THE HISTORY WE ARE SKETCHING is one of boundaries double crossed between India and the West and between periods of the South Asian past. On one level our story is about an historical irony, how late nineteenth-century Orientalism resuscitated the romantic mystique of the eastern dancer in the West just as South Indian dancers were being repressed in their homeland by Indian reformers influenced by western mores. Within that history there is another dynamic that is less about crossing than about shifting boundaries, boundaries between the sacred and the profane and their expression in colonial law. We will be looking at these movements and transformations within the context of current scholarship that is historicizing even those elements of Indian culture conventionally understood to be most ancient and unchanging.

From the eighteenth century forward, there was one assumption about India that was shared by Orientalists like Sir William Jones, who revered Indian civilization, and Anglicists like James Mill and Thomas Babington Macaulay, who notoriously thought one shelf of a good European library to be worth more than the entire native literature of India and Arabia. To both camps, India was an ancient and essentially static civilization whose chief glories, whatever their relative merit, belonged to the eastern equivalent of classical civilization with Sanskrit in the place of Greek and Latin. Although of great moment, the conflict between the Orientalists and Anglicists took place within an evolving imperial ideology with more disagreement about means than ends. The shared assumption was that while the West too had sunk into its middle ages, a renaissance had followed leading to a progressive modernity, whereas India had fragmented politically and stagnated culturally. Civilization was not many, but one. It was a hierarchy with Europe at the top, and by European standards India was in severe need of long-term, moral and material uplift (Cohn; Blaut).

India became British while liberalism was on the rise, and as the British expanded the territory under their control they quite naturally looked for equivalents of the liberal social order's ideological building blocks: private property, individual liberty, a legal code that sustained property and liberty, and education in western categories of knowledge, and found them wanting–or, excepting education, forgotten. In the case of the law, for example, were there not Sanskrit codes that could be recovered and doubly translated first into English, and then into a system that would counter Oriental despotism with written laws? Unlike the judgments of local pandits, this textual authority would be at once native and intelligible to the British, a set of rules upon which case law could be built. As Thomas Metcalf notes, the idea that there existed “original texts, and that these could be taken as representing an enduring Indian reality inevitably meant that any code based on these texts would devalue India's historic experience,” experience like that of the local pandits, who functioned in the place of the common law that the British could not find (28–39).

Even the caste system, that prime symbol of eternal India, has now been historicized. Effective caste status, which had been, says D. A. Washbrook, subject to multiple influences and flexible interpretations, became more rigid because the British accepted the Brahmanic varna caste schema as descriptive and consequently formalized in law a system of the elite that had been more flexible in practice than in theory. Colonial law, in turn, affected everything from inheritance, property, and freedom of movement, to religious practices and the rules governing marriage (412–15). The evolving British construction of the “Aryan,” whether through Orientalist philology, archeology, or ethnography, allowed the British to see themselves as successors to original, assumptively white, conquerors of India and to racialize and naturalize the caste system as the residue of successive conquests (Dirks, Part I–Part III; Trautmann).

Unlike earlier conquerors, the British ruled India for the benefit of a home country, restructuring its economy and creating a class system on top of its caste system. In the process they altered social relations, fractured the subsistence economy that was the traditional hedge against famine, and cut incipient, Indian modernities off at the root. Post-colonial political and economic history now refers routinely to the traditionalization, peasantization, decapitalization, and Sanscritization of British India, to which we would add secularization. Looking past a stagnant social order to an Indo-European, Aryan civilization selectively linked social and moral progress with the recuperation of an idealized past, a construction that resonated not only with imperialists, but with Indian reformers, nationalists, and Indophiles like the Theosophists, who came to support Indian self-rule. The transformative effect of this historical construction of an essential India was as profound in the arts as any other aspect of Indian life.

Ideological investment in an idealized, originary civilization figuratively folded time, bringing the horizons of the present and the Vedic past close enough to be bridged by Sanskrit texts and archeological studies while creating an abyss. It was not just the modernities that might have been that were consigned to that abyss, but repressed social practices, abjected ways of life reduced to traces. This loss is the negative side of hybridity, the repressed term of its dialectic, that which is understood but goes without saying in accounts of hybridity as resistant and a form of subaltern agency, whether as “intervention in the exercise of authority,” the practice of “sly civility,” or colonial dialogism.¹

It is the realm of specters that the sociologist Avery Gordon in Ghostly Matters calls “the lost subjects of history–the missing and lost ones and the blind fields they inhabit” (195). Our ghost is that of the devadasi, the Indian temple dancer.

As a student of Bharatanatyam dance, Avanthi Meduri absorbed the received history that bridged the abyss, the tale of a once threatened, but never-the-less continuous art form whose postures were sculpted on ancient temples and whose gestures were codified in the second century Sanskrit of the Natyashastra of Bharata.²

As a scholar, however, she discovered that the oral transmission of the dance revival she was trying to document through fieldwork was speaking an already textualized history. Although referred to as if always already there, the long-lost Natyashastra was rediscovered piecemeal in European archives between 1865 and 1890 and published in English in 1890 and in Sanskrit in India in 1894. Its recovery was crucial in linking aesthetic history to a nascent Indian nationalism that took pride in the Vedic heritage, but its specific effect was to Sanskritize the dance by undercutting “variant practices that had evolved locally in the long period of its loss” (32). The story of how the devadasi, the temple dancer of South India, came to be abjected, consigned to the abyss, while her Vedic ancestor was being celebrated, is part of an ironic interplay between eastern and western ideas about the dancer and her dance.

Of all the Hindu practices that the British invoked to mark their moral superiority to their Indian subjects, “temple prostitution” may well have been the most notorious after the predations of the so-called “thugs” and the self-immolation of widows (sati).³

The West had no conceptual category for women who were at one and the same time unchaste and holy. The temple dancer's combination of religious and sexual expression reminded Europeans of that abomination of the ancient Near East, ritual prostitution. (“Thou shalt not bring the wages of a kedesha…into the House of Jehovah,” Deut. 23.18 [Metcalf 102]).

The devadasis–literally the slaves or bond-servants of god –were not, strictly speaking, a caste. Rather they had, as Saskia Kersenboom-Story (179) and Amrit Srinivasan note, a way of life or professional ethic [vrtti, murai, not jati]. One could be born into their community, but some girls were formally adopted after being offered, or even sold, to the temples by their families. They were trained in performance from childhood by male teachers from the community and, unlike members of a caste, could not officially perform without their ratification and approval. In contrast to the conventional, patriarchal system of arranged marriages, it was mothers in the community who dedicated their daughters in childhood to the temple god. A traditional marriage ceremony was performed with a sacred object standing in for the god husband.

After the dancer had completed her training and was ritually accepted into the service of the deity, a bodily consummation might be performed with a temple Brahman or a wealthy patron who would, ideally, then support the devadasi and her attendant musicians while her auspicious presence blessed his house and entertained his guests. She was an “other” woman with a recognized place in a society that sanctioned polygamy. Her children were legitimate. She alone among women could inherit property, adopt children, and lead a relatively autonomous sexual life. She was the vital link between the god, the temple, the priest, and the street. Her domestic duties were for the god only; conventional domesticity was forbidden.

As the tradition came forward from the seventeenth and eighteenth centuries, she was nityasumangali, the evergreen bride of god, the woman whose bangles were never broken because she could not be widowed.⁴

When she died her body was covered with the god's own cloth; her pyre was lit by temple fire, and the idol himself observed mourning. So despite being a liminal figure, a dangerous supplement to the conventional domestic order, and an ambiguous presence in the long South Asian literary traditions, she was an auspicious being. She performed the arati ceremony that warded off the evil eye from gods and human beings alike, blessed marriages, and performed other rite of passage ceremonies for families that could afford her services and were of appropriate caste status (Gaston 31, 42; for her own rites of passage, Kersenboom-Story, 185–93).

Devidasis were a presence in traditional South India from the village to the major temples and royal courts, which maintained large numbers of musicians and performers (dance is really too restrictive a category). Not every devadasi did everything, of course, but taken together their dancing, performing, singing, and playing of instrumental music ranged from temple ritual and strict devotion, to seductive nautch dances, to satirical and comic numbers, even feats of gymnastic dexterity. They sang in Sanskrit, local languages, and even in English.

The ritual function of the temple devadasi was only visible to westerners during car festivals (urkolums) when the deities went out from the temple to see and be seen by their devotees. Preceding the divine images, the devadasis expressed through their constant movement the female principle of embodied energy (prakriti). Through the language of hand gestures (hasta mudras), mime, and facial expression they narrated divine stories, animating the passive male principle (purusha) of the idol. Though they deflected the evil eye even from the gods themselves, to the English they were just another form of nautch girl.⁵

Nineteenth-century sources list as many as six terms in various Indian languages for ranks of female dancer ranging from the most skilled in dance, music, writing, the composition of songs and erotic poetry and in some cases painting–for in traditional India these were the only classes of women with systematic education–down to the level of common prostitutes. Each group associated with caste-appropriate clients. There were local distinctions by language and custom. There was no single term for the devadasi's dance, and “devadasi” itself only became a pan-Indian word for the temple-dancer in the twentieth century (Marglin 313). Moreover, while all female dancers from the isai vellala community of dancers and musicians were referred to as devadasi, or the equivalent local term, not every dancer was a dedicated temple dancer. These non-dedicated dancers were closer in their practice to the parallel but secular courtesan tradition among Muslims, often grouped under the term “tawaif”–not to mention the “baiji” in Bengal who could be either Hindus or Muslims. All of these women danced and sang to entertain the guests of the high born or wealthy; all were referred to by the English as DG's (dancing girls), or nautch girls.⁶

Preeminent among the southern kingdoms in music and devadasi performance, the Kingdom of Tanjore was a key ally of the British in their struggle against Tipu Sultan, supplying rice for the Company army. The Rajahs of Tanjore were reformers tutored by westerners. Although Tanjore was a client state, and British exactions drastically reduced their revenue, the Rajahs remained inventive patrons of the arts and themselves writers of songs and dramas. They introduced western learning, instruments and performers, experimented with western music and musical notation, and, in short, promoted an indigenous modernity. It was at the court of Serfoji II that the famous brothers known as the Tanjore Quartet developed a concert form of the Tamil devadasi's dance that became the ancestor of Bharatanatyam. This formation was not simply the codification of traditional form but an instance of an indigenous modernism that in some aspects was influenced by European music and ballet.

However important the work of the brothers was aesthetically, it was also an indigenous displacement, part of the trend that was shifting the balance of power in the isai vellala community away from the women by putting their signature on the dance. The rift between the devadasis, who traditionally held sway as the chief source of revenue for the community, and the musicians in particular gained irreversible momentum after 1856 (Srivinasan, “Reform or Conformity?”). In that year Raja Shivaji II died before completing the adoption of a male heir, and the British, citing their codified Hindu law, refused, in the name of Queen Victoria no less, to accept a female heir, declared the throne empty and the kingdom forfeit–an act of monumental bad faith with a loyal ally and a disaster for the devadasi. Hers was an expensive patronage tradition based on support of the temple, the royal court, and the emulation of the court by the wealthy. Overnight the royal patron was gone and, as the century waned, urbanization took population away from the old temple cities where most devadasi were based. Private patronage declined. She became subject to the English codification of Hindu domestic law. She was sliding into the abyss.

The key legal question was whether under colonial law devadasis were to be essentially holy women whose devotion was expressed artistically, and therefore only accidentally public women, or were they to be essentially prostitutes, and therefore only accidentally artists. It was, after all, the women's religious status that distinguished the devadasi institution from the courtesan traditions of elsewhere. The judge's ruling deciding the fate of a seventeen-year old who sued because her temple prevented her from inheriting the estate of the mother who had adopted her by refusing her final ceremony, the tying of a bottu-tali, is unambiguous: “inadmissible, as being in effect a claim by the plaintiff to be enlisted as a public prostitute” (Thurston 149). Early in the nineteenth century the British often attended and even provided Indian entertainment. But as the century progressed, the authorities began to discourage British attendance at nautch parties, finally forbidding it. A full-fledged Indian anti-nautch movement–its very name an insult to its target, the devadasis–sprang up in the 1890s fueled by the moralizing of missionaries from without, and a campaign for women's rights, and against Brahmanic privilege, from within South India. For both missionaries and internal reformers the dedication of girls to the temple for “immoral purposes” was a wedge issue to attack child marriage generally as well as to promote education for all women, the right of widows to remarry, and the prevention of venereal disease (Srnivavasn, “The Hindu Temple Dancer”).

Conceivably, the devadasis might have better resisted changes in patronage, population, and legal status had they not been accompanied by a broader process of secularization that was for the devadasis a desacrilization. Temples began to expel devadasis, making them beggars at their own gates. The musicians of their community, the nagaswaram, petitioned to be recognized as a separate sub-caste and were able to go onto the urban concert stage as classical musicians free from an association that had become a contamination. While the sacred status of the devadasis declined, the taboos associated with them persisted. Outside of a few high profile marriages with reformers, the devadasis were transformed from women outside the system of domesticity who could not have a mortal marriage, to women judged by the standards of domesticity whom no one would marry. As the opportunities to practice their traditional art declined, the accusation of prostitution became self-fulfilling prophecy for many who could not find work as actresses or find another means of livelihood. As for the dance, it was an embodied practice, part of who the devadasis were, and not an art form that could be taken up by others without a change in cultural context. Hence the judgment made by reformers that if the abolishment of temple dedication meant the loss of the dance, they would, however regretfully, accept that loss.

By the time the French romancer Pierre Loti (Julien Viaud) asks to see the famous bayaderes of Pondicherry on the trip detailed in his 1904 travelogue India, there are none there to be found. Devadasis have to be sent for from a Siva temple further south to give command performance. Loti admires the figures of the dancers, particularly the mime of the principal dancer that first seduces, then reproaches. The “creature glittering in gold and jewels” mocks him with laughter, accuses him, and treats him with disdain while the crowd jeers good naturedly. Afterward he sees his favorite dancer wiping sweat from her face as she receives his complements with cool bows. “What thoughts can there be in the soul of a bayadere of the old race and pure blood?” he wonders, “the daughter and granddaughter of bayaderes, one who has been trained through descent, that has lasted for hundreds and thousands of years, to be a creature of naught but phantasy and pleasure” (135–40).

Even more than the western moralistic and ethnographic accounts of the devadasi, the dancer's mute presence before Loti as he conjectures her interiority, translating her into “a creature of naught but phantasy,” marks her as a subaltern.⁷

Even before British colonization, the devadasi was not an autonomous subject, but a generic one, a symbol, an icon, a creature of the temple and the court. Her privileges were inscribed within a larger patriarchal system of exchange. What she did have was her art. But her performances were not expressions of individuality; they were “inscribed in performative histories and genealogies known as paramparas…with their own subjectivities.” Her gestures, her songs, the seductive movements of her eyes “supplemented her subalternity and facilitated the misrecognition of dancer with dance, self with performance” (Meduri, Nation, Woman 21–23). As an artist she performed a seductive sexuality that she did not practice, but that distinction between performer and person faded with the loss of patronage and her sacred status.

Elite devadasis tried to defend themselves in the new context. In 1910 the devadasi Bangalore Nagaratnamma reprinted the classic Telugu poem Radhika Santwanam (Appeasing Radhika) by her eighteenth-century sister Muddupalani, one of the world's great expressions of the erotic from a woman's point of view, a key feature of the Telugu devadasi's bhakti tradition, only to have the book suppressed as obscene (Tharu and Lalita 1: 1–12). As late as 1932, devadasis, still trying to recontextualize themselves, wrote to the Madras Mail objecting to the equation between public performance and “leading a vicious life” concluding: “As a class we enjoy the freedom which the Women's Indian Association is supposed to be struggling for…. We live according to the ideals of our caste following a definite code of morals. We wonder if the social reformers understand what they are doing when they talk ill of us as a whole for the lapses of the few” (Meduri, Woman, Nation 226). But with the loss of their sacred status, the devadasi lost the real, if constrained, subject position they once enjoyed. Without the power to define themselves, they were reduced to simple subalternity, and “the subaltern is not heard” (Spivak, “How to Read” 138). There was no place for the devadasi within the value structure of the conventional Hindu domestic (grhasta) system. Her day as a dangerous supplement to that system was gone; the formal banning of temple dedication in 1947 merely a coda.

Although some traditional dancers and teachers survived long enough to participate in the revival of the dance when a new artistic context enabled “respectable” women to take up the concert form of the dance in the 1930s as a classical art, they had long since bent with the times. Asked for a courtesan song, the greatest survivor of the devadasi tradition, T. Balasaraswati, taught Jon Higgins a padam of erotic promises with the refrain “But only if you have the money,” but insisted emphatically that she did not dance it (155). We know what has survived, but the random evidence of lost performances in the archive, the empty names of dances, the occasional drawings and descriptions, merely hint at what fell into the abyss without leaving a trace behind.

With our account of the constrained and abjected figure of the subaltern devadasi dislodged from the temple, out of fashion in the great houses of the zammandars, a pawn in the political struggles between anti-Brahman Tamil reformers and Brahman traditionalists–exposed to western eyes, whether the hostile gaze of the missionary, the analytic eye of the ethnographer, or the idealizing vision of latter-day Orientalists – we have reached the limit of discursive history.

How then to represent the being of a figure who comes down to us first through western accounts of an eastern Other, and then as the contested colonial site of a social morality play? What are we to make of a signifying, dancing body, whose significance changed not only under the impact of the western gaze, but within Indian tradition, a body named, renamed, reformulated in multiple discourses, who is even today a spectral aura in the classical dance that was abstracted from one aspect of her traditional performances. Although the subaltern devadasi did not occupy a subject position from which she could speak and be heard while she was being discursively figured, her situation can be performed.

In God Has Changed His Name, Avanthi Meduri has created a post-colonial, metaleptic drama, a multilayered performance in which the devadasi appears as a recalcitrant ghost-woman, a mute figure haunting the imagination of a dance historian trained in the West, demanding through gesture that her story be told. The drama itself is in the popular South Indian Koothu tradition, which symbolically returns the devadasi to the street, to the element of popular culture that was lost when her dance became a classical stage performance like ballet in the West.⁸

The traditional Kattakaryhan (narrator) and chorus of representative caste figures join in a series of failed attempts to tell and stage the devadasi story from 1856 to the 1920s, thereby endowing this muted historical figure with a kind of performative agency that is not autonomous but is at once collusive and subversive. The play incorporates the fate of the historical “body” of the devadasi, objectified and textualized under the disembodied gaze of alien eyes and reconfigured in the competing, multilingual discourses of colonialism, Orientalism, and indigenous nationalism. These discourses produced new heuristic frames for the “dancing girl” and reproduced her within the new technologies of print capitalism.

As part of the British inventory of Tanjore after the annexation of 1856, Capt. Linnaeus Tripe walked into the Brihadishwara temple and photographed its presiding deity, the ancient dancing Shiva, and his “vehicle,” the giant Nandi bull before whom the temple devadasis danced. The objectifying eye of the camera is a key trope in God Has Changed His Name, and 1856 the crucial year. The camera comes into the temple. They say “hey you, devadasi, put your body down. We want to take a photograph of your god.” It's a tragic moment in the life of the traditional storyteller and wife of god as the camera doubles the image of her divine husband. It is dristi, the evil eye. She sees historical signatures on her speaking body overwriting her marginal place as a subject and says: “Woman, the site of desire. Bayadere, narthaki, temple-prostitute, nautch girl, they gave me all these names. They wrote one name, they cancelled it, they wrote another again. They wrote in Tamil, Telugu, Sanskrit, English…. Am I a woman, a map, a page, what am I to you? What am I to me?” She carries the Orientalist Sanskrit legacy, the colonial English language legacy, and the vernacular language legacy on her semiotic body. “Where is the source? Multiple beginnings, talking alphabets, this is the source of my creation. Am I a woman, a temple, a page?…. Here is a body that is totally imprinted by history.”

The Ghost-Woman protests at the house of each Master of a discourse, but they do not respond to her, only to the kapita hasta mudra that the historical devadasi bore as a semiotic sign of her particular womanhood, and which the Masters hyphenate with their own designation of “bayadere,” or “nautch girl,” or “narthaki.”⁹

She leaves each house wearing its definition on her costume until there are three different figures on her own shriveled body. She beholds her transformation in the mirror of time. Unable to recognize herself, split between an “I” and the names she is called, she undertakes an epic journey, a padayatra of fifty-six years out of historical sight.

When, in the second act of the play, she emerges into the public discourse surrounding the 1890s anti-nautch movement, it is as a limping grotesque. She has become that monstrous body that manifests history, the body against which, Hayden White insists, the normal body must be defined, the normal being “a double negation, which is to say, a negation of whatever it is that the totality of monstrous bodies is presumed to lack” (233). Following White, it is only through the imagined history of the monstrous body, like the overwritten body of the Ghost-Woman devadasi in God Has Changed His Name, that a normal body, lacking what the other lacks, can be redefined. Only after the body of the devadasi was abjected could the Brahman girl take on a “purified” form of her dance.

In the West, while the history represented in God Has Changed His Name was unfolding, the Oriental dancer, whether Middle Eastern or Indian, was a pervasive, exotic image, and the seed for the twentieth-century search for the originary Indian dancer was being sown. Sir William (Oriental) Jones was not only a jurist and Sanskrit scholar but also a very popular poet. Through his journal Asiatick Researches, his translations, particularly of Kalidasa's Sanskrit play Sakuntala, and his own verse, he spread a broadly Neo-Platonic, and so comprehensible, vision of Indian culture and aesthetics that influenced the Romantic movement in England and continental Europe.

In dance as in painting, a fantasy Orient, featuring the roles of harem women and temple dancers was favored “not only because of the aura of glamour and intimations of promiscuity that surrounded them, but because of the extra piquancy the characters' status as chattels gave to their bravery and resourcefulness” (Jowitt 55). While western Orientalism in the arts was a self-referential tradition and the seed bed of modern dance, it also piqued periodic curiosity about the originals. In 1838 a troupe of Indian musicians and devadasi dancers from Tiruvendipura near Podicherry signed a contract to perform at the Variety Theater in Paris. In what was considered a theatrical coup at the time, Frederick Yates, the impresario of the Adelphi Theatre, brought them from Paris to London where they were a huge success, even giving a command performance for members of the royal family. (“The spectator is rapt in wonder and admiration at the extraordinary development of genius and elegance so lavishly bestowed by capricious nature on her heathen offspring” [Theatrical Register October 20, 1838].) The chief dancer, Amany, was celebrated by Theophile Gautier, who later wrote the libretto for a Sakuntala ballet choreographed for the Paris Opera in 1858 by Lucien Petipa ([Guest]).

After Marius Petipa in Moscow created The Talisman in 1889, a fantastic ballet featuring bayaderes and Hindu danseuses, he famously revived and revised Tagione's La Bayadere of 1830, which became the first great solo role of Anna Pavlova. “Throughout her life as a dancer [Pavlova] had been surrounded by varying suggestions of Indian dance,” but when she toured India in 1923, eager to see the real thing, such traditional dancing as had survived was already excised from the public sphere. She could only find the “real” India in the ancient carvings and (restored) frescoes of the Ajanta caves, and in contemporary marriage ceremonies. Her subsequent work with Uday Shankar gave their representation of Krishna and Rhada a Manipuri inflection, but she influenced the revival of traditional dance in India most directly by her encouragement of Rukmini Devi Arundale to abandon ballet and study the dance of her own country, the dance that, despite her fame, Pavalova had been unable to find (Money 313–25).

The Nautch Girl, or the Rajah of Chutneypore (1891), the first post-Gilbert and Sullivan comic opera at the Savoy, was an Indian reprise of The Mikado, including a dance travelogue concluding with the loosely orientalizing skirt dance as pioneered by Kate Vaughn and her followers. When actual South Asians appeared as part of ethnographic displays like the Indian Village put up by Liberty and Co. at the Albert Palace in 1885, the dancers seemed very tame compared to what the illustrated papers and the reports of missionaries had led the public to expect. They were not, in the parlance of the music hall, very “nautchy.”

Given the declining status of the dance in India after 1856, it is not surprising the most significant factor in the late-century idealization of Indian dance in England did not occur on stage. Rather it was the fashion for Oriental spirituality, in particular Theosophy, that created an idealized image of the Indian dancer. Annie Besant, who succeeded Madame Blavatsky as the head of the Theosophical movement, contended that the devadasis were once as chaste as nuns, but had degenerated. The theosophist poet Edwin Arnold, however, acknowledges the sexuality of Gunga, the vina playing “Nautchni” of his poem “In an Indian Temple” (1887) while making her an intellectual equal of the “Saheb” and the “Pandit” as they philosophize in verse (3–54).

In the 1890s a spate of dance poems followed in the wake of the fashion for painted and dancing Salomes and the swirling draperies of Loïe Fuller. In London the young Sarojini Naidu responded to her friend Arthur Symons's “Javanese Dancers” with her own poem “Indian Dancers” in which she tried, she told Edmond Gosse, to capture their voluptuous movement.¹⁰

But Naidu's dancer is an idealized figure dancing under the stars. Her “wail of desire” has no object (Naidu 39–40). It is not surprising that later, as a Congress Party activist, she was both one of the enthusiasts who welcomed Ruth St. Denis and her imitation temple dancers to India and a supporter of the anti-nautch movement hoping to see the dance transferred from the debased devadasi to the unsullied body of the middle-class woman.¹¹

Among western dancers it was St. Denis who first fixed on Indian rather than Middle Eastern dance as an aesthetic model to emulate. The young vaudeville skirt dancer from New Jersey was already a Christian Scientist and aware of the triumph of Swami Vivekananda at the Chicago World Parliament of Religions in 1893. Looking for a turn that would set her apart from the legion of fashionable belly dancers and Salomes, she was transfixed by the nautch dancers featured at the Durbar of Delhi, the spectacle that marked the grand reopening of Luna Park on Coney Island in 1904. She read Loti's India, sat at the feet of the Orientalist actor Edmund Russell as he read from Edwin Arnold, and researched India at the Astor Library. Encouraged by the small New York Indian community, she discovered Rhada, if not rasa. She was encouraged by support from a wealthy Parsee merchant family in New York. After a London performance, to her great delight, the Rani of Baroda told her she “acted like a high-class Indian woman” (55). Despite such adulation, her dances were not constructed on Indian aesthetic principles. As Suzanne Shelton notes, St. Denis's oriental dances were essentially art nouveau performances built from Delsarte exercises movements while incorporating some Indian gestures. They generally followed a spiritual pattern of descent into sensual temptations, resistance, and finally transcendence without embodying any specifically Hindu practices (49–58).

It is often said that when the Denishawn dancers finally went to India in 1925, they, like Pavlova before them, could not find an authentic temple dancer. That is not so. A genuine devadasi named Kamalambal danced for Shawn for several hours at a Madurai temple. Shawn said she was “technically the best dancer we saw in India,” but influenced by revivalists like Krishnaswamy Rao and reformers like Sarojini Naidu, he saw the erotic aspect of her devotion as part of the history of decline, a “dry rot.” He dismissed the argument of a Hindu pandit that the erotic themes in the dance should be interpreted like “The Song of Solomon” when it is read as expressing “The Love of the Church for Christ” (132–33).

Just as the moral condemnation of the nautch girl and the misrecognition of the devadasi under western eyes contributed to a process of Indian modernization that abjected the traditional public women and their artistic traditions while making public space for once-sequestered women, so the renewed western interest in eastern religious practices, revaluation of non-western dance traditions, searches for authentic native dancers, and attempts of dancers like Isadora Duncan to revive or reinvent lost classical traditions, contributed to the revival of Indian dance. Indian dance traditions not only had potential as positive and unifying symbols of the emerging nation on the international scene, Indian revivalists had something the West lacked, a descriptive classical text, the Natyashastra.

The dance revival is usually treated as a localized event within the discursive boundary of dance history and the national boundary of the emerging Indian nation. But as our narrative implies it is better understood in Arjun Appadurai's terms as “deterritorialized” and “transnational.”¹²

Its chief architect, Rukmini Devi Arundale, was not an insular revivalist, but an international figure whose perspective was both Indian and global. She was Brahmin by birth, the spiritual daughter of Annie Besant by adoption, the wife of Dr. George Arundale, third President of the Theosophical Society, and a leader in that international organization in her own right. She did not go to the traditional, male dance teachers in their own sites to learn the dance, but brought them into her cultural institution, Kalakshetra, and became herself a teacher, deterritorializing and finally displacing them. She reunited the idol, the traditional musicians and the dancer, but not in the temple, rather on the proscenium stage with the dancer again in the leading role.

While the male teacher, the nattuvanar, had traditionally stood behind the dancer keeping time, Devi honored the teacher, as she had the musicians, with a place on a fine rug at the edge of the stage, thus isolating the dancer as the focus of attention. The dancing body on the stage of Devi's Anna Pavlova Theatre had regained its primacy, but it was no longer a “hereditary” body, and her yearning for union with the divine when expressed in the erotic mode, the sringara rasa, was made more “refined.” The sadir or dassi attam of the devadasi was thus transformed into Bharatanayam, The Dance of India. The new name invoked the aura of the Natyashastra, associating the dance with the originary civilization postulated through colonial Sanskritization, but Devi did not, as has been argued, herself Sanskritize the practice of the dance because she still honored traditional methods of teaching and left room for innovation.¹³

She did not instruct according to the rules of Natyashastra. Rather, she used Sanskrit chanting to invoke the spirit of the text while retaining the traditional guru-shisya (teacher-pupil) method of teaching.

While Kalekshetra and the other institutions Rukmini Devi established on the grounds of the Theosopical Society at Adyar outside Madras have generally been understood in nationalist terms as local and indigenous, they were from the start part of a much broader program of cultural revival and purification in the arts, crafts, Tamil literature, and education. The program had no single method. It was carried out in the traditional forms of guru-shisya and guru-kul (teacher-community), in Theosophical and Montessori pedagogies, and even Sanskrit scholarship. Hers was at once an attempt to recover an idealized past by bridging the temporal abyss, and a transnational, avant-garde, aesthetic, cultural, educational, and public culture movement going well beyond dance history and the scope of this essay, which will conclude by returning to our title and looking at what happens to the famous question that concludes Yeats's “Among School Children” when East meets West.

“O body swayed to music, O brightening glance,/How can we know the dancer from the dance?” (214). At the time of its composition in the West in the 1920s, the question was rhetorical, and the image of the dancing body transcendent, even ecstatic: the perfected symbolist image with dance, dancer, and meaning indivisible. While our argument does not require that Yeats's generalized image have a specifically oriental origin, Kathleen Raine does postulate an Indian source for his linking of the cosmic images of tree and dance, citing Yeats's knowledge of the the fifth-century Sanskrit poet Kabir and of Indian classical dance. Indeed, images from the drafts of “Among School Children,” “blazing foot,” and “glittering glance,” recall the “glittering jewelry” of the bayaderes noted by Loti. To return to South India of the 1920s is to reverse the abstraction of the dancing image in western Orientalism that culminated in Yeats's famous image. There the dancer's “glance” was part of contested expressive vocabulary, and the question of dancer and dance not rhetorical, but an embodied, social, colonial/nationalist question with its terms reversed. The question about the devadasi and her tradition was not how to know the dancer from the dance, but was it possible to know the dance from the dancer? If so, if the devadasi's dance could be distinguished from her body and transferred to another's body, what was to become of those bodies?

In the liminal space of the dressing room, the modern dancer transforms herself from an individual into a vehicle for traditional stories, staring into the mirror, applying kohl to her eyes. The costumed dancer's last act before going on stage honors her instructional lineage with a numinous gesture that repeats in a secular context the worship of her ankle bells that linked hereditary dancer to her god. As she touches the bells to her eyes, there, for the blink of an eye, a ghost-woman looks back.