Women's oppari can be a vehicle of their agency. As women perform oppari at the funeral, they express their fears and concerns, and in this way appeal for sympathy. Through oppari a woman can protest unfair social conditions. Women in rural Tamil Nadu live lives circumscribed by tight codes of modesty which usually prevent their voices from being heard. They are not allowed to negotiate their wages, and in Icaikurichi are paid around 60% of men's salaries for the same period of work in the rice fields. Although domestic work is also very difficult, they are often unappreciated for their long hours (see also Jeffery, Jeffery, and Lyon 1989:43).  

In Icaikurichi there are few recourses for domestic abuse. Under most circumstances, a woman must channel her complaints through her parents' household, and hope that they are taken up by her father, uncle, brother, or cousin. And to become a widow is to suffer even greater hardships. A widow is considered inauspicious and dangerous, and is therefore excluded from most ritual functions. In Icaikurichi, she is expected to wear white (i.e., uncolored) saris and not to adorn herself with jewelry or anything that might make her attractive or might draw attention to herself. She is usually not expected to remarry, although I was told there were exceptions among Kallar widows. Without a husband, she typically suffers a big drop in her household earning power and general standard of living. At a funeral, the widow of the deceased has a special opportunity to voice all these concerns and more through her oppari, to cry out about the woes of being suddenly thrust into widowhood. 

Remarkably, a women's oppari can express a grievance against someone of higher status. This is in part because, in many Tamil folk tales, a good, chaste woman who suffers is attributed a special moral status (Egnor [Trawick] 1980:16). Egnor [Trawick] (1986) finds that Tamil crying songs can successfully communicate grievances up the caste hierarchy, and reach the ears of wrong-doers of higher castes. Such grievances are often not explicit, but rather implicit, suggested through metaphors spun out in the performance. Crying songs, although considered intimate and private, may therefore be intended to be overheard. Funerals are semi-public events, and even "private" oppari at home is likely to be overheard since most homes in rural Tamil Nadu are open-air structures and sounds naturally carry beyond the walls of a home. Egnor [Trawick]'s findings are supported by my field data. Not only do the women at the funeral succeed in appealing for sympathy to those around them, but they also lay claims, through oppari, to the dead man's possessions. In their crying, several women mention in some detail specific interactions they had had with the deceased, in which he has shared or loaned various objects to them. These publicly-voiced cries can later help to substantiate their competing claims to the man's possessions. In general, the more poignantly one is able to express one's closeness to the deceased, the stronger one can lay claim his or her possessions. 

In contrast, in M.G.R.'s oppari I detect no evidence of an agenda to lay claim to the dead man's possessions, nor to protest wrong-doing, nor to ascribe special sympathy to one or another of the grievers. The full text of a re-performance of this oppari is given in the Appendix. Although no one present actually says so, I can see that, as he performs, he actively draws attention away from the grieving women. He is a very prominent, emotionally expressive stand-in for grieving women, eclipsing or at least marginalizing them. In this vein, it is worth pointing out that Dumont's informants at a Kallar funeral indicate that an important role of the professional musicians is to "keep the mind occupied, or distract it" (Dumont 1986:272). Perhaps also significantly, M.G.R. is hired by male relatives of the dead man, not female ones. Although no one explicitly says so, I find that M.G.R.'s role at the funeral, and the traditional roles of professional funeral musicians generally, involves shaping, guiding, or marginalizing the volatile emotions of grief, especially of women. This is not to say that M.G.R.'s performance is in some way "inauthentic," rather that it fulfills a traditional role, one of the results of which is that women's emotional expressions are further marginalized. 

Reynolds' research (1980) on Tamil women and widowhood are pertinent here. Reynolds finds that one reason Tamil women are so carefully controlled by men is that they are believed to be extremely powerful. Tamil daughters bring prosperity to their brothers and fathers. Tamil mothers and mother-like goddesses bring order to life. According to countless Tamil folk epics, a woman who is wronged, especially if she is wrongly killed, can become a deity and exact a terrific revenge. The power of a Tamil woman best serves the family and the community if she is controlled, both symbolically and in concrete ways, by a father or husband (Reynolds 1980:43, 46). An unmarried goddess is worshipped because, in the absence of male control, her female power, or shakti, makes her unpredictable, dangerous, but also potentially a fiercely protector of her devotees (Reynolds 1980:43-44). Such unmarried goddesses are among the most important deities in rural Tamil Hinduism, exemplified in the village of Icaikurichi by the fierce and very popular Muttumariamma.  

When a woman becomes a widow, male control and binding--both symbolic and actual--are broken. The tali, or marriage necklace, is removed, and a Tamil woman becomes unbound from male control. Like an unmarried goddess, a Tamil widow represents capriciousness, a lack of control (Reynolds 1980:36). At least one man who has exercised control over her is now absent. But unlike village goddesses, a widow is especially dangerous, because she is suspected of sinning, by not maintaining self-control in this or a previous life. Were she not a sinner, her husband might still be alive (Reynolds 1980:56; Wadley 1980:155). Thus, a widow's presence is considered inauspicious not because she is weak, but because she is dangerous. Following Wadley's research, one may theorize that one reason male control over women's oppari at this funeral was indirect--through the centralizing and showcasing of M.G.R.'s less dangerous oppari--may be that there are no other, more direct ways to control women at this cultural moment. A more direct approach might even be dangerous. 

Death 
Although not all opparis are about death, most are, and in rural Tamil Nadu, contact with death is something to be avoided. A shunning of death is evident in the fact that, after the immolation, all participants immediately made arrangements to leave the site, return home, and bathe. Bathing is an act of purification, believed to remove the association with death and with other undesirable substances or beings, and is therefore also an important part of preparations for sacred activities. Icaikurichi villagers preparing for pilgrimage to Sabari Malai, upon encountering death, are expected to bathe (Greene 1995:283-295). Whether or not the villagers actually conceive of death as a kind of "ritual pollution," the grievers respond to it in ways that are compatible with Madan's (1985:12-13) understanding of ritual pollution in Hindu culture. Madan describes pollution as a tarnishing of one's soul which can be corrected through acts of purification, such as bathing. In any case, the Paraiyars' association with death is the primary reason they are given the lowest social standing in the village of any caste community. It is the job of the Paraiyars to work with dead carcasses, process cow leather, and watch over cremation sites. As a result, they are separated from other caste communities through taboos of food sharing, intermarriage, and physical contact. Paraiyars may be understood as a kind of buffer between people of other caste communities and death. 

M.G.R. like many professional oppari performers, is of the Paraiyar community, a Harijan caste. Oppari performance, the tappu, and the Paraiyar community are therefore associated with death. Unlike other participants at the funeral, then, it is in some ways more difficult for him to shake off the association with death. His association with death is the source of his employment, but it also keeps his social standing--and that of his caste community--at a low level. This means he is in a double bind. He must genuinely perform and embody the emotions brought about by death, and he must perform the persona of a widow, because this is his job. His contact with death is highlighted in performance by the fact that, to perform oppari properly he must sing to the dead man, cry out to him, recounting so many things she (i.e., M.G.R.'s performed persona of a woman) had done for him. "I took such good care of you, but you died anyway!" is a common oppari outcry (compare to similar laments in North India noted by Henry [1988:103-104]). But at the same time he must attempt to distance himself, inasmuch as he can, from death and widowhood. Caught in this double bind, he cultivates a kind of professionalism that allows him to put some distance between himself and his own performance. This professionalism is taken up below and in the performance analysis in Section 4. 

Emotion 
Interviews in Icaikurichi lead me to the conclusion that oppari is looked down upon not only because of its association with death and with widowhood, but also because it involves or represents the expression of spontaneous emotion. Several days after the funeral, the son of the dead man explains to me, "At the funeral I must be strong. A man must show no emotion or sadness, even at a funeral of his father. Women cry at funerals because they are weak. But a man must not show weakness in public." He then admits, "When I am alone, perhaps then I can express my sadness and cry." (I have also heard that men are sometimes known to cry in the brief, informal period before the funeral when family members are beginning to arrive.)  

I find that many people in the village also make the distinction between "strong" and "weak" actions or emotional expressions, although villagers often disagree about how to classify any given action or expression. Brenneis (1990) finds a similar classification system of emotions in a community of Indians in Bhatgaon. He finds that some spontaneous personal emotions are considered appropriate only in solitary settings. Although the dead man's son does not say that M.G.R.'s performance is "weak" in this sense, a few others from higher caste communities in the village do: public emotional displays, whether deliberate or otherwise, are not considered fitting for a person of prestige. M.G.R.'s critics posit that a better-educated person simply would not express such intense and personal emotions in public. They also point out that M.G.R.'s performance is very physical and body-based, not lofty or spiritual, and not sufficiently self-controlled. 

The ideals voiced by M.G.R.'s critics echo those described in classical rasa theory, principles of aesthetics and emotional expression articulated in the great texts of the Indian high arts, and a part of high-caste education. Although it is debatable the extent to which rasa aesthetics are part of one's cultural upbringing in rural India, or among all but the social elite, several researchers, such as Appadurai (1990), Brenneis (1990), and the author (1995:79ff) have observed many cultural practices in everyday life that seem to embody rasa-like ideals, whether or not the practitioners were schooled in the classical rasa principles. According to rasa theory, emotion should be cultivated, or "built" (bhavan) through stylized gestures, which should be contemplated by a meditative audience. Emotional expression may become intense, but it should never be spontaneous. Ramanujan and Gerow describe the rasa ideal as follows: "In a play, what the actor acts is not the central mood of love or grief. He acts out the conditions that excite the mood and the responses that follow from it: he shudders or faints or sweats, he weeps and his voice cracks. The Indian theorists spelled this out in great detail, prescribing for each of the rasas the correlative consequents, the kinds of dramatic personae, the gestures and scenery and kinds of diction, thus analyzing content into forms" (1974:128). They further observe that "the emotion itself ... is never real; it can only be suggested. Paradoxically, any eruption of real emotion, which is by its nature grounded in individual awareness, would terminate the process of suggestion and therefore terminate the drama as well" (1974:133). This understanding of emotions seems to underlie critiques of women's oppari: grieving women are sometimes critiqued as expressing "real," personal emotions rather than performing the expressive trappings of emotions in specific, prescribed, stylized ways. According to M.G.R.'s critics, his own performance also fails to exhibit what could be called rasa ideals, because it resembles that of grieving women too closely.⁸ 

Therefore, the oppari performer, whether a non-professional woman or a professional male musician, is caught in a double bind by the problematic nature of spontaneous, intense emotions. On the one hand, it is essential to the effectiveness of oppari performance that it at least seem to be rooted in genuine emotion. Women's appeals for sympathy or protests of grievance must sound genuine to be taken seriously, at some level. Professional musicians must work especially hard at funerals to offer expressions that can be accepted by those present as appropriate to the event. If they do not succeed, they are reprimanded. Indeed, M.G.R. was reprimanded at one point during the funeral for deviating from this aim as he sang the pop folk song "Rasathi." On the other hand, these intense emotional expressions are taken as evidence of one's low social standing by some high-status villagers. 

In my analysis of emotions, I follow Abu-Lughod and Lutz in suggesting that the most fruitful ethnographic approach is not to examine whether emotional expressions are actually spontaneous or premeditated, actually rooted in emotion or style. Instead, I find it more fruitful to examine what various people say about the expression (Abu-Lughod and Lutz 1990:1). And I find that one of the ways that women's oppari is marginalized is through discourse that dismisses it as mere emotional outburst. Men's professional oppari, in contrast, is more frequently accepted in village discourse as a form of stylized, deliberate, and musical expression. This is not to say that professional oppari does not involve "real" emotion. But since it is very difficult and problematic to attempt to study the nature of inner, personal emotions, I focus analytic attention instead on the social dimension of emotional expressions, and on the cultural discourses surrounding them. It would be intriguing to know how M.G.R.'s inner emotions, as he moved himself to tears, were similar to or different from those of the grieving women, but this is beyond the scope of this article. 

Professionalism of Male Performers 
The tension between the contradictory goals of being accepted as genuine and as prestigious shapes the way professional oppari performers describe what they are doing, and the way oppari is performed. I ask M.G.R. how he is able to perform with such emotional conviction, and he responds by bringing up the concerns of being a "strong" performer, as the dead man's son had before. He strongly denies that his performative skill comes from grief experiences in his own life, probably in part because this would make him vulnerable to a discourse of "weakness," as described above. Instead, his skill comes from his musical experience: from his long apprenticeship, the many kinds of music he has learned to perform, and the many years he has performed as a professional musician in the area. To him, performing oppari is not primarily about being emotional, but rather a matter of musical craftsmanship.  

It is through the artistry of musical expression that he arrives at intense emotional expression, not the other way around. Through careful control of pitch, vocal inflection, and rhythm, he explores the various possible musical embodiments of sorrow, and weaves these sonic embodiments together into an unfolding musical form. Because he is a musician and a professional, he argues that his performance should be considered prestigious and "strong."⁹ In at least two other ethnographic accounts of Tamil funerals, professional musicians likewise claim that their craft requires training in specific musical skills (Dumont 1986:272 n28) as well as discipline, sacrifice, and years of study with the elders of their communities Moffatt (1979:197). 

Some other villagers confirm M.G.R.'s claims. Although M.G.R. embodies in his performance some of the outward characteristics of a "weak" person, of a person moved spontaneously by powerful emotional forces, he is not considered by many villagers to be performing from a position of emotional "weakness." In fact, one of the reasons he was hired was that, in his career as a professional musician, he has demonstrated an ability to control his emotions. Also, it is important that, although he knows the deceased man, he has not been too close to him during his life. In fact, I am told that such closeness would make him ineligible to perform at the funeral.  

In any case, the presence of musicians at a Tamil funeral is generally believed to be "indispensable" (Dumont 1986:272). Moffatt even finds in his account of a funeral near Madras [Chennai] that the role of Harijan (Paraiyar) musicians is so important that members of other caste communities tolerate the musicians as they recount a Paraiyar origin myth--in which Paraiyars have dominance over other castes--as part of the funeral ritual. It is said that, at the funeral "we must nod our heads to these Harijans" (Moffatt 1979:195-196). Further, the funeral is a rare opportunity for a Harijan musician to take on the persona of a person--a widow--of a higher status caste community than his own. Moffatt observes in Endavur that funeral musicians perform special rhythms at funerals of higher-caste people, suggesting that the substitution of a professional performer from one caste for a griever of another is reflected in a kind of musical "code-switching" (1979:200). 

Ironically, then, M.G.R.'s role is a source of income, prestige, and significant social importance, but it also puts him dangerously close to death, widowhood, and emotional spontaneity. Perhaps, much as the villagers shun death and are therefore dependent on Harijans to handle death for them, villagers may also refrain from publicly expressing emotion yet still wish their grief to be expressed. This is the double bind of the professional oppari performer: he derives social importance and income as he expresses the unexpressible, and he is also thought less of by some for doing so.¹⁰  

Moreover, the men who hire M.G.R. are also in this double bind, since they are responsible for public expressions that they consider to be characteristic of low status. As Tamils of different social standing and class position themselves differently with respect to the oppari professional, they are situating themselves vis a vis the dangerous associations of widowhood, death, and emotional spontaneity, which could damage their own social standing. It seems, although this warrants further ethnographic examination, that M.G.R.'s professionalism and the musicality of his performance seem more suspect to those of higher social status, who have more to lose from association with his oppari. 

Professional oppari performance, then, is situated in an area of contention in cultural discourse about which kinds of public expressions are to be considered "strong"--forms of reputable artistry--and which kinds of expression are to be considered "weak"--emotional indulgences. Although women and their opparis are more often classified as "weak," I find that they, too, are concerned about the deliberate patterning of their expressions. Women structure their opparis not only for cathartic release, but also for communication, to convince the listener of the importance of their messages, and to get the listener to see the world from their point of view (see also Egnor [Trawick] 1986:302).  

The whole genre of oppari, whether performed by male professionals or women, is situated in this area of contention. The project of identifying the distinctive musical features of the oppari genre, undertaken in the following section, must be informed by contradictory definitions and standards of artistry and good musicianship which are found in rural Tamil discourse surrounding oppari. More is at stake in this contentious discourse than the status of oppari as an expressive genre. Oppari performances, and the discourse surrounding them, have to do with the status of women, the reception of their message, and the status of Harijan professional musicians. At an "official" level, oppari strengthens and confirms social hierarchies because it evidences the emotional weakness of female and low-caste performers and connects these performers to death and to widowhood. At an unspoken level, oppari performances exhibit good musical craftsmanship and, in the case of many women performers, are exceptional vehicles of personal protest and empowerment. 

3. Musical and Expressive Features
of the Oppari Genre