The Ethical Ambivalence of Resistant Violence: Notes from Postcolonial South Asia

In the face of mounting militarism in south Asia, this essay turns to anti-state, ‘liberatory’ movements in the region that employ violence to achieve their political aims. It explores some of the ethical quandaries that arise from the embrace of such violence, particularly for feminists for whom political violence and militarism is today a moot point. Feminist responses towards resistant political violence have, however, been less straightforward than towards the violence of the state, suggesting a more ambivalent ethical position towards the former than the latter. The nature of this ambivalence can be located in a postcolonial feminist ethics that is conceptually committed to the use of political violence in certain, albeit exceptional circumstances on the basis of the ethical ends that this violence (as opposed to other oppressive violence) serves. In opening up this ethical ambivalence — or the ethics of ambiguity, as Simone de Beauvoir says — to interrogation and reflection, I underscore the difficulties involved in ethically discriminating between forms of violence, especially when we consider the manner in which such distinctions rely on and reproduce gendered modes of power. This raises particular problems for current feminist appraisals of resistant political violence as an expression of women's empowerment and ‘agency’.

introduction Political violence and militarization have become an indelible part of the narrative of postcolonial south Asia. The political violence of genocidal proportion that marked the birth of nations in the region has continued to structure everyday life in the post-independence decades, becoming more and more militarized in several instances. The militarization of civil societyobservable, for instance, in the public enthusiasm towards the nuclearization of India and Pakistan -exemplifies not simply the normalization of military ideology but a wide-ranging consent to militarized forms of life (John, 2004: 305). Such consent is produced and sustained not only by the structures of the state but also by those that fall outside its ambit, including non-state or even anti-state actors. Although less acknowledged, the resistant politics of revolutionary movements have contributed their fair share to the wider militarization of these societies (as elsewhere). The countering of statist militarism with forms of revolutionary militarism that tend to become indistinguishable from the former raises pressing but unanswered questions to do with political ethics.
The issue of ethics, particularly the ethics of political violence, is what is at stake in this essay, which takes as its point of departure resistant political struggles in contemporary south Asia. Ethical norms and values lie at the heart of any discussion of violence, whether to do with a narrow consideration of its use value or broader issues around its legitimacy and justification. These concerns are arguably more complex when we turn to the politics of resistant violence, usually justified as a response to the violence of the state, and for the sake of democratic ends. In fact, the moral economy in which violence is located has not only precluded, in many instances, its outright condemnation but has also prevented a serious consideration of its underlying ethical presuppositions and ambiguities. While a recipe for establishing an ethical response to political violence may not be possible (or even desirable), the ethics of violence, whether state sanctioned or not, must be taken seriously even if only to acknowledge the difficulty that such a task involves.
In the context from where I speak -India -a discourse, however provisional, on the ethics of resistant violence assumes a degree of urgency in the face of a rapidly deteriorating standoff between the state and revolutionary groups, and its spin-off effect on ordinary people, everyday life, and progressive politics. I also speak with the theoretical resources of feminism, which, as politics and knowledge, is deeply committed to the struggle against violence and militarism. In south Asia, as Manchanda (2004) notes, both the state and oppositional movements have mobilized women, and subsumed their interests to larger nationalist and socialist projects inscribing a 'patriarchal containment' (de Mel, 2001) of women in the political domain. Notwithstanding this critique, feminists have responded to the violence of anti-state movements with a degree of ambivalence: one that cannot be found in their engagement with other forms of (state/rightwing) violence. This might stem from the 'sustained, if qualified, support' (John, 2004: 305) that postcolonial feminism has tended to offer anticolonial and other 'liberatory' struggles, even those that deploy violence to achieve their ends.
In the contemporary context of women's (often violent) political mobilization in south Asia, the ethical significance of political violence raises contextually specific as well as broader questions to do with women's place in 'men's wars', whether just or unjust. Should feminists argue for women's equal rights to take up arms against repression and injustice or should they champion the democratization and demilitarization of such struggles? Should they celebrate the unexpectedly large numbers of women that have joined the ranks of south Asian insurgencies as a measure of their 'agency' or should they challenge the militarization of women's identities therein? Like Virginia Woolf, should they seek to liberate men and masculinity from militarism or insist on the expansion of the sphere of war and combat to include women? At the heart of these ethical quandaries lie certain assumptions about violence, power, and gender that need to be interrogated or at least reflected on. These include, I show in this essay, assumptions to do with the nature and practice of violence including the possibility of ethically discriminating between forms of violence. Such distinctions tend to rely on and further entrench gendered norms and hierarchies, especially when we consider how militarized discourses of resistance employ gender as an 'ethical shorthand' (Hutchings, 2007a: 101) to justify their use of violence. The assumption that resistant violence might be 'empowering' for women also tends to reproduce a victim-agent binary in thinking about postcolonial subjectivities from a feminist perspective. The fact that gender plays a crucial role in producing certain kinds of violence and its subjects means that ethics cannot be approached from the perspective of feminism alone but from an understanding of how ethics itself is fully imbricated by issues of power, including gendered modes of power (cf. Hutchings, 2007a). Such a feminist ethics must also stem from an appreciation of the deeply ambiguous nature of political action to which an autonomous agent cannot be readily ascribed.

feminist ethics and political violence
In a recent article, Kimberly Hutchings (2007a) delineates three dominant trends within feminist ethics and their implications for assessing the ethical legitimacy of the use of violence for political ends. Given that this is one of the few recent feminist reflections on the ethics of violence to include revolutionary violence, it provides a useful starting point for some of the tensions that this essay seeks to explore. Very briefly, Hutchings identifies an enlightenment feminist ethics that upholds universal standards of justice and freedom, a care ethics that draws on women's distinctive but universal capacity for motherhood to argue for a unique female moral perspective, and postcolonial feminist ethics that rejects the ethical universalism of both these traditions for the recognition of difference and plurality or the particular social and cultural context in which ethical dilemmas are situated. In paying close attention to issues of context and their meaning (including the context of an unequal world order), postcolonial feminists acknowledge that 'the same ethical norm may have very different implications in different places at different times' (Hutchings, 2007a: 94).
From the perspective of a postcolonial feminist ethics, questions around war, violence, and militarism (as with the categories of 'woman' and 'gender') cannot be dealt with a priori but within specific cultural and political contexts and the ethical challenges that lie therein. Such an ethical position opens up the possibility for feminism in a postcolonial mode to discriminate between forms of 'good' and 'bad' violence, and to support the use of violence for certain ethical ends (such as the self-determination of an oppressed people). To this extent, postcolonial feminism is not entirely different from an enlightenment feminism that is conceptually linked to the possibility of legitimizing the use of violence in certain circumstances such as humanitarian intervention against genocide. Both traditions (unlike care feminism that Hutchings associates with pacifism) are, in this way, close to just war thinking in that they legitimate the use of political violence in highly exceptional circumstances. What Hutchings does not underscore but which is more than obvious in feminist appraisals of militant anti-colonial and nationalist struggles as 'empowering' for women (cf. Cock, 1994;Omolade, 1994;Mama, 1997) is an endorsement of violence not only for politically progressive but also feminist ends. Violence is justified, in other words, not only as a 'necessary evil' on behalf of an oppressed group but also for the sake of an invisible minority therein, namely women. Thus Cock (1994) argues that unlike liberal forms of western feminism, revolutionary feminism in a third world context is often militarist: it asserts women's equal rights with men to take up arms against repression and injustice. Beyond the call for equality, the revolutionary nature of these struggles is seen as affording women the opportunity for politicization, especially in military or combatant roles that have been historically unavailable to them. Such an ethical position legitimates the use of revolutionary violence in order to fulfil not just antioppressive ends, but even feminist ones. What it tends to ignore however is the role of gender 'in the fabric of ethical values, judgment and action in both principle and practice' (Hutchings, 2007a: 91). For Hutchings, ethics cannot be understood as a category that is distinct from or unaffected by issues of politics and power. In order to seriously engage with the ethics of political violence from a feminist perspective, what is required, then, is an appreciation of the degree to which ethics is rooted in (and not outside of) gendered relations of power. Before moving onto an exploration of this problematic, it is necessary to place it in the context of contemporary south Asia.
contextualizing resistant politics 'Resistant politics' is a broad, unwieldy category, whether applied in the context of south Asia (itself a heterogeneous category; cf. Kumar, 1994) or elsewhere. My usage of the term encompasses fairly, but not entirely, disparate political struggles that have 'particular and general histories' (Banerjee et al., 2004: 128) in the post-independence decades. These are primarily anti-state 'liberatory' struggles that have mobilized groups on the basis of identity categories like religion and ethnicity for the establishment of national liberation or secession from the state, or freedom from class oppression within the nation-state (or a combination of these factors). The 'armed' or militant nature of these struggles is a major point of commonality given that the organizations leading these movements are frequently identified as military, guerilla, or even 'terrorist' ones, especially by the state. The most prominent of these is the decade-long ethnonationalist conflict in Sri Lanka between the state and the Tamil-led militant Liberation Tigers of Tamil Eelam (LTTE), and equally by separatist groups struggling for an independent Kashmir. Less known are the ultra left 'people's wars' currently waging across at least eight Indian states (recently declared by the Indian Prime Minister to be the 'gravest internal security challenge' faced by the country), separatist groups in India's north-east and in Bangladesh, and the recent Maoist insurgency in Nepal. Religious nationalisms such as Islamization in Pakistan and hindutva in India -what some have referred to as pro-state as opposed to anti-state nationalisms (Alison, 2004) -are further militarizing civil and political society through a refashioning (rather than a rejection) of the state.
Indian feminists have had to urgently respond to the rightwing mobilization of women in both leadership and rank-and-file positions in the new hindutva movement (cf. Butalia and Sarkar, 1995). In contrast, much less has been written of women's participation in anti-state revolutionary movements (with the exception perhaps of the Sri Lankan case). This in spite of the fact that 'ultra left' militant groups in India, varyingly called Naxalites or Maoists, have seen a steady increase in the ranks of their female cadres, some even engaged in combat. 1 Women's participation in the recent Communist Party-led insurgency in Nepal -one of the most reported aspects of this allegedly 'pro-woman' people's war -has grabbed some feminist attention (cf. Gautum et al., 2001;Manchanda, 2004;Pettigrew and Shneiderman, 2004). The initial enthusiasm with which feminists welcomed images of 'young, gun-toting guerrilla women' as a measure of the leadership's commitment to the 'woman's question' seems to have waned 1 A complex and contradictory formation, the 'far left' in India includes mostly underground Maoist (or Naxalites as they are locally known) groups that draw on in the face of the lack of any visible advancement of women's rights by the newly formed communist government. This is also true of the LTTE in whose ideologies and practices women play a prominent role, constituting between a third and a half of its membership, including a well-established Women's Military Wing (Basu, 2005). In spite of the visibility afforded to women as combatants and suicide bombers -the most famous example being that of 'Dhanu' who killed the Indian Prime Minister, Rajiv Gandhi and herself -feminists have lamented the tendency of women being reduced to mere 'cogs in the wheel' given their containment through regulative patriarchal ideals 'at the very moment of their most innovative empowerment' (de Mel, 2001: 212). In general, as de Alwis (1998) rightly observes, the feminist debate around women's participation in militant movements has been framed in binary terms: perceiving them as liberated or subjugated subjects; as victims or (as increasingly the case) agents of violence.

ambivalent agency
Political violence, particularly that of the state, has been a primary point of address for feminist theorizing, critique, and mobilization. In India, for instance, a self-conscious 'feminist' women's movement emerged as a response to the institutional violence of the state and its agents such as the police. In more recent times, feminists have turned to the violent underpinnings of the hindutva movement, viewing it as equally oppressive and deplorable as the state's exercise of violence, notwithstanding the possibility of politicization that it might offer women. When we turn to the political violence of anti-state, revolutionary struggles, the picture becomes slightly more complex. It would be fair to say that unlike the state or the religious right, there has been no comparable assessment and critique of resistant violence from a feminist perspective. The feminist attitude to such violence is also much more ambivalent.
The left legacy of feminist politics might explain this more ambivalent stance towards the violence of oppositional, including left-inspired, movements. Most women's movements in south Asia have socialist roots; the left constitutes an important legacy for feminist mobilizing in the region. As Omvedt (1993: 216) notes of the Indian women's movement, feminists have always identified with the left, with specific implications for the question of political violence: 'particularly in India, where the tenacity of the Naxalite effort to build a force based on the rural poor made them romantic heroes to so many, an endorsement of violence seemed to provide the distinction between paternalistic Gandhism and the dividing line in the communist movement between ''revisionists'' and ''revolutionaries'''. Besides the political alternative that it provided, revolutionary violence was morally acceptable because of the appeal of the 'heroism of variations of Marxism-Leninism and Maoism, united perhaps only in their commitment to the rallying cry of 'people's war' or the armed overthrow of state power. While not constituting a serious challenge to the authority of the state or even the hegemony of the organized left, the political might of the far left cannot be underestimated either, given the significant degree of national presence it commands today. It also has effective international links, such as with the Nepalese Maoists who, after nine years of insurgency and underground operations, have recently formed a coalition government with their leader as the new prime minister. guerrilla warriors' but also owing to the popularization of images of 'the woman with a gun in her arms and a baby on her back' (ibid.). This image of the motherwarrior is one that has a longstanding presence in the imaginary of 'liberatory', especially nationalist struggles in the 'third world', and is an acknowledged part of a revolutionary femininity. The combined force of these symbolic structures meant that the question of women's empowerment could readily be decoupled from that of violence in the context of liberation struggles (cf. Omvedt, 1993: 216).
The separation between emancipation and violence is also implicit in the historical recovery of women's participation in peasant and working-class struggles such as in Telengana and Tebhaga in India (see Stree, 1989 andCusters, 1987, respectively). Feminists sought to recover these histories in 'celebrating a lineage of resistance' (Stree, 1989: 19) as well as countering the longstanding orientalist 'myth of passivity' with respect to Asian women (Trivedi, 1984: 38). Violence, in this mode of feminist history writing, is identified in the leadership practices of these movements (the 'Party') that thwarted the possibility of emancipation that revolutionary upheaval -that 'magic time' (Kannabiran and Lalitha, 1989) -held out for women. In general, these works are limited to clearly identifiable 'women's issues' such as patriarchy and sexuality, and do not extend to broader concerns of political violence and its implications for feminist ethical aspirations and ends.
Discussions on resistant politics in contemporary south Asia have acknowledged (if not always engaged with) their violent and increasingly more militarized nature. The recent visibility of women as perpetrators and not strictly as victims of armed conflict is partly responsible for this more reflexive stance toward political violence. 2 Women's active involvement in militarized political cultures such as the LTTE in Sri Lanka and Maoist groups in Nepal (and to a lesser extent, in India) has raised specific concerns for the feminist project; one of acknowledging the possibility of women's empowerment in the 'crucible of a militarized, hierarchical, authoritarian culture of violent politics' (Manchanda, 2004: 237). Manchanda's research on south Asian conflicts is typical of recent studies that point to the ambivalent nature of the 'agency' that violent 'liberatory' struggles offer women, and beyond that, to the inadvertent and 'ambivalent gains' of conflict itself (de Mel, 2001;Alison, 2004;Pettigrew and Shneiderman, 2004;Rajasingham-Senanayake, 2004). The female combatant is viewed as an important instance of this 'agency', one that belies the 'natural' association of women with peace but also the presumed passivity of women (particularly third world women) within a narrative of war (cf. Manchanda, 2001a;Alison, 2004;Coomaraswamy and Fonseka, 2004). 3 Yet the continuing predominance of a discourse of victimization in south Asia's war zones has not only impeded, it is said, a serious consideration of such ambivalent and 2 The increased participation of women in armed movements is not limited to south Asian countries; recent studies offer the basis for a good comparative approach to a transnational feminist ethics of political violence. See, for instance, Hasso (2005) and Hamilton (2007).
3 Titles such as 'Women as Agents of Political Violence', 'Women, War and ambiguous forms of agency, but it has also limited concrete attempts towards the establishment of peace (Rajasingham-Senanayake, 2004).
I want to suggest that the ambivalence that these studies speak of with regard to armed conflict is an instance of the more general ambivalence towards political violence that we have already noted of postcolonial feminism. The underlying tensions of the feminist stance towards resistant violence can be attributed to its reliance on the left but also to transnational articulations of revolutionary feminism. They also stem from a wider assessment of anti-state 'liberatory' nationalisms as being more politically empowering for women than pro-state nationalisms that invariably restrict women's expressions of autonomy (Alison, 2004). 4 This ambivalence suggests the following: that the use of violence could be politically and socially enabling for women, even as such violence contributes to the wider militarization of society, and is generally deplorable to feminists. For some, this ambivalence is read as a feminist failure to acknowledge women's agency in times of armed conflict and war. On the one hand, for secular feminists, 'women's political violence is often the uncomfortable black hole wherein women's agency, because violent, becomes a male patriarchal project' (Rajasingham-Senanayake, 2004: 151). On the other hand, feminists have failed to perceive 'the unintended transformations brought by war, of seeing positives in violence, lest we be branded ''war-mongers''' (ibid: 147). The tension that Hutchings outlines in a feminist position that keeps open the use of violence in certain circumstances is exemplified here. For here we have a feminist position that condemns violence but acknowledges, even endorses violence that might have unintended positive outcomes for women. The mobilization of gender to justify violence is also, I argue in what follows, at the heart of the problematic of resistant violence for a feminist ethics.

ethical ambiguities: means and ends
Together with Hutchings, I have identified a distinction between forms of violence on ethical (and feminist) grounds as part of the postcolonial feminist response to violence in general and resistant violence in particular. Hutchings (2007a: 100) notes that a prominent way of critiquing such positions relates to the ability to distinguish between ethical and unethical uses of violence. Such distinctions, based on an instrumental understanding of means and ends and the ability to accurately calculate means in relation to the ends or outcomes of actions, are hard to sustain in practice even if they might make sense in principle. Generally speaking, even though our actions might be oriented towards and determined by known ends, we cannot control (by calculating/knowing in advance) the outcomes of our actions. The tension between ends and outcomes (one that Hutchings seems to conflate in the distinction between means and ends) makes all action lacking in clarity, predictability, and definiteness. Elsewhere Hutchings  Giles, 2001;Moser and Clark, 2001;Hamilton, 2007;Coulter, 2008).
4 See, however, Sen's (2007) recent study of the women's wing of the Hindu rightwing party, the Shiv Sena in western India, which, she argues, provides women a degree of power and agency as well as a way of coping with a life of poverty and uncertainty.
(2007b) draws on Simone de Beauvoir's discussion of the ethics of resistant political violence, where she argues that human existence or human action (including violent action) is inherently ambiguous because it lacks certainty and predictability. De Beauvoir criticizes new left violence on the grounds that instrumental justifications of political violence, grounded in unsustainable divisions between means and ends, deny this ambiguity by positing certainty and calculation in its stead. 5 The fact that the impact and implications of the use of political violence cannot be calculated in advance makes it hard to control the kinds of violent excesses that have become near synonymous with ultra left politics in south Asia, for instance. These include the manner in which the exercise of resistant violence invariably mirrors and leads to further oppressive violence, besides the high probability that violence against oppressors may be used against those who are not directly responsible for oppression (Hutchings, 2007b). The dominance of internal killings as a way of dealing with 'informants' is paradigmatic of the former while the routine killing of civilians caught in the crossfire between the state and guerrilla forces is an instance of the latter. Less obvious (and certainly less acknowledged in feminist scholarship/activism) is the manner in which such left-inspired violent politics actively produce (rather than simply enable) other violences, including violence against women. In my own research on the 1960s and 1970s extreme left Naxalbari movement of West Bengal, 6 I noted how the political violence employed by the state and responded to by the revolutionary was continuous with other forms of everyday, especially gender-based and sexual violence (Roy, 2008). To take a very small example, the construct of the state (a rapist state, in particular) aided the creation of an illusion of safety within the movement; it magnified the violence that existed 'out there', in the public domain, while rendering invisible the gender-based violence (including rape) that middle-class women activists faced within the community, at the hands of their own comrades. Not surprisingly, such violence was rendered largely unspeakable within the movement. Suffice to say that the manner in which political violence not only implicates but also produces other violence (as well as the possibilities of witnessing and resisting violence) suggests that 'distinctions that make sense in principle are unsustainable in practice' (Hutchings, 2007a: 100).
It should be obvious from this discussion that the decision to use political violence cannot be a matter of calculation or moral certainty, nor can it rely on ethical distinctions that are grounded in a narrow instrumental rationality (as with most 'liberatory' movements). It must be made on the basis of engaging with the ethical basis of violence. And ethics, Simone de Beauvoir (1976) notes, is grounded in uncertainty. It involves an avowal of our own ambiguous (as opposed to socio-historically transcendent) nature, and the necessarily limited nature of all our actions. Thus we can justify recourse to violence not on the basis of moral certainty or by positing unrevisable ends, but on the basis of the 5 It is also because we cannot know in advance the outcomes or consequences of the use of violence that violence (unlike power), according to Arendt (1970), can never be legitimate, but calls for specific justification in particular instances. Like Simone de Beauvoir, Arendt is critical of new left violence on the basis of the impossibility of knowing in advance the outcomes or consequences of the use of violence for political ends. For both, means cannot be clearly distinguishable from the ends they serve. In the context of political struggles, means also tend to overtake ends such that violence becomes an end in itself.
6 The Naxalbari movement began as a peasant uprising in northern West Bengal in 1967 led by a dissident group of the Communist Party of India (Marxist) who later formed the first of many Maoist parties in India. Popularly known as 'naxalites', they declared a 'people's war' against the Indian state structured on the Maoist model of protracted armed struggle. Middleclass students, who left homes to 'integrate' with the peasantry in the necessarily ambiguous nature of human action and the revisable nature of human ends. However, such an ethical position -an ethics of ambiguity -makes it far harder to justify the use of violence in advance, or even to raise the question about whether violence can be justified in such and such circumstances (as with certain postcolonial and enlightenment feminist positions). For underlying this question is an assumption that violence is an external other that needs to be overcome as opposed to a set of practices that is 'rooted in the world and in subjects' (Hutchings, 2007b: 127), including a world that is necessarily structured by gendered relations of power.

gendered ambiguities
The ambiguity that de Beauvoir speaks of is very often denied in ethical discriminations of violence, and not only to do with distinctions between means and ends. Like de Beauvoir's 'serious man', the revolutionary acts with a high degree of moral certainty with respect to actions and ends that are, per force, judged to be right. There is thus a denial of ambiguity, and violence becomes for the revolutionary a matter of course rather than an ethical choice (Hutchings, 2007b: 126). The denial of ambiguity is not limited, however, to the revolutionary agent alone. The reading of resistance often mirrors the moral economies within which resistance is located, inviting us to view individuals and actions on the basis of absolute moral binaries 'as either all bad or all good, sinful or virtuous, noble or ignoble' (Bourgois, 2002: 222). Such polarized understandings have meant that the battle lines between the revolutionary and the state get drawn all too starkly, engendering a series of splits between just and unjust wars, legitimate and illegitimate violence, and the 'good' and 'bad' man. What tends to get lost in this need to take sides is the grey nature of revolutionary (and indeed all) violence, and the ambiguity under whose sign revolutionary relations are often lived in the course of such struggles (see Roy, 2008).
It is important to mention that a scholarly (including feminist) investment in ethical divisions in violence has also precluded, in some instances, a full assessment of the multiple forms and meanings that violence assumes within revolutionary movements. This is especially true of the feminist engagement with contemporary Maoist and Naxalite groups in India. An underlying belief in the moral worth of revolutionary resistance seems to have generated an evasion of the question of violence and militarism (and their underlying patriarchal assumptions) even as feminists have debated issues such as sexuality, the politics of housework, and female political representation in the context of left political cultures (see Kannabiran and Kannabiran, 2002). The possibility of emancipation has largely been debated, in other words, in separation from that of violence. There is also a tendency to view sexual violence as the dark villages, emerged as an unlikely support base of the movement. The politics of violence that the movement professed cost it the initial popular support it enjoyed besides precipitating a brutal onslaught by the state that it could not withstand. underside of progressive politics, its perverted form rather than a product of violent political cultures. 7 What should be of greater concern to these feminists (and postcolonial feminism at large) is that the ethical divisions that justify forms of violence tend to rely on and reproduce a complex gendered economy of power. As we shall see in what follows, discourses of gender and sexuality are often key to the repertoire of rationalizing 'good' political violence and producing its subjects (the revolutionary martyr, for instance). The fact that gender plays a major part in representing such violence (and even war) as 'legitimate, honourable or desirable' (Treacher et al., 2008: 1) poses, Hutchings (2007a argues, a deep challenge to feminists, particularly those that defend the use of violence in certain conditions. For the postcolonial feminist, the challenge would be one of demonstrating how ethical distinctions between violence can be sustained in theory and in practice, and that too, without producing and reproducing the patriarchal positionings on which such violence and its legitimacy rely. A postcolonial ethics of violence would, thus, have to be cognizant of the role and significance of gender in producing particular ethical positions and subjects in sustaining certain forms of violence. As Hutchings (2007a: 101) herself acknowledges, there is a vast body of feminist scholarship that considers the mobilization of gender in ideologies and practices of political violence, be that nationalist political projects (Chatterjee, 1989;Yuval-Davis and Anthias, 1989;Jacoby, 1999;Sarkar, 2001), military discourses, war, and wider practices of militarization (Enloe, 1989;Chenoy, 1998), discourses of humanitarian intervention (Young, 2003(Young, , 2007, rightwing nationalisms (Banerjee, 2003;Bacchetta, 2004), and ethno-nationalist struggles (de Alwis, 1998;de Mel, 2001;Tambiah, 2005;Haq, 2007). The patriarchal underpinning of violence and militarization are known to have entrenched male power and privilege even in the case of anti-colonial, liberation struggles in the 'third world' (see most recently, White, 2007). The major themes uncovered by these studies have enormous bearing on the gendered logic of resistance, which, at least in the immediate context of south Asia, requires closer consideration.
The 1960s Naxalbari movement is paradigmatic in its mobilization of the category 'woman' in order to justify the use of revolutionary violence as key to class struggle. The sexual victimization of peasant women at the hands of the state and the 'ruling classes' was a major justification for armed struggle, itself discursively configured as a battle for honour or izzater lorai. Peasant women were central to this project given the cultural (and middle-class) construction of izzat or honour as located in the female body that sanctions, in turn, violent masculine protection and a limited degree of female agency (see Roy, 2007). Middle-class women, for their part, entered the revolutionary imaginary largely through the discourse of motherhood, whether as an enslaved mother/land that needed liberation or as a warrior-mother that aided the struggle against the 7 To take one recent example, Manchanda (2001a, b: 81-82) mentions cases of rape and abduction in the context of the armed militant movement in Kashmir. Rape, she says, is a common way to coerce marriage or to punish state informers. Although she roots these atrocities against women in the steady 'corruption' of the militancy, there is less of a sense that such oppression is internal to the logic of militarist political cultures, and not merely an accidental product of their perverse forms. Such a distinction between a 'true' militancy and a 'corrupted' version (which mirrors the larger distinction between 'good' and 'bad' violence) might also explain her more positive evaluation of the Maoist movement in Nepal.
'bad' violence of the state. 8 At the same time, middle-class women activists were constructed in terms of their violability or 'rapability' (Marcus, 1992) at the hands of the state by which the Party could draw them within its protective, paternalistic care. An appeal to protection thus provided a 'righteous rationale' (Young, 2007: 126) for violent class struggle besides reinforcing male superiority and female subordination in the political domain. As with protectionist discourses elsewhere (cf. Young, 2003), 9 the 'good' woman who deserves male protection is distinguished from the 'bad' woman who is undeserving of it. This is salient in the manner in which the Party dealt with sexual violence at the hands of its own members and sympathizers. As I have detailed recently (Roy, 2008), women's testimony to sexual violence by male workers and peasants was routinely disqualified on the basis of their own middle-class status and the working class status of their aggressors. Class was thus key to distinctions between 'good' and 'bad' women within the discourse of protection. 10 At least in the context of Naxalbari, ethical distinctions between good and bad violence and its subjects and objects were secured through particular mobilizations of 'woman', and presumptions of femininity and masculinity. As Hutchings (2007a: 101) notes, it is hard to see how this rhetoric could work without relying on normative understandings of male and female, and a gendered division of labour. It is important to underscore, however, that the legitimation of political violence relied upon a complex network of power relations, including those of class and caste, and not gender or sexual difference alone. Both the female and subaltern subject were 'useful' to the construction of Naxalite identity and to the validation of its politics, although their iconic usefulness largely required them to 'stay put' in the discursive domain of the movement (Gedalof, 1999: 60). Yet the utility of 'woman' -situated across and productive of multiple categories of identity including political identity -cannot be underestimated either. Unlike the familiar characterization of women as always victimized and marginalized by violence and war, the gendered politics of resistant violence points to the manner in which women are also foregrounded, made visible, and useful in the legitimation of violence and the constitution of its subjects (Gedalof, 1999: 49). Yet, as Gedalof is quick to remind us, the particular forms of visibility through which women emerge useful 'make some forms of agency more acceptable than others within prevailing relations of power' (ibid.: 49).

problematizing power
In this final section I want to briefly consider why positing the possibility of women's agency in the context of resistant violence might be problematic for feminism. What has emerged throughout this discussion is a tendency to view 8 Politicized motherhood is, of course, central to the revolutionary imaginary world over, from religiopolitical struggles to ethno-nationalist ones to secular leftled class struggles. In these varied contexts, the use of motherhood tends to underline the retraditionalization of women within the public domain. See in the context of south Asia, de Alwis (1998); de Mel (2001);and Haq (2007).
9 Young (2003Young ( , 2007 has recently unpacked the 'logic of masculine protection' in US security discourse in which a feminized 'protectee' is to be protected by the good violence of a masculine 'protector' against the bad violence of an aggressor. The rhetoric of protection also resonates with just war thinking in which women are 'beautiful souls' that need to be protected by 'just warriors'. For a recent feminist appraisal of just war thinking, see Sjoberg (2008), and for an older critique of the language of protection particularly in relation to the state, see Sunder Rajan and Pathak (1992).
10 In a recent piece, Sarbani Bandyopadhyay (2008) observes how contemporary Naxalite-led movements in central India fail to question and thus uphold the patriarchal ideology resistant politics as a space of emancipatory and agentive potentialities for women (even if these are rarely actualized). In spite of an acknowledgment of the 'ambivalent' nature of such agency given a wider context of violence and even war, feminists have made it a point not to entirely discount the possibility of empowerment and agency (often used interchangeably) either. The question of agency -understood in this context as autonomy or power -assumes all the more importance once we consider the orientalist, colonialist, and neoimperialist economies of thought within which third world woman have figured, of which 'western' feminism is itself a product (cf. Mohanty, 1991). The orientalist imagining of third world women as passive, powerless objects has propelled, in turn, the project of recovering and affirming the autonomy and subjectivity of non-western women; of reconstructing them as 'active, autonomous subjects in their own right' (Stephens, 1989: 100). This quest for historically denied agency and subjectivity -often pursued as an unqualified good (Sunder Rajan, 2004)needs to be located, I believe, in a wider mode of rationalizing and legitimizing violence on the grounds of the self-determination and liberation of the subjects of feminism; that is, women.
The idea that violence could give rise to self-determination and agency (whether for women or for any other group) is deeply problematic; the recurrent feminist claim of 'ambivalence' with regard to women's agency in conditions of war is itself testament to this. Implicit in such a claim is recognition that agency does not usually involve the use of violence; 'agency' is rather the freedom from violence or force. The use of violence to be free from violence is thus contradictory, and not at a conceptual level alone. There are enough historic instances that illustrate how violence tends to breed more violence in ways that undermine freedom and power. Violence and power are, for Hannah Arendt, opposites: violence is based on strength and force while power is a function of human relations; it involves collective action to bring about collective ends. Yet the tendency to conflate violence with power is pervasive. When feminists presume that certain forms of violence can restore power that has been historically denied to women, they too are complicit in this conflation.
If we learn from Arendt that power and violence are opposites, we also know that modes of power are not always empowering or enabling for their subjects. Underscoring much of the feminist discussion on women's empowerment is an understanding of power (like agency) as a positive social good, an 'inherently radical force or attribute of women and other subordinated groups' (Sunder Rajan, 2004: 327). As with a particular liberal feminist strand, such an understanding of power calls for women's greater participation 'in and control of the existing structures of political power' (ibid.: 327). The problem with such a redistributive model of power (as Young calls it) is that it tends to view questions of empowerment and agency largely in abstraction, outside of their imbrication in wider 'historically specific networks of power relations' (Gedalof, 1999: 53). of feminine modesty in discourses of community honour even when considering the sexual abuse of women. Such unquestioned patriarchal assumptions blunt the edge of their radical politics. While this article is exceptionally scathing of the patriarchal ideology of the Naxalites, it fails to unpack the patriarchal assumptions on which their defence and use of violence relies. As with most feminist critiques, the question of violence remains separate from that of women's emancipation.
It thus asks that women be included in structures and institutions of power, leaving their normative and gendered nature largely unchallenged. In the immediate context of militarized resistant struggles in south Asia, feminists have observed that the entry and presence of women, while initially empowering, is regulated by rigid conformity to the masculinist and patriarchal ideologies of these organizations (Menon, 2004: 62). The historically overdetermined nature of women's victimology has, at the same time, made it incumbent upon feminists not to rest content with such narratives of victimization alone. It is, however, one thing to 'remain fixed to a one-dimensional conceptualization of women as victims of war' (Coomaraswamy and Fonseka, 2004: 3), and quite another to assume that situations of widespread violence and conflict are actually favourable and even 'liberatory' for women.
One of the major contributions of feminist discussions on women's mobilization in rightwing hindutva politics in India has been to complicate our understanding of power by suggesting that not all modes of power are equally liberatory for women or desirable for feminism (cf. Sarkar, 1991, Sangari, 1993Gedalof, 1999;Sunder Rajan, 2004). Rather than a straightforward association of power with empowerment, these discussions draw on an understanding of power as a modality of subjection or subjectification, in a Foucauldian sense, enabling and at once constraining the subject. The manner in which LTTE rhetoric, for instance, produces 'woman' as a non-traditional, masculinized subject and as a carrier of national culture/tradition is one instance of the productive and repressive capacity of power (see de Mel, 2007). The current feminist stress on 'women's ambivalent agency' (Rajasingham-Senanayake, 2004: 143; emphasis added) in the realm of armed conflict can be seen as opening up the possibility of exploring the paradoxical nature of power as subjection. This is a largely lost opportunity given the underlying construction of women as perpetually marginalized, silenced, and excluded from political projects, and of power, even military power, as a mode of restoring visibility to women. 11 Such an argument effectively reproduces the victim-agent binary that it is at pains to deconstruct. On the one hand, it overestimates women's exclusion from the realm of political conflict in ways that ignore how 'woman' and gender are central to the moral economy of resistant political violence, as explored previously. On the other hand and in a liberal feminist mode of rendering hitherto invisible 'woman' visible, it problematically associates women's visibility in cultures of violent politics with a foregrounding of women's rights, empowerment, and 'agency'. A way of moving beyond conceptions of power as either empowerment or subordination (the woman-as-victim or woman-as-subject/agent division) would be 'to understand how the category of ''woman'', the subject of feminism, is produced and restrained by the very structures of power through which emancipation is sought' (Butler, 1990: 2).
Finally, and extremely problematically, the feminist association of militarized revolutionary struggles with 'women's empowerment' condones rather than 11 On the limitations of binary thinking especially around power-asexclusion and power-as-inclusion in Indian feminist thought, see Gedalof (1999). questions the use of political violence and the militarization of civil society. It contributes to the idea that such violence is morally permissible not only to achieve its own legitimate political ends, but also feminist ones. In the Indian context, it is particularly striking how feminists have shied away from condemning the routine recourse to violence by Maoists and Naxalites so much so that violence has become an end in itself. It has thus become possible for feminists to engage the 'ultra left' on issues of political representation and leadership, for instance, without addressing the politics and ethics of violence and militarism and even to declare such politics as 'the most radical in the country' (Kannabiran and Kannabiran, 2002). The problem with such a feminist position is at least twofold. By reducing political ideology to its practice, it contributes to the perpetuation of the gendered and patriarchal images on which this ideology depends (such as the feminized protectee or the warrior-mother). A second problem with this feminist stance, as Young (2003: 230) has noted of the use of 'woman's liberation' in legitimizing humanitarian intervention, is that it does not 'have principled ways of distancing itself from paternalist militarism'. For Young (2003: 231), a concern for the well-being of women is not a sufficient condition of feminism, which also requires a commitment to democratic values and citizenship on a global scale. Likewise, a feminist ethics of resistant political violence cannot be limited to narrowly construed 'women's issues' to do with the politics of sexuality or the question of widening participation in resistance movements. It cannot, in other words, be concerned with gender or sexual difference alone at the cost of addressing a more complex moral economy of rationalizing and reproducing violence and war to which 'woman' is central.

conclusion
Even such a limited appraisal should make obvious the complexities involved in articulating a feminist ethics of resistant political violence in contemporary south Asia. While a singular vision of such an ethical position may not be possible or even desirable, it must at least grasp the underlying gendered politics of revolutionary violence. This includes, first, the manner in which conceptual distinctions between forms of violence (based on the ethical ends that they seek to achieve) are largely unsustainable in practice, especially when we consider how political violence tends to give rise to other (sexual, gender-based) violence. Second, I have argued, after Hutchings, that normative ideas of gender, masculinity, and femininity are central to the legitimization of revolutionary violence as a 'good' violence against the 'bad' violence of an other. Distinctions between forms of violence on ethical grounds thus reinforce rather than disrupt existing gendered norms and hierarchies. Finally, a feminist ethics of resistant political violence needs to consider the modes of agency and subjectivity that such violence makes available to women on the understanding that not all modes of subjecthood or power are equally desirable or 'liberatory' for women. The 'gender politics of political violence and its mode of justification' (Hutchings, 2007a: 103) should caution against the enthusiasm with which many have embraced the image of the female combatant as an index of women's power and equality. Constituting women as agents of violence, in this manner, entails making them a part of a moral paradigm that relies on and reinforces normative gendered hierarchies in defending violence in advance. How does such a mode of subjectivity -one that is testament to society's deep commitment to patriarchal norms of femininity and gender, if anything (Hutchings, 2007a: 101) -fulfil the postcolonial feminist agenda of 'rescuing women from an underprivileged position in both knowledge and society' (Kumar, 1994: 8)? Yet to reject such an image is not necessarily to fall back on a historically overdetermined narrative of objectification. One way of transgressing the dichotomy between resistance and domination is available, I have suggested, in a more complex theory of power that addresses the ambiguous, even paradoxical nature of power in constituting subjects by constraining them to wider power relations. This might effectively free up feminists from viewing revolutionary moments as spaces of liberation or further subjugation to map, instead, how the rhetoric of revolution (and war) articulates itself through gendered power. A theoretical expansion of the feminist agenda might also address the ethical challenges involved in resistant political violence; challenges that cannot be met in the context of feminist goals alone but in a wider context of the gendered politics of political violence and militarism in south Asia today. That ethics does not furnish recipes and is a kind of a wager makes it even more imperative for feminists to partake in the 'painfulness of an indefinite questioning' (de Beauvoir, 1976: 133) in which morality and power may reside.
acknowledgements An early version of this paper was presented at the Silver Jubilee of the Indian Association for Women's Studies, Lucknow, 2008. I am grateful to those present for providing me with feedback and support, in particular, Disha Mullick. My thanks to Rafael Winkler for his perceptive reading of the piece at various stages. Without the intellectual generosity and editorial patience of Amal Treacher, however, the ideas in this article would not have come to fruition.

author biography
Srila Roy is a Lecturer in Sociology at the University of Nottingham. Her research/ teaching interests include gender and feminist theory, cultural memory, violence and trauma, and the affective economies of social movements and everyday life more generally. Her PhD research on the cultural memory of the extreme left Naxalbari movement of West Bengal has appeared in Feminist Review, South Asia Research, the Journal of the Royal Anthropological Institute, and is forthcoming in anthologies on violence, trauma and social suffering within and beyond the bounds of south Asia. Supported by the University of Nottingham and the British Academy, Srila's current research explores issues of institutionalization, temporality, and historiography in the Indian women's movement.