1
Introduction: Gender Representation and the Problem of Language and Resistance in India

This book focuses on women's oral traditions and women's use of language in rural Uttar Pradesh and Rajasthan, northern India. We examine stories, ritual songs, personal narratives, and ordinary conversations from the villages of Pahansu, Hathchoya, and Ghatiyali and reflect on the ways in which these speech genres may be implicated in women's self-perceptions and self-fashioning, and the ways in which they may be understood as constituting a moral discourse in which gender and kinship identities are constructed, represented, negotiated, and contested in everyday life.

Several theoretical concerns inform our ethnographic analyses. At the most general level, we wish to position our arguments about gender in South Asia in such a way as to comment on current attempts to rethink the idea of culture in anthropology and in the social sciences, and attempts to understand the politics of representation in these disciplines. Second, we situate contemporary ethnography from western Uttar Pradesh and eastern Rajasthan in relation to the work of the Subaltern Studies historians, insofar as they are concerned with recovering the voices of those whose subjectivity and agency are generally obscured by most historical writing, and insofar as they offer theoretical perspectives on the interpretation of power and subaltern subjectivity. We share with those scholars an interest in what James Scott (1985) has called "everyday forms of resistance" to systems of ideological or material dominance. Our interests lie in exploring the "hidden transcripts" (Scott 1990) implicit in women's speech and song, the often veiled, but sometimes overt and public, words and actions through which women communicate their resistance to dominant

North Indian characterizations of "women's nature" (triya charitra ) and of kinship relationships. While our primary aim is to understand women's language—the words in which they construct and communicate alternative self-perceptions and alternative vantage points on their social world—we wish to comment also on the relationship between women's moral discourse and everyday resistance, on the one hand, and forms of power in North Indian social relationships, on the other. Third, we focus on the interrelationship between kinship and gender in northern India, specifically on the ways in which a consideration of women's voices and women's agency necessitates a rethinking of standard anthropological conceptualizations of marriage and patrilineality in South Asia. And finally, our consideration of women's speech genres speaks to broader issues of the relationship between language and gender, particularly those connected with the pragmatic aspects of language use, with speech play and verbal art as forms of discourse (Sherzer 1987a) and with women's communicative devices as loci of potentially subversive speech.

Culture and the Politics of Gender Representation in India

The anthropological objectification of a social practice often takes the form of positioning that practice within a single determinate discourse, a single interpretive frame, which becomes, then, a token of a coherent and totalizing "culture." Within the terms of such a positioning, culture is envisioned either as a "mode of thought" that "incarcerates" the native in a fixed and definite "way of thinking" (Appadurai 1988: 37-38) or as a set of "laws" or "rules" that transforms social chaos into social order, with the result that all human behavior becomes either unambiguously "normative" or "non-normative" within a specific cultural system (Bourdieu 2977: 1-15; Das 1989a: 310; Rosaldo 1989: 102). The construction of such an understanding of a particular social practice has a number of further consequences. Any interpretive strategy that views culture as a completed process, a coherent, bounded, and internally homogeneous "whole," tends also to view experience as given directly in that coherent culture, and thus the necessity of examining experience, and the voices of particular

thinking subjects, is rather dramatically obviated.^[1] Second, attempts to understand a social practice as the typification of an ordered and knowable cultural totality focus attention almost exclusively on what Clifford Geertz referred to as "scope," the degree to which the determinate meaning of a particular pattern of thought or social practice does indeed resonate throughout a cultural whole (1968: 111-12). Geertz contrasts the scope of a cultural pattern with its force, "the thoroughness with which such a pattern is internalized in the personalities of the individuals who adopt it, its centrality or marginality in their lives." As Renato Rosaldo points out, in privileging the wholeness and coherence of culture while denying the import of particular positioned subjectivities, anthropologists have not generally attended to the "force" of a particular way of thinking or a particular social practice. If, however, we follow Rosaldo in recasting the notion of force to encompass an attention to the positioned subject (1989: 225-26 n.1), and if we begin to view culture not as a single totalizing discourse but as a universe of discourse and practice in which competing discourses may contend with and play off each other (Kelly 1988: 41), compose ironic commentaries on or subvert one another, or reflexively interrogate a given text or tradition or power relation (Das 1989a: 312; Ramanujan 1989b), we might then begin to interpret experience and subjectivity not in terms of a single, incarcerating mode of thought, but in terms of multiply voiced, contextually shifting, and often strategically deployed readings of the social practices we seek to explicate.

In speaking of gender and oral traditions in South Asia, we find that it is of critical importance to stress the multiplicity of discursive fields within which social relationships are constructed, defined, and commented upon.^[2] The Indian woman has all too frequently been portrayed as a silent shadow, given in marriage

[1] The limitations of this conception of stable and seamless cultural "wholes" have been explicitly discussed from a number of vantage points in contemporary anthropological writing. See, for example, Appadurai 1986, 1988; Clifford 1988:63-64.

[2] For some of the many ethnographic examples of the interplay of multiple and sometimes competing discourses in South Asian social life, from a variety of theoretical perspectives, see Egnor 1986; Gold 1988a; Grima 1991; Holmberg 1989; March 1984; Oldenburg 1990; Prakash 1991; Raheja 1988b, 1993; Wadley n.d. (forthcoming).

by one patrilineal group to another, veiled and mute before affinal kinsmen, and unquestioningly accepting a single discourse that ratifies her own subordination and a negative view of female-ness and sexuality. Such a unitary representation of feminine passivity in India has a very long history. Colonial reports on the practice of sati , for example, often stress women's submissive, unquestioning obedience to the dictates of "religion," and their identity as passive bearers of a rigidly circumscribed "tradition" (Mani 1985, 1989). Such colonial documents also tend to infantilize women, in speaking of the widow as a "tender child," even though most satis were undertaken by women over the age of forty (Mani 1989: 97-98; Yang 1989). These representations of female passivity played a double role in British efforts to construct a moral justification for colonial rule in India. As Partha Chatterjee has pointed out, the representation of Indian women as voiceless and oppressed provided a rationale for British colonial intervention:

Apart from the characterization of the political condition of India preceding the British conquest as a state of anarchy, lawlessness and arbitrary despotism, a central element in the ideological justification of British colonial rule was the criticism of the "degenerate and barbaric" social customs of the Indian people, sanctioned, or so it was believed, by their religious tradition. Alongside the project of instituting orderly, lawful and rational procedures of governance, therefore, colonialism also saw itself as performing a "civilizing mission." In identifying this tradition as "degenerate and barbaric," colonialist critics invariably repeated a long list of atrocities perpetrated on Indian women, not so much by men or certain classes of men, but by an entire body of scriptural canons and ritual practices which, they said, by rationalizing such atrocities within a complete framework of religious doctrine, made them appear to perpetrators and sufferers alike as the necessary marks of right conduct. By assuming a position of sympathy with the unfree and oppressed womanhood of India, the colonial mind was able to transform this figure of the Indian woman into a sign of the inherently oppressive and unfree nature of the entire cultural tradition of a country.
(Chatterjee 1989: 622)

The "protection" of weak and passive Hindu women became, then, a strategy of colonial domination, and gender characterizations became vehicles for moral claims on the part of colonial

administrators, missionaries, and so forth (Mani 1989; O'Hanlon 1991).^[3]

Apart from attempts to legitimate the colonial enterprise as a civilizing mission that would secure a greater degree of freedom for Indian womanhood, nineteenth-century British colonial authors also, as Ashis Nandy (1983) persuasively argues, attempted to differentiate themselves maximally from the colonized by articulating an ethic of "hyper-masculinity" (involving aggression, control, competition, power, and so on) for the West, and characterizing India as the feminine antithesis, as radically "other." Embodying colonially devalued "feminine" qualities, Indian society was seen as unfit to rule itself and as morally inferior to the masculine West. Representations of women's silent submission to the dictates of religious tradition became, then, politically strategic metonymic tropes for the passivity of India as a whole.

Chatterjee has written of the way in which late nineteenth- and early twentieth-century nationalism responded to the colonial critique by constructing yet another discourse about women and Indian society, in terms of a redefined "classical tradition":

The social order . . . in which nationalists placed the new woman was contrasted not only with that of modern Western society; it was explicitly distinguished from the patriarchy of indigenous tradition, the same tradition that had been put on the dock by colonial interrogators. Sure enough, nationalism adopted several elements from tradition as marks of its native cultural identity, but this was now a "classicized" tradition—reformed, reconstructed, fortified against charges of barbarism and irrationality. . . . The new patriarchy was also sharply distinguished from the immediate social and cultural condition in which the majority of the people lived, for the "new" woman was quite the reverse of the "common" woman, who was coarse, vulgar, loud, quarrelsome, devoid of superior moral sense, sexually promiscuous, subjected to brutal physical oppression by males. It was precisely this degenerate condition of women which nationalism claimed it would reform.
(Chatterjee 1989: 627)

Sumanta Banerjee (1989) has analyzed the profound effects of this nationalist redefinition of the ideal woman on women's ex-

[3] See Suleri 1992:69-74 for an example of a similar portrayal of the pathos and vulnerability of Indian women, albeit with a different political agenda, in a play by Richard Sheridan first performed in England in 1799.

pressive traditions. In nineteenth-century Bengal, songs and other forms of women's popular culture were often critical of women's position in the society of the time. "Often stark and bitter in expressing the plight of women in a male-dominated society, the poems and songs popular among the lower social groups were, at the same time, tough, sensuous or bawdy, in an idiom specific to women" (Banerjee 1989: 131-32). From the mid-nineteenth century onward, Bengali men, influenced both by colonial education and by nationalist sentiment, attempted to arouse public opinion against these expressive genres, and there were concerted efforts to denigrate and suppress them as "corrupting," indecent, and unworthy of proper Hindu women.^[4] In the early twentieth century, the singing of such songs was viewed as a serious feminine shortcoming, and women's lack of formal education was cited as the cause of such moral fallings, in women's didactic literature of the time (Kumar 1991: 21).

As Chatterjee observed, Indian nationalism came to view women as the guardians and preservers of tradition.^[5] Ketu Katrak (1992) has argued that Gandhi further essentialized "tradition" and female sexuality through his appeals, in the nationalist cause, to the "female" virtues of sacrifice, purity, humility, and silent suffering that could be deployed in the service of the independence movement. His evocation of the mythological figures of Sita, Draupadi, and Savitri as exemplars of noble forbearance and contained sexuality led him to ignore, Katrak suggests, the more defiant and less passively submissive women of Indian history and legend who could equally have served as models of female identity.^[6]

The contours of a similar politically motivated deployment of a discourse concerning women, propriety, and "tradition" are evident in Julia Leslie's discussion (1989) of the Sanskrit text Stridharmapaddhati . Written by Tryambakayajvan in an eighteenth-cen-

[4]   Tharu and Lalita (1993:187-90) provide translations of the songs of two female folksingers of this period.

[5] This continues frequently to be the case in contemporary India. Dulali Nag (1990) has vividly drawn the contours of such a discourse on woman and "tradition" as it is deployed to market saris to middle- and upper-class Bengali women in advertisements in popular women's magazines.

[6] The Gandhian legacy for women's movements in India is vastly more complex than this brief survey can comprise. The issue is further debated, as Katrak points out, in Jayawardana 1986 and Kishwar 1985.

tury Maratha court, the text outlines the duties and dispositions appropriate to the pativrata , the ideal wife who is devoted to and subordinate to her husband. Leslie suggests that the author's defense of an orthodox Hindu tradition concerning women represents a response to the challenges to that tradition posed by growing Muslim domination, by Christian missionary influence, and by the increasingly popular devotional movements (bhakti ) that claimed, at least in their poetic traditions, that social distinctions between men and women were religiously insignificant (4). Leslie also points out that the polygamous milieu of the Maratha court presented certain political problems when queens and mistresses became involved in succession disputes and court intrigues. In such a milieu, she writes, "a work prescribing the proper behaviour of women might well have appeared to both kings and ministers to be a project of vital importance" (20). Thus, although the Stridharmapaddhati presents itself as a treatise on the inherent nature of women grounded in the timelessness of the dharmasastrik texts, the internal and external political compulsions for its composition suggest to us some of the ways in which particular representations of gender are deployed for particular purposes in particular historical contexts (see Chakravarti 1991).

Further, the very existence of a text like the Stridharmapaddhati indicates, perhaps, the existence of contrary discourses on gender and "women's nature." As Uma Chakravarti writes,

A close look at the Stridharmapaddhati indicates that the powerful model of the pativrata thus plays a crucial role in the "taming" of women. Once internalised by them it also makes them complicit in their own subordination. Ultimate social control is effectively and imperceptibly achieved when the subordinated not only accept their condition but consider it a mark of distinction. What the eighteenth century Stridharmapaddhati also unwittingly indicates is that not all women at all times accepted their condition nor considered it a mark of distinction: Hence the need for repeated reiteration of the duties of women, including the exhortation to women to mount the husband's pyre. The Stridharmapaddhati was a complete manual on the way women "ought" to behave, written in order to counteract the potential or actual "recalcitrance" of women.
(Chakravarti 1991: 185)

Chakravarti's commentary here adumbrates an important aspect of the argument we attempt to make in these pages: that

a discourse of gender and kinship found in certain admittedly authoritative texts and practices does not by any means exhaust or fully define women's subjectivities in South Asia. This seemingly obvious point has not often found its way into writing on the lives of ordinary women in South Asia.

May not other such discourses, whether indigenous or composed by outside observers, be strategically and politically motivated representations that grossly oversimplify and misinterpret women's consciousness? We do not wish in any way to minimize the difficulties and inequities experienced by women in northern India today and in the past; we wish only to highlight the fact that there are no women's voices in many of these representations (Spivak 1985a: 122), and the fact that, as Lata Mani has pointed out (1989: 90), these colonial, nationalist, and "traditional" discourses, and some similar contemporary Indian political discourses as well (Pathak and Rajan 1989), are not primarily about women; they are, rather, political commentaries on the authenticity and moral worth of a tradition.

Chandra Mohanty (1984) has written of the tendency, within Western feminist writing, to define women of the "third world" as victims of male control and of an unchanging "tradition," as unresisting objects in relation to ahistorical and uncontextualized images of "the veiled woman," "the obedient wife," and so forth. Such representations, she argues, define and maintain postcolonial relations between the first and third worlds by positing, implicitly or explicitly, the moral superiority of the West and the moral degradation of the "patriarchal" third world.^[7]

We cannot of course claim to have extricated ourselves entirely from these colonial and postcolonial ways of thinking. We are painfully aware of the perils involved in presuming to speak with any authority about women whose lives are so different from our own. Edward Said (1979) exhibits a quote from Marx as an ani-

[7] It is not only representations of gender that are implicated in postcolonialism; the often harsh realities of women's economic and political situations in India are of course produced in the context of postcolonial national politics and the global political economy. Mies (1982), for example, analyzes the location of Indian housewives' lace-making work within the contemporary global economy, and Spivak (1992) explicates Mahasweta Devi's short story "Douloti the Bountiful" as a critique of the production of women as bonded laborers and bonded prostitutes within the caste and class dominations of postcolonial Indian capitalism.

madversary epigraph to his Orientalism : "They cannot represent themselves; they must be represented."^[8] It is precisely because of our conviction that Indian women can indeed represent themselves that we struggle to translate their words in these pages. Said explicitly castigates those totalizing and essentializing Orientalist discourses that privilege historically unchanging and univocal characterizations of social life in which human agency is radically deemphasized and a mute submission to "tradition" assumed. In attending to the multiple perspectives, shifting purposes, and reflexive and ironic commentaries evident in North Indian women's songs, stories, personal narratives, and everyday talk, it is our hope that the authority of such normalizing and essentialist discourses may begin to disintegrate as we come to understand both the heterogeneity and the resistance evident in women's speech. We do not intend, of course, to speak with any presumption of closure or totality with respect to the subjectivity of the North Indian women we know. When Indian women represent themselves in their own words, no single unitary voice is heard; we have only begun to listen to a few of these voices.

Trawick has recently commented that representations of Indian women as "repressed" and "submissive" are but half-truths (1990: 5). They are half-truths in the sense that they may not define the only discourse of selfhood and feminine identity available to women in India. Many such scholarly representations have involved assumptions concerning Indian women's "ideological self-abasement" (Spivak 1985a: 129), that is, the degree to which they have assimilated "traditional" devaluations of womanhood and female sexuality and values stressing women's submission to male kinsmen into their conceptions of selfhood. Thus Sudhir Kakar has argued that the identity and self-perceptions of Hindu women depend heavily on a set of male-authored mythic themes concerning the ideal woman condensed in the figure of Sita, the virtuous and faithful wife of Rama in the Ramayana . Kakar perceives a "formidable consensus" in India as a whole, rural and

[8] For discussions of some of the specific guises Orientalist discourse assumes as it constructs its representations of Indian society, see Cohn 1968, 1984, 1985; Inden 1986a, 1986b, 1990; Mani 1985; Pinney 1989; Prakash 1990; Spivak 1985b; Suleri 1992.

urban, "traditional" and modern, in folklore and in myth, concerning the image of the ideal woman that Sita represents:

The ideal of womanhood incorporated by Sita is one of chastity, purity, gentle tenderness and a singular faithfulness which cannot be destroyed or even disturbed by her husband's rejections, slights or thoughtlessness. . . . The moral is the familiar one: "Whether treated well or ill a wife should never indulge in ire."
(Kakar 1978: 66)

Thus male psychological development, which is in fact Kakar's central focus, is profoundly affected by the mother's "aggressive, destructive impulses" or emotional claims directed toward her son as a result of her inability to challenge the ideals represented by Sita and to demand intimacy and a recognition of her worth as a woman from her husband (87-92). In Kakar's view, women either uphold a univocal normative order or deviate from it; there seems to be little recognition of a multiplicity of culturally valued strategies or perspectives for constructing selfhood and moral discourse.

A similar set of representations often defines the relationship between gender and sexuality in anthropological writing on South Asia, and a similar set of assumptions about women's subjectivity is often entailed. Both Sanskrit texts and vernacular oral traditions contain positive images of women as mother and as ritual partner and exhibit disdain and reproach for women as wives who are seen as sexually treacherous, sexually voracious, and polluting by virtue of their association with menstruation and childbearing (Bennett 1983; Kakar 1990: 17-20; Vatuk and Vatuk 1979b). These attitudes have undeniably serious implications both for women's self-perceptions and for the material and ideological relations of power in which they live their lives. But to assume that such characterizations define the limits of women's self-understandings and moral discourse is to ignore or silence meanings that are voiced in ritual songs and stories from Pahansu and Ghatiyali and in gestures and metamessages in ordinary language throughout northern India (Das 1988: 198), the existence of which indicates "the inadequacy of official kinship norms to give an exhaustive and definitive understanding of the sexuality of women" (202). If Indian women do unequivocally internalize certain ad-

mittedly prevalent South Asian views of female sexuality found both in texts and ordinary talk, as Patricia Jeffery, Roger Jeffery, and Andrew Lyon (1989) seem to assert, we would be justified in speaking of an ideological self-abasement. When, however, there are contextually shifting moral perspectives within a set of cultural traditions, the question of the relative force, the relative salience, and the relative persuasiveness of these discursive forms is of critical significance in the anthropological representation of the ideologies of gender and selfhood held by women and men. In the expressive traditions that we examine in chapters 2 and 4, women confront and voice their rejection of those devaluations of female sexuality, and they begin to comment critically on their implications for men's control over women's bodies and women's lives.

Characterizations of South Asian women as repressed and submissive are also half-truths in the sense that, at times, submission and silence may be conscious strategies of self-representation deployed when it is expedient to do so, before particular audiences and in particular contexts. There may often, in other words, be something of a discontinuity, a schism, between conventional representations and practices, on the one hand, and experience, on the other; certain discourses may be invoked and employed in particular circumstances, without exhausting the explanatory and evaluative possibilities within a given way of life (Jackson 1982: 30-31). But to say that a particular practice or a particular way of speaking is strategically deployed need not imply that it is motivated by individual sef-interest (de Certeau 1984: xi), that it is directed only toward political or economic ends, as Pierre Bourdieu's use of the term (1977) would suggest, or that it is disconnected from a set of culturally informed moral valuations. In his characterization of poetic discourse as "strategic," Kenneth Burke writes that literary texts may be thought of as adopting "various strategies for the encompassing of situations. These strategies size up the situations, name their structure and outstanding ingredients, and name them in a way that contains an attitude toward them" (1973:1).^[9] Burke suggests that the strategic aspects of poetic

[9] My understanding of the relevance to anthropology of Burke's theoretical position owes much to conversations with Richard Herrell and to his work on gender identity in the United States (Herrell 1992).

discourse might be understood as exhibiting several of the properties of proverbs. A proverb takes its meaning not only from its overt imagery, its semantic content, the shared understandings it presupposes, but also from the context in which it is deployed, and the communicative functions it fulfills within a specific speech situation (Briggs 1985; Gossen 1973; Raheja 1993; Seitel 1977). Burke writes that "proverbs are strategies for dealing with situations " and that strategy here should be understood as an attitude, an evaluation, a perspective on a situation (1973: 296). Burke's critical writing continually stresses the multiplicity of strategies, of attitudes toward recurrent social situations that are created in literary discourse, paralleling the manner in which a proverbial utterance, in Hindi as in English, may be countered with another equally compelling proverb that evaluates the situation at hand in radically different moral terms. Like proverbs, the larger discursive forms we analyze here are invoked in particular circumstances and by particular positioned actors, as strategies in the construction of selfhood and relationship, gender and kinship, in Pahansu and Ghatiyali.^[10]

In the preface to Women, Androgynes, and Other Mythical Beasts (1980), Wendy Doniger O'Flaherty writes that although her book is about images of women in the Hindu tradition, the relevant texts have all been composed by men. "If women composed their own mythologies," she goes on to say, "we do not have them." It is just this sort of image of the silent Indian woman that allows Kakar and Gayatri Chakravorty Spivak to see her as a passive assimilator of a one-dimensional and monologic set of cultural premises and moral perceptions.

In analyzing women's oral traditions and women's commentaries on them, it soon became apparent to us that women have in fact composed their own mythologies. Though they may not carry the authoritative weight of certain versions of the Ramayana ,^[11] women's songs and stories consistently compose ironic

[10] De Certeau (1984:18-21) has also pointed out the analogies between proverbial enunciations and ways of using, manipulating, and recreating systems of meaning and valuation.

[11] In fact, the many different Ramayana texts and oral renditions are not unanimous in depicting Sita as totally submissive to male authority. Variation on this point exists among the texts composed and performed by males (Lutgendorf 1991), and even more so in the Ramayana texts, in the form of song traditions, composed by women (Narayana Rao 1991). A sixteenth-century version of the epic composed in Telegu by the female poet Atukuri Molla also appears to speak of the vitality and strength of Sita more consistently than the male-authored renditions (Tharu and Lalita 1991:94-98). This heterogeneity of "tradition," and the alternative visions of female virtue to be found in these many Ramayanas , were ignored in the tremendously popular televised serial production of the Ramayana aired in India in 1987-88. Many Indian feminists have protested against the images of female subordination and female passivity depicted in the televised epic.

and subversive commentaries on the representations of gender and kinship roles found in the epic texts, in male folklore genres, and in a good deal of everyday talk. This is not to suggest that women always and everywhere subvert North Indian kinship structures or discourses that place them in a subordinate position, or that they are equally inclined or empowered to do so. Rather, our work points to the polyvalent nature of women's discourse, and to the multiple moral perspectives encoded therein. When one listens to women in rural North India, Kakar's "formidable consensus" dissolves into a plurality of voices. In this book we attempt to discover how these varying interpretations of gender and kinship are situated in women's complexly figured identities, and how they shape the tenor of their everyday lives.

Gender and the Subaltern Voice: Resistance, Acquiescence, and Women's Subjectivity

The work of the Subaltern Studies historians has raised a series of critical questions concerning the existence of subtle modes of resistance to hegemonic forms of social and political hierarchy in South Asia. Their investigations focus on recovering the subaltern voice that has not generally been recognized or represented in historical writing. Although the six volumes of essays edited by Ranajit Guha and published under the title Subaltern Studies are concerned primarily with caste and class subalternity and resistance to colonial rule, several of the papers address the issue of gender and women's subjectivity from this historical perspective (e.g., Das 1989a; Guha 1987). These essays are subtle and provocative analyses. Yet several methodological and theoretical exigencies become apparent in these attempts to represent a gendered subaltern perspective.

The first is simply the inherent difficulty of retrieving the sub-

altern woman's voice from the South Asian historical archive, wherein there was little concern to preserve a record of such voices, the voices of ordinary and often illiterate women. Indeed, insofar as they did not speak in a voice agreeable to powerful males, women's voices were often quite literally erased from the historical record. In the early years of the twentieth century, Lt. Col. Charles Eckford Luard, an administrator serving in Central India, commissioned Indians serving under him to record women's songs "from the lips of the local people." In the margins of one of these handwritten records in the India Office Library, I found the following notation, made by an Indian serving under Luard: "This is quite obscene. I have therefore used pencil that this should be struck out if considered befitting" (Luard n.d.: 160). But, as we shall see in chapter 2, it is precisely through such supposedly "obscene" language that women voice potent critiques of prevailing gender ideologies. By shifting attention from the historical archive to ethnographic inquiry, then, we hope to reposition some of the theoretical and epistemological concerns of the Subaltern Studies scholars. We hope to grasp the relation between hegemonic discourses on kinship and gender, on the one hand, and women's subjectivity and agency, on the other, within the context of lives we have come to share, however marginally and intermittently, in Pahansu, Hathchoya, and Ghatiyali, and with reference to the language these particular women speak, a language we have struggled to recognize and translate in these pages, a language that is alive and not yet erased from all memory.

Such a repositioning is almost essential methodologically, given the limitations of the historical record, but it is also significant from a second, theoretical, standpoint. Spivak (1985c), Rosalind O'Hanlon (1988), and Veena Das (1989a) discuss a number of critical epistemological issues embedded in the subalternist project. Das, for example, points out that the term subaltern cannot refer to distinct and well-defined social groupings; rather, the term should be used with reference to certain kinds of perspectives on a cultural discourse or a set of social relationships (1989a: 324), perspectives that may employ reflexive devices to interrogate or subvert that dominant discourse (312) but that may not necessarily coalesce into a dosed, unified, discrete, and

knowable totality. In offering her critique of an essentialist view of subjectivity, Spivak has raised several caveats concerning the tendency of the Subaltern Studies scholars to posit a unified, consistent, and pristine subaltern consciousness, the tendency to create self-determination and continuity out of what may actually be heterogeneous determinations and discontinuous discourse (1985c: 10-15). Similarly, O'Hanlon suggests that the attempt to recover and represent the subjectivity of the subaltern and his resistance to hegemonic discourse has been predicated on an insufficiently problematized notion of "the self-originating, self-determining individual, who is at once a subject in his possession of a sovereign consciousness whose defining quality is reason, and an agent in his power of freedom" (1988: 191).^[12] She argues that "the demand for a spectacular demonstration of the subaltern's independent will and self-determining power" has meant that the continuities between hegemonic discourse and subaltern culture have generally been ignored (213). She goes on to suggest that Guha (1983: 13), in speaking of the "sovereignty," "consistency," and "logic" of subaltern consciousness, has inadequately documented the limits of resistance, and the fact that the subaltern may at times speak from within the dominant discourse and at times stand outside and comment critically upon it (O'Hanlon 1988: 203, 219).^[13]

[12] These observations on the difficulties inherent in attempts to recover a subaltern subjectivity are analogous to those considered by de Lauretis (1984, 1986) with respect to the understanding of female subjectivity. The central problem is to articulate an understanding of subjectivity and experience that avoids, on the one hand, a totalizing and essentialist reading that sees subjectivity as determined by gender and, on the other hand, an overemphasis on free, rational intentionality. We share with de Lauretis a view of subjectivity as "a fluid interaction in constant motion and open to alteration by self-analyzing practice" (Alcoff 1988:425). The particular appropriateness of such a formulation for the understanding of South Asian selfhood and emotion is brilliantly illustrated by Trawick (1990).

[13] But Guha has more recently, following Gramsci, written of the contradictory and fragmented nature of subaltern consciousness (Guha 1989). On this point, see also Arnold's discussion of the relevance of Gramsci's view of the contradictory nature of subaltern culture to peasant society in India (1984).
In a recent publication (1991), O'Hanlon has written explicitly of women's resistance to male-authored gender ideologies in nineteenth-century colonial India, and she has made the point that this resistance both critiques and perpetuates the dominant discourse. In the text she examines, a book written by Tarabai Shinde in 1882, she finds that this female author at times simply reverses essentializing male characterizations of female nature, instead of critiquing the very structure of patriarchy itself. I do not find this to be the case in the women's songs we consider in this volume. Rather, women contest the very notion of essential natures and the idea that domination arises from, say, the viciousness of male nature rather than from intricate structures of power that shape both male and female sensibilities. I have elsewhere made this argument explicitly and at greater length (Raheja 1991).

Our ethnographic inquiries situate themselves within a similar set of epistemological concerns. We concur with Das's point that a subaltern perspective cannot be inextricably linked with concrete and invariable categories of persons, and with Spivak's and O'Hanlon's observations on the possible disunity and heterogeneity of subaltern subjectivity. The commentaries on kinship and gender evinced in women's oral traditions are dramatically different from the perspectives most frequently found in textual traditions and folklore genres performed by and for males. Yet, in everyday talk, women frequently speak from within the dominant discourse, and men may speak in terms set by the more muted discourse associated with women's speech genres.^[14] It is precisely because men and women may shift from one perspective to another that we speak of strategic deployments of these varying discourses. Women may tend to use gender characterizations and valuations differently from men, but we do not find a unitary female voice opposed in all respects to a male discourse. Yet, in listening to women's voices in song, in story, and in ordinary talk, we become aware that there is a subversive discourse more likely to be invoked by women than by men. Our repositioning of the concerns of the Subaltern Studies scholars within an ethnographic inquiry permits us, then, not to "recover" a unitary subaltern voice unrecorded in the historical archive; it allows us, rather, to begin to recognize the discontinuity, the interpenetration of the hegemonic and the subversive, and their varied deployments, from moment to moment in everyday life. It permits us to ask questions concerning contextual shifts in meaning and value, sometimes indexed only by an ironic tone, a gesture, a pattern of rhyme in a wedding song. Sociality often finds its meaning in such evanescent subtlety, the embellishments and improvisations that provide the ground for creativity and resistance and

[14] On the distinction between dominant and muted discourses and its relationship to issues of gender, see Ardener 1975. In these chapters, we use the word muted not to describe women's supposed silences but to allude to the fact that they may often be constrained from speaking overtly or publicly of their resistance to dominant ideologies.

that can only rarely be recovered from the historical archive. Ethnography is perhaps more likely than history to bear witness to such fluid and contextual creativity on the part of women who have all too often been denied the powers of literacy, and whose words have been erased from historical records when they proved unpalatable to those who did the writing. If in this book we direct our attention more to the resistance exhibited in expressive traditions than to the power relations that place women in subordinate positions, it is because we wish at first to counter colonial, and some anthropological and feminist, assumptions about the passivity of Indian women, just as the Subaltern Studies historians have generally had as their first priority the study of subaltern resistance, in order to question pervasive assumptions about peasant inertia, the internalization of structures of domination, and the "harmony" of Indian village society (Arnold 1984: 175).

Kinship and Gender

As anthropologists have come to reflect seriously on Levi-Strauss's observation, which he comes to only in the last paragraphs of The Elementary Structures of Kinship , that "woman could never become just a sign and nothing more, since even in a man's world she is still a person, and since insofar as she is defined as a sign she must be recognized as a generator of signs" (1969: 496),^[15] we have become increasingly aware of the limitations of a perspective on kinship that attends only to the trails left by women as they move from natal home to conjugal home, weaving alliances between groups of male kinsmen (Das 1989b: 276; Kolenda 1984;

[15] This isolated observation is of course at variance with Levi-Strauss's structural approach to the study of kinship systems, wherein systems of relationship are understood to construct themselves, as it were, behind the backs of actors, and wherein there is little concern with native experience and native theories of experience (Bourdieu 1977:4), multiple discursive possibilities, or individual intentions and strategies. Indeed, Spivak has rightly commented that if women had in fact been recognized as users of signs rather than as signs of a relationship between men, Levi-Strauss's theoretical edifice would have been seriously compromised (Spivak 1988:291n.45). With respect to South Asian anthropology, similar limitations characterize the work of Louis Dumont on caste (1980) and kinship (1957, 1966, 1986). For commentaries on such limitations insofar as the study of caste is concerned, see Raheja 1988a, 1988b, 1989; and with respect to kinship, see Trawick 1990:115-84.

March 1984; Meeker, Barlow, and Lipset 1986; Rubin 1975; Spivak 1985c: 356-60). We are reminded that as they are "given" or "exchanged" among kin groups, women seldom act or speak as if they were commodities, or signs rather than generators of signs (Holmberg 1989). They actively construct representations of marriage, gift giving, and kinship relationships, and these representations may often diverge dramatically from those generally invoked by men; and further, these alternative representations are frequently realized in particular kinship practices—marriage strategies, patterns of gift giving, terminological usages, deference behavior, and so forth—in which both men and women are involved. But in the totalizing projects of much ethnographic writing on kinship, these divergent discourses and divergent practices have often been ignored.

The Pahansu and Ghatiyali ethnography provides little evidence for unitary male or female perspectives on kinship relations (cf. Yanagisako and Collier 1987: 25-29). We find, rather, universes of discourse and universes of practice that tend to be drawn upon and used in different ways by men and women (Bourdieu 1977: 110, 122-23). Standard anthropological accounts of North Indian kinship speak in terms of a structured system of unequivocally defined relationships and an internally consistent and homogeneous cultural ideology stressing, first, the importance of patrilineally defined categories of kin (e.g., Mayer 1960; Parry 1979);^[16] second, the hierarchical nature of the relationship between bride givers and bride receivers, its persistence through time to the following generation, and the importance of gift giving in relation to this enduring tie between groups of male agnates (Dumont 1966; Vatuk 1975); third, the idiom of marriage as kanyadan (the gift of a virgin), entailing the "complete dissimilation of the bride from her family of birth and her complete assimilation to that of her husband" (Trautmann 1981: 291; Inden and Nicholas 1977); and fourth, the necessary subordination of the bride to her husband and conjugal kin. In interview situations in

[16] Vatuk, however, has presented exemplary ethnographic examples documenting the bilateral definitions of certain categories of kin in northern India (1975) and the uses of "matrilateral assymmetries" in urban North India (1971), and Jacobson (1977a) points out some of the ambiguities surrounding the rights of natal and conjugal kin with respect to a married woman. These studies provide important points of departure for the analyses of kinship set forth in these pages.

the villages in which we have worked, both men and women tend overwhelmingly to affirm the importance of these cultural constructions and in doing so seem to speak in a single voice. The ethnography from Uttar Pradesh and Rajasthan that we present in this study does not lead us to question the existence of such propositions in popular understanding; they are critical and pervasive aspects of local ideology and of power relations in vast regions of northern India, and they are embodied in many kinds of social practice.^[17] If one fails to ask questions concerning the pragmatic deployment of such a discourse in everyday life, in ordinary speech, one might come to agree with Spivak's pronouncement that "the subaltern [as woman] cannot speak" in a language other than that of patriarchal authority (1985a: 130). Or one might with Spivak view North Indian women as having no identity apart from this dominant discourse:

The figure of the woman, moving from clan to clan, and family to family as daughter/sister and wife/mother, syntaxes patriarchal authority even as she herself is drained of proper identity. In this particular area, the continuity of community or history, for subaltern and historian alike, is produced on (I intend the copulative metaphor—philosophically and sexually) the dissimulation of her discontinuity, on the repeated emptying of her meaning as instrument.
(Spivak 1985c: 362)

In this book, we raise questions concerning the force, the persuasiveness, and the salience of dominant cultural propositions about patriliny, hierarchy, and women's subordination in everyday experience, and questions concerning the contexts in which they are likely to be invoked and who is likely to invoke them. We find that although neither men nor women would normally dispute these understandings of North Indian kinship relations in interview situations, they are clearly open to ironic, shifting, and ambiguous evaluations in the rhetoric and politics of everyday

[17] For a discussion of some of the mechanisms through which these dominant propositions produce women as gendered subjects within the patrilineal and virilocal milieu of Indian society, see Dube 1988. But Dube also points out, at the end of her essay, that although the values and practices she considers are made to appear as part of a given and hence "natural" order of things, women are nonetheless able to question aspects of this order and their position within it.

language use, in strategies of marriage arrangement, and in certain genres of oral traditions, particularly those performed by women. Thus we must pay close attention to the contexts in which words are spoken. If we record only women's responses to our own questions, we may all too quickly come to the conclusion that they cannot speak subversively and critically, that their voices are muted by the weight of male dominance and their own acquiescence in the face of "tradition."

At least some of the speech genres used by women in Uttar Pradesh and Rajasthan tend more frequently than men's to stress the desirability of disrupting patrilineal unity in favor of a stress on conjugality; they speak of neutralizing hierarchical distinctions between bride givers and bride receivers, of the emotional significance to women of affinal prestations, of the enduring nature of a woman's ties to her natal kin and the shifting evaluations of marriage that this entails, and of the moral obligation to reject sometimes a subordinate role vis-à-vis one's conjugal kin. From the vantage point of such a discourse, patriliny, hierarchy, female subordination, and so on are seen not as aspects of a fixed and reified cultural system but as strategically invoked idioms in an ongoing negotiation of personhood and relationship that, like proverbial utterances, may often be countered with other contrasting idioms that evaluate the situation at hand in quite different moral terms. Indeed, as Rena Lederman points out, women's acts of resistance to men's definitions of social order "raise questions about the extent to which male ideology can be understood fully without appreciating how this ideology is an argument against women's ideas, rather than simply a positive, independent statement" (1980: 495-96).

While it will be obvious that the efficacious manipulation of these varying discourses in northern India depends to a great extent on one's position within a set of power relationships in which women may not inherit land, must move from natal village to conjugal village at marriage, and are generally expected to maintain a subordinate position relative to their husbands and conjugal kin, it is nonetheless apparent that women's self-perceptions are not reducible to the terms of the dominant discourse on kinship. And those self-perceptions, and the discourses in which they are constructed and negotiated, may subtly but

distinctly alter the widely ramified networks of relationships in which both men and women live their lives.

Language and Resistance: What Do Women Do With Words?

The shift in anthropology from a focus on culture as a fixed, internally homogeneous, and logically ordered totality to an emphasis on decentered, heterogeneous meanings negotiated and contested in social praxis is paralleled in linguistics by the shift away from a concern primarily with language as a formal abstract system, a structured grammar and semantic system (what de Saussure called langue ) opposed to and distinct from speech (parole ), the actualization of language in naturally occurring social and cultural contexts.^[18] A "discourse-centered approach" to language and culture has been proposed as a perspective capable of encompassing attention to the agency of particularly positioned speakers and actors, to the relationship between culture as a system of shared meanings and culture as a set of contested and negotiated meanings, to the relationship of forms of discourse to the ongoing constitution of social life and social processes, and to the strategic use of forms of expression in everyday interactions. (See particularly Bauman 1986, Sherzer 1987a, and Urban 1991.) The linguistic significances of the term discourse relate in fact to a notion of praxis:

Discourse is a level or component of language use, related to but distinct from grammar. It can be oral or written and can be approached in textual or sociocultural and social-interactional terms. . . . Discourse is an elusive area, an imprecise and constantly emerging and emergent interface between language and culture, created by actual instances of language in use and best defined specifically in terms of such instances. . .. Discourse includes and relates both textual patterning . . . and a situating of language in its natural contexts of use. Context is to be understood in two senses here: first, the social and cultural backdrop, the ground rules and assumptions of language usage; and second, the immediate, ongoing, and emerging actualities of speech events.
(Sherzer 1987a: 296)

[18] For an overview of the shift in anthropological theory, see Ortner 1984.

Attention to the emergent properties of praxis and discourse opens cultural and linguistic meanings to the circumstantiality of everyday life, the multiple interpretations, indeterminacies, and ambiguities therein, and the differing speech strategies of variously positioned speakers.

We wish here to begin to view North Indian women's relationship to language from such a perspective. We are less interested in the positivist enterprise of ascertaining the manner in which specific forms of speech behavior ";reflect" gender differences than in the ways in which gender identities are created and negotiated in discourse.^[19]

Joel Sherzer (1987a) suggests that it is in verbally artistic speech genres such as poetry, myth, magic, song, verbal dueling, and political rhetoric that the interface between language and culture is most salient and the creative possibilities of language most fully realized. He argues that to assume that features of social structure are somehow congealed into linguistic structures would in fact ignore the ways in which language use creates and recreates culture and the social parameters of everyday life. Sherzer in fact discusses several examples of such language use from northern India (drawing upon Tiwary 1968), pointing out that even in a society exhibiting overt and enduring social hierarchies verbal play may constitute a tactical negotiation of status and identity (Sherzer 1987a: 300-302). This discourse-centered approach, like praxis theory, views societies not primarily in terms of structural fixities but in terms of the processes through which relationships are constructed, negotiated, and contested.

How do these perspectives on the creativity of verbal play in everyday social life help us to understand women's relationship to language in northern India? Linguistic forms are strategic actions, and verbal interactions are often sites of struggle about gender, kinship, and power (Gal 1991: 176). In relation to questions of language and gender, Susan Gal has also pointed out that we need to focus not only on women's powerlessness and their "mutedness" but on the "processes by which women are ren-

[19] For discussions of the limitations of the former approach to the study of language and gender, see Briggs 1992, Ochs 1992, and Sherzer 1987a, 1987b. Such an approach would, of course, be subject to the epistemological quandaries concerning the overdetermination of women's subjectivity discussed by de Lauretis (see n. 12 above).

dered 'mute' or manage to construct dissenting genres and resisting discourses" (190).

Roger Keesing (1985) provides an analysis of the micropolitics of women's speech about themselves and their society among the Kwaio of the Solomon Islands. He warns us that an anthropologist's failure to elicit from women reflective accounts of themselves and their place in a cultural tradition cannot be taken as evidence of their "muteness" or their uncritical acceptance of male ideology and male political hegemony. He argues that women's speech, like all speech, is produced in specific historical and micropolitical contexts, and that what women will say reflects the power relationships implicit in the elicitation situation, and their own perceptions of what their speech will accomplish. If we rely only on women's interview statements, or on our observations of women's public adherence to the norms of silence and submission, we run the risk of assuming that women are incapable of using verbal strategies to oppose that dominant ideology.

Precisely because the nature of power relations in northern India often prevents women from speaking in many political contexts and in the presence of senior male affines, women inevitably come to recognize that they must, as Das puts it, "learn to communicate . . . by non-verbal gestures, intonation of speech, and reading meta-messages in ordinary language" (1988: 198); and such communication often subverts the official language of the dominant discourse and subtly articulates a contrapuntal reading of gender and kinship relations.

Ordinary conversations in Pahansu and Ghatiyali are filled with such subversive communication. Women tend to deploy a distinctive set of proverbs, for example, as tokens of an alternative moral discourse (Raheja 1993). In the ethnographic descriptions contained in the following chapters, we are primarily interested, however, in certain relatively formal expressive genres performed by women in northern India. Ramanujan has written that the cultural traditions and verbal genres of South Asia are indissolubly plural and often contradictory, and that this plurality is organized both through context-sensitivity and through various forms of intertextuality (1989b: 189; see also 1986, 1989a). He goes on to suggest that a consideration of this context-sensitivity and reflexivity will enable us to go beyond grossly conceived

frameworks (such as the distinction between a literate and written Great Tradition and an oral or folk Little Tradition) for the understanding of diversities, and to go beyond monolithic and totalizing conceptions, toward a comprehension of the points of closure and of openness in a textual or ethnographic domain (1989b: 191). Before his death, Ramanujan began to explore the reflexive worlds and the "counter-systems" he found in women's tales from South India (1991a). We find his observations to be a useful starting point for our own investigations of women's expressive traditions in North India.

While Ramanujan speaks of the plurality of verbal genres in South Asia, and Sherzer and Gal of the tactical uses of such discursive forms in the creation and negotiation of meaning in social life, it has also been suggested that the "muted" and often subversive social discourse of women is frequently embedded in such formal speech genres as songs, laments, jokes, poetry, and so forth (Warren and Bourque 1985; for some specific ethnographic examples, see Abu-Lughod 1986, Briggs 1992, Harrison 1989, Karp 1988, Narayan 1986, Seremetakis 1991, and Egnor 1986). Both T. O. Beidelman (1986) and Ramanujan (1986) stress the fact that such imaginative genres of folklore may enter into and transform the more dominant discursive frameworks of social life, and Margaret Trawick Egnor (1986) sets forth some provocative suggestions concerning the changes in cultural patterns that might potentially be instigated in songs sung by untouchable Paraiyar women of South India. Similarly, in emphasizing the interconnections between the social and the poetic, Richard Bauman (1986) attempts to understand verbal art as socially meaningful action.

We might contrast these positions and our perspective on the creative power of women's discourse in the negotiation of meaning with Stanley Tambiah's comment that critiques posed in North Indian women's songs might simply be "context-restricted 'rituals of rebellion' that leave the dominant male ideology more or less intact" (1989: 418). It was Max Gluckman (1963), of course, who argued that inversions of everyday behavior found in many "rituals of rebellion" has a socially cathartic significance, functioning to ensure that "conflicts" were enacted only in well-defined ritual contexts, allowing social unity and the dominant ideology otherwise to prevail in everyday life.

Guha has written at some length concerning ritual and linguistic inversions in India in Elementary Aspects of Peasant Insurgency in Colonial India . He shares Gluckman's perspective on these inversions and speaks of their predictability, conventionality, and recursivity as buttressing, rather than undermining, dominant patterns of traditional authority. "Ritual inversion," he writes, "stands for a continuity turned into sacred tradition by long re-cursive use under the aegis and inspiration of religion. As such, it represents the very antithesis of peasant insurgency. . .. For if the function of prescriptive reversal is to ensure the continuity of the political and moral order of society and sacralize it, that of peasant insurgency is to disrupt and desecrate it" (1983: 36). Guha goes on to speak of an "unquestioning obedience to authority" as characteristic of peasant "traditionalism" in India, and of outright rebellion as the only real semiotic break with the "traditional" codes governing relations of dominance and subordination. This sort of dichotomous thinking, which posits a rupture, an antithesis, between a fixed and invariant "tradition," on the one hand, and radical social transformation, on the other, essentializes and unifies what in fact seem to be polyphonic discursive formations within the "tradition" itself. The reflexivity and multiplicity of South Asian language, text, and ritual are thus elided, and the analysis does not capture the ways in which imaginative genres may provide alternative moral perspectives, serving as a ground for possible social transformation or for the contextual and strategic deployment of varying discourses by particularly positioned actors. Resistance and tradition may not inevitably be at odds with each other. Though the women we know in Uttar Pradesh and Rajasthan do in many ways assent to the dominant ideologies of gender and kinship, they also sing of their resistance to these ideologies, and we have tried to comprehend some of the ways in which they may come to insert this stance of resistance into their everyday lives.

In her discussion of ritual and festive sexual inversions in preindustrial Europe (1975), Natalie Zemon Davis has observed that while such reversals in some ways renewed and reinforced established hierarchies, there was at the same time some transformative spillover into everyday life. Comic inversions and play with gender imageries, she suggests, prompted new ways of

reacting to systems of power and kept alive alternative ways of thinking about family structure. The women's expressive traditions we translate in the chapters that follow have similar overall functions; they may at some levels serve to perpetuate gender inequalities, but they also render conceivable and may indeed sanction women's active resistance. The "hidden transcript" found in women's speech genres can, we think, be viewed "as a condition of practical resistance rather than a substitute for it" (Scott 1990: 191). The active rebellion that may at one moment be impractical or impossible may at another moment become plausible precisely because the idea of social transformation has been nourished in proverbs, folk songs, jokes, rituals, legends, and language (Scott 1985: 37-41).

These perspectives on the multiplicity of cultural discourses, on the strategic uses to which they may be put, and on verbally artistic speech genres as potential loci of a language of resistance are critical vantage points for our interpretations of many aspects of women's language and women's oral traditions in rural North India. Far from speaking only in a language dominated by males, the women we have come to know imaginatively scrutinize and critique the social world they experience and give voice to that vision in a poetic language of song and story; as they do so those songs and stories may enter their lives and shift, however slightly or however consequentially, the terms in which their lives are led.

As we translated the stories and songs told and sung by the women we knew in Pahansu, Hathchoya, and Ghatiyali, we began to see that these expressive genres vividly and articulately depicted the most poignant experiences in women's lives— leaving one's natal home and going off to the unknown house of a new husband, the difficulties of submitting to the authority and domination of the husband's senior kin, the joys of intimacy and sexuality, giving birth, a husband's betrayal, the pleasures of adorning oneself with jewelry and fine cloth, and so forth. But we also began to see that despite this diversity of genre, region, language, and specific ritual or seasonal context in which a song is sung or a story told, these expressive genres converged in addressing several interrelated sets of contradictions within the dominant gender and kinship ideologies of northern India. The

contradictions that women's speech genres especially comment upon are those that entail an implicit or explicit splitting of female identities, splittings that in every case work to the disadvantage of women. This book examines these various contradictions concerning women's identities, and the dissident, ironic, and resistant stances that women take up in relation to them.

Chapter 2 focuses on one of the most commonly postulated and widely ramified split images of women in Hindu South Asia: the deep and pervasive split between the destructive and threatening sexual potency of women as wives, on the one hand, and the beneficent, procreative, and positive capacities of women as mothers, on the other. Women as mothers contribute to the continuance of the male patriline, while women as sexual partners are seen to disrupt patrilineal solidarity among men. This ideological splitting has a number of practical and material consequences in women's lives. Because it is viewed as dangerous and volatile, female sexuality must be constantly controlled and brought under the surveillance of male kin; women must assume a posture of sexual reticence and shame and of withdrawal from any situations in which they might come into contact with men who are not close kin.

Several recent studies of women's lives in India assume that women have internalized this ideological split and are thus shameful of their bodies and their sexuality. Though women are certainly silent on the subject of sexuality and the pleasures of the body in many situations of everyday life and when the anthropologist elicits in an interview their views on these matters, the words of Rajasthani women's songs tell us, in genres created by women, of another perspective. In these songs there is evidence of an exuberant sexuality, a positive valuation of sexual pleasures, and a conjoining of eroticism and birth giving that undermine and resist that split between sexuality and fertility posited in the dominant ideology. In subverting such split images of female identity, women are at the same time questioning that authoritative discourse of male control over female sexuality and female lives.

A second contradiction within North Indian gender ideologies concerns women's transfer from natal home to conjugal home at marriage. In her natal home, a woman does not veil her face and

remain silent in the presence of men, she need not show particular deference to senior kin, and she is not obliged to perform arduous domestic labor. This comparative freedom ends when she is married and moves to her sasural , her husband's home. Though she is expected to transfer her loyalties from natal home to conjugal home upon her marriage, she is nonetheless often viewed as an outsider there, one who has the capacity to alienate her husband from his natal kin, precisely because of her sexual powers. Thus attempts may be made to curb a too rapidly developing conjugal intimacy, in the interests of maintaining the solidarity of the husband's patrilineal kin. The abrupt ending of the pleasures of her life in her natal home, coupled with this devaluation of the conjugal bond, may place the young wife in a situation of emotional turmoil and powerlessness. No longer "one's own" to her natal kin nor yet "one's own" in her sasural , she may come to feel a profound sense of isolation. In talk about the gifts given to her by a woman's natal kin, in songs, and in women's readings of genealogically ambiguous kinship relations, the analyses in chapter 3 discern a critically ironic female commentary on the split in a woman's identities as daughter and sister, on the one hand, and as wife, on the other, and on the contradiction in a kinship ideology that makes her "foreign" (parayi ) to her natal kin and yet still alien to her husband's.

This ironic stance is one voiced specifically by women speaking as daughters and as sisters, though all women in a neighborhood or kin group sing the songs together. But there are other songs that give voice to the woman speaking specifically as wife. Chapter 4 examines such songs, which stress not the persistence of a woman's natal ties but the necessity of establishing intimate conjugal bonds even if this entails a deemphasis of patrilineal solidarities and a curtailment of the power of a husband's senior kin over the wife. These songs are more openly subversive than those translated in chapter 3; they also involve powerful critiques of the norm of wifely silence and submission that are found in so many other South Asian expressive traditions. Taken together, then, the songs presented in chapters 3 and 4 are seen as interrogations of that dominant discourse that severs the continuity between a woman's life as daughter and her life as wife and daughter-in-law and that renders her powerless in her sasural .

Chapter 5 is a study of a single Rajasthani oral narrative, the

tale of the "jungli rani," which is related by women as part of worship activities. The tellers of this story are concerned with false assumptions about female nature, an externally imposed split between female virtuosity and female virtue that is grounded in the idea that independent and powerful women are intrinsically dangerous and destructive. The tale of the jungli rani challenges this supposition and rejects such dichotomizing of power and virtue, which all too frequently succeeds in silencing and disempowering women.

Chapter 6 tells the story of a storyteller. Shobhag Kanvar of Ghatiyali is a skilled teller of tales and a woman possessing knowledge of ritual and healing that rivals that of any man in her community. She is the narrator of the tale of the jungli rani translated in chapter 5, and just as that story reads as an oppositional commentary on a discourse that imposes untenable splits in female identity, so too does the story of Shobhag Kanvar's life read as a series of strategies that successfully negotiate the whirlpools and the shoals of this discourse. In rejecting the imposed dichotomization of female identity that says that a woman cannot simultaneously pursue her own aims as an independently acting religious adept and remain a virtuous woman, Shobhag Kanvar has challenged masculine authority along with the discourse that views female agency and female independence with suspicion and with fear. We see her and the other women whose strategies we have tried to explicate in these pages as inserting the resistance found in song and story into their everyday lives and practically (as well as poetically) challenging the cultural discourse encoding female subordination.

The words that we have attempted to translate and to understand are diverse and heterogeneous. There is no single South Asian female voice, no single female consciousness, that unequivocally rejects or accepts prevailing North Indian ideologies of gender and kinship. There are many voices, of women in varied kinship, class, and caste positionings. Yet the expressive genres from Uttar Pradesh and Rajasthan that we have tried to understand do converge in viewing the imposed splits between sexuality and virtue, between natal loyalties and conjugal ties, and between female power and female virtue as essentially contestable cultural forms.