Thinking Woman-to-Woman Rape: A Critique of Marcus’s ‘‘Theory and Politics of Rape Prevention’’

This article uses the empirical fact of woman-to-woman rape as a lens to critique Sharon Marcus’s ‘‘Fighting Bodies, Fighting Words: A Theory and Politics of Rape Prevention.’’ To the extent that any theory forecloses this fact, we can assume it is erroneous. While Marcus’s work is promising in its intention to deconstruct binary views of gender, it largely reiterates the very dualism it seeks to destabilize. I explore two different deconstructive arguments that can be drawn from the piece, each of which has been adopted by some thinkers. The ﬁrst forecloses woman-to-woman rape while the second makes theoretical room for it. The second argument has the potential to deconstruct the ﬁrst. Following the logic of Judith Butler’s thoughts on gender transgression, I suggest a synthesis of these two arguments. Finally, I explore ways the self-defense strategies Marcus promotes can be made to accommodate survivors of gender transgressive assaults.

''Fighting Bodies, Fighting Words: A Theory and Politics of Rape Prevention'' (henceforth TPRP) to ascertain its capacity for accommodating woman-to-woman incidents. TPRP is of particular interest because it is taken by many to decouple womanhood and victimhood (e.g., Hesford 1999;Herberle 1996;Gatens 2000;McCaughey and King 1995;Robson 2007;Mazurok 2010;Binswanger et al. 2011;Lichtenstein 2005;Binswanger et al. 2011). In some senses, Marcus does open discourse to women's violence. Yet in important respects her piece reiterates the discourse it attempts to disrupt, foreclosing woman rapists. To the extent that any theory forecloses the possibility of woman-to-woman rape it can be empirically demonstrated wrong.
The first section below presents some key terminology used here. The presentation of ''rape'' as an essentially contested concept lends support to the term ''gender paradigmatic rape,'' which underlines the constructedness and contestedness of this category. Then, inspired by Butler, the term ''gender transgressive rape'' applies to non-paradigmatic possibilities in the context of hegemonic heterosexuality. It inheres particularly useful theoretical implications. In stark contrast to gender paradigmatic rapes, it is far from obvious to all that womanto-woman rapes occur. To demonstrate that they do, the second section provides an empirical sketch of the phenomenon.
Next begins our theoretical discussion proper. Radical feminist thought is particularly relevant to rape theory because of its foundational relationship to current academic, legal and popular perspectives. Further, proponents of radical feminist views constitute Marcus's principal foils and so are essential to contextualizing her work. A critical summation of some efforts in this vein will be presented in the third section. The fourth section constitutes a close reading of TPRP. Specifically, it explores two distinct perspectives drawn from the text under consideration. These will be referred to as Arguments One and Two. Argument One highlights the normativity of man-to-woman rape, a point which should by no means be abandoned. Notably, where this argument is concerned, Marcus's perspective is largely consistent with the very views she critiques. Like them, she exaggerates the strength of gender norms to the point of foreclosing woman-towoman rape. On the other hand, Argument Two accommodates woman-to-woman rape very effectively, deconstructing the monolithic heteronormacy of Argument One and acknowledging myriad possible forms of sexual violence, something which cannot be done so long as we adhere strictly to the standard gendered paradigm. We shall see that TPRP has been taken up by a variety of authors in ways that are consistent with one or the other of these conflicting perspectives and as such, with quite different results. My approach is unique not only in proposing a synthetic resolution to this conflict, but in recognizing this conflict in the first place. The synthesis proposed will bring us back to Butler's performative theory of gender. Argument One will be retained to account for gender paradigmatic rape's discursive and statistical dominance. However, this will be done in a way that simultaneously makes sense of gender transgressive incidents and that does not conflict with Argument Two. Finally, a modified approach to the physical selfdefense strategies Marcus suggests will be advanced in accordance with the synthesis proposed.

Terminology
Following Reitan (2001), I understand rape to be an essentially contested concept (ECC). ECCs are terms that inhere value judgments. They are based on complex sets of characteristics. Each ECC has paradigm cases about which there is general agreement. However, no such general agreement exists on which aspects render the paradigmatic cases paradigmatic. Thus, ECCs are the center of ongoing debate about which non-paradigmatic cases might belong under the rubric of the term in question. As Reitan (2001) argues, where rape is concerned, [T]he paradigms involve, on the part of the perpetrator, physical violence, coercion, control, disregard for the woman's wishes, a clear intent to overmaster the woman's will, a divorce of the sexual act from feelings of intimacy, the objectification of the woman, etc.; on the part of the victim the paradigms involve active resistance, the lack of desire, the lack of consent, powerlessness, suffering, a feeling of violation and dehumanization, etc. It is also clear that different theorists emphasize different aspects of the paradigms as being significant or essential, such that while they agree that the paradigms are instances of rape, they disagree about what makes them rape. (p. 49) Notably, paradigmatic gendering appears in the above citation not as an element in Reitan's list, but as an aspect of its structure. Thus, Reitan reiterates marginalization of gender transgressive rape rather than pointing to that marginalization as a politico-discursive opportunity. However, this issue bears only on his deployment of the ECC, not on its potential. Crucially, pushing the boundaries of an ECC is ''part of the proper use of the term'' (p. 45, original italics). Consistent with the ECC's structure and to underline how heterosexual framing both limits thought and invites contest, I term rapes that are committed by a man against a woman ''gender paradigmatic''. I refer to other rapes, which in their otherness may contest the male aggressor/female victim paradigm, as ''gender transgressive''.
In using the term ''gender transgressive,'' I appeal to Butler's (1999) performative theory of gender. That framework associates the Derridian observation that marks signify by referencing past uses in new contexts with the way we enter the world of hegemonic heterosexuality when declared girls or boys. From that moment on, we are socially shaped so that very often we come, generally if never completely, to identify with and embody those initial attributions. We embody and are animated by our genders through processes of subjectification, for gender attributions do not refer to nature but reiterate an illusion of it. Disciplinary acts which are only one element of this order range from subtle to violent. But order notwithstanding, these norms are inconsistently instantiated. Butler describes her project by saying, ''I'm interested in the problem of cross-identification; I'm interested in where masculine/feminine break down, where they cohabit and intersect, where they lose their discreteness'' (Cheah et al. 1998): 24. Thus, her groundbreaking work has sought to theorize how conventions are inevitably used in gender troubling ways to address unexpected circumstances, giving rise to new subjectivities and collectivities. To varying extents and for a multitude of reasons, we internalize and perform gender counter-normative comportments. Because normative genders are as reiterative as counter-normative ones, when girls learn boy things they own them as much as boys do and vise versa. Such transgressions create conflict which may reinforce gender norms, for example through punishment that makes an example of the offender. Alternatively or even simultaneously, transgressions may produce a new fold or opening in the social fabric. Butler's (1999) discussion of butch/femme in Gender Trouble is a much cited example of this process. Since butch identification cites masculinity ''against a culturally intelligible 'female body''' it does not simply reproduce heterosexuality; it ''is neither some decontextualized female body nor a discrete yet superimposed masculine identity, but the destabilization of both terms'' (156). As distinct from butch/femme, gender transgressive rape is a traumatic and wrongful transgression. Yet, analogous to butch/femme, it destabilizes the gendered norms for sexual violence that monopolize popular and scholarly understandings. Convention is at once deployed and transgressed. Despite the breadth of this term, rather than attempting to explore all of the many ways in which non-paradigmatic rapes can occur, I focus specifically on woman-to-woman incidents.
Even as I endeavor to extend the conceptual boundaries of the term ''rape'', its use remains a risk insofar as for many it evokes only paradigms (Stewart et al. 1996). For example, some evidence suggests forced non-penile penetration and penetration of the mouth or anus remain unlikely to be labeled rape by survivors (Kahn et al. 2003;Young and Maguire 2003). Nonetheless, I privilege ''rape'' over ''sexual assault'' due to its greater force and also because, after struggling to name their experiences many survivors themselves come to adopt it (Girshick 2002a;Young and Maguire 2003). Rather than alienating survivors who don't identify with this term, I hope to help broaden its scope and legitimizing power.
Encouragingly, a number of state bodies have recently displayed movement in this same direction. For instance, Rumney (2008) points to progress made under Washington and Michigan State laws. These jurisdictions have ''adopted an expansive definition of sexual intercourse that includes penetration of the vagina, anus or mouth with a penis, hand, tongue or inanimate object'' (p. 143). Michigan State defines penetration as ''sexual intercourse, cunnilingus, fellatio, anal intercourse or any other intrusion, however slight, of any part of a person's body'' (Criminal Sexual Conduct Act in Rumney 2008, p. 143). Both of these states cover circumstances where the perpetrator forces the victim to penetrate her or him. This year, the definition of rape used by the Federal Bureau of Investigation's Uniform Crime Report (UCR) has been brought much closer to these usages. Police departments have been asked to temporarily offer two sets of numbers, one under the old UCR definition and another under the new one, in order to make sense of the statistical surge the change is expected to generate (Savage 2012).

An Empirical Picture of Rape Between Women
While such changes are enormously important, it should be kept firmly in mind that reliable statistics for any form of rape are unlikely to exist. All rape researchers face well founded survivor distrust of the legal system, embarrassment and reluctance to Thinking Woman-to-Woman Rape 363 name perpetrators. Women assaulted by other women may fear homophobia or disbelief on the part of legal authorities, feel shame about their sexual orientations, or wish to protect their marginalized, queer communities by remaining silent. Underreporting almost certainly results (Renzetti 1988;Girshick 2002a;Ristock 2002;Burke and Follingstad 1999). Nonetheless, existing data is more than sufficient to convince that woman-to-woman rape must be accounted for in theory and practice. Because researchers of intimate lesbian violence frequently fail to look for sexual forms of aggression, Ristock (2002) was particularly struck by the number of lesbian informants reporting rapes by their partners. Girshick's (2002a) research reveals many reports of rape between women in and outside intimate relationships. Her monograph covers testimony about forty-nine assaults which took place in domestic situations and 42 outside such circumstances. The incidents Girshick (2002a) describes include sexual harassment, statutory assaults, violations of authority by therapists, teachers, and a doctor, date and acquaintance rapes, ex-partners turned stalkers and what she calls ''S/M sex gone awry'' (p. 109). This final category points to cases where abusers manipulate, misrepresent or ignore typical or agreed upon practices around consent with respect to S/M sexuality. Further accounts of womanto-woman rape outside intimate relationships are offered in Marlowe's (1999) autobiographical report and Morrissey's (2003) text on female killers, including Karla Homolka and Valmae Beck. The sexual sadism of these perpetrators is met by a silence that has led Morrissey to deem them ''feminism's limit cases'' (p. 134).
The National Violence Against Women Survey of 1999 reported that 11 % of 79 American women who had lived with a female partner had been raped or physically assaulted by her (Tjaden et al. 1999). Defining rape as forced and unwanted vaginal, anal or oral penetration with an object, weapon, hand or finger, Ristock (2002) found that of 102 lesbian-identified women reporting intimate violence, three survived rape attempts, twenty rape, nineteen sexual coercion and thirty-two emotional sexual abuse (for example, derogatory comments about sexual body parts). Renzetti (1988) reports that 48 % of 100 victims of intimate lesbian abuse experienced incidents of forced sex, four involving vaginal insertion of a gun or knife. Waterman et al. (1989) indicate that of 36 lesbian students and activists, 30.6 % reported forced sex in their most recent relationships and 8.3 % admitted forcing sex on their partners. Turell (1999) found that 12 % of 499 gay, lesbian and transsexual respondents reported same-sex sexual abuse with 9 % reporting forced sex. Irwin (2008) also offers testimony about rape between female partners.
In the coming sections, I will discuss two broad feminist approaches to the theorization of rape: the radical feminist approach and then Marcus's response to it. I will demonstrate that these foreclose the empirical fact of woman-to-woman sexual violence before suggesting an approach sufficiently nuanced to accommodate this reality.

Radical Feminist Theorizations of Rape
Radical feminism has been and remains enormously influential in academic, legal, and popular conceptualizations of rape. There is no doubt this branch of feminist thought and activism has extended protections and autonomy for women in important ways. These steps, however, have been based on a gender-paradigmatic framework. In the opening section of her groundbreaking 1975 text Against Our Will, Brownmiller mused: In the violent landscape inhabited by primitive woman and man, some woman somewhere had a prescient vision of her right to her own physical integrity, and in my mind's eye I can picture her fighting like hell to preserve it….Fleet of foot and spirited, she would have kicked, bitten, pushed and run, but she could not retaliate in kind. (p. 14, original emphasis) Such dualistic thinking is what Marcus (1992) seeks to oppose. As the latter puts it, Brownmiller represents ''rape as an inevitable material fact of life and assume[s] that a rapist's ability to physically overcome his target is the foundation of rape'' (p. 387). In terms articulated elsewhere in TPRP, such ideas enact ''a gendered polarization of the grammar of violence in which the male body'' is understood as invulnerable and dangerous while ''the female body is predicated by this grammar as universally vulnerable [and] lacking force.'' (p. 395).
Feminists since the 1970s have concluded that rape confirms and congeals women's subordination. Shafer and Frye (1977) argue that this form of violence reminds the victimized woman ''she is not a person'' (p. 342). Gender paradigmatic rapes indeed confirm in Foa's (1977) words, ''what we have all been trained to expect'' (p. 355), at least in the west. 1 Best known for her radical perspective on rape, MacKinnon (1982) characterizes sexuality as ''the primary social sphere of male power'' (p. 529). On this view, constituted by the heterosexual requirement, men are dominant, women submissive (1982; 1983). MacKinnon (1983) holds ''Our sexuality, meaning gender identity, is not only violable, it is (hence we are) our violation'' (p. 656, original emphasis). This perspective clearly cannot account for the experiences of people who know they have or have not consented (Moore and Reynolds 2004;Henderson 1991) let alone for female aggressors. Interestingly, MacKinnon (1983) demonstrates a sort of awareness of female perpetration by quoting Shafer and Frye: Rape is a man's act, whether it is a male or a female man and whether it is a man relatively permanently or relatively temporarily; and being raped is a woman's experience, whether it is a female or a male woman and whether it is a woman relatively permanently or relatively temporarily. (p. 650) This poses a logical and political difficulty. If heterosexual intercourse shows rape is occurring while rape constitutes men and women, it becomes impossible to identify men, women or rape. Such circularity sanitizes womanhood, demonizes manhood and flattens the multiplicity of each into one dimension. The defense of a favored perspective via such circular logic is always an invitation to pursue greater rigor through deconstruction. As emphasized by Marcus (1992), if womanhood is coextensive with victimhood, feminism is a hopeless enterprise. As to my present concern, MacKinnon forecloses transgressive rapes by forcing them back to her paradigm.

Theory and Politics of Rape Prevention
This section is a close reading of Marcus's (1992) ''Fighting Bodies, Fighting Words: A Theory and Politics of Rape Prevention,'' a text which works to loosen the implicit determinism of radical feminist theory. The subsections below extract two distinct arguments from the text. The first (Argument One) proves antithetical to theorization of woman-to-woman rape whereas the second (Argument Two) proves amenable to it. From the start, this piece indicates its postmodern affiliation by taking issue with the view that language-based postmodern theories are incompatible with feminist anti-rape action. Rather, Marcus (1992) insists ''rape is a question of language, interpretation, and subjectivity'' (p. 387). Therefore, a feminist fight against rape requires on the one hand a language for this violence and on the other an understanding of this violence as a language. The underdetermined nature of this second conclusion is a concern beyond my present scope. I am more interested in how Marcus follows up on this position than how she reaches it.
TPRP suggests the language of rape can only be based on ''political decisions to exclude certain interpretations and perspectives and to privilege others'' (p. 387). It aims to replace the focus on ''rape and its aftermath'' with a ''focus on rape situations themselves and rape prevention'' (p. 387, original emphasis). Its means to this end is to see rape as ''a process to be analyzed and undermined as it occurs'' (p. 388). So far, so good. Marcus proposes we see rape as a process in two ways which will be crucial to this discussion: (1) by treating rape ''as a linguistic fact'' (p. 388, original emphasis) enabled by cultural scripts and imposed by changeable ''narratives, complexes and institutions'' (p. 389) and (2) by differentiating various rape and attempted rape situations to ''develop the fullest range of prevention strategies'' (p. 388).
My criticism rides on the observation that TPRP largely drops the second mode of seeing rape as a process to focus on the first, which, ironically, becomes rigidly gendered as it is developed. This occurs through the logic of what I will call Argument One. In the remainder of the current reading, I first explicate this logic then continue to the rest of Marcus's text before offering examples of how Argument One has been used by other thinkers. Then I return to Marcus's second mode of reading rape as a process, which despite its underdevelopment, is emphasized by certain interlocutors.

Argument one: End Gendered Assumptions, End Rapes
Argument One holds that our assumptions about gender and sexual violence make men rapists and women victims. Essentially, the argument says, ''end gendered assumptions, end rapes.'' It is based on the insight that gender paradigms bear on what we do. Here, the male perpetrator/female victim paradigm subject to TPRP's critique remains rape's cause and effect. However, women's capacities for violence constitute fissures in the rape paradigm that open it to deconstruction.
I shall now continue, tracing this argument as TPRP goes on to propose three ways of taking a linguistic perspective. It suggests we attend to (1) cultural representations of women as wanting, provoking or susceptible to rape (2) socially structured and internalized language that restricts the sexes differently, producing non-combative responses to rape attempts in women and (3) rape's language-like structure which shapes verbal and physical gestures of those involved in a rape attempt.
Rigidifying of gender really starts to set in when Marcus defines rape ''as a scripted interaction which takes place in language and can be understood in terms of conventional masculinity and femininity as well as other gender inequalities inscribed before an individual instance of rape'' (p. 390). Later, rape is again defined, this time as ''a sexualized and gendered attack which imposes sexual difference along the lines of violence'' (p. 397). The metaphor of the script conveys three meanings. The first is narrative, suggesting rape is ''a series of steps and signals whose typical initial moments we can learn to recognize and whose final outcome we can learn to stave off'' (p. 390). The second is structural, referring to the ''social structures [that] inscribe on men's and women's embodied selves and psyches the misogynist inequalities which enable rape to occur'' (p. 391, original emphasis). Rape is here a scripted cause and effect of gender inequalities. The last is structural and deconstructive: the rape script as a framework for intelligibility which may be legitimized or exploded on the level of action. Evidently, Marcus intends her text to be explosive as opposed to legitimating. The rape script's form is ''a gendered grammar of violence'' (p. 392) which, like Brownmiller's passage cited above, ''predicates men as the subjects of violence and the operators of its tools, and predicates women as the objects of violence and the subjects of fear'' (p. 393). In a nutshell, Marcus's concern with the feminist approaches she critiques is that they have too often reproduced the linguistic conditions for rape she identifies in Argument One. Her views advanced so far display no doubt, however, that normative gendering exhausts the possible conditions for rape. Now something happens that really gets deconstructive sensibilities tingling. The article imposes a suspiciously rigid distinction separating violence between men from sexual violence. It reads: ''In subject-subject violence, each interlocutor expects and incites violence in the other, whereas in sexualized violence women are excluded from this community of violence…. Although on one level men are opponents, on another level they cooperate in their agreement to play the same game'' (p. 396). These remarks delegitimize men as victims, thus enabling the binary being advanced. Contra Marcus, male civilians do not agree to be attacked. To argue otherwise is to reinforce the feminization of victimhood. Strangely, this last passage follows the recognition that woman are ''neither in fact the sole objects of sexual violence nor the most likely targets of violent crimes'' (p. 394). These facts, which belie the opposition between sexualized and subject-subject aggression, are in no way taken into account by the theory advanced; the article continues as before to foreclose them. 2 To change the script, Marcus (1992) recommends that we ''develop a feminist discourse on rape by displacing the emphasis on what the rape script promotes-male violence against women-and putting in place what the rape script stultifies and excludes-women's will, agency, and capacity for violence'' (p. 395). She thus urges us to ''place ourselves as subjects who can…respond to aggression in kind'' (p. 397). Finally, she states, ''While the ethical burden to prevent rape does not lie with us but with rapists and a society which upholds them, we will be waiting a very long time if we wait for men to decide not to rape. To construct a society in which we would know no fear, we may first have to frighten rape culture to death'' (p. 401). This ethically unburdened ''we,'' contrasted with ''rapists'' and ''society,'' is presumably meant to signify women as a group. In effect, TPRP here insists on women's moral purity where rape is concerned. By the same token it ousts us from the society to which we belong and contribute, granted on what is generally speaking a decidedly unequal footing. Whereas men were earlier disqualified as victims, women are here disqualified as perpetrators. If any doubt remained, the logic of the gendered rape paradigm is neatly closed and completed.
Again, TPRP is particularly important because it has been adopted by many authors as an alternative to rigidly gendered perspectives; as a welcome relief from radical feminist binaries. The authors who draw on this piece generally do so by appealing either to Argument One or to Argument Two, but not to both. Along the lines of Argument One, Gatens (2000) contrasts Marcus's article with the works of ''Andrea Dworkin or Catherine MacKinnon who tend to define women in terms of their 'violability''' (p. 70). She defines rape as ''a 'specific technique' through which sexual difference is created and maintained'' (p. 70). From a Deleuzian perspective, her article seeks new ways for women to be and act. But despite this opening to gender transgression, defining rape as a technique for the creation and maintenance of binary sexuality obscures all rapes which transgress rather than uphold this structure.
Similarly, Herberle (1996) has drawn on Scarry and Marcus to contend that, like torture, rape solidifies men's phantasmic dominance: ''On the politicized terrain of gender relations the reality of women's pain may translate into the reality of male power'' (pp. 67-8). She urges deconstruction to reveal ''the naturalized social truths about gender and victimization that are embedded in the events of sexual violence itself rather than see[ing] only the immutable and singular 'reality' of sexual violence for all women in common'' (pp. 69-70). Although not shared by all women, these ''naturalized truths'' remain rape's monolith. Thus, although Herberle encourages a proliferation of narratives to represent women's diversity; the major thrust of her article is consistent with Argument One.
Finally, of particular interest is McCaughey and King's (1995) article, which documents the translation of Marcus's ideas into a specific practice. As instructors, they have observed that teaching about gender issues by showing videos depicting violence against women disempowers female students, even when those videos are shown in the interest of critique. Further, they worry that male students ''manage their behavior'' according to these depictions (p. 377). These concerns in mind, the authors have developed Mean Women, a collage of scenes from popular films in which women physically assault men for defense, revenge or fun. Marcus's thinking is used to ''imagine the female body as subject to change, as a potential agent of violence, and object of fear'' (p. 378). The authors explain, ''Both 'dangerous men' videos and our 'mean women' video offer sets of assumptions about male and female natures, each of which can shape behavior in complex ways. They differ mainly in that one is a male dreamworld and the other is a male nightmare' ' (p. 377). McCaughey and King (1995) have collected responses from students after screenings. Most men have been enthusiastic, suggesting they may ''be able to accept women as equals in this way'' (p. 381). However, some worry that the images reflect an increasingly violent society or suggest that the reverse roles would be received with disgust. Certain colleagues have wondered about negative influences on women's morality. McCaughey and King respond with certainty their video will not undo students' feminine socialization or ''make women inappropriately violent'' (p. 384). I quote: ''Only male viewers regard the 'mean women' fantasy as a danger to be avoided. Women do not express fears about making the world a more violent place because they already know too well that the world is violent'' (p. 381). The enthusiastic responses of female students suggest Mean Women may be taken up in various ways. One comments, ''I think watching females take action and defend themselves is a good thing. In a way it empowers women, makes them feel more competent.'' Another remarks, ''It means that women are finally taking their place as violent, bloodthirsty, savage, testosterone-fuelled egomaniacs next to the men. I love women with guns!'' (p. 380).
McCaughey and King's (1995) intervention is based on women's abilities to escape stereotypical feminine subjugation and fight back. Yet its cavalier attitude toward women's violence reveals an assumption that women are effectively socialized to femininity and so never really experienced as dangerous. As in TPRP, violence of the really dangerous kind-the problematic kind that we should counter theoretically, politically and physically-remains the exclusive providence of men.

Argument Two: End Gendered Assumptions, Reveal Rapes
Unsurprisingly given the textual trajectory just traced, when I first read TPRP, the contours of Argument One leapt immediately to my attention. Given my research interest in woman-to-woman rape, this naturally generated the foregoing critical response. I was thus initially puzzled to come across Ristock's (2002) positive reference to this piece in the context of her work on violence between lesbians. Ristock suggests that in striving to ''deconstruct limiting categories and discourses'' (p. 21), Marcus enables the recognition of violence between women. My puzzlement only grew when I encountered articles by Hirsch (1994) and Buss (2009) which drew on TPRP in formulating perspectives that were in fact quite amenable to my work. Looking more closely, I identified the logic of the second argument.
Argument Two is distinct from Argument One in form and aim. It holds that our assumptions about gender and sexual violence render gender transgressive rapes invisible while gender transgressive rapes in turn constitute the deconstructive fissure in the gendered rape paradigm. Whereas in Argument One, structures of paradigmatic gender bore on doing, in Argument Two, they bear on knowing. Conformity with the paradigm is not the necessary cause of rape events because manhood and womanhood do not necessarily map onto aggressor and victim,  respectively. Argument Two is consistent with Marcus's (1992) first mode of envisioning rape as a process, namely, the differentiation of rape situations aimed at establishing a battery of rape prevention responses. Basically, the argument goes, ''challenge (gendered) assumptions, reveal rapes.'' This deconstructive view is equipped to see women's self-defence and women's unjustified violence as possible forms of gender transgression. For on this view, the invisibility of paradigm transgressing incidents sustains the paradigm rather having the gendered paradigm reproduce itself in all incidents we could or should label ''rape.'' Along closely related lines, Giorgio (2001) suggests the existence of lesbian battering ''disrupts the overarching domestic violence discourse of gendered binaries and offers us the potential to attend to intimate relationship violence in its multiple formations'' (p. 252). Although Hirsch uses paradigmatically gendered language, her article foregoes heterosexist assumptions about rape's causes and effects. Hirsch (1994) uses TPRP to compare American to Kenyan media constructions of the St. Kitizo Mass Rape. She focuses on how ''media accounts of rape reproduce 'culturally appropriate' stereotypic images of rape victims and rapists'' (p. 1054). American representations elide important cultural and political elements of the Kenyan context. For Hirsch (1994), attending to rape's various forms and contexts illuminates hidden aspects of women's oppression while encouraging innovative prevention practices. She applies the second mode of rape as process to the first of Marcus's three angles on rape as linguistic fact, i.e., cultural representations of women: As Marcus and other scholars from feminist and critical perspectives demonstrate, media representations of women and violence are potential sites for contesting cultural ideas about deviance and law and about gender, race and class relation…Following Marcus, I argue that the uncritical assumption that rape ''experiences'' are identical is precisely the means through which scripts of rape are reinscribed and thereby depict women as ''always, already '' victims. (p. 1025) By insisting on the various forms this violence can take, Hirsch (1994) follows Marcus in ways that implicitly open the door to woman-to-woman rape. More precisely, rather than moving from rape's gendered representations to its gendered facts, she exposes overgeneralizations to reveal obscured forms of sexual violence. Buss (2009) deploys TPRP in discussing the Rwanda Tribunal to argue that viewing rape as a genocidal tool blindingly frames only those offenses committed against Tutsi women. She thus situates herself against Brownmiller's and MacKinnon's analyses of genocidal rape. In her words, ''just as Hutu women are hard to see as victims of the genocide, so too are men subjected to rape and sexual violence'' (p. 159). Consistent with Argument Two, her article concludes favouring a focus on specific rape situations: ''It is at this level of detail, with all the inconsistencies and complexities revealed, that it becomes possible to imagine a situation where rape is not inevitable'' (p. 161). Hirsch's (1994) and Buss's (2009) analyses remind us of what Crenshaw (1989) has called ''the complexities of compoundedness'' (166). Racism as well as other axes of oppression such as ethnocentrism, classism and ablism, intersect with gender as rape incidents and social responses to them unfold. Thinking in such intersectional terms, as Crenshaw (1989) does in her already classic piece, ''Demarginalizing the Intersection of Race and Sex: A Black Feminist Critique of Antidiscrimination Doctrine, Feminist Theory and Antiracist Politics,'' is another potential of Argument Two that Marcus does little to pursue. For example, in a war context like the one Buss (2009) discusses, where men and women alike are martial rape victims, we would do well to counter Marcus's gendered logic by remembering that racism or ethnocentrism may be more central to any given sexualized aggression than sexism.

Synthesizing Arguments One and Two
Suggesting only an unspecified proliferation of preventative measures in reaction to myriad forms of rape, Argument Two offers less in the way of clear solutions on the level of structure than Argument One. Encouragingly, as Buss (2009) suggests, this variability does make it easier to imagine a society free of rape by denaturalizing the phenomenon. Consistent with Marcus's first mode of seeing rape as a process, Argument Two makes room for woman-to-woman rape by focusing on detail and inconsistency. At the same time, Argument Two deconstructs Argument One by illuminating what Argument One implicitly denies. Argument Two therefore appears far more promising when it comes to theorization of woman-to-woman rape. Ironically, Argument One depends upon the male aggressor/female victim paradigm and thus to a great extent re-inscribes ''the gendered polarization of the grammar of violence' ' Marcus (1992), p. 395 sets out to deconstruct.
However, with some nuance these two arguments could synthesize in a way that would advance sophisticated, contextualized theorization of woman-to-woman rape. Argument One would be more accurate if replaced by the slightly weaker claim that the rape script entrains paradigmatic rape's statistical and discursive dominance. Socialization has effects; sexual violence is clearly more normative for men. Crucially though, the suggested reformulation doesn't render gender transgressive rapes invisible. On the contrary, it could frame a discussion about how discursive dominance creates the illusion that statistical dominance is statistical monopoly. This in turn could be contextualized by a broader discussion, initiated here, regarding the differences and interactions between deconstructing knowing with doing on the one hand and deconstructing doing with knowing on the other.
A Butleresque opening to gender transgression would be illuminating. Butler (1996) has used the terms ''heterosexual matrix'' and more recently ''heterosexual hegemony'' to refer to our culturally dominant gender norms. These terms suggest the ways one's sex, associated gender characteristics, sexual desires and behaviors are expected to correlate such that masculine men desire feminine women and vice versa. On this thinking, what appears natural is rather a repetition of norms. Butler's terms also refer to the social punishments and exclusions regulating this order. To rape and to be raped are elements in normative repertoires of manhood and womanhood, respectively (Butler 1992). Consistent with Argument One, sexual violence is often simultaneously a way of doing gender and a way of enforcing it. Along similar lines, Lichtenstein (2005) uses Butler's performative theory of gender Thinking Woman-to-Woman Rape 371 as a framework for her research on domestic violence. In keeping with Butler's view as described thus far, Lichtenstein sees domestic violence as an enforcement of gender hierarchy that aims to preserve hegemonic heterosexuality. ''This temporality suggests that gender absolutes in terms of hyper-masculinity and hyperfemininity can (and will) be enforced to maintain the status quo'' (p. 702). However, Butler's is not a static view, but a diachronic one. Essential to her theorizing is the perfectly reasonable insight that, to varying degrees, in ways that are culturally circumscribed, girls will internalize and perform normatively masculine ways of doing while boys will internalize and perform normatively feminine ways of doing. Repetition is inextricably connected in her work to opposition ''from within the very terms by which power is reelaborated'' (Butler in interview with Meijer and Prins 1998, p. 278, original emphasis). Resignification occurs as the terms and categories we inevitably use are deployed in unexpected ways under unexpected circumstances. Where gender is concerned, this process constitutes ''gender trouble'' (Butler 1999). Woman-to-woman rape may well be a copy of a copy of a stereotypically masculine copy, but we mustn't forget that the same is true of its man-to-woman analogue. Like other aspects of heterosexual hegemony, we should expect the male aggressor/female victim paradigm to be transgressed, consistent with Argument Two. If gender trouble is, in Butler's (in interview with Breen et al. 2001) words, ''a way to lend a language of description to what has been foreclosed from normative discourse for too long'' (p. 22), it is precisely what is needed in the effort to theorize woman-to-woman rape. Any assumption that transgressions will only come in the form of women's self defense and never in the form of rape itself can only be a counterfactual reinscription of binary gender. 3

On Physical Self-Defense
In the words of Perilla et al. (2003), ''The idea that any person, regardless of gender, has the capacity of committing violent acts is a sobering thought'' (p. 34). Physical prevention strategies advanced by Marcus (1992) and McCaughey and King (1995) could be developed in a manner mindful of gender transgressive and gender paradigmatic sexual violence. Writing under a pseudonym, Marlowe (1999) describes a self-defense course which she joined to address post-traumatic stress eight years after being raped by another woman. When students were asked why they registered, Marlowe remained silent. Only with great trepidation and because she was paired with a friend did she complete an exercise where another woman ''simulate [d] an attempted rape'' (p. 400).
In Marlowe's (1999) words, How much easier it would have been if I had been able to talk through my fears beforehand. There seemed to be no place in the group for that. It was as though rape was only ever a question of a male aggressor, female victim, so a female-to-female simulation could not possibly trigger any traumatic memories. (p. 401) Importantly, I am not taking issue with self defensive strategies per se. Extending TPRP to domestic violence, Lichtenstein (2005) finds support for Marcus's supposition that fighting back constitutes an effective response to this problem. Very encouragingly, her study on domestic abuse and HIV risk in the American Deep South found women who fought back were most successful in reducing or ending abuse. ''In this way, they were fulfilling Marcus's (1992) challenge to women to exert will, agency and the capacity for violence'' (p. 712).
Given the dominance of gender paradigmatic rape, I even accept that, as McCaughey (1997) makes clear, classes for women only are a good thing. But whatever the gender composition of services, it is important that they be explicitly open to and mindful of non-paradigmatic survivors (Sokologg and Dupont 2005). Modification of the gendered language used in and about self defense courses would constitute an opening to such individuals and by the same token raise awareness about women's capacities for unjustified violence. If an important part of selfdefense culture is the non-violent values it inculcates, these lessons, in combination with the acquisition of self-defensive abilities, are worthwhile for all of us.

Conclusion
Marcus has rightly been credited with producing one of the most politically forward and gender transgressive theoretical renderings of rape. Yet looking at her effort through the lens of woman-to-woman rape illuminates certain weaknesses in the form of gendered assumptions. I have shed light on these matters by detailing two separate arguments that have been drawn from TPRP. Argument One is consistent with Marcus's (1992) treatment of rape as a culturally enabled and contingent ''linguistic fact'' (original emphasis, p. 388). It holds that rape culture is sustained by our assumptions about gender and rape. Although assuming that the male perpetrator/female victim paradigm is deconstructable, it maintains this paradigm as essential to any given rape situation. Argument Two draws on Marcus's (1992) suggestion that we differentiate among rape situations to establish multiple modes of rape prevention. It suggests that transgressive rapes, rendered invisible by the dominant paradigm, constitute the latter's deconstructive fissure. Thus, we err in believing that conformity with the male perpetrator/female victim paradigm is rape's necessary cause. Argument One obscures what Argument Two might reveal; Argument Two is equipped to open Argument One to woman-to-woman rape survivors. It could also foreground issues of racism, ethnocentrism or other forms of oppression, as mentioned above. I have suggested that a synthesis of Arguments One and Two in Butleresque form would better adhere to the empirical realities explored here. Where self defense strategies are concerned, modification of language along with an awareness of gender transgressive violence could put such a synthesis into practice.
Thinking Woman-to-Woman Rape 373 At its best, seeing rape as a linguistic fact highlights how survivors of gender paradigmatic rapes line up on the side of heterosexual hegemony while survivors of gender transgressive rapes regularly find themselves shut out of our texts. Since rape, gender and language are deconsructable, it stands to reason that, despite its cultural power, the male aggressor/female victim paradigm fails to fully enforce the normative articulation of these three phenomena. We therefore need a new capacity for thinking about the all too expected rapes of Argument One without foreclosing recognition of the jarringly unexpected rapes Argument Two reveals. Gender paradigmatic rapes and gendered assumptions mutually reinforce each other, teaching women that ''rape has always already occurred and women are always already raped or always already rapable'' (Marcus 1992, p. 389). Deconstruction of such assumptions in a way that reveals a full range of women's violent capacities may temper such life-inhibiting beliefs, encourage women to defend themselves and thus discourage attacks against women in general, as Marcus hopes. Gender transgressive rapes reside in the shadow of the male aggressor/female victim paradigm, a silencing fact some female assailants take advantage of (Hassouneh and Glass 2008;Girshick 2002a;Giorgio 2001). As Girshick (2002b) notes, while all rapists tend to deny culpability, ''Female perpetrators' denial is supported by our society'' (p. 1508). A more complete deconstruction of the rape script than the one offered by Marcus would cease to support such denial because it would recognize female perpetrators. Defining rape ''as a sexualized and gendered attack which imposes sexual difference'' (Marcus 1992, p. 397) will not help with this, regardless of the fact that paradigmatic cases adhere to precisely this logic. A more appropriate definition would follow Washington or Michigan State models (Rumney 2008) in delineating a range of forms of forced contact definable as rape while remaining neutral with respect to assailant and victim sex. At least in determining whether an assault has occurred, the focus should be on behaviour over identity (McPhail et al. 2007). Consistent with Argument Two, this would create space for focus on the minutia of specific rape situations including their ''inconsistencies and complexities'' (Buss 2009, p. 161). In light of the reality of woman-to-woman rape and other gender transgressive sexual assaults, nothing less will do.