Search the Site

Late Style in Balanchine

Photo: Unity Phelan and Adrian Danchig-Waring in George Balanchine’s Agon. Photographer: Erin Baiano. Courtesy of the New York City Ballet.

Individual items of a repertory mutate happily in the hands of the individuals who perform them. In Of Late Style, Edward Said described this process as akin to rhetorical inventio, which he discussed in relation to music, where it means “the finding of a theme and developing it contrapuntally so that all its possibilities are articulated, expressed, and elaborated.”[1] Said summed it up in this way: “Invention is therefore a form of creative repetition and reliving.”[2] With the classic repertory of New York City Ballet something similar can and does happen. Choreographic work must not only be recognized, it must also be rediscovered; it must fit within a known pattern but also be restored to life under the changed conditions of the present. In Said’s felicitous phrasing, all repetition must be relived, which is why we go back to the theater to experience it again.

A case in point is the duet in Balanchine’s Agon (1957), which was recently performed with striking authority by Unity Phelan and Adrian Danchig-Waring (May 10th and May 27th). This interpretation of the fabled Pas de Deux evoked both a known and an unknown landscape within the ballet. Danchig-Waring excels at dancing Balanchine’s late Stravinsky pieces, and Phelan has been dancing Agon for some time, although she was only recently promoted to principal. In the Pas de Deux they keep direct eye contact to a minimum and portray a sense of severe formality in their relationship, combined with a mysterious urgency. I had not seen Danchig-Waring in this ballet before, but I think he and Phelan are now its major exponents.

In the May 10th performance, at the end of the first section of Agon in the tableau vivant, where the dancers look out at the audience like a little court society representing itself to the public gaze, Danchig-Waring’s pose was particularly striking. As is well-known, the ballet was based to some degree on a 1623 French dance treatise. Danchig-Waring’s haughty demeanor embodied French noble heroism and the taste for glory of the period prior to the consolidation of absolute monarchical rule, which is precisely when the generative treatise was penned. His theatricalization of this gesture was not only clarifying for the choreography’s historical intertext (which is a minor point), more importantly, it freed him to inhabit the Pas de Deux with an immense feeling of integrity, making the two dancers of equal weight and as though absorbed in a contest (echoing the ballet’s title). This sense of character I have just described may run counter to the received philosophy for performance of Balanchine as committed only to the movement. But the rebellious and recalcitrant noble context is very present in Agon, and the title itself may bear a trace of this context in the minds of Balanchine and Stravinsky.

The score is itself an example of the composer’s late style, as Jonathan Cross notes: “Stravinsky had in later life found a new musical direction, a new, much more rarified kind of neoclassicism, an even more refined stylization of past forms and traditions.”[3] And Said points out that much of modernism is itself a late-style phenomenon: “Modernism has come to seem paradoxically not so much a movement of the new as a movement of aging and ending.”[4] The term “agon,” meaning contest, may refer to the contest between noble and monarchical rule, one that took place in the first half of the seventeenth century. Through micro-movements of his torso that accentuated his body’s volume in space rather than uniquely its line, Danchig-Waring gave sustained attention in the Pas de Deux to transitional moments: this focus opened the choreographic score to a more contrapuntal reading, just as history may play a contrapuntal role in the ballet’s modernism.

Another important intertext for Agon relates to the ballet’s original casting, when Arthur Mitchell partnered Diana Adams. The duet for a black male dancer and white female dancer brought race into the purview of neoclassical ballet in 1957. Even though Mitchell said Balanchine was interested in the black and white skin color contrast as a purely formal element, his presence also introduced a sense of character at a time when audiences were not used to seeing black and white dancers working together. I use character here to mean character on stage: white, middle-class audiences were asked to consider what it meant to see black and white characters interacting in a ballet.

Character is not a very popular way to read dance movement in what is often glibly considered to be abstract ballet (ballet without plot). But there are other ways to read dance movement, one of which can be thought of as volumetric or stereometric and the other linear, or at its most extreme, purely optical. My impression of Danchig-Waring’s performance is that the character revealed in the tableau vivant unleashed the dancer’s volumetric potential. In 1973, philosopher David Michael Levin claimed Balanchine superseded the stereometric perception of the dancing body by replacing it with “a purely optical reduction of corporeal mass and weight in accordance with the modernist aesthetic.”[5] For example, in Agon the moment of the ballerina’s right angle arabesque penché supported from a position of the male dancer on his back could be experienced as just such a purely optical moment. The sense of a vertical line intersecting a horizontal line can dominate our visual perception of this moment. One can have this impression in Balanchine’s work because his choreography tends to play with optical illusion.

Building upon this idea, Levin also posited that Balanchine moved beyond poise in dance (which he considered an historically outmoded aristocratic quality) to arrive at grace, “a sublime and sacred presence” which, however paradoxical this may sound, made ballet modernist.[6] However, what I pointed out in Danchig-Waring’s pose was not poise but arrogance. Levin evacuates the ballet’s historical subtext while retaining the historical connections between grace and the sacred in ballet’s monarchical heritage. He associates the purely optical experience with the sublime and the sacred as well as with grace in a quasi-religious sense. In doing so, Levin followed the consecrated narrative of ballet as the product of royal absolutism in the second half of the seventeenth century. Here the king appeared as the sun, which made vision possible and bedazzled the eye. More importantly, the purely optical moment transcends bodies existing in history. In the context of the seventeenth-century court ballet, grace and sacredness evoke belief in monarchical absolutism and the king’s historical use of his own dancing on stage as a sign of political force. Levin’s purely optical moment derives, like it or not, from the political use of spectacle.

But this idea—that the primary aesthetic characteristic of modernist art is the body itself being sublated with grace—is based to my mind on a flawed reading of historical ballet aesthetics (and perhaps of modernism as well). For it identifies the modern by removing it from the historical, whereas in fact neoclassical ballet (ballet modernism) is deeply embedded in its historical lineage. It is therefore disingenuous and reductive to equate court ballet, as Levin does, with superficial courtly manners, given that ballets often served important political purposes as ritual representations of political belief. In the French classical era, the second half of the seventeenth century, court ballet was in fact instrumental in presenting the king’s body itself on stage in a new political theology. Ballet had thus already advanced well beyond the superficial qualities of poise in the seventeenth century. Why then is ballet today—whether you choose to call it modernist or not—still burdened with impressions linked to sacralization, understood in more modernized terms as the suspension of the body’s weight and mass? In Phelan and Danchig-Waring’s performance this moment, while perfectly executed, did not create the awaited effect of suspended corporeal materiality. They fully realized but also desacralized Agon.

When we think today of grace and the sacred in the context of modernist aesthetics, the example of Balanchine points toward the importance of historical context, whether that of seventeenth-century French history or of race in the United States of the 1950s. In the case of Agon, the two contexts intersect because Balanchine’s decision to cast Mitchell in Agon in 1957 resonated with a resistant and individualistic gesture from the first half of the French seventeenth century, a gesture Danchig-Waring brought forth in his performance. And, at the same time, Balanchine in 1957 was making a politically charged statement while apparently using the most aestheticized means. As ballerina Allegra Kent remarked: “Balanchine was very devoted to that black and white thing, not for political reasons, but because he liked that contrast.”[7] Yet it is precisely the politics of aesthetics that plays a decisive role in how we read dance. The neoclassical articulates modernist formalism within a late-historical intertext, and this needs to be accounted for in an adequate dance criticism.

Said’s notion of late style also refers to work created late in the career of certain artists that counters the expectations we have of them. “[N]ear the end of their lives their work and thought acquires a new idiom.”[8] Based on Adorno’s understanding of Beethoven’s late works, we come upon descriptions of the late works as a form of exile or “the alienated masterpiece.”[9] In the spring 2023 season I would consider Balanchine’s Kammermusik no. 2 (1978) as such a late work. The Hindemith score (1924) follows two distinct orchestral lines that Balanchine interpreted as a gendered division of labor, one for an all-male corps de ballet and the other for two female soloists (May 10th). As it is perhaps the most conceptual of Balanchine’s ballets, it is difficult work to take it in at first viewing, and it shows him at the top of his game in terms of manipulated stage space through choreographic contrasts and mutating compositional configurations. Of its revival in 1984 dance critic Anna Kisselgoff said: “the music remains as startling and jarring as ever . . . It will never become an audience favorite.”[10] With the male dancers taking on what would normally be the ornamental role of the female corps de ballet, the word alienation comes to mind, particularly given how the male-female relation is starkly modified with the two genders following distinct musical lines of Hindemith’s score. The result is that the two genders are handled strictly contrapuntally, which leaves room open for future combinations of gender identity. Despite Balanchine’s personal obsession with heterosexual relationships, his final work augurs liberation from this binary. The outcome of the two heterosexual duets—the only place where male and female meet—are fundamentally dystopic.

The men are frequently linked to one another, holding hands at arm’s distance, and thus create extended sculptural shapes. This framing presence of the corps de ballet, placed at a distance behind the main action, worked stereometrically rather than optically because of the accumulated mass and weight of the male bodies, not simply by virtue of their gender alone. Balanchine was clearly interested in this quality, which may be why he gave them dramatic stances and grouped them in ways that accentuated their massive occupation of space as a group. The corps de ballet usually multiplies the presence of the principal ballerina, whereas here the prima ballerina is herself shadowed by another dancer—somewhat behind her; she is in a sense being undermined—while the ornamental frame usually designed to magnify the ballerina is taken on by men who do not interact with either of them. The women do occasionally partner two men, who break off from the corps only to return to it. While the women are not defined separately in relation to their sometime male partners, they are defined chiefly in relation to each other, like alter egos. In a sense, Kammermusik no. 2 is an inverted Concerto Barocco, with its female corps de ballet and its two female soloists accompanied only by the brief appearance of one male dancer. It is therefore particularly interesting to see these two ballets on the same program. Concerto Barocco is baroque by virtue of its contrapuntal qualities and its choreographic shading (repeatedly falling off balance to the side), whereas Kammermusik is, at least for the men, more choral and gestural. Without wishing to get too psychological, one can almost read the women’s frenetic movement as a hysterical reaction to the male chorus.

The repertoire has been exceedingly mixed during Winter and Spring seasons, and New York City Ballet’s approach to programming, despite some excellent curation, tended occasionally to veer off course. The way we receive works is very influenced by the programmatic sequence of an evening. Despite the packaging of programs by content, some ballets performed together on one evening can generate challenging contrasts. Balanchine’s high-modernist Episodes (1959) paired with Robbins’ musical-comedy oriented Fancy Free (1944) is one example. Although both are beautifully performed in themselves, they deal with relationships from such clashing perspectives that one cancels the other out. They are best seen separately. Or consider the first all-Balanchine program this spring, which combined Concerto Barocco (1941), Kammermusik no. 2 (1978), and Raymonda Variations (1961), in that order. Concerto Barocco, Balanchine’s first fully neoclassical ballet set to Bach, was followed on this program by his highly unusual late-career work, set to Hindemith. The program closed, however, with Balanchine’s nostalgic evocation of the Russian classical tradition (Glazounov). The pairing of the first with the second piece is fascinating: it marks the limit points of an important choreographic trajectory commonly known as neoclassical but that could also be called baroque. Raymonda Variations, however, came across in this context as a curious throwback to nineteenth-century classicism. Since the chronology of the three pieces was not observed, why did Raymonda not open the program? As it was, we went to the limits of Balanchine’s late style only to rebound back, as it were, to the nineteenth century.

Another jarring programming effect was produced by Balanchine’s La Source (1968) paired with Ratmansky’s Namouna (2010). Both works use music by nineteenth-century French composers (Leo Delibes and Edouard Lalo, respectively) and both refer to dance history. But they clashed conceptually and stylistically. In her recent book, One Dead at the Paris Opera Ballet: La Source 1866-2014, Felicia McCarren discusses the original ballet conceived by Charles Nuitter and choreographed by Arthur Saint-Léon as concerned with gender, ecology, and Orientalism. One Muslim character, Naila, is considered “the incarnation of spring water” (which is the meaning of the French term source).[11] And, as McCarren argues, “[T]he ballet reads not as abstract romanticism but as colonial commentary, not poetic reflection but strategy for the management of water and women.”[12] She also discusses the Paris Opera remake of La Source in 2011 choreographed by Jean-Guillaume Bart, a version the publicity claimed took an “eco-feminist approach,” although McCarren finds the issues of race and the environment to have been ignored. She does not discuss Balanchine’s La Source, which is an extraordinarily sophisticated pas de deux mirrored by the all-female corps de ballet. But it is safe to say that, once the plot has been scrapped as in Balanchine, any such concerns evaporate. The abounding energy of the piece as personified in the ballerinas—Erica Pereira (May 3) and Emma von Enck (May 6) were both excellent—could itself be thought of itself as a spring or a source. In this sense, Balanchine was close philosophically to Merce Cunningham, who said: “Movement comes from something, not from something expressive, but from some momentum or energy.”[13] Balanchine has used the idea of the spring with all its associations to animate the work.

Ratmansky has done quite the opposite with Namouna, where the only link to the original 1882 ballet of that name is the musical score. The original libretto was also penned by Nuitter, and Lucien Petipa’s choreography featured a slave-girl character Namouna in a ballet that, like La Source, was originally rife with exoticism. Questions of race and colonialism were part of the history of some ballets presented this season. Modernist formalism can, however, mask an intertext rather than articulate it. Serge Lifar used the Lalo score for his Suite en blanc, an abstract neoclassical ballet created in 1943 during the German Occupation of Paris that highlights French academic technique. Working with the score while discarding the original plot is certainly one way to start from scratch with a nineteenth-century ballet. But, unlike Lifar, Ratmansky fills his new ballet with plot, or rather, with free-floating plot fragments. The hodgepodge of plot fragments does not reference the nineteenth-century narrative of Namouna but instead quotes motifs from iconic nineteenth- and twentieth-century ballets. To begin with, he cites the famous solo known as “La Cigarette” in Lifar’s Suite en blanc where a female soloist uses her port-de-bras overhead to suggest wafting cigarette smoke. Ratmansky has the dancer (Georgina Pasoguin) smoke a cigarette to the same music. This motif is woven throughout the ballet. There are soloists whose headgear make them look like moving statues (another reference to neoclassicism) but they also suggest bathing caps as in Nijinska’s Le Train Bleu or Balanchine’s original Apollon Musagète. We bear witness to a judgement of Paris scenario that can be found in Apollo, but the male figure in a sailor outfit seems to have emerged from John Cranko’s Pineapple Poll (1951), a very popular piece of post-war British repertory.

The whole ballet, in other terms, is a mash-up of ballet references. We find every conceivable movement trope of ballet repertory from Nijinsky’s Rite of Spring, where women assemble in a circle only to ostracize one of their own, to Martha Graham’s Primitive Mysteries, when a press into the hips with head up and hands and feet rooted in the ground is performed by the same group of women. There is the marriage plot (Swan Lake) with a celebration motif from the end of Firebird, and there are references to Giselle and Bolero as well as some zany gesticulations and pedestrian gestures that poke fun at the clichés on view. The problem is in each instance the quote is gratuitous and sometimes presented crassly. At no moment does any of this add up to a point of view or a thoughtful critique of balletic dramaturgy or ballet tradition.

Ratmansky’s Namouna is, in short, postmodern pastiche as parody, while the dancing is characteristic of neoclassical ballet performed as virtuosity pushed to its limits of endurance. Yet, this is a neoclassicism that envisages movement itself as theatrical paraphernalia. In 1911, Akim Volynsky foresaw neoclassicism and its attraction to the audience when he suspected the audience could see through all the detritus of a traditional production: “This attraction relates not to the ballet’s plot, which in most cases is mediocre; not to its music which is almost always irritating by virtue of its simplicity; not to the vulgar apparatus of its stage sets and costumes but rather exclusively to the dancing itself . . . and the brilliant illusion of its forms.”[14] Ratmansky has created a tongue-in-cheek choreography out of ballet’s most spectacular effects, but in the process, he has created a piece that is about ballet as effect. Namouna reminds me a bit of Forsythe’s Artifact (1984), except that the latter was grounded in a questioning of the origins of ballet in the seventeenth century, and thus more provocatively conceptual. The comparison between these two works is not flattering to Namouna. Whereas La Source, now over fifty years old, is an incredibly inventive look at the conventions of classical partnering choreographed (and danced with Anthony Huxley partnering both female leads) in an extremely sophisticated manner, Ratmansky’s postmodernism seems decidedly out of date and lacking in intellectual and emotional energy despite the commitment of the dancers. And Balanchine’s work, which earlier may have come across as slight because of its light classical touch, now seems all the more profound by comparison with an empty rewriting of history.[15]


[1]Edward W. Said, On Late Style. Music and Literature Against the Grain (New York: Vintage Books, 2007), 128. I wish to thank Tanya Jayani Fernando for her impressive theater piece Late Style or a Double Fugue, which brought this book back to my attention.


[3]Jonathan Cross, “Stravinsky’s Theaters,” in The Cambridge Companion to Stravinsky (Cambridge: Cambridge University Press, 2003), 145.

[4]Said, On Late Style, 135.

[5]David Michael Levin, “Balanchine’s Formalism,” in What is Dance? Readings in Theory and Criticism edited by Roger Copeland and Marshall Cohen (New York: Oxford University Press, 1983), 137.

[6]Ibid., 133.

[7]“Agon at Sixty,” New York Review of Books, June 26, 2017

[8]Said, On Late Style, 6.

[9]Said, On Late Style, 8.

[10]Anna Kisselgoff, “Dance: City Ballet’s Kammermusik,” in The New York Times (February 13, 1984):

[11]Felicia McCarren, One Dead at the Paris Opera. La Source 1866-2014 (New York: Oxford University Press, 2020), 1.

[12]Ibid., 3.

[13]Merce Cunningham and Jacqueline Lesschaeve, The Dancer and the Dance (New York: Marion Boyars, 1985), 68.

[14]Akim Volynsky, “Dance as Solemn Ritual,” in Ballet’s Magic Kingdom. Selected Writings on Dance in Russia, 1911-1925 translated and edited by Stanley J. Rabinowitz (New Haven and London: Yale University Press, 2008), 3.

[15]I wish to thank Susan Jones for her input during conversations we had as I wrote this article.


MARK FRANKO is Dance Editor of The Massachusetts Review and Laura H. Carnell Professor of Dance at Temple University. A revised edition of his Dancing Modernism/Performing Politics was published this spring by Indiana University Press.

A PDF of this essay is available here.


Join the email list for our latest news