ‘calm’: the ballet reconciles the on–off seamstress/suitor relationship amongst the celebratory festivities of a ball, denying any sombre connotations or sense of struggle.

What are we to make of these similarities? What do they reveal of Thérèse, its creative methods, motivations and general historical aura?²⁹ My next section, on the ballet’s structure and stylistic shifts, will cast light on these questions with specific reference to Hahn’s score.

Backbone: the score

On a most basic level, a musical score can reveal a ballet’s structural divisions, its succession of ‘numbers’ or seamless, through-composed form. But a score can also indicate movement – by means not only of the mimetic and cartoon-like synchronization of contour, punctuation and rhythm (known as ‘mickey-mousing’), but of a more general coincidence of musical style and movement style. In Thérèse, two musical styles prevail: one is characterized by periodic phrasing, rhythmically distinct themes (often repeated, occasionally in different keys), solo obbligato textures, on-the-beat and metrically regular bass-lines, punctuated chordal endings or endings softer in dynamic and articulation, suggesting the winding-down of stage movement and the dispersal of dancing groups (see Example 2.1³⁰); the other is through-composed, changing continually in phrasing, metre, theme and key (see Example 2.2, of which more later). The former corresponds to scenes of dance – often set dances, extraneous to the plot and entitled ‘danse’ or some more precise descriptor (‘valse’, ‘menuet’). Here, metrical and phraseological stability tends to anchor dance movements to a codified aural pattern; easy-on-the-ear melodic material encourages a focus on the moving body. The latter style signifies scenes of mime: annotations include stage directions and, occasionally, words to be mimed by the dancers. This style follows the unfolding of drama on stage, as well as the shifting nuances of character and emotion.

Example 2.1 Hahn, La Fête chez Thérèse, Act One, scene 2, ‘La Contredanse des grisettes’.

Example 2.2 Hahn, La Fête chez Thérèse, Act One, scene 5, ‘Pantomime’.

An oscillation of these two styles in a ballet score of 1910 was not exceptional. Although the Ballets Russes, in their first season of 1909, had tended to dance to musique pure (music not written for dance), ballet at the Opéra was almost always accompanied by music of bipartite categorization: discrete and formulaic ‘numbers’ contrasted with passages of wandering affect. Yet what stands out in Thérèse is the concentration of each style – hence the concentration of dance and mime, even the concentration of ‘story’. In brief, of the Opéra’s ballets premiered between 1909 and 1914, Thérèse had the greatest proportion of through-composed music – that is, music to accompany mime. Just over half (52 per cent, by my reckoning) of Hahn’s score was devoted to music for miming, compared to roughly two-fifths of Lucien Lambert’s La Roussalka, one-fifth of Alfred Bruneau’s Les Bacchantes, and none of the orchestrated Chopin that accompanied Suite de danses.³¹ Moreover, in their own ways, La Roussalka, Les Bacchantes and Suite de danses all promoted dance: Suite de danses, as the title suggests, comprised a succession of dances with no intervening mime; La Roussalka featured characters, the Roussalki water spirits, whose supernatural movements resembled ‘classical’ gestures; and Les Bacchantes offered an act-length sequence of diegetic dances performed as part of Bacchus’s spiritual rite. This last ballet, in fact, appears less a mimed musical drama with dancing than a danced work with a modicum of miming. Of the twenty-one ‘numbers’, only four are labelled ‘pantomime’: ‘story’ is subordinated to the dancing spectacle, is almost entirely submerged.

One might wonder whether this narrative evanescence relates to changing dance practices elsewhere (the evolution of ‘modern’ dance, the famously non-narrative Les Sylphides) or to the now infamous literary conceptions of dance as a non-figurative ‘Idée’, an abstract sign without signified.³² But this would be to stray from the topic at hand. Thérèse, clearly, was devoted to storytelling: it offered roughly equal amounts of ‘ballet’ (set dance) and pantomime, thus devoting roughly equal time to virtuosic display and to action. Yet if this structural composition marked a point of departure from contemporary practice, it harked back to the 1840s genre introduced earlier. Thérèse, it appears, is a literal ballet-pantomime, a formal equivalent of the genre typified by its titular split of dance and mime.³³

Back-chat: music and drama

Another angle from which to consider Thérèse’s score and its historical significance involves the coincidence of music and dramatic expression. Exploring the role of ballet music, needless to say, is a complicated business. There is little context for discussion, little archival material and little sense of how music closely allied with dance might sound, might mean. Nonetheless, the inevitable complications may be less severe with Thérèse: even as the ballet escapes gestural reconstruction, its scenes of pantomime, complete with stage directions and dialogue to be mimed, hint at the relation between music and the drama unfolding on stage.

One such scene has been mentioned already: the episode from the end of Act One in which Théodore, mesmerized by the sight of the duchess in déshabillé, snubs Mimi and her amorous declarations (Example 2.2). In this episode, two distinct musical figurations are obvious, at least in the opening thirty-seven bars. The first comprises off-beat semiquavers in duple time, oscillating thirds (at the start of each entry) and simple, top-heavy textures. The second, in a slower tempo and compound time, is textually thicker and more lyrical, and consists mainly of conjunct (and often chromatic) melodic motion. These two styles alternate throughout the scene: moreover, they do so in line with text and character. According to annotations in the score, the frenzied semiquavers mark an exclamation, question or action by Mimi; the more expressive music accompanies Théodore.

There is, then, a clear sense of musical dialogue, of characters linked to musical styles – styles that are themselves dramatically suggestive. Mimi’s off-beat snatches depict not only the agitation of her outbursts, but the aural impression of outbursting: the on-the-beat rests break the musical flow, creating short bursts of instrumental flurry; the undulating contours, widening intervals and occasional syncopation approach the inflections of speech, giving the music a ‘talking’ quality. (In Mimi’s first entry, the number of musical snatches equals the number of phrase she is to ‘speak’/mime: ‘verbal’/gestural and musical phrases line up.) As for Théodore, whilst his music lacks that ventriloquistic tendency (he is largely silent, after all), it helps communicate his mood. Not only is the soupy lyricism appropriate to his wistful, dream-like state, but it tends towards the two identifying themes of Thérèse, the subject of his reveries. The first of these themes – which both recur throughout the ballet – appears in bar 12 in an inner voice: this is a lilting melody, rising in contour and seemingly heading towards A major, until its interruption five bars later by an implied diminished seventh (on D♯) over a bass E. This last chord, which seems to highlight Mimi’s disturbance of Théodore’s reverie, is left unresolved as the tonality shifts to C♯ minor for Mimi’s next ‘speech’. The second identifying theme, beginning in bar 22, is an upper-voice modified-triplet figure, the final chord of which (a dominant ninth on A♭) also fails to resolve: Mimi’s fast notes recur (for the third time) in the next bar, again wrenching the tonality towards an unrelated minor key.

Théodore’s music thus conjures not only his dramatic mood, but the focus of his thoughts; Mimi’s music reflects her disposition as well as her mode of address. This loading of the score with narrative content – with metaphorical and real-time ‘voices’, aural reminiscences of character and other pictographic traces – continues throughout the scene: Mimi’s gesture to Thérèse’s portrait in an attempt to convince Théodore of the duchess’s noble status is accompanied by one of Thérèse’s identifying themes in a heavily articulated, chordal and fanfare-like guise (bar 46); Mimi’s collapse, following her pursuit of a departing Théodore, is synchronized to a low timpani roll breaking out of a rising chromatic line over four octaves higher (bar 74); and Mimi’s solitary crying at the end of the scene is attended by a sparse orchestral texture and a staccato melody played pianissimo by solo oboe and piano (bar 78). This last moment was singled out as one of the ballet’s highlights by critic Édouard Risler:

Risler’s words are worth noting; clearly, the critic was impressed by the music’s seeming poeticism, its ability to bring silent, moving images to life. He noted as much earlier in his review:

Others spoke similarly of the music’s dramatic resonance. Robert de Montesquiou went so far as to describe the score as an active force, coaxing the characters into movement:

André Rigaud summed up:

These comments endorse the impression of carefully wrought musical-dramatic effects; but they also strike a peculiar note. In the context of early twentieth-century ballet criticism, commendations of music’s dramatic effectiveness were scarce. For the most part, reviewers described French ballet music in unexceptional terms, often lamenting a perceived superficiality or surface gloss. According to Pierre Lalo, for example, Henri Maréchal’s score for Le Lac des aulnes followed the action precisely, but without energy, vivacity or life:

Five years later, critic Julien Torchet invoked a similar mediocrity in Alfred Bruneau’s score for Les Bacchantes, noting that the music would add nothing to Bruneau’s reputation.³⁹ Another critic even suggested a reason for Bruneau’s troubles:

Here, of course, we are some distance from the lauded pictoriality of Thérèse. Shared amongst the above quotations is a sense of ballet music’s arbitrariness, a sense that runs contrary to earlier comments about Thérèse’s musical poetry and animation. The ballet’s incongruity in a contemporary context comes to the fore once again, inviting us to compare Hahn’s score with a couple of others.

Comparing scores

Let us begin with Maréchal’s Le Lac des aulnes and a scene, chosen almost at random, in which a magician’s daughter, Lulla, attempts to introduce her father to a sprite, Elfen, visible only to Lulla herself (see Examples 2.3a and 2.3b). A first point to note is the repetition of mime music at separate dramatic moments. Consider, for example, the G major motif of the ‘Largo assai’ (Example 2.3a, bar 2), used to accompany Lulla’s examination of Elfen. This motif recurs (albeit altered melodically) a few moments later (bar 8) when, following her father’s disavowal of the sprite’s existence, Lulla begins to play with Elfen; it is heard again, doubled at the octave, in the ‘Largo maestoso’ (Example 2.3b, bar 1). Such a musical recurrence may seem of only minor importance, perhaps reflecting an imperative of musical economy rather than any musical-dramatic coherence. Nonetheless, it may also alert us to a potential non-pictoriality about the music (what does it mean that the same motif accompanies actions of bodily inspection, play and dance?), and, more important, to a kinship between music for mime and music for dance. This is a second point to note: that mime music tends to borrow from dance, and vice versa. The motif of the ‘Largo assai’ is exemplary. Even on its first ‘mimetic’ appearance, this music has something dance-like about it, owing to those undulating flourishes, a general melodic expressivity and, not least, a lilting, waltz-like feel. (The time signature, taken slowly, is likely to suggest a triple metre.) When the motif recurs at the ‘Largo maestoso’ it is granted full-scale waltz treatment: the entire mimed passage of the ‘Largo assai’ is arranged as a set dance, ‘ironed’ into a regular, periodic structure, complete with fanfare-like accompaniment.

Example 2.3a Maréchal, Le Lac des aulnes, Act One, ‘Largo assai’.

Example 2.3b Maréchal, Le Lac des aulnes, Act One, ‘Largo maestoso’.

This blurring of styles may be significant: because there are plenty of occasions in the ballet when a similar motivic continuity is apparent (or else occasions when mime music approaches dance, if not becoming the subject of a danced scene itself); and because this blurring contrasts acutely with the musical specificity of Thérèse. An initial comparison of Maréchal’s and Hahn’s scores suggests a marked disparity of dramatic effect. Whilst Hahn’s music tends to be shaped by the immediate needs of text and action, belonging firmly in the diegesis, Maréchal’s participates less directly in the diegetic goings-on (here I disagree with Lalo, quoted earlier). The latter remains less a raconteur than a scenic cloth suggestive more of a general style than a specific substance – a cloth stitched from idiomatic snippets, gesturing to a mode of performance.

As for Les Bacchantes, what of its supposed ‘misunderstanding’ between ‘Dance’ and ‘Music’? Turning to the first ‘Pantomime’ scene of only thirteen bars, one finds a consistency of metre, a thick, chordal texture, regular harmonic progressions, question-and-answer phrases and a rounded form, with introductory staccato octaves and concluding sustained chords (Example 2.4). These features may well be dramatically suggestive: the chorale-like texture may evoke the ballet’s religious overtones, whilst the très lent tempo, the plodding rhythms and heavy, semiquaver articulation may conjure Bacchus’s nobility of character. But more intimate, moment-by-moment effects are absent. There is no ‘speech’ music (à la Mimi in Thérèse) and no local musical-dramatic synchronicity; ‘voice’ and ‘words’ are submerged. Indeed, this idea finds literal realization in a handful of passages from the ballet in which characters sing. Whilst one might expect these sung sections to help advance the action, to state plainly the words to be mimed, they tend to do the opposite. Singing occurs not in scenes of pantomime, but scenes of dance; vocal lines offer personal reflections on the stage goings-on rather than narrative exposition. Moreover, in the final sung passage (the penultimate number of the ballet), the singing voice is used purely for its sonic effect. There are no words at all, but merely the tuned murmurings of characters.

Example 2.4 Bruneau, Les Bacchantes, Act One, ‘Pantomime’.

This ‘neutralization’ of musical narrative could certainly generate further discussion – of the neutralization of narrative itself (mentioned earlier) and the development of more subtle means of musical characterization. For the present, though, it will be sufficient to register the contrast – between, on the one hand, Les Bacchantes and Le Lac des aulnes and, on the other, Thérèse. The last offers a musical-dramatic intimacy, a planned and patterned synchronization, that suggests little in the way of similarity with the other ballets of the period. Yet, and perhaps unsurprisingly, this synchronization may be reminiscent of an older balletic tradition: Thérèse draws from the same stock of musical-dramatic crutches as the 1840s ballet-pantomime.⁴¹ As Smith describes in her study of the genre, composers went to great lengths to tell a story, to provide a literal and transparent musical text made up not only of identifying themes and familiar signifiers for mood, feeling and geographic location, but of specifically balletic devices, notably the instrumental recitative.⁴² This device, designed to imitate the speech rhythms and contours of words to be mimed (written in libretti), features prominently in Thérèse. Mimi’s music, for example, constructs a straightforward signifying system that parallels verbal language: the music appears to ‘talk’, to ‘envoice’ the silent dancing character.

A second musical device favoured by ballet-pantomime composers was also employed by Hahn: his score is notable for its borrowings, especially from music with familiar connotative value – music thus able to directly facilitate spectators’ understanding of the plot.⁴³ One of these borrowings may be of particular dramatic significance. Prior to the dress consultations of Act One, the seamstresses spot a familiar face amongst the clientèle; it is Carlotta Grisi, beloved étoile of the Opéra. Gathering around the ballerina, the seamstresses request a private performance. Grisi acquiesces, performing the waltz from Giselle to original music by Adolphe Adam. The seamstresses burst into applause.

From an interpretive standpoint, this scene offers a crowning stroke. Thérèse not only participates in an act of borrowing typical of the ballet-pantomime; it borrows from one of the most illustrious ballet-pantomimes of all. In other words, and to pull together the several strands of these pages, Thérèse thematizes the genre – literally, in its quotation of Giselle – whilst simultaneously offering a structural, narrative and musical restoration. What is more, following the seamstresses’ applause, Grisi teaches the Giselle waltz to Mimi, and the seamstress goes on to repeat the dance in full. This repetition not only highlights the historical signifier; the dramatized didacticism extends off stage and towards the spectators in the theatre. The audience also receives instruction in the ways of the waltz, reminded once again of Grisi, of Giselle and of those illustrious days of ballet at the Opéra.

It is but a small step from here to a ‘cultural-political’ reading of the ballet, one that speculates about motivation, agenda and reception. Thérèse’s references to the ballet-pantomime may be configured as tactics of edification, designed to moralize on a national balletic history; the ballet itself may be an exercise in self-promotion, a product of an Opéra seeking roots in its past, historicizing its own history as a means of self-glorification. Indeed, in its references to the 1840s, Thérèse evoked an era when ballet – its technique, terminology, aesthetics, practitioners and products – was explicitly French. The ballet-pantomime, as noted earlier, was an exemplary French species, created and developed by French practitioners; although it gained acceptance on many European stages (in Milan, Stuttgart and Vienna), its home was the Paris Opéra.

Of course, this ‘cultural-political’ argument might be made firmer and more compelling by recourse to the historical context of early twentieth-century ballet – to the ‘décadence’ at the Opéra, not to mention the success of the Ballets Russes. One might ponder the fact that, just as the Russians captivated Paris with their exotic dancing and decor, the national company withdrew into the traditions that enfolded it. The historical coincidence of Thérèse and the Russian seasons, then, may encourage us to think again about old oppositions: oppositions between companies and the men at their helm, but also between creative aesthetics, dramatic models and musical styles.

There is an additional opposition to consider, one that may signal the complexity of visual and musical signifiers, as well as the possibility of historical cross-currents. To explain, I shall turn to another of Hahn’s musical borrowings.

Back to reality? A counter-narrative

After her turn as Giselle, and her warm reception by Palmyre’s workers, Carlotta Grisi singles out the prettiest: ‘Toi, qui es-tu, petite? Tu es la plus jolie!’ The seamstress (played by Carlotta Zambelli) replies with the following mimed ‘speech’:

The words are not those of Palmyre’s apprentice; the accompanying music is not that of Hahn. Quoted here is a well-known poem by Alfred de Musset, to music by Frédéric Bérat – a poem that confirms the identity of the seamstress as the blonde-haired grisette of 1840s Parisian subculture, a certain Mademoiselle Pinson (see Example 2.5).⁴⁴ What is offered here, then, and on a most basic level of character, is another reference to the period of Thérèse’s assumed heritage, a reference ‘realized’ through one of its most iconic characters: the Mimi of Musset, of Henri Murger, Puccini and Leoncavallo.⁴⁵

Example 2.5 Hahn, La Fête chez Thérèse, Act One, scene 3, ‘La Chanson de Mimi Pinson’.

Yet Mendès’s grisette shares few characteristics with her historical ancestors: with Musset’s thoughtful, insouciant heroine; with Murger’s pert, wilful coquette (and Leoncavallo’s similarly drawn gold-digging flirt); and with Puccini’s fragile muse.⁴⁶ The Mimi of Thérèse – incidentally, the real protagonist of the ballet – is a different creation: confident and self-assured, she demonstrates a mature initiative, winning back her lover through practical pleas and, unusually, surviving the ordeal. One might even describe her narrative portrayal as forgiving. Not only does Mimi forgive her lover’s wandering affections, but we, the audience, tend to forgive her more irksome characteristics: her spectacular balletic prowess (she learns the Giselle waltz in record time, trumping Grisi’s version with her own, six times the length); the aggressive, if sincere, language of her plea to Thérèse (‘Oh! Don’t love him! Tell him not to love you! Give him back to me!’); and her nagging interrogation of Théodore at the end of Act One.

Mendès may have modelled this balletic Mimi, if not on the character’s historical ancestors, then on her seamstress sisters of 1910: the ‘Mimi Pinsons’ of Gustave Charpentier’s Parisian night school, the Conservatoire Populaire de Mimi Pinson, founded by the composer in 1902. These namesakes – young, underprivileged working women (mainly shop-girls and seamstresses) – were offered classes by Charpentier and his cadre of distinguished colleagues in piano, popular song, dance, pantomime, declamation and, interestingly, fencing. By January 1910, roughly five hundred ‘Mimi Pinsons’ attended the school on the Élysée des Beaux-Arts, performing in public concerts at the Tuileries, the Place des Vosges and the Place de la Nation, and even attending national events (such as the unveiling in the Paris suburb of Suresnes of Émile Derré’s bust of Zola). The students became symbols of social advancement and equality; Charpentier’s school, to quote scholar Mary Ellen Poole, represented ‘the last gasp of Fourierist “art for the masses” as a didactic, morally uplifting, and pleasure-giving force’.⁴⁷

Mimi, too, may represent an ideal of social parity – an upward mobility that resonates with Charpentier’s institution. In a manner not dissimilar from that of the composer’s best students, Mimi triumphs over her domestic and social circumstances. More specifically, she triumphs over Thérèse: as one critic noted of the ballet, ‘C’est le triomphe de la grisette sur la grande dame.’⁴⁸ (Incidentally, this triumph owes to a musical coercion: Mimi’s identifying theme, granted an extra-diegetic quality, is literally ‘heard’ by Théodore and Thérèse during their passion, prompting the duchess, out of sympathy, to return Théodore to the seamstress.) Thérèse herself loses her lover, her clothes and even her dignity. Humiliated over Théodore’s presence during her undressing in Act One, scene 3, she continues to dance in only a petticoat. Critic Robert de Montesquiou was appalled: characters of noble rank rarely danced on stage (owing to librettists’ concerns for verisimilitude), let alone in their underwear.⁴⁹

As well as this broadly socialist resonance, the character of Mimi may also evoke contemporary politics of gender and identity. On this subject, Hahn’s score is once again evocative, in terms of both the amount of material allotted to Mimi and this music’s linguistic authority. The mime scene between Mimi and Théodore examined earlier will help illustrate (Example 2.2). Clearly, there is a marked disparity between the quantity of music assigned to each character (Mimi receives roughly twice as many bars as Théodore) – a disparity that might simply reflect Théodore’s dreamy, faraway state and Mimi’s desperate resolve. But Mimi’s music is by far the more idiomatic, as well as the more rhetorically persuasive. She has a distinct and distinctly recognizable identifying theme (the Bérat melody), as well as characteristic semiquaver ‘chatter’ – ‘chatter’ that, often beginning with a tonal wrench (suggesting the relentlessness of her discourse), overwhelms Théodore’s mimetic attempts. Théodore’s music, on the other hand, is hazy and vacuous. Here and earlier in the ballet, he is granted no identifying theme and tends instead merely to borrow music from one of his women. He seems disinclined to articulate himself clearly or individually, to enter into dialogue or to address the audience in monologues; he seems disinclined to ‘talk’ at all.

So much is evident – and the ensuing gender implications equally so: Mimi represents feminine authority, assertiveness, even aggression; Théodore embodies masculine submissiveness and docility. These gendered representations, which inverted the essential divisions of bourgeois life, are characteristic not only of the ballet in question, but of the early 1900s, a period that witnessed the rise of feminism and feminist periodicals, easier access for women to higher education and professional careers, a declining birth-rate and new divorce laws.⁵⁰ This general historical condition shored up the image of the femme nouvelle (the bourgeois ‘she-man’), a woman newly visible and equal to men (in social and professional station, as well as physical presence), independent and educated, and keen to campaign for her rights. One example of such a woman may be Mlle Stichel, the Opéra’s ballet-master and the choreographer of Thérèse. Previously engaged at the Châtelet, Paris, and the Théâtre de Monte-Carlo, Stichel was one of few women to occupy the much-vaunted position.⁵¹ She was also one of few to contest her authorial rights, taking legal action against the Opéra in a bid for an equal share (with librettist and composer) of Thérèse’s royalties and authorship. The bid, pleaded in front of the Tribunal Civil de la Seine on 9 February 1911, was successful, Stichel claiming victory for choreographers – and for women – nationwide.⁵²

A second example may be Mimi herself. The character – her actions, her fate and not least her music – may be representative, a theatrical incarnation of the social and gendered ideals of bourgeois life. Even her costume is suggestive. To rewind to the dresses with which the ballet began: all the dresses on display had long skirts, the length being historically authentic in suggesting 1840s fashion (see Figure 2.1). That Mimi’s skirt was long, though, was something of an irregularity. As Zambelli described in an interview with Le Miroir:

Zambelli went on to describe how one male spectator failed to recognize her backstage following the premiere and made the mistake of asking her name.

Figure 2.1 Zambelli as Mimi Pinson, La Fête chez Thérèse (1910).

Functioning as a protective shield, the skirt was not only exceptional on a balletic body more accustomed to knee-length (and above) gauze; it screened the body itself, concealing Zambelli’s identity – especially her sexual identity as an object of the male gaze. Yet the skirt also lent the dancer, and the character she played, some of the physical modesty and sobriety associated with the image of the femme nouvelle, an image that overturned the traditional type of decorative, eroticized femininity. This new, desexualized image even struck a chord with the dancer’s offstage social status and reputation. As Zambelli made plain in an article for the journal Je sais tout, the Opéra’s étoiles of the early twentieth century were far from the sexual trophies of the mid-nineteenth; they were rarely offered the lavish gifts, salaries and attention received by their historical counterparts. Instead, the Opéra’s brightest stars thought of themselves as mere civil servants, self-made working girls working hard to make a living. Like Mimi herself.⁵⁵

Backing up (conclusion)

Considered together, these ideas project a striking image of Thérèse and its resonance with contemporary culture – with the French feminist movement, with social reform and even with the practical matter of dancers’ social status. The ‘contemporary’ should be underlined, for all this discourse on ‘real life’ c.1910 leads us far from the original historical signifier – the French ballet-pantomime of the 1840s. A new line of argument emerges, one that may well be of signal importance to our understanding of the institutional project of the Opéra. Talk of didacticism, for example, might refocus on social instruction, on the amelioration of the female working class; previous notions of nationalism and nostalgia might cede to contemporary ideals of womanhood and the equality of classes and sexes.

Yet these two historical signifiers – the one retrospective, the other specific to the belle époque – can in fact coalesce: in other words, the contemporaneous currency of the ballet can endorse its generic heritage. As Marian Smith explains, the ballet-pantomime was prone to a mirroring of social relations, particularly gendered ones. Plots tended to uphold notions of masculine power and feminine submissiveness, to teach object lessons in social and sexual relations.⁵⁶ Thérèse, then, fits the bill. Like the ballet-pantomime of the 1840s, the ballet leans heavily on contemporary gender ideology. Only now, of course, this ideology has been inverted, the image of the decorative and eroticized ballerina replaced by that of the long-skirted Mimi, merely one of a crowd.

This argument – almost too convenient – should not make for a miserable conclusion. The position of Thérèse within an institutional and historical context throws up an important question about the relationship discussed at the start of this chapter – between the Paris Opéra and the Ballets Russes. Both companies were active in a period of intense debate about ballet, tradition and influence, a period that saw ballet-makers strive for self-preservation, self-aggrandizement and the wholesale capitalization on what was then a new and prominent public and critical interest. The Ballets Russes are often considered in this context. Historians have long described – and celebrated – the troupe’s innovative aesthetic, its relation to tradition and its ‘uncanny’ awareness of contemporary trends and tastes. But what would it mean to discern these same characteristics in the French ballet, the official and supposedly ‘conservative’ company? What if Thérèse was as self-conscious and strategic as the Russian offerings? The broadness of the ballet’s frame of reference, the number of onstage/offstage analogies, the multiply-allusive and iconic characterization, the composite score: these features may be contingent not only on a pre-established generic mould, but on a contemporary artistic climate in which ballet, for the first time in a good few decades, was actively seeking to make itself relevant, to communicate with a newfound audience.

These ideas run deep into the complexities of the historical scene, as well as the historiographical assumptions of ‘innovation’ and ‘Otherness’ that saturate the scholarly literature. In airing the ideas here, I hope simply to encourage a more wide-angled view of early twentieth-century ballet. For if the usual story of the period glorifies (foreign) individuals and initiatives, then revisionism might value continuities and cohabitation, might consider the French and the Russians (the self and the ‘Other’) in the same frame.