The promise and peril of agency as motion: A feminist new materialist approach to sexual violence and sexual harassment

Abstract

Organizational scholars have established that sexual harassment, the most studied kind of sexual violence, is an organizational problem. Extending this work, we analyze two critical events regarding sexual violence in the United States—one in the military and another at a university—in which discourse detracts from understanding the problem in this way. We draw upon feminist new materialism and its primary method—diffraction—to track ‘cuts’, the practices that simplify and pause agency’s complex, perpetual motions. Our analysis shows that agency moves in discussions about the aftermath of violence. That momentum highlights the organization’s capacity to respond to rape. Even so, during discussions about enacting violence, the perpetual motion of agency congeals around discrete humans, thereby maintaining assault as an individual act. These cuts, whereby agency pauses on individual perpetrators, obscure how organizational dynamics make sexual violence more or less likely to occur. We suggest that a focus on agency’s kinetic qualities can help feminist scholars continue to highlight how the systemic aspects of harassment and other forms of violence become hard to notice.

Decades ago, Ferguson (1994) invited organizational scholars to advance more theoretically robust, rigorous analyses of gender and inequity. Since then, others have echoed the call. Townsley (2003) and later Harding et al. (2013) cited the need for more feminist organizational scholarship. Specifically, Townsley (2003) suggested that scholarship could rely more on Donna Haraway’s work. Heeding this call, we turn to feminist new materialism (FNM), a theory that emerged from Haraway and related thinkers and is gaining traction in interdisciplinary communities (Coole and Frost, 2010; van der Tuin, 2011b). We adopt tenets of FNM—specifically its approach to agency—to extend research on sexual violence, a central topic for feminist academics and activists. For purposes of this article, sexual violence includes harassment, stalking, rape, and unwanted touch. As we proceed, we caution that when agency appears to pause on individuals, those pauses obscure the organizational and systemic dynamics of sexual violence. By focusing on agency’s motion, scholars who are attentive to the risks of an FNM approach to agency can bring those organizational dynamics to the fore.

To arrive at this argument, we examine how agency circulates in organizational discourse about sexual violence. We focus on the aftermath of two high-profile incidents at US organizations where sexual assaults are at some of the highest rates: higher education and the military (Fisher et al., 2000; Pacatte, 2013). The first occurred when a lawsuit drew widespread public attention and outrage to the historically silenced problem of military sexual violence. The second occurred after a judge found a US university had violated federal anti-discrimination laws regarding sexual violence, and, almost a decade later, the payout to plaintiffs remains one of the largest sums in all similar cases. Both of these incidents occurred well before #metoo gained global momentum, and scholars have argued that these incidents helped spur the current activism (e.g. Heldman et al., 2018). Focusing on discourse surrounding these incidents, we assess what Barad (2003) called cuts, the processes whereby agency—which FNM scholars assume is in perpetual motion—comes to settle on select entities or components of an organization. We show that, although agency often moves fluidly in discussions of organizations’ responses to sexual violence, it rarely moves in discussions of who or what can commit or contribute to sexual violence. We suggest that cuts, which pause agency on individuals, though seemingly necessary for holding sexually violent perpetrators accountable, also obscure the organizational dynamics of sexual violence. Moving forward, our purpose is to amplify existing feminist efforts to understand sexual violence as an organizational phenomenon.

Sexual harassment as an organizational process

Among organizational scholars, sexual harassment, the most studied type of sexual violence, is well established as an organizational phenomenon (Chamberlain et al., 2008; Keyton and Rhodes, 1999). Rather than frame harassment and related forms of sexual violence as isolated events within an organization, Hearn and Parkin (2001) argued that violation is part of any organizing process. Similarly, Harris (2013b) asserted that violence is a component of organization, not something that simply happens in an organization. Given this framework, scholars have studied not only individual targets (e.g. Kensbok et al., 2015) and perpetrators (e.g. Lucero et al., 2003) of harassment, but also organizational processes including policies (Dougherty and Hode, 2016), cultural norms about service work (Brunner and Dever, 2014; Liu et al., 2014), dialectics or paradoxes (McGuire et al., 2006), and reporting practices (Harris, 2017). These organizational studies resonate with interdisciplinary feminist scholarship and activism that, for decades, has underscored the importance of locating violence in systems, not individual ‘bad apples’ (Brownmiller, 1975; Enck-Wanzer, 2009; Wieskamp, 2018). In sum, much of the organizational scholarship on sexual harassment focuses on the organizational processes that support or undermine sexual violence.

Despite the availability of this scholarly framework, varied organizational processes obscure organizations’ contributions to sexual harassment and other forms of sexual violence. In other words, organizational aspects of violence frequently ‘disappear from view’ (Catley and Jones, 2002: 34). For example, organizations expect individuals to immediately label their experiences of sexual harassment (Taylor and Conrad, 1992), even though raced and gendered organizational processes deter people from identifying violence as such (Harris, 2017; Richardson and Taylor, 2009). When harassment remains unlabeled, organizations can undercount its prevalence, stall investigations, and assign too few resources to prevention and response programs (Harris, 2013b). Eyre (2000) showed that organizational discourse (e.g. discussions of academic freedom in universities) can deflect an organization’s attention from harassment. Organizations also maintain harassment as a private problem, not an organizational process, by trivializing harassment or framing it as a mere misunderstanding (e.g. by saying harassment was a joke, not something serious; Clair, 1993a, 1993b). These processes make it difficult to notice the organizational dynamics that support harassment.

Compared with other topics in organizational research, such as rituals (Koschmann and McDonald, 2015) or branding (Mumby, 2018), the literature on sexual harassment is somewhat unusual because it rarely makes questions about agency explicit. Scholars often draw upon theories of agency when they establish organizational, not individual, explanations for phenomena (Castor and Cooren, 2006; McPhee, 2004). Brummans (2015) declared that agency is ‘fundamental’ to organizational processes. Although agency is central to many organizational explanations for phenomena, and sexual harassment has been established as an organizational phenomenon, existing theories of agency and research on sexual harassment rarely if ever overlap in this academic community. We assert that the disconnect occurs because of disciplinary silos that segregate a rich tradition of feminist theory from prevailing conceptualizations of agency in organization studies. Others have made similar observations. Broadfoot et al. (2018) argued that organizational scholars have not yet accounted for the gendered-raced implications of prevailing theories of agency. In addition, Broadfoot and Munshi (2015) asserted that the field’s understandings of agency are rooted in colonialist logic that frames some humans as non-agentic objects. Without incorporating a feminist intellectual tradition, to use prevailing theories of agency to study sexual violence could be problematic.

The potential problems are well established outside the field. They revolve around a subject/object split that locates violence in either perpetrator or target but not in organizational processes. Though sometimes implied, these difficulties are not explicit in existing organizational scholarship on sexual harassment. Under prevailing theories of agency, ‘a subject of violence acts on an object of violence’ (Marcus, 1992: 399). In this equation, if an individual experiences violence, they become a passive object (MacKinnon, 2017). Violence attaches the categories ‘subject’ and ‘agent’ to men who enact violence while the people who most often experience violence—White women, people of color of all genders, and LGBTQ individuals—remain outside those categories. This ‘zero-sum’ approach (Phillips, 2000), in which an individual is either an active, agentic subject or a passive object, leaves no space for resisting a sexually violent culture: if a person experiences violence, it means they have no agency. Worse, to acknowledge agency anywhere other than in the individual who commits violence risks suggesting that people who experience violence were active participants in their own victimization (Mardorossian, 2014). This approach maintains a heteronormative gender binary (Patterson, 2016). Moreover, responses to violence that emerge from this framework often urge potential targets to stop behaviors that could ‘invite’ assault (Hall, 2004). Under such understandings of agency, even when communities hold perpetrators accountable for wrongdoing, they rarely consider how broader systems and processes support sexual violence (Chen et al., 2016; INCITE!, 2016). Thus, for feminists who study sexual harassment and other forms of sexual violence, prevailing understandings of agency—ones that focus on the rational choices and behaviors of individual humans and retain a subject/object split—offer a theoretical and political dead-end. To understand agency in a manner useful for resisting the organizational dynamics of sexual violence, feminists must develop theories that undo divisions between subject–agency and object–passivity. FNM, a theory just beginning to influence organizational scholarship (see Orlikowski and Scott, 2015; Scott and Orlikowski, 2014), does just that. We use it to consider how agency both supports and detracts from understanding sexual violence and sexual harassment as an organizational problem.

Feminist new materialist approaches to agency: metaphors of movement

Feminist new materialists’ discussions of agency depart from dominant understandings that locate agency exclusively in human decision-making and behavior (Coole and Frost, 2010). Noting that this dominant approach reifies the passive-object, active-subject split, feminist new materialists developed an approach to agency that relies on metaphors of movement. Dolphijn and van der Tuin (2012) said that FNM is ‘fascinated by … movement as it travels in all directions’ (p. 113). Drawing on this language of motion, Barad (2003) described agency as ‘dynamism’ (p. 818) and an ‘ongoing flow’ (p. 817) that results in ‘reconfigurings of the world’ (p. 818). Relatedly, Bennett (2010) established agency as ‘a movement away from some initial condition or configuration and toward something else’ (p. 457). This FNM approach to agency begins from the premise that individual humans and nonhumans are not separate from everything else in the world. Indeed, Barad (2003) relied on physicist Bohr who ‘rejects the atomistic metaphysics that takes “things” as ontologically basic entities’ (p. 813). Scholars employing this framework—one strand of an emerging relational ontology—abandon the terms ‘agent’, ‘actor’, and ‘actant’ because they separate and segment agency and reify the problematic subject–object bifurcation previously described (Dolphijn and van der Tuin, 2012). As Hultin and Mähring (2017) put it, ‘In this flow [of agency], we cannot distinguish any actors or entities that move’ (p. 571). Instead, agency is movement.

Agency’s movement is obscured when agency appears to temporarily ‘congeal’ (Coole, 2005: 140) and settle in a ‘provisional concentration’ (p. 132). When agency appears settled it may look as though discrete actors exist. Rather than assume these entities as given, Barad (2007) focuses her analytic work on the processes through which seemingly separate actors appear. She calls these processes boundary-making practices. Like agency, these ‘boundaries do not sit still’ (Barad, 2003: 817), but unlike agency, the boundary-making practices produce ‘exclusions’ (Dolphijn and Van der Tuin, 2012: 54). They leave out some of agency’s complex and ongoing movements, and some boundaries endure more than others. The boundary-making practices that persist, ones that have a ‘long and ossified’ history (Warfield, 2016: 2), are dubbed ‘cuts’. As Dale and Latham (2015) explained, cuts are consequential because they ‘produce inequalities and hierarchies, relations of domination and exploitation’ in organizations (p. 171). In their study, the ‘cut’ that separated ‘agents’ from the flow of organizational agency required humans to be self-sufficient. When humans were not clearly separate from the flow of agency and depended on technology for their bodily needs, as in the case of a person who was disabled, they became objects rather than agentic human subjects. That classification was the basis for discrimination. Because FNM carefully attends to these relations of exploitation, it pursues this central question: ‘Which cuts are made, when, and for whom?’ (van der Tuin, 2011a: 37). It tracks the implications of cuts, the consequences of those abiding pauses in agency’s motion.

Given these assumptions about agency, Bennett et al. (2005) argued scholars should ‘take the cascade as the unit of analysis’ and focus on movement (p. 457). Dale and Latham (2015) added that ‘we need to explore how organisational processes are involved in the “cuts”’ because those cuts produce ‘inclusions and exclusions, inequalities and hierarchies’ (p. 179). As we continue, we connect these concerns regarding moving agency to sexual violence.

Methods and methodology

Our grounding in FNM and the metaphor of movement informs not only our theoretical approach to agency, but also our methodology and methods. Haraway (1997) first began discussing diffraction, the most common feminist new materialist method, more than two decades ago. Since then, the method has been adapted by scholars in both the social sciences and the humanities (e.g. Lenz Taguchi and Palmer, 2013; Wurth, 2014).

The inspiration for diffraction comes from physics (Barad, 2014), and an example using waves and water illustrates some of diffraction’s core methodological principles. If a stone falls into a pond, ripples in the shape of concentric circles spread over the water’s surface from the point where the stone dropped. If a second stone falls near the first, the two sets of circular ripples, one from the first stone and one from the second, eventually encounter one another. When the two sets of ripples meet, some ripples cancel each other out and some amplify each other. This process, in which a new pattern of conical-shaped ripples emerges, is diffraction. Whereas more traditional qualitative methods would generate insight about the original stone (e.g. an organization or a social phenomenon) based on the pattern of concentric circles, diffraction focuses instead on the relational process whereby the ripples meet and produce the cone-like pattern. In this diffractive methodology, knowledge claims derive from different elements of the world co-mingling, or, as Barad (2007) said, intra-acting. Accordingly, the stones need not be similar, and the number of stones is not particularly important so long as more than one is involved. The stones might be made of different materials or be differently sized. Diffraction could still be performed if the original concentric circles emanated from dissimilar objects like a stone and a twig or a stone and a set of statements.

Adapting this metaphor to scholarly work, feminist new materialists use diffraction to draw together ostensibly ‘separate entities (and separate sets of concerns) with sharp edges’, things akin to the stones, in order to ‘promote their mutual consideration’ (Barad, 2003: 803). Diffraction is also sometimes described as ‘re-reading’, drawing together texts, ideas, and objects from disparate times or traditions (Dolphijn and Van der Tuin, 2012; Grosz, 2010). This principle manifests in FNM scholars’ reliance on seemingly incommensurate theoretical assumptions (e.g. Harris, 2016), texts (e.g. Sehgal, 2014), disciplinary locations (Geerts and Van der Tuin, 2016), and writing styles (e.g. Barad, 2014; Minh-ha, 1988). For example, Ulmer (2016) interviewed teachers about educational policy and used two distinct theoretical frameworks to analyze the data. In a different example of diffraction, Davies et al. (2013) authored a biography about a collective of people. Instead of focusing on one human’s life story, these authors considered how their individual stories amplified and differed from one another. The resulting patterns formed the basis for their collective biography. Each of these studies started from the assumption that boundaries around entities—including academic communities, humans, and organizations—are enacted, not solid. In other words, they are in movement. For purposes of the present study, these methodological principles influence our focus on critical events in higher education and the military, our selection of data, our analysis, and our writing style.

We employed diffraction in this study first by drawing together high-profile events from two different US organizational contexts, the military and higher education. Following Kaiser and Thiele (2014), we moved away from ‘discrete, given entities as units of analysis’ (p. 166), an approach that would consider either the military or higher education separately as if they were stones. Instead, we considered the impact of high-profile incidents at these organizations through one another, at the point where the concentric circles of ripples meet.

The first incident happened in the US military. In 2011, a lawsuit was filed against then Secretary of Defense, Robert Gates, and his predecessor, Secretary of Defense Donald Rumsfeld, for failing to address military sexual violence (Cioca v. Rumsfeld, 2013). The lawsuit helped draw widespread public awareness to what had historically been a silenced problem (Wieskamp, 2018). This catalyzed additional advocacy efforts: additional lawsuits were filed against the Department of Defense, increasing numbers of service members began demanding that the military address the issue, legislators began speaking out about the problem, and the critically acclaimed documentary, The Invisible War (Dick, 2012), featuring stories of several claimants in the initial lawsuit, was released. These efforts exposed both the high rates of soldier-on-soldier sexual violence within the military, as well as the low rates of reporting, trial, and conviction (Gillibrand, 2015). The sudden media spotlight on the military sexual violence left its leaders scrambling to address the problem.

The second incident happened at a US university, the pseudonymous Western University (WU), a large public institution. Per the federal civil rights law, Title IX of the 1972 Education Amendments, ‘even a single instance of rape or sexual assault’ qualifies as discrimination (Students Active for Ending Rape, 2008: 2). In the early 2000s, a judge found this particular university had been deliberately indifferent to the risk of sexual violence and had therefore violated Title IX and ordered the university to pay several million dollars to the people who had been violated. As an early case of its kind, it garnered extensive media attention. A few years later, after extensive changes to their processes, WU became known for having some of the most robust violence response policies among US universities. They were one of the first schools to be scrutinized in this way. Shortly following that incident, similar Title IX cases started to occur more often (Reynolds, 2019). In 2010, years after the case at WU was decided, three universities were under federal investigation. By late 2018 more than 300 cases were open. The incident at WU was a precursor to the current attention to Title IX.

Our decision to consider these two high-profile events together emerged when we began conceptualizing this study. Each author had already conducted independent research on sexual violence in the US military and US universities (Harris, 2013a; McFarlane, 2015; Wieskamp, 2015). We were struck by the similarities in our findings despite the disparate organizational contexts and methods we had used. The first two independent studies contextualized the military’s responses to The Invisible War and Cioca v. Rumsfeld through rhetorical analyses of policies, laws, historical facts, leaders’ statements, and media coverage about sexual assault. The last offered a grounded theory analysis of interviews, field notes, and organizational documents to investigate how Title IX impacted WU in the wake of the judge’s decision that the university had violated the law. We wondered what insights might yield if we considered the high-profile incidents in these organizations simultaneously. In line with feminist new materialist methods, we decided to ‘make connections between entities that do not appear to be proximate’ (Barad, 2007: 74), including organizational contexts and our originally separate studies.

We also enacted diffraction in decisions about which data we would re-analyze. The combined documents from our original research amounted to more than 560 single-spaced pages. They included court cases, newspaper articles, policies, press releases, field notes, statements from organizational leaders, and transcripts of interviews the researchers conducted with organizational members. In traditional rhetorical scholarship in our home discipline, analysts rely exclusively on already available, public documents: court cases, newspaper articles, policies, press releases, and public statements from leaders (Jasinski, 2001). Qualitative researchers in our home discipline, by contrast, rely much more on interviews and field notes (e.g. Malvini Redden and Scarduzio, 2018). In keeping with diffraction, we retained both kinds of data and considered them together, even though we did not have every kind of data for both high-profile incidents. We are influenced by Endres et al. (2016) who established rhetorical field methods, a lauded fusion of rhetorical and qualitative techniques. Though not grounded in FNM, these authors emphasize that drawing together these two traditions generates new insights for both, a principle that resonates with diffraction.

We continued diffraction while analyzing our data by considering how an existing theoretical concept developed in one setting, the university, would produce insights when considered in relationship to our collective data, a practice akin to the re-reading we mentioned earlier. Harris (2013a) developed the term ‘organizational handicapping strategies’ to describe how organizations deflect responsibility for sexual violence. Social psychologists have long observed ‘self-handicapping strategies’, or those actions that individuals take ‘to externalize (or excuse) failure and internalize (reasonably accept credit for) success’ (Berglas and Jones, 1978: 406). Noting the term’s ableist connotations, Harris adapted the term to show how organizations, not only individuals, excuse their own failures and claim credit for their successes.

Beginning with this idea, we re-read data from the original studies. We identified instances where each organization credited itself for success or excused itself for failure in responding to sexual violence. These instances often involved debates about who or what could act. In other words, these organizations’ strategies to position themselves as capable respondents to sexual violence seemed related to agency. We again re-read the original data in its entirety, focusing on who or what was said to act and noting resonance or dissonance within and across the three existing studies. Given the pervasiveness of liberal, rational actors across many social sectors in the United States, we expected to find mention of clear agents. Instead, we found complex, contradictory, and fluid attributions of agency. At that point, we began to think about metaphors of movement, and that emergent metaphor led us to FNM as a theoretical frame.

Finally, we enacted diffraction by retaining elements of the divergent modes of analysis and writing from our original studies which included both rhetorical and qualitative traditions. The result is a ‘plastic’ analysis that is ‘unscripted and malleable’ (Ulmer, 2015: 1105). Following Davies (2014), we emphasized motion in our own practice by trying not ‘to fix the analytic process so that it can be turned into a methodic set of steps to be followed’ (p. 734). We focus more on ‘complexity [and] competing arguments’ (Ulmer, 2015: 1104) than on ‘sorting’, ‘grouping’, or ‘coherence’ (Jackson, 2013). To provide this, we adopt Braidotti’s (2006) feminist new materialist interest in writing that offers ‘creative links and zigzagging interconnections between discursive communities’ (p. 7). The resulting analysis draws forth the constant boundary-making practices in which ‘cuts’ separate individual humans from agency’s movement in ways that risk supporting a violent status quo.

Agency in motion promotes organizational responsiveness (but not always)

In many instances, organizational members relied on the idea that agency is in motion. They did so to offer organizational explanations for how the military or the university responds to sexual violence. These explanations appeared almost exclusively in discussions of the aftermath of sexual violence, and not in every instance. On the contrary, in some discussions of the aftermath of violence, a discourse of moving agency preserved the status quo and supported subtle forms of victim-blaming.

At WU, members of the organization relied on the flow of agency to emphasize the efficacy of organizational systems. Ashley, ¹ a WU official responsible for the university’s sexual violence response processes, reviewed an informed consent form the first author created. She made this comment in response to a sentence attributing ownership of a sexual violence policy to her office:

On the second page [of the form] you talk about, ‘If you’re aware of the [Office of Violence Response’s] policy—’. You know, one of the things that we try to remind people of is that the policy is not ours, it’s the university’s. We’re the office that enforces it. So I would say, you know … it gets a little bit complicated actually, … [One] policy was signed off by our Chancellor. The other one is signed off by our President. So, you know, it’s accurate to refer to them as the [Western University] policy.

Ashley challenges the interviewer’s attribution of policy ownership by tracing the flow of agency. She allows that her office enforces the policy but resists the suggestion that her office created that policy. In so doing, Ashley pushes back on a ‘cut’ the first author made, one that assigned certain capacities to the office alone, thereby separating the office from the flow of agency. Instead, Ashley makes organizational processes intelligible. Though she mentions discrete entities (e.g. university leaders), she refuses to center them. In effect, Ashley is affirming agency’s motion and challenging the idea that agency can rest on any one component of the organization. By resisting the researcher’s cut, Ashley offers an organizational, not individual, explanation for WU’s responses to sexual violence.

This discourse about agency’s motion, one that Ashely’s comments exemplify, was so influential that some WU members doubted whether university members could make autonomous decisions. Of the multiple instances where this doubt appeared, we found Sarah’s comments particularly illustrative. She talked about the university’s mandate—one linked to federal guidance—that anyone who hears about sexual violence has to file an official report, even if that goes against the wishes of the person who experienced assault. Because Sarah worked in the university’s assault advocacy program, she was exempt from that mandate. People would contact her for advice about the reporting obligation. Asked to describe the conversations she had with WU members who were uneasy with making required reports, Sarah said this:

We all can make choices. We can all break the law and walk across the street without—you know, when it says not to walk, and I do that the same … I let [WU members] know what the policy is and that that’s what the university expects of them, but it’s still their—I don’t like to say that because I would say someone from [the office that enforces the policy] [would say], ‘It’s not their choice! They don’t have a choice. That’s the policy’.

Sarah initially acknowledges that policies cannot compel people to action, a stance consistent with prevailing understandings of agency. Before uttering the word ‘choice’, she interrupts herself and expresses discomfort in indicating that WU members can decide whether to comply with the reporting policy. Sarah suggests her hesitation occurs because the enforcement office that Ashley works in interprets agency differently. She implies that for the enforcement office, the moving agency of organizational systems, including university policy, eclipses personal decisions. Though Sarah assumes humans choose how to respond to a policy, the university offices suggest that people cannot make such choices. She surfaces a tension in two boundary-making practices at the university: one of them, exemplified in Ashley’s comments and referenced in Sarah’s, emphasizes the ongoing flow of agency; the other, the one Sarah prefers, cuts human agents apart from the flow in order to prioritize their decision-making capacity. Sarah’s comments suggest that a discourse about flowing agency, not individual human agency, dominates WU’s discussions about its responses to sexual violence.

The tension between these two boundary-making practices also surfaced in the military, but with one important distinction. Whereas the discourse about flowing agency seemed to encourage WU’s responsiveness to sexual violence, in the military it sometimes did the opposite. For example, it discouraged organizational responsiveness in Cioca v. Rumsfeld. The case began when Kori Cioca, along with 14 other women and two men who experienced sexual violence filed a class action lawsuit against former Secretary of Defense Donald Rumsfeld and then-Secretary of Defense Robert Gates. Plaintiffs argued that these two people violated their Constitutional rights by failing ‘to prevent Plaintiffs and others from being raped and sexually assaulted’ (Cioca v. Rumsfeld, 2013:4). They emphasized the decisional capacities of two people, Rumsfeld and Gates, and their lack of appropriate action. Despite this cut, in which agency momentarily paused on two leaders, the plaintiffs overwhelmingly emphasized agency’s motion. In their complaint, they focused less on the actions of discrete individuals and more on problematic processes: ‘perpetrators were promoted’, the military ignored ‘Congressionally-mandated institutional reforms’, and ‘victims were openly subjected to retaliation’ and prohibited ‘from telling anyone about the criminal acts’ (Cioca v. Rumsfeld, 2011: 3). Their description parallels Ashley’s comments about the flow of agency at WU. Like Ashley, the plaintiffs resist centering any one entity and mention individual humans only in passing while highlighting systems and processes. By emphasizing the flow of agency, the plaintiffs framed sexual violence as an organizational problem, one the military should do something about.

Judge O’Grady, who heard the case, also emphasized the flow of agency to frame sexual violence as an organizational problem, not an individual one. Unlike the plaintiffs, however, he implied that agency had so much momentum that the military could do nothing to stop sexual violence. In his decision to dismiss the case, Judge O’Grady relied on the Feres Doctrine, a precedent that resulted from three court cases, each pertaining to physical but not sexual injuries active duty service members sustained due to the negligence of other military members. ² O’Grady (2011) noted that the Feres Doctrine circumscribes the court’s interventions and ‘counsels against judicial intrusion’ so that ‘not even the egregious allegations within the Plaintiffs’ Complaint will prevent dismissal’ (p. 2). Because of the doctrine, he could not advance the case. In making this claim, the judge undermined the notion that humans are sole agents. Had O’Grady instead emphasized his unmitigated decision-making power, a boundary-making practice that would let agency pause on him alone, he would have appeared indifferent to ‘egregious’ sexual violence. Instead, O’Grady made a different cut. He focused on the flow of agency between only the text and himself, downplaying his choice. That boundary—which put limits on the flow of agency—led the courts to excuse the military from taking further action in response to sexual violence.

O’Grady’s temporary cut mitigated his ability to choose a particular course of action, but he nonetheless acknowledged that agency moved around the organization. He ruled that the court cannot compel the military to act if ‘injuries arise out of or are in the course of activity incident to service’ (Feres v. United States, 340 US: 135, 138, 146) (1950). By focusing on the inevitable dangers of the job, Judge O’Grady implied that sexual assault is endemic to the military. He does not cut human perpetrators or other actors apart from the flow of agency. Instead, he suggests that sexual violence is embedded in the work itself, an ‘occupational hazard’ (Purchia, 2012). This is an organizational, not individual, framing of the problem. Even so, O’Grady’s decision denies that an organizational response—one that would involve moving agency—could fix the issue. The ruling implies that a different cut—one that separates human actors from agency’s flow—governs responses to sexual violence. Military laborers must decide not to join the military or, upon enlisting, accept that ‘being raped by co-workers’ is ‘just part of the job’ (Purchia, 2012, para. 3–4). As in Sarah’s comments about WU’s processes, O’Grady’s ruling also indicates competing boundary-making practices in discussions of military sexual violence. The university denies that individuals can make autonomous decisions, but for O’Grady, those autonomous decisions—ones made by those who risk experiencing, not perpetrating assault—are the only available remedy for sexual violence. At WU, the flow of agency prompted an organizational response to sexual violence, but in Cioca v. Rumsfeld, O’Grady argued the flow of agency meant an organizational response would be ineffective.

In sum, in both high-profile incidents we analyzed, various organizational members suggested that agency was in motion during the organization’s response to violence. Contrary to feminist arguments that organizational frameworks for sexual violence and harassment prompt transformation, at times, the movement of agency appeared to stall responses to violence that feminists might advocate. To further complicate this argument, we next focus more fully on how pauses and cuts operated in both of these high-profile incidents.

Agency at rest obscures organizational explanations for violence

Although prevailing understandings of agency—in which agency pauses and settles on one human—were neither exclusive nor dominant in these critical incidents, they still operated in the discourse we analyzed. Notably, agency often rested on one person during discussions of the act of violence. Whereas agency moved in discourse about each organization’s response to these incidents, that movement did not characterize discourse about violent events themselves. As we continue to argue, these pauses detracted from framing sexually violent acts as organizational problems.

Despite WU’s insistence that the organization and its policies compel and eclipse human choice during the response to sexual violence, thereby emphasizing agency’s motion, that understanding was not universal. In particular, university officials resisted the suggestion that agency moved during the enactment of violence. For example, WU’s campus violence policy specifies that individuals can be sanctioned for violent acts. The language indicates that discrete humans, not systems or processes, incur the consequences for committing violence. If the policy allowed that violent acts themselves could be sanctioned, it would retain some flow of agency. Instead, in this policy, only individual humans can commit violence, and agency is fixed to a single, discrete entity. By implication, for violence to occur, the flow of agency must pause on one person. This pattern occurred repeatedly in the data from WU. In every discussion of committing violence, humans were cut apart from the perpetual motion of agency. That boundary-making practice sharply contrasts those that surfaced in discussions of the university’s responses to sexual violence that emphasized agency’s perpetual motion.

This repeated cut, in which agency pauses on a human when violent acts occur, hindered the university from conceptualizing sexual violence as a systemic, organizational problem. The first author was a participant observer and took field notes at a training session about university sexual violence policies. At that meeting, an employee asked how the university would respond to repeated, hostile, sexual comments in university publications and other public forums at WU. The trainer replied that the university would not consider these actions sexual violence, even though it would consider repeated, hostile, sexual comments from one organizational member to be sexual violence. The difference depended on the presence or absence of a particular cut: when an individual human exhibited harassing behaviors, those actions were considered sexual violence. That human could be separated from the flow of agency. But when identical behaviors could not be associated with one discrete entity that could be ‘cut’ apart from agency’s motion, the university’s policy did not classify those actions as sexual violence. If, as implied in the policy and reinforced in this training, sexual violence occurs only when agency settles on one entity, WU need not frame sexual violence as an organizational problem. The contrast between these explanations and those in the last section is notable. For example, Sarah said that the university claims its systems and processes remove a human’s choice about how they respond to sexual violence. Yet, in university policy and training sessions, the university never suggested that its processes and systems could remove a human’s choice about whether they commit a violent act. Moving agency allowed the university to frame its response to sexual violence as an organizational problem. By contrast, the university relied on paused agency to insist that the enactment of violence is not an organizational problem.

This notion that agency flows during the organization’s response to sexual violence, but pauses in the moment when violence occurs, also appeared in the military. In 2012, Secretary of Defense Leon Panetta issued a directive to move adjudication of assault complaints up the military hierarchy (Panetta, 2012b). According to Panetta (2012a), ‘at the local unit level sometimes these matters are put aside, they’re not followed up with’ (para. 7). Panetta implies that the military responds poorly to sexual violence when agency stops moving and congeals around the choices of lower-level commanders who can decide to ignore sexual violence. To remedy the problem, Panetta’s directive ‘requires that any time a complaint is received that it is referred up the chain of command’ (para. 7). Panetta’s solution for the military emphasizes the flow of agency: he implies that movement through the military’s systems and processes, signaled in the phrase ‘up the chain of command’, is crucial for an effective organizational response. As at WU, Panetta implies the military can eliminate or curtail individual decision-making about whether and how to report sexual violence. By undermining the possibility of human choice, thereby refusing that agency congeals in an individual, Panetta offered an organizational explanation for the military’s response to sexual violence.

Despite sometimes relying on flowing agency to advocate a better response to sexual violence, Panetta’s rhetoric also cut individual humans from the flow of agency. These cuts occurred in discussions of who or what enacts violence. Panetta (2012a) states, ‘The most important thing we can do is prosecute the offenders, deal with those that have … committed this crime’ (para. 13). As in WU’s campus violence policy, this statement focuses on individual humans who commit violence. Panetta does not suggest that it is most important for the military to prevent sexual violence from occurring. Instead, he suggests that the organization must increase accountability for perpetrators after assault has happened. Although Panetta’s comments about the chain of command alluded to flowing agency, in this statement, Panetta makes a cut that sutures the act of violence solely to individuals. In allowing agency to pause on the ‘bad apples’, Panetta overlooks that acts of violence, not simply responses to them, have something to do with organizational systems and processes. As at WU, Panetta understands the response to sexual violence as organizational, but he makes cuts that detract from framing sexual violence itself as the result of organizational problems. Indeed, Panetta (2012a) asserted, ‘Sexual assault has no place in the military. It is a violation of everything that the US military stands for’ (para. 2). Panetta suggests that sexual violence is antithetical to the military, not embedded in it. This move reinforces the boundary-making practice whereby sexual violence is understood to be the independent act of one human, not the result of flowing agency.

Panetta’s cuts are notably different from those that O’Grady made in Cioca v. Rumsfeld. Whereas Panetta claimed that rape violates military values, O’Grady argued that sexual violence is an unavoidable ‘occupational hazard’ and ‘incident to service’. Panetta emphasized that the organization can utilize moving agency to develop a better response to sexual violence, but O’Grady claimed that the military was unable to stop violence. Panetta established that violence occurs because of individuals who commit assault, but O’Grady accepted the plaintiffs’ arguments that violence emerges from the military’s systems and processes. In sum, for Panetta, agency flows during the military’s response to sexual violence but not during the enactment of violence by ‘the offenders’. By contrast, for O’Grady, agency flows during the enactment of violence, but not during the military’s response. Indeed, for O’Grady, the only viable response is for individuals to decide not to enlist, a framework that relies on agency congealing once again in single humans.

Taken together, the critical events we analyzed suggest that boundary-making practices in which agency congeals around individual human choice do not simply prevail at these organizations. Moreover, discourse about flowing agency did not consistently frame sexual violence as a systemic, organizational problem. Instead, something more complicated is happening in the boundary-making practices we identified. In the next section, we draw out the implications of these complexities and how they extend existing research on harassment.

Implications for scholarship on sexual violence, sexual harassment, and agency

By tracking the movement of agency, we extend scholarship on sexual harassment. In this discussion section, we consider how debates about the political utility of agency in motion matter. We focus first on the promise of moving agency, then on its peril, for shifting understandings of sexual violence and harassment in organizations.

The promise

Conceptualizing agency as motion has promise for framing sexual violence as an organizational problem. Although our analysis identifies some ways in which moving agency upholds organizational problems, inert agency—when located only in individual humans—also prevented these organizations from seeing violence as the outcome of organizational systems. Clair (1993a) suggested that taken-for granted meanings maintain harassment. For instance, she said that when organizations advise people who experience harassment to ‘say no’, the organization makes two assumptions: the person has not already said no and more assertive responses to harassment are effective. Both assumptions are dubious.

Our analysis surfaces another taken-for granted meaning: violence requires the action of a discrete agent. If, as Brummans (2015) as well as Putnam and Cooren (2004) argued, agency is the motor of organizing, then when the motor stops running, no organization remains. Left with a mere collection of violent individuals, it makes no sense to consider how agency constitutes a sexually violent organization. At WU and in Panetta’s remarks, a focus on discrete agents detracted from understanding sexual assault as an organizational, not individual problem. If organizations continue to assume that violence occurs when agency consolidates in one human, the most viable recourse is to expunge a member here or there, and the system (i.e. the organizational dynamics of sexual violence) stays intact. Undoing this type of thinking is particularly important as the #metoo movement gains traction around the globe. If the focus stays on ousting those individuals who have committed assault, the movement will miss an opportunity to shift public consciousness about the societal and cultural mechanisms that maintain sexual violence, regardless of which individuals are in power.

Agency as motion may thus be crucial for identifying the organizational aspects of any given phenomenon, and it can add nuance to existing descriptions of assault in organizational research. For example, Cooren (2006) described a case in which a bishop from the Roman Catholic diocese sexually assaulted a child. Cooren emphasized that, when the Supreme Court of Canada ultimately faulted the diocese, they made an ‘act of selection’, a ‘matter of decision’ that ‘consists of delimiting where one should end in the chain of agents’ (pp. 87–88). In Cooren’s metaphor, people make artificial, somewhat arbitrary choices so that agency appears to stop at a single point. This process is akin to the feminist new materialist ‘cuts’ we described in our literature review: boundary-making process whereby agency’s complex motions are simplified and contained. Unlike FNM, Cooren’s metaphor retains the idea that discrete agents exist, but nonetheless FNM illuminates that seeking an end to the chain serves those who believe punishment drives social norms. For example, in our analysis, agency only appeared to stop when the organizations were trying to identify who committed violence, and often those discussions revolved around appropriate sanctions for the offending individual. Agency circulated again when organizations were determining next steps. Given these stopping and starting points, we suggest that the impulse to find an ‘end’ to a chain of agents reveals more about assignments of blame (in the form of liability and wrongdoing) than it reveals about agency (the actions that can change organizations). Thus, emphasis on agency’s motion may keep organizational discourse about sexual violence from becoming what Bennett et al. (2005) calls a ‘politics of condemnation’ in which the focus is on rape’s aftermath, not its prevention.

The peril

Although conceptualizing agency as motion has some utility for the study of organizational sexual violence, this approach also bears some risk. Clair (1993a) identified organizational discourses that maintain sexual harassment, and we add organizational discourse about agency to that list. She asserted that strategic ambiguity masks ongoing power relations and inhibits systemic change. We add that organizational discourse about agency can also be strategically ambiguous and stall meaningful responses to rape and other forms of assault. In our data, the flows of agency made the organization responsive to sexual violence in some instances but not in others. The military suggested assault was inevitable. By this logic, no decision or action would stop sexual violence, yet Panetta also suggested that moving agency—the organized connections among officers, directives, and procedures—would ensure effective action against sexual violence. These ambiguities about who or what can enact and respond to violence have the potential to absolve the organization of the need to do something different. For instance, at the WU training session, we discussed earlier, when an individual harassing agent was unclear, the trainer suggested that no formal organizational response was required. In these instances, moving agency enabled the organization not to change, thereby allowing sexual violence to continue. If agency as motion predominantly supports this status quo, it is a perilous framework.

In this sense, our analysis lends some support to arguments against locating agency beyond individual humans. As Caldwell (2007) summarized, this kind of theorizing—in which the movement of agency makes its location ambiguous—appears to undermine ‘any hope of change’ because an intentional human is no longer able to make politically transformative decisions (p. 770). In the context of scholarship on sexual harassment, agency as motion risks obscuring the action of perpetrators in a climate where sexual violence is still pervasive and not unequivocally seen as wrongdoing. Noting lack of accountability, Dougherty et al. (2011) expressed frustration with how rarely perpetrators of harassment are reprimanded. In their study of medical organizations, they found it unusual for a patient to be ‘held responsible for his actions’ when he harassed a nurse (p. 276). The phrase ‘his action’—which pauses agency on an individual human—seems a necessary precursor for assigning responsibility. When agency is not located in one individual perpetrator but instead moves constantly, organizations could make cuts in which agency—and therefore the responsibility to prevent violence—would rest on entities where it should not: those individuals who experience assault and harassment. Indeed, we saw this cut occur in the military when O’Grady acknowledged that agency moved fluidly in the organization. Instead of assuming that this motion would allow the military to take multiple different actions, he suggested that individual service members had the responsibility to avoid the ‘occupational hazard’. Because of this possibility, we suggest that agency in motion is not a panacea for entrenched organizational patterns of victim blaming and deferred accountability that perpetuate rape culture.

In sum, we find both promise and peril in agency as motion. Entitative thinking (Chia and Holt, 2009)—which occurs after ‘cuts’, depends upon immobile agency, and focuses on individual perpetrators—can obscure the organizational dynamics that our academic field works to illuminate. In contrast, moving agency highlights those organizational dynamics, but in our analysis, it introduced ambiguity that allowed organizations to continue not to change. Feminist new materialist scholars are aware of this trouble. Bennett (2005) reminded us that mobile agency

does interfere with the project of blaming, but it does not thereby abandon the project of identifying … the sources of harmful effects. To the contrary, such a notion broadens the range of places to look for sources. (p. 463)

Rather than undermining systemic transformation, for Bennett, agency in motion proliferates options for change. Bennett implies that concerns about moving models of agency—ones we share—can be mitigated by decoupling agency from blame. The peril of agency as motion emerges from an elision of these concepts, and distinguishing them can complement and extend existing scholarship on sexual harassment. For example, Barad (2012) argued that FNM focuses not on sanctions and retributions, but on actions that alter the conditions of organizing. This shift, one that relies on agency as motion, multiples ‘the possibilities of mutual response’ (Barad, 2012: para. 14). This approach, in which identifying multiple avenues for change and punishing criminal acts are not identical processes, can further establish sexual harassment as an organizational phenomenon.

The perils we identify in this article are not evidence of political limitations in conceptualizations of agency as motion. Instead, they demonstrate the persistence of cuts that suture agency to individual humans. Under this sedimented approach to agency, when a perpetrator is not blamed, a target is, and the organizational dynamics that support violence remain unchanged. But Warfield (2016) argued that the capacity to make change emerges ‘not from entities but from small topological reconfigurations—small changes and shifts—that move and relocate the boundaries, demarcations, and marks that have historically separated entities’ (p. 6). By highlighting the ways in which agency in motion is already infused in organizational discourse about sexual violence, we provide a record of these emerging topological reconfigurations. These small shifts depart ever so slightly from the individualist and entitative styles of thinking and being that pervade not only extant theories of agency, but also numerous justice systems and other organizing practices.

A feminist new materialist approach to agency and sexual violence

Organizational scholarship on nonhuman agency enjoys a growing presence in the field (e.g. Brummans, 2018; Carlile et al., 2013; Cooren, 2004; Wood, 1998). This trend usefully expands and complicates what counts as agency, and it advances Taylor’s (2013) call to ‘break with [a] tradition … that automatically takes the individual as the starting point of theory building and analysis’ (p. 208). Indeed, Ganesh (2015) noted that conceptualizing agency beyond individuals is particularly important for advancing social justice movements. Some work that engages these concerns is part of an emerging relational ontology (e.g. Cooper, 2005; Kuhn et al., 2017). In this analysis and discussion, we joined and extended these ongoing efforts to nuance theories of agency by (a) highlighting the consequences of particular agential cuts for organizational sexual violence and (b) drawing on FNM and diffraction to do so.

We are not surprised that, in this study, during discussions of the capacity to enact violence, agency settled exclusively on individual humans. Other scholars have observed the difficulty of noticing systemic forms of violence. Catley and Jones (2002) argued, ‘violence takes many forms, but when it appears in a “sovereign” act of individual physical violence it is obvious, and recognised as being violent. When violence is symbolic or structural it tends to disappear from view’ (p. 34). The violence built into organizations is hard to recognize. As we tracked moving agency, we revealed one of the mechanisms that obscures systemic violence, enabling it to ‘disappear from view’: the cuts that make agency appear to rest on discrete individuals, supposedly in isolation from the organized flow of agency.

As Enck-Wanzer (2009) argued, an exclusive focus upon individual perpetrators prevents feminists from intervening in the systems of power that scaffold gendered violence. Furthermore, interdisciplinary feminist scholars have long noted that situational and contextual factors, ones ‘that make rape thinkable and doable’, must be centered in order to undermine assault (Hall, 2004: 13). We extended this argument by suggesting that when agency appears to be discrete and unmoving, it prevents organizations from examining the situational and contextual factors that exacerbate sexual violence. Furthermore, complicating this argument, we also highlighted that the movement of agency does not automatically lead to productive change.

Given our contention that when boundary-making practices render agency motionless they obscure systemic violence, we are cautiously optimistic about conceptualizing agency as motion. Contrary to the contention that individualized understandings of agency are necessary for organizational change, we showed how unmoving understandings of agency allowed organizations to deny that their own processes support sexual violence. Agency as motion promises to imbue the whole organization, not simply an individual, with the capacity to prevent sexual violence. This framework is especially salient in the contemporary political moment, in which activists aim not just to oust powerful individuals who have committed assault, but to transform a broader climate and culture that lives in higher education and the military. As recent news has shown that climate and culture also permeates Hollywood, North American and European governments, global media, medical practices, and churches.

Thus, we assert that tracking the movement of agency can help feminist scholars highlight the organizational dynamics that support sexual violence, including sexual harassment. Moreover, understanding how agency moves and appears to pause can not only expose systems of violence, but also highlight the multiple options not only for addressing assault after it happens, but also for preventing sexual violence before it begins. Indeed, we suggest that organizational scholars might adopt a similar focus on agential movement in order to show how other systemic violence, including but not limited to racialized incarceration and reproductive injustice, becomes difficult to notice.