Dancing Through Laws: A History of Legal and Moral Regulation of Temple Dance in India

No abstract


Introduction
In 1947, in the state of Tamil Nadu in South India, an Act was passed, "The Tamil Nadu Devadasis (Prevention of Dedication) Act," which among other things banned the dancing of women in front of Hindu temples. The Act was to target prostitution among the so-called devadasis that were working as performers within and beyond Hindu temples, and who, according to custom also were ritually married or dedicated to temple gods. The Act was the culmination of decades of public and legal debates centred on devadasis, who had come to symbolize what was considered a degenerated position of women within Hindu society. Concurrent with this debate, the dance of the devadasis which had developed through centuries was revived and reconfigured among the Indian upper class; and eventually declared one of Indian national dances, called bharatanatyam (which can translate as Indian dance). Today, while parts of the devadasi tradition have been banned, bharatanatyam is a popular activity for young girls and women among the urban middle and upper classes in all parts of India.
The aim of this article is to examine moral boundaries tied to the female moving body in India. I do so by looking into the ways in which the regulation of a certain kind of dancers has framed the moral boundaries for contemporary young bharatanatyam dancers. A focus on legal and moral interventions in dance highlights the contested role of the female body in terms of gender roles, religious ideology, and moral economy. I argue that young bharatanatyam dancers deal with an ambiguity associated with the devadasis, as they act as both devotees and performers, for godly as well as human desires, motivated by tradition as well as independence. The article consists of two sections. In the first section, I give an overview of the history of the devadasis, in order to understand the background for the legal debates. This section also describes the concurrent revival of the dance of the devadasis in the form of bharatanatyam. The insight from this section has evolved out of the part of my research on dance, where I have focused specifically on 17th century temple women in Tamil Nadu and how foreigners encountered these women (Puri 2009(Puri , 2014. The second section of the article is based on fieldwork started in 2004 among dancers at a bharatanatyam dance institution in New Delhi. In this section, I focus on how the moral boundaries of the dance are negotiated through stage practice, personal narratives and visions relating to issues of gender, religion and economy.

A short history of the Devadasis
The history of the devadasis is a contested one. To begin with, it is questioned whether there is such thing as a history of the devadasis (Orr 2000, 5). The term devadasi is a Sanskrit version of the Tamil term tevaratiyal, which means a slave or servant of god. It was not until the 1920s that devadasi came to define a pan-Indian identity with a common heritage beyond local histories. This classification grew out of the legal debates targeted at the banning of women's ritual dedication to temples. Despite the fact that the devadasi identity is a constructed one, consisting of the history of multiple communities of performers and sex workers with ties to Hindu temples, I will provide a narrative of a development in the role of female dancers, based on existing in-depth research by Leslie Orr, Saskia Kersenboom-Story, Davesh Soneji as well as myself. This is in order to give the reader a sense of the historical and geographical contexts out of which the devadasi identity developed, which illustrates the ambiguous position associated with devadasi at the time the act was introduced. In between a devotee and a courtesan, the story I am going to tell starts around the 9th century in Hindu (at the time Agamic) temples situated in South India in the present day Tamil Nadu.
Based on research on temple inscriptions, according to American historian of religion Leslie Orr, during the Chola dynasty (approx. 850-1300), women could gain a special position inside the temples as devotees (Orr 2008). Women of low social status were able, through the giving of donations to the temples, to obtain the right to do services, such as cleaning, making garlands and taking care of the statues of the Hindu gods inside the temple. Also, a more prestigious task was to dance and sing in front of the statues of temple gods. This was the period in which the bhakti movement within Hinduism became more dominant in this area. The bhakti movement shortly described was a movement where salvation was sought not through priests, but by the establishment of a direct relationship with the gods. This relationship could for example be established by being seen by the gods and by serving them in their presence embodied in the statues. At the time, women could primarily gain status through marriage, as well as caste. However, these female donors and/or devotees through their gifts and ritual participation, gained social recognition for their honourable tasks and could also acquire land given by the temple patrons in return for their services. For widowers, divorcees and unmarried women from the lower castes, becoming a temple woman could thus be a way to climb the social and economic ladder (Orr 2000, 75). Dancing was a privilege that also enabled women of lesser means to dress as a wealthy woman or a goddess, carrying the expensive costumes owned by the temples, thereby improving their position in society.
Based on textual sources as well as the oral history of devadasis today, Dutch anthropologist Saskia Kersenboom-Story (Kersenboom-Story 1987), unfolds the development of the temple women where Orr left it, approx. 1300 and onwards. Being a temple woman at this time was not marked by her donations, but by a ritual through which she becomes married to the temple god, and also becomes a property, or at least some kind of permanent employee, of the temple. According to her, during the Vijayanagara Empire (1336-1646, also in present day Tamil Nadu) temple women had a number of roles, in addition to that described by Orr, which took the women out of the temple complex. At village ceremonies, such as weddings or naming ceremonies, they could protect the people from the divine forces of dangerous village gods. Also they protected the temple gods when their statues were taken outside for processions. Furthermore, they enhanced the god-like status of kings by dancing at their processions, at a time where the kings' power was maintained through their connection to the temple institutions (Kersenboom-Story 1987, 36). According to Kersenboom-Story, the temple women were present at all these different functions because they were considered auspicious, or "Nityasumangali" (Kersenboom-Story 1987, xix). In line with Amrit Srinivasanan (Srinivasan 1985), she emphasizes that these dancers were highly respected in the local communities (Kersenboom-Story 1987, xix).
During the time of the Nayak Dynasty of Thanjavur (1532-1673), temple dancers became increasingly connected to the royal court. The rulers of the Thanjavur Nayak Dynasty were known for their patronage of music and dance (Appadurai 1981, Asher andTalbot 2006), as well as for their harem. At this time there was a transmission of dancers between village temples, court temples, and the royal court in Thanjavur. Davesh Soneji, in his research has focused on the history of those women who were attached to the Thanjavur court from the 17th century and onwards (Soneji 2012). At this time, the ritual ceremony (khumbhahahrathy) in which they became married to gods did not mark a position of religious dedication, but was rather tying women to a sexual economy and courtesan lifestyle (Soneji 2012, 36). This was also the time when the dance which today is known as bharatanatyam was stylized. At the beginning of the nineteenth century, movements were choreographed and the 'Tanjore Quartet' composed music incorporating lyrics for dance at the Thanjavur court (Soneji 2012). The women were trained both in dance, music and in the languages of the lyrics. The dance consisted both of pure movements and of more narrative elements, where stories of the Hindu gods and goddesses where depicted. One of the dance pieces that were formatted at the time was the varnam, a central piece for today's performances, where devotional and erotic love to a god, often in the form of Krishna, is expressed. Kersenboom-Story argues that concurrent with the sexual economy, devadasis were further sacralised through the development of these devotional dance items, while disconnected from their more pragmatic ritual role in the rural communities (Kersenboom-Story 1987, 178). Soneji furthermore maintains that the devadasis tied to the Thanjavur court were in an ambiguous position. On one hand the dancers of this time possessed a degree of freedom from the norms of patriarchal society as they lived in quasi-matrilineal communities, had non-conjugal sexual relationships with upper-caste men, and were literate as opposed to most other women at the time. On the other hand they were commodities acting to fulfil male desires in the roles of concubines, mistresses and "second wives" (Soneji 2012, 3).
This sexual economy developed in a cosmopolitan setting of colonialism, where both the nayaks (kings) and their foreign guests at the court were enjoying their artistic and erotic services. Sources by foreigners encountering dancing women both at temples and in the courts show fascination as well as a judgement of a religious institution that takes part in the ritualization of a sexual economy (Puri 2009;2014, Bor 2010. When the British were exposed to the courtesans, they became known as the nautch girls. This is an anglicised version of the Sanskrit word for dance, natya, which nevertheless carries the connotations of harlots. The British saw these women, because of their ritualized connection to the Hindu gods, as proof of an inferior religion and a culture of loose morality. As India came under the British crown in 1857 a new phase in the British presence in India started. India was no longer simply a place of trade and enjoyment; India was to be governed. At this time the legal system became central for discussions on how to restructure society, not just through economic regulations, but also through the possibilities of removing the obstacles of "customs" to economic development and dignified lives (Birla 2009). One such obstacle was embodied in the devadasi dancer.

Dance legislation
Before 1857, the legal system was divided between Hindu, Muslim, and Christian law. Hindu law was based on Sanskrit texts, such as the Manusmriti, central to the upper-caste Brahmin interpretations of law, as well as customary law of specific castes or groups. However, after 1857, common law -and as part of this also criminal law under which issues of prostitution was to be dealt with -was established along with the apparatus of High Courts in the British presidencies. Issues that related to family such as inheritance, marriage, adoption etc., however, remained subject to Hindu, Muslim or Christian law. According to Jordan Kay, between the 1860s and 1880s when the cases concerning devadasis was first brought to High Court, it was the legal status of adoption in the community, which was in focus (Jordan 2003, 55). At the time unwanted girls from poor families were handed over to some devadasi communities. Devadasis, as per customary law, unlike other women of society, were not only allowed to adopt children as single women, but also to give their property on to daughters instead of sons. In these ways, the devadasis were not fitting into the social structure of the surrounding society. In strictly legal terms supported by custom, the devadasis enjoyed certain economic privileges that other women under Hindu, Muslim and Christian law did not. However, these rights also enabled more young girls to be integrated into communities involved in sex work.
These concerns, however, did not in the first place lead to any legislation. British officials maintained that dancing girls dedicated to temples and performing at private family celebrations were to be considered a custom and also that the devadasis was to be considered a distinct caste. In those presidencies where there were no records of devadasis (Delhi, Punjab, Central Province and North-West Province), they were not so afraid to legislate against the devadasi custom. But for those presidencies where there were records of devadasis, they were opposing a legal interference based on the argument that banning the practice could cause local discontent. It was argued that banning adoption among devadasis, would not be the right way to eliminate the prostitution of minors, which was already banned in criminal law of 1861 (Jordan 2003, 43-61). Rather than introducing legislation targeted specifically at the devadasi "customs", the British led government encouraged the Indian Westernized elite to stimulate public support for social change in these communities, through social reform movements (Jordan 2003, 65).
In 1892 the devadasi tradition was for the first time critiqued by the just established Madras Hindu Social Reform Association. The association was part of a larger project in which reforms relating to Hindu women were instigated, which also involved the formation of an Indian feminist movement (Srinivasan 1985, 140;Hubel 2010). They organized campaigns against devadasi customs, along with the practice of sati (burning of widows on their husband's funeral fire), and child marriages, as well as other practices that had earlier been condemned by Christian missionaries (Bates 2007). The reform movements brought devadasis back into the courtroom, and instigated the so-called 'anti-nautch movement' focused specifically on the devadasis. This movement consisted of several Brahmin but also non-Brahmin Hindus, British missionaries, journalists and doctors, all part of the middle and upper class in India. The anti-nautch movement sought to illegalize all dancing in temples as a way of targeting prostitution that could be linked to Hindu temples. It was argued that customary law was protecting local practices, which corrupted a "pure" form of Hinduism. One of the arguments raised in court was that the devadasi custom was not a Hindu religious custom, even though they were dedicated to the Hindu temples, and therefore they ought not to be protected by customary law (Jordan, 2003). While following a brahmanical influenced value system, the reform movement was also attempting at carving out a space for a more modern vision for India, where universal norms, yet defined from a Hindu ethos, ruled out the more "dirty" vernacular traditions. As part of the debate, questions were raised concerning what was to be considered "the real basis of religious life", and the values of chastity and purity were contrasted to "persons of loose character" (Jordan 2003, 83). In the debates, it was expressed that the devadasi tradition may have developed out of a dedication, but that this foundation had been corrupted. The devadasis thus became symbols of what could have been a unique Hindu temple practice, which nevertheless, some way along the line went wrong (Jordan 2003, 84).
In 1927, Muthulakshmi Reddi, the first female legislator (also part of the Women's India Association), proposed a bill, which was to illegalize females dedicated to temples: "This council recommends to the government to undertake legislation to put a stop to the practice of dedicating young girls or young women to Hindu temples which has generally resulted in exposing them to an immoral life" (Reddi 2010(Reddi [1927, 115).
Reddi emphasize the morality of these women, as a part of the problem (which we will see becomes a central issue for the further development of the dance of the devadasis). As a response, a number of devadasis founded an association that was to represent their often over-heard voice in the debate. In 1928 the association published a document, signed by eight women, which represents the first record of a devadasi statement. Up till then, the history of the devadasis had been written by non-devadasis. In the document they argue against the proposal, and among other things highlight that: These devadasi women emphasized that devadasis could not simply be considered one kind of occupational group working simply as prostitutes. They foregrounded their role as women doing religious services in and beyond temples, while also suggesting that illegalizing devadasis in temples would not necessarily help those women within their community who were sex workers. According to Judith Whitehead, the threat of the devadasi community not alone came from the reform movements, but also from within the community itself. The devadasis at the time had become increasingly polarized between those who were recruited to the community as sex workers and those trying to uphold its ties to dignified services (Whitehead 2001). Part of the problem of the Devadasi Act was that it took part in the construction of a singular devadasi identity, when in real life they covered a very diverse group of women.
In 1934 an Act was passed in the Bombay presidency, which illegalized the ritualized connection of devadasis to Hindu temples. As the devadasi tradition was one associated with South and East India, and not Bombay, the Act at the time had little impact. The devadasis remained a contested issue in the Madras Presidency, with Thanjavur and its surrounding area in present day Tamil Nadu being a central area for the devadasi institution. Finally in 1947, not long before the birth of Independent India, the "Madras Devadasi (Prevention of Dedication) Act of 1947" was passed in an area with a large number of those categorized as devadasis, building upon the Bombay Act. In the Act it is stated as follows (here I have stressed those sections which specifically involve dance): "An Act to prevent the dedication of women as devadasis in the Province of Madras: Whereas the practice still prevails in certain parts of the Province of Madras of dedicating woman as "devadasis" to Hindu deities, idols, objects of worship, temples and other religious institutions; And Whereas such practice, however ancient and pure in its origin, lead many of the women so dedicated to a life of prostitution; And whereas it is necessary to put an end to the practice; It is hereby enacted as follows (…) [Section 3, subsection](3) Dancing by a woman, with or without khumbhahahrathy [someone ritually married to god], in the precincts of any temple or other religious institutions, or in any procession of a Hindu deity, idol or object of worship installed in such a temple or institution or at any festival or ceremony held in respect of such a deity, idol or object of worship, is hereby declared unlawful (…) [Section 4,subsection](2) Any person having attained the age of sixteen years who dances in contravention of the revisions of Section 3, subsection (3), or who abets dancing in contravention of the said provisions, shall be punishable with simple imprisonment for a term which may extend to six months, or with a fine which may extend to five hundred rupees or both" (Soneji 2012, 235-36).
With this Act, the 'dedication' of females to temples was made illegal in order to target those involved in prostitution. But not only that, all dances by any female was made illegal around religious institutions and in the presence of Hindu deities. This meant that if the devadasi dance was to survive at all, it would have to be staged far from the temple settings -and perhaps it would even have to be danced by completely different women.

The Staging of Bharatanatyam
Parallel to the anti-nautch movement that had led to the above Act, there was another movement of people of the Madras elite; known as the 'revivalists', who advocated for a revival of the dance of the devadasis disassociated from prostitution (Schechner 1985;Srinivasan 1985;Meduri 1988;Gaston 1996;O'Shea 1998;Parker 1998;Chakravorty 2000;Jordan 2003;O'Shea 2007;Peterson and Soneji 2008). In line with the mood at the time leading up to Indian independence, both the anti-nautch movement and the revivalists were concerned with identifying an Indian cultural heritage and national identity. Arts had become of great political significance, since regional artistic traditions were primed as representation of a common Indian culture. In this context, dance became one of these signifiers of a national culture and identity, which came into the spotlight among the Indian elite, who took a leading role in defining the new India.
In 1935, at a stage in the Theosophical Society in Madras, Rukimini Devi was the first Brahmin woman who performed what was then advertised as nautch dance. This marked the beginning of what may be seen as a reinvention of the devadasi dance tradition (Schechner 1985, 69), which was to take the dance from the temples to the stage of auditoriums; from the devadasi-community to the cultural elite; and from Tamil Nadu to the whole of India. Rukimini Devi was a Brahmin woman from upper-class Madras, who as part of her overseas travels, had become fascinated not only with ballet, but also with the orientalist representation of Indian dances (such as those of Ruth St. Denis) (Allen 1997). Furthermore, Devi was inspired by the theosophical society and involved in the neo-Vedanta movement of the 19th century, which puts emphasis on universal concepts such as unity and Hindu spiritual practices above localized belief systems. In the years to come, Rukimini Devi established a dance institution where she taught non-hereditary dancers as well as teachers, she designed a sari-based dance costume; she systematized the dance and she gave the dance a new name: bharatanatyam (which, as mentioned, can be translated into "Indian dance"). She was inspired by the fine art of ballet (Allen 1997); but also she framed bharatanatyam to be a highly spiritual practice, which had unity with god as its ultimate goal.
During this development, dancers of different backgrounds debated what the dance was and should really be about, and disagreed especially concerning what the role of eroticism was in the dance. The female dancing body thus continued to be central to debates on proper Indian womanhood, though not in a legal but rather in an artistic setting, as different dancers communicated and danced their visions of a revitalised temple dance. Tanjore S. Balasaraswati (1918Balasaraswati ( -1984 was born in Chennai into a family with ancestors dancing at the Thanjavur court. Her first performance was at a temple site in Tamil Nadu in 1924. As the devadasi institution had lost its respect, neither Balasaraswati's mother nor grandmother had been dancing, yet her mother was trained in Carnatic music used for dance. In Chennai, however, Balasaraswati eventually entered the stage that was being set up for the revived bharatanatyam. Unlike most other devadasis, Balasaraswati had the resources to engage in a debate where she was challenging the image of the devadasi as an immoral victim. Mathew Allen has examined the opposing visions of Devi and Balasaraswati of the 1930s and 40s. The difference between Devi and Balasaraswati concerned the space for sensuality in the dance, and the connection between spirituality and eroticism. Whereas Devi was developing a dance, which with controlled movements and downplayed emotions was embodying respectability; Balasaraswati was emphasizing the importance of sensuality in the dance, as a central aspect of sringara-bhakti or devotional love part of the bhakti movement. The debate was simultaneously about the source, and thereby also ownership, of the dance. As a Brahmin, Devi emphasized the relevance of Sanskrit texts for the dance, and positioned the dance within Vedanta philosophy based on the Upanishad texts. Balasaraswati in contrast, as part of a devadasicommunity, emphasized its connection to the region of Tamil Nadu in particular the Thanjavur court (O'Shea 2007). Out of the debate developed different dance styles within bharatanatyam, thus Rukimini Devi has become an exponent of the Kalakshetra (also the name of her dance institute which means "temple of art") style of Bharatanatyam, and Balasaraswati represents the Thanjavur style. Nevertheless, it was the discourse of Devi, which came to frame the significance assigned to bharatanatyam at the stages around the art centres of India.
Both the legal debates of the anti-nautch movement working for juridical interventions in the devadasi institution, and the concern of the revivalists developing a new stage tradition for the dance, illustrate the way that moral lines were drawn around the female body in public space. These moral lines meant a polarisation that deemed lower caste dancing women immoral, while upper caste dancing woman were considered "pure" and idealized as exponents of "classical" culture. To maintain this moral divide between the proper and improper dancers, the morality of the female dancer remains in focus for bharatanatyam dancers today.

Moral boundaries of contemporary bharatanatyam dancers
As we reach the present and the second section of the paper, I move from a historical approach, to an anthropological one. Anthropologists of dance and performance emphasize how dance can both incorporate, communicate and challenge basic ethical codes of society, and deals with dance both as a performative practice communicating to soceity and as an embodied practice central to the experience of the individual (Royce 2004, Spencer 1985, Reed 1998, Kaeppler 2000, Cowhan 1990, Thomas 1993 In bharatanatyam moral boundaries tied to the female body and gender are thus encoded and negotiated (Hanna 1998). Apart from the formal rules of law that concerns dance highlighted above, there are also informal moral "laws" on how the female body ought to move. These two layers of regulation are interconnected, as I suggest that the moral codes framing the dance of contemporary young bhartanatyam dancers have developed out of the juridical debates conerning the interventions in the devadasi institution.
At the dance school, where I did fieldwork in 2004, more than 100 female dance students were enrolled between the age of 7 and 30. Most of the younger students came once a week for class after school, whereas a number of the slightly older students practised at the school daily. As part of my participant observation, I took dance classes, conducted interviews, attended performances, and spent time with the dancers in their homes and other contexts. I was particularly interested in the dance practice for those women who were between the ages of 16 and 25, who were dancing daily with hopes of taking up dance as part of their careers. The various fieldwork methods enabled me to explore the contradictions between these women as subjects, dancing for themselves; and as objects, aware of how others perceived their dance. In the following, I look into how the dancers face moral challenges in their dance. I do so by focusing on their perspective on devadasis, the interpretive framing and staging of the dance, their family relations, and financial background, which puts into focus questions that concern religion, gender and economy.
Overall, the dancers were aware of the history of the dance, as sketched out above. Yet when talking with the dancers, they insisted that the devadasis were foremost acting as devotees, and not as prostitutes. In fact, they did not clearly differentiate themselves from the devadasis. They claimed to be continuing the devadasi tradition and insisted on the respectability of their past. Some of the dancers had read the same historical texts as those referred to in this paper, such as the work of Kersenboom-Story, who emphasized the devadasis as auspicious and respected. What was interesting to notice was that the dancers in their imagination of the devadasis, often looked to the far past rather than from the 19th century into the present. Thus, I never heard mention of the ban during my fieldwork (and when I later asked a dancer who was also a law student, she too was unaware of the legislation). This is because the dancers are mainly engaged with an idealized past in which the devadasis lived and studied in the temple complex, a past in which the dancers found inspiration. Even though it was in the royal courts that the dance became refined as an art form, the dance students are more interested in searching for a connection to the past through the temples. This is part of their focus on the devotional identity of the devadasis, in contrast to the part of the history of the devadasis that involves sex work. For some of the dance students, the sexual liberty of the temple women was not simply encountered as a question of loose morality, but of personal freedom. Some of the dance students in their late teens and early twenties, several of whom had relationships prior to marriage without their parents knowledge, emphasised the freedom that the devadasis had both in relationship to dance and to men, as opposed to other women of their time. They also dreamt of a life where dance would be a source of income in such a way that they would not have to give up their dance practice after marriage, which was otherwise in the cards for the majority of dance students. They even fantasised about a life of dedication to dance only, beyond familial obligations, where they would have the freedom to develop erotic relationships with men beyond the family sphere. Even though this might be a glorified version of a devadasi past, they nevertheless, through their personal narratives challenged the idea of the devadasi as a mere prostitute. The devadasis are mainly sources of inspiration at a subjective level, as the connection between the dance and the devadasis is not emphasized or verbalized when the dance is staged.
Noticing how the dance is framed when on stage brings insight into the moral boundaries that are set up around the dance in public. During performances, Janet O'Shea has noticed that since the 1990s, it has become common practice to introduce the dance items in English, as a way to gear the performances more towards a Western audience (O'Shea 2003). The Western audience in Delhi are made up of expats or those in the city on official visits that are encouraged to witness part of 'India's cultural heritage'. Also, there are travellers who have come to know of performances through listings in the city's weekly guides, and sometimes one fourth of the audience at the bharatanatyam performances in Delhi are Westerners. The English presentations of the dance, I suggest, not only invites this audience to follow the narrative of the dance items, it also downplays the erotic aspects by setting up an interpretive framework that emphasizes spirituality. The audience is told that when the dancer shows us how she erotically longs (in all the limbs of her body) for Krishna, it is a representation of the longing for a union with god. Having seen many dance performances in Delhi over the years, I have noticed that there are often very similar explanations, even similar sentences used to express the significance of otherwise different dance items. These English dance translations have further solidified the kind of interpretive framing of the dance that was highlighted by Rukimini Devi, where the erotic aspects left in the dance are exclusively attributed to spiritual aspirations. In interviews with dancers, they told, with almost identical sentences as those presented on stage, that dance is a spiritual art, oriented at the unity with the divine. I am not indicating that this is not true, however, I am suggesting that there are many layers of significance of the dance for the dancers, several of which are not part of how the dance is framed on stage. During my fieldwork it took me a long time to get beyond these identical sentences, when asking about the significance of the dance for the dancers. However, when I experimented with using videos of their daily dance practices followed up by talks on their dance while watching the video together, I got closer to subjective experiences of dance. I came to know also from hanging out with the dancers, that for them, the dance was something spiritual, but it also represented opportunities for a career and was a way of exploring their own sexuality.
One dancer told me of how, when she was performing a dance where she was showing her devotion to Krishna, she was imagining her boyfriend, in order to portray authentic emotions (Puri 2011). In bharatanatyam, they can express parts of their sexuality in a public space without becoming too vulnerable, partly because of this framework that defines their sexuality as devotion. Religion thus provides them with a kind of symbolic, protective veil on the stage. The stage is marked as sacred space not only through language, but also through various symbols such as the statues of the Hindu gods, and practices such as the ways that those on stage move in-between the dance items (Puri 2011). Anne Marie Gaston has gone as far as to claim that the desire for the spirituality of bharatanatyam is actually a desire for social acceptance (Gaston 1996). I would rather suggest that there is space for many kinds of desires. But despite dancers' explorations of their sexuality through dance, where they for example take a dance item that involves Krishna to explore their fantasies about other males, many acted as strict moral judges in front of other dancers. When I attended dance performances with other dancers in the audience, I noticed how important it was for them that the stories were clearly represented as stories of devotional and not human love. If desires were shown in a 'too human' way they would judge them as 'vulgar' as opposed to the dancers who were 'subtle' and 'cute'. Furthermore, they would not only evaluate their relation to their dance but also their assumed intentions with the dance. Some would critique dancers whom they thought was motivated by fame and money, rather than respect for the art and spiritual development. Despite the differing narratives dependent on context, the dancers themselves took part in the discourse through which the dance is defined as a spiritual practice, rather than a form of entertainment, a career opportunity or personal enjoyment -especially when the dance is staged. As was the case in the 30s and 40s (with Devi and Balasaraswati, as the main performers) there was off course a difference of opinions, however, here I have stressed those perspectives, which clearly show how moral boundaries are set up around the dancers; boundaries with ties to the history of the dance.
Despite the clear representation of the dance as purely spiritual, the stage is an ambiguous space. On stage, as opposed to in dance class, the dance is not simply a form of cultural education, but a space where the women invite others to observe them in their moves. Dancers explained how they felt that people generally looked on them with a great deal of respect as dancers. The same people, however, would not want a performing dancer as a wife or daughter-in-law. It is not dancing as such that is considered against the moral code for women, but dancing in public. Bharatanatyam is mostly respected as a kind of cultural education for young females, but not as a performing career path. For most bharatanatyam dance students today, the dance is an expense rather than a source of income. The economy of the dance is highly patronised by state institutions as well as by parents, and most dance performances can be attended for free. The closest dancers get to a compensation are flowers for the dancer handed over on stage, yet most often it is also the dancers presenting gifts to her teacher/guru on stage, and sometimes also serves food for the audience. Dance is thus emphasised as a donation rather than a source of income tied to material needs, not unlike the first signs of temple women, described by Orr. Dancing today shows that one has the money to dance (tuition alone was 500 rupees a month), rather than the need to dance for money. However, as middle class women are increasingly expected to join the work force, these dancers dreamt of a career path that involved dancing. One exception to the familial problems dancers might be facing when desiring a career in dance, was if they wanted to work as a dance teacher. The bharatanatyam teacher makes money off stage in the company with other women, rather than on stage. One dance student explained to me that her boyfriend had told her that in case they got married, he would support her in opening a dance school where she could teach and perform. However, he would not allow her to dance outside her dance school. Thus the problem with dancing as a profession is not simply dance as a means of income, but dance as a means of income in a space where her sexuality is the object of the male gaze.
Today, Hindu temple complexes form the background of a number of dance festivals, where dancers from all of India come to perform the various classical dances including bharatanatyam, for example, at the Mamallapuram Dance Festival, organized by the Department of Tourism of the Government of Tamil Nadu. With these resourceful women dancing it seems that the Act of 1947, which bans dancing around temples, is not relevant, as it was really directed towards the devadasis. The bharatanatyam dancers I spoke to saw the temple setting as the ultimate setting for their dance, since they felt a heightened energy around the temples. Furthermore, some of these temples have images of female dancers carved into their walls, images from which present day dancers draw inspiration from for their own dance movements. This setting for the dance also takes part in framing the dance in a particular way that emphasis the dance as a spiritual practice with ties to the temple more than to the royal courts. In comparison, some dancers told me of experiences with dancing in other secular public places, where the stage had not been clearly demarcated. Here they had felt uncomfortable, as the audience was noisy and clapping, as opposed to the more demure audience of the temple dance festivals and in the national auditoriums. Here the ambiguity of the dancing woman, by many still associated with a loose character, became emphasized.
The above moral dilemmas mostly concern the dance as a performance practice. Performing dance is only a small part of the dance practice of these young women. For the most part, the dancers express how their daily or weekly dance practice empowers them both physically and emotionally. In dance, they feel a sense of freedom. Dancers thus simultaneously embody moral codes and explore individual freedoms through their dance.

Beyond the movements of bharatanatyam
In the above, I have showed how the ban against temple dance has had an effect on the foundations for bharatanatyam and its moral framework. But even more so, it has had an effect on the devadasis of various Dalit communities. Legislation involving the devadasis did not end at Indian Independence; however, it ceased to be a central political issue. The most recent update of the legislation was in the state of Maharashtra in 2005 "The Maharashtra Devadasi System (Abolition) Act, 2005". This Act focuses on the state's responsibility for offering rehabilitation schemes for devadasis. Furthermore, it pushes the devadasi towards the conventional marriage institution, as it states that co-habituation between a devadasi and a man "raise the presumption of legal and valid marriage subsisting between the two". Some of the problems of the devadasis, it seems, still concerns the difficulty of classifying them along dominant family patterns, as they in legal terms remain unmarried when married to god. Furthermore, the bans of dancing at temples have been left out from the most recent Act, perhaps leaving a legal space for bharatanatyam dance performances at some of the impressive temple complexes in Maharashtra.
It can be complicated to determine the effect the various stages of legislation have had on devadasis. One reason is that it is not that simple to figure out whom the devadasis are. Among those women working as sex workers, it is difficult to determine whether and when sex work can actually be linked to a hereditary devadasi institution tied to a temple ritual (Orchard 2007). Existing research shows that some devadasis express that the conditions of their community has worsened with the legislative interference, as they have become stigmatized and do not perform the same variety of roles within the temple complex as they did in the past (Soneji 2012). Unfortunately, I do not have the same kind of ethnographic material on present day devadasis. In that sense, my own study supports the bias of history and of dance scholarship where the loudest voice has been given to the cultural elite and not the devadasis themselves (Soneji 2012).
The question is whether the laws against dance in one context have actually carved out a space for dance to flourish in another context. Judging from the number of bharatanatyam dance institutions located in the larger cities around India, their number today perhaps exceeds that of the devadasis, and those communities that do work similar to that associated with devadasis. Whereas the law meant a marginalisation of devadasis, whose occupation was banned while they were not initially offered alternative life paths, the cultural coded part of the "laws" for bharatanatyam dancers has meant a specific framing of the dance as a spiritual practice beyond economic and erotic realms. The dancers nevertheless move in an ambiguous space, similarly to many of those who have become classified as devadasis, where they balance a role of independent erotic career-seeking women, with that of a dependent traditional religiously devoted woman -which continues to put them in the spotlight of moral negotiations.