“ I Didn ’ t Want To Be ‘ That Girl ” ’ : The Social Risks of Labeling , Telling , and Reporting Sexual Assault

This article deploys ethnographic data to explain why some students do not label experiences as sexual assault or report those experiences. Using ideas of social risks and productive ambiguities, it argues that not labeling or reporting assault can help students (1) sustain their current identities and allow for several future ones, (2) retain their social relationships and group affiliations while maintaining the possibility of developing a wider range of future ones, or (3) avoid derailing their current or future goals within the higher educational setting, or what we call “college projects.” Conceptually, this work advances two areas of sociological research. First, it expands the framework of social risks, or culturally specific rationales for seemingly illogical behavior, by highlighting the interpersonal and institutional dimensions of such risks. Second, it urges researchers to be more attentive to contexts in which categorical ambiguity or denial is socially productive and to take categorical avoidance seriously as a subject of inquiry. Substantively, this work advances knowledge of why underreporting of campus sexual assault occurs, with implications for institutional policies to support students who have experienced unwanted nonconsensual sex regardless of how those students may label what happened.

T HIS article outlines the factors at play when students do or do not label an experience of sexual assault as "assault," tell others about that experience, and/or report that experience to an authority.Empirically, we exclusively focus on those who are subject to assault; in other parts of our work, we address how those who are accused or admit to committing assault understand this experience (Hirsch and Khan 2019;Walsh et al. 2018).We argue that the acts of labeling, telling, and reporting mean something different to victims than to authorities or policymakers.Labeling, telling, and reporting are an actualization for victims; they formalize ambiguous experiences, making them concrete and "real."We show some of the social risks of such actualization for undergraduate students; we underline how for many students, ambiguity-not labeling, telling, or reportingcan be socially productive, allowing for a greater range of (imagined or real) future identities, relationships, and cultural projects.Labeling, telling, and reporting concretize thoughts, feelings, relationships, affiliations, and experiences in ways that many students believe can negate or harm their current identity and limit access to current or future opportunity structures.Although some undergraduates find the label of sexual assault "survivor" empowering, for many others, not labeling, telling, or reporting can help them (1) retain their current identities and allow for several future ones, (2) maintain their social relationships and group affiliations and the possibility of developing a wider range of future ones, or (3) avoid derailing their current or future goals within the higher educational setting, or what we call "college projects."This suggests a potential misalignment between those who would encourage reporting assaults, both for the sake of public knowledge and to set in motion processes that might punish a student who has allegedly caused harm, and what individuals who have experienced what we would label an assault feel that they need or want in terms of making sense of and managing their experiences.
We develop these arguments from an analysis of more than 150 interviews with students at two interrelated institutions-Columbia University's undergraduate schools (co-educational) and Barnard College (women only), both located in New York City-18 months of ethnographic observation, and a random population survey of the student body.Our analytic framework explores the individual, interactional, and cultural contexts for why people act in the world in the ways they do; this focus on multiple levels of influence allows us to better understand the social risks people experience when labeling, telling, or reporting a sexual assault.We categorize social risks in three ways: (1) identity risks, or those experienced as potential threats to an individual's sense of self; (2) interpersonal risks, which include fears about tensions or fractures within relationships, friend groups, and social networks; and (3) college project-related risks, which reveal what is at stake in terms of students' goals as college students.
We define "labeling" as categorizing an experience as an assault, "telling" as communicating that experience to someone else, and "reporting" as communicating that experience to an official who holds a position that carries a responsibility to investigate or act.The person who is assaulted is not necessarily the only one who labels, tells, or reports.When a person is assaulted, someone else may be the one to label that experience as an assault (e.g., when a friend says to someone who had an experience that they aren't sure how to label, "I think you were raped") even if the person rejects that description of the experience.A person who is assaulted can choose to report, but they can also report unintentionally (e.g., by sharing the story with a mandatory reporter, who then faces an institutional obligation to make an official report, or when they tell a peer or someone else who either choses to report what they were told or inform a mandatory reporter about it).

Background
Sexual assault, which refers to nonconsensual sexualized touching (e.g., fondling private parts), attempted penetration (e.g., oral, anal, or vaginal sex), or completed penetration, is a significant problem on college campuses, with 20 to 25 percent of women and 8 percent of men reporting exposure (Banyard et al. 2007;Fisher, Cullen, and Turner 2000;Krebs et al. 2007).Reporting those assaults to college officials is extremely rare, with research showing that anywhere from one in 20 to one in 50 students report their experiences (Cantor et al. 2015;Mellins et al. 2017).
Public attention to sexual assault has focused intensely on improving processes of adjudication on college campuses, with a great deal of discussion about standards of evidence, right to representation, and other questions about due process; the implication is that improving those systems would (1) lead more people to report, thereby (2) allowing victims to access important resources beyond adjudication, (3) giving victims, through adjudication, a sense of resolution, and (4) reducing the likelihood of assaults, as punishment would serve as a disincentive.Yet the vast majority of assaults go unreported, most reported assaults go unpunished (Cohen and Kyckelhahn 2010), and, as we shall show, the few victims of assault who did report in our study experienced the process negatively.Rather than take a stance on what students should do, we focus here on what they actually do and on the social forces shaping behavior, understandings, and experiences.Examining how context shapes sexual behavior has become a standard approach within public health research (see, for example, research on HIV and family planning: Dixon-Mueller 1993; Parker, Easton, and Klein 2000), yet it has been relatively underutilized in the study of sexual violence.
Research shows that many who have experiences that meet the definition of sexual assault do not label it as such (Harned 2005;LeMaire, Oswald, and Russell 2016;Orchowski, Untied, and Gidycz 2013).Labeling an experience that meets the behavioral definition of assault as such is influenced by attitudes towards and beliefs about sexual assault, by a person's level of intoxication and relation to the assailant, and by a feeling of victimization (Kahn et al. 2003;Peterson and Muehlenhard 2011).Research on reporting finds that (1) feelings of shame, guilt, or embarrassment; (2) fears of not being believed; (3) perceptions of not having sufficient evidence to support a claim of assault; (4) fears of retaliation; (5) not wanting an assailant who is a friend or family member to be prosecuted; and (6) not wanting family or friends to know can lead those who do label an experience as sexual assault to not make an official report to police, campus authorities, or friends and family (Ahrens 2006;Fisher et al. 2003;Sable et al. 2006;Zinzow and Thompson 2011).As with most sexual assault prevention research, however, the focus has been on the individual attitudinal or interpersonal-level factors shaping labeling.We contribute to the list of specific reasons that people eschew the label, but more importantly, we provide a more unified conceptual framework for understanding the labeling, telling, and reporting processes.

Social Risk and Productive Ambiguities
The social scientific concept of social risk highlights the social reasons that people engage in behaviors with detrimental consequences, fail to take actions that might protect themselves, or otherwise act in ways that seem illogical (Hirsch et al. 2010).Our focus on social risk underlines the ways in which maximizing one's health is not necessarily a person's paramount objective, calling attention both to the reasons that people might engage in behavior that is detrimental to their own wellbeing and to the reasons that people might refrain from practices that appear logical and beneficial to an external observer.A social risk framework foregrounds the culturally specific rationalities and socially situated goals that shape people's actions.In the case of this article, we highlight the reasons why students may choose not to label or report sexual assault even in cases in which reporting might have helped them access supportive resources (emotional support, administrative aid with classes or housing, resolution with the assailant, etc.).Our analysis shows variation within the experience of assault, including its labeling and the decision to discuss it.Our use of the concept of social risk is meant to call attention to what is at stake for students, things that lead them not to label an experience as sexual assault, report it, or talk about it.Understanding those concerns is vital to shaping adjudication systems that are sensitive to the needs of those who have experienced assault as well as institutional designs that help members who are experiencing suffering.
Prior work has examined how social risk can be gendered in relation to sexual behavior (Hirsch et al. 2010), focusing on how men use extramarital sexual relations to build gendered selves in ways that are shaped by peer groups, ideologies of masculinity, and gendered labor markets.Our point in this article about gender and social risk is subtler: it is not about what people do but rather about the gendered social risks in how people label what they do (or what is done to them).Although we point to a number of instances in which heterosexual women refused to use the label "sexual assault," all other things being equal, heterosexual women are far more likely to deploy that label, and queer students likelier still, than are heterosexual men.In part, this reflects population-level patterns of assault, yet it also highlights the dynamics of categorization.Using our social risks framework, we suggest that men more acutely experience an identity risk in thinking of experiences they've had as assault.In our discussion, we think more broadly about this gendered character of categorization.
Although the social risks framework points to the more agentic dimensions of individual deliberation or choice (e.g., things that people decide to report or discuss), the notion has relational and institutional dimensions.By relational, we mean how relationships influence or structure available actions; we think here of the structuring influences of something like the subject's relationship to the person who assaulted them, how their friends help them process and label their experience, or the potential impact of telling on valued social relationships.On a more institutional level, structuring forces are what the subject has learned about reporting and what they believe happens when people do report (irrespective of what actually happens).Prior work on social risk has primarily focused on ways in which the concept illuminates what people may weigh as they navigate their lives (Parker et al. 2017).Our more expansive approach opens up paths for thinking about institutional responses that extend beyond a focus on changing individual behaviors and perceptions.
This social risks framework foregrounds the power of bureaucratic formalization and of activist mobilization to define experiences and identities in particular ways.Decisions about labeling, telling, and reporting may also be based on managing social relationships (Tilly 2006).The instability of the category of sexual assault has a conceptual implication; although many classic variables (years of completed education, number of children, etc.) fit squarely in a positivist framework, quantification can at times obscure important aspects of how categorical variables are experienced and managed.To be clear, our mixed-method project fully embraces the value of quantitative research.Our argument is that augmenting this work with ethnographic research (and vice versa) reveals important dynamics.In our quantitative work, we have used latent class analysis to show that sexual assault is not one thing but many different things (Khan et al. 2018; see also Mellins et al. 2017).In this ethnographic work, we shift this point in a different direction, arguing that the category itself can be a problem for those who experience assault and that we need to pay attention to and explain why people want to avoid that category.An implication of this work is the importance of "categorical avoidance" or "categorical embrace" as objects of inquiry.
With this as a framework, we argue that reporting is not just about a dyadic relationship between a victim and a set of bureaucratic authorities; it also represents a potential intervention into a person's network ties.Whereas being reported to helps authorities meet institutional goals and satisfy legal requirements, the act of labeling and reporting, although potentially beneficial, incurs risks for those who have experienced assault.Those risks are at the individual level (in terms of their identities), at the interactional level (in terms of group affiliations and social ties), and at a broader cultural level (in terms of the college projects and trajectories that are culturally desirable).Students make choices that help them maintain empowering self-identities either as people who are capable of forgiveness or as people who are in control of their lives.Not labeling or reporting allows them to psychologically downplay the experience, to avoid confronting it regularly, to refuse to see themselves as victims, to persuade themselves that a very difficult thing didn't happen, or to continue to understand the person who assaulted them (often someone they know well and about whom they care) as a good person, friend, or partner.This last point highlights how labeling, telling, and reporting is not simply about one's own identity but has identity implications for the other person and relationship implications for both; labeling an experience as an assault makes the other person a perpetrator.Not labeling or reporting helps the person who experienced an assault maintain social ties with groups that can often include both the victim and the person who assaulted them.Although to authorities and policymakers, reporting often is about justice, adjudication, sanctions, and assistance to victims, for victims, it is also about making a difficult incident real and turning a possibly ambiguous thing into an assault, with all its attendant consequences.For many victims of assault, even though denial can result in harm, as those who avoid difficult experiences may also not get the help they need, ambiguity can feel overall more desirable or productive than reporting and adjudication.
Such "productive ambiguities" conceptually suggest analogies between the bureaucratic formalization of an identity such as "victim" or "survivor" and a range of other identities.We can take, for example, the process of transforming a man's act of having sex with other men into the identity of being gay.Such formalizations create conditions of possibility, community, support, resonance, at least for some people in some social contexts, but there is a vast literature on the multitude of reasons that, for example, some black and Latino men as well as genderqueer men have either avoided or been excluded from claiming a gay identity (Boellstorff 2011;Epstein 1999;Ford et al. 2007;Garcia et al. 2016;Muñoz-Laboy 2008;Parker et al. 2017).Just as not every man who experiences same-sex desire and has same-sex experiences embraces a gay identity, not every person who experiences a sexual  (Israel et al. 2001), two advisory boards informed this study (Wolferman et al. 2018).The 20 students on the Undergraduate Advisory Board (UAB), all of whom were excluded from study participation because of their advisory role, met with researchers weekly throughout the planning and data collection phases, providing input on instrument development, recruitment, maximizing sample diversity, and other aspects of research implementation.After data were collected, the group met approximately monthly to discuss and interpret findings.UAB members were each paid $750 for each semester of participation.The Institutional Advisory Board, which was comprised of key institutional actors such as deans, student life professionals, and student health providers, met with researchers approximately twice each semester during planning and data collection then more frequently during data analysis.
The ethnographic research team, led by two faculty members (Hirsch and Khan), consisted of seven team members, with up to five researchers being in the field at one time and at least two working full-time as research assistants.The three men and four women represented a range of racial and class backgrounds as well as sexualities and religions.With support from the university's Office of General Counsel, the researchers were granted an exemption from mandated reporting requirements in their role as researchers on this project.The ethnography team met weekly for at least two hours to discuss fieldnotes, observations, and interviews.Hirsch and Khan also met weekly, organizing interpretations, discussion points, and directions for the broader team meeting; they also met regularly with the quantitative researchers to coordinate practices and integrate insights.
After interviews were transcribed, checked for quality, and stripped of identifying information, two research team members independently coded interviews for  2014) for research laying the groundwork for multilevel approaches to sexual assault prevention, the project as a whole was conceptualized and executed in ways consistent with Bronfenbrenner's ecological model (1977).Drawing on the senior ethnographers' prior work, from its inception, the ethnographic research was grounded in attention to gendered prestige structures, sexual projects, sexual geographies, and social risk as concepts potentially relevant to understanding how students navigate the social and sexual landscape (Hirsch et al. 2010;Khan 2011).Ethnographic data analysis used an inductive approach; for this article, that entailed describing the breadth of reasons why students did not report experiences of sexual assault.Analysis drew on the coding, team members' fieldnotes, and a document that included a lengthy description of incidents of sexual assault that young people experienced.
Importantly, for this article, we coded experiences as assaults based on whether described behaviors met the definition of an assault (unwanted and nonconsensual sexual behavior) and not on the basis of whether or not subjects labeled such behaviors as assaults.This is consistent with the practice of survey researchers who work on assault, for whom the standard is to ask behaviorally specific questions, such as, "Has someone had sex with you without your consent or agreement?"rather than, "Have you ever been assaulted?"The reason for this distinction is that many people who have had sex without consent would not label that experience as sexual assault whether because of the stigma associated with that label, because of the weight that such a label carries, or because of the seriousness that it conveys.We include here, for example, instances from the interviews in which a student described having sex when they exclaimed "no" to a partner with whom they did not want to have sex.However, they did not label this experience as rape or even assault; explaining why this happens is one of the aims of this article.Regardless of how the person subjectively understood this kind of experience, we categorize it as an assault because it meets the behavioral definition of unwanted sex without consent.Because of this decision, we categorize some experiences as assault even if subjects themselves do not use that label.In total, across our 151 interview subjects, we categorized 89 sexual experiences as assault; several interviews contained more than one assault story.In total, 66 of our interviewees told at least one assault story.
This article draws lightly on our random population survey to assess some of the applicability of our claims against a representative sample of the Columbia community.The SHIFT survey was conducted during the same period as when our ethnographers were in the field.Survey participants were selected via stratified random sampling from the March 2016 population of 9,616 undergraduate students ages 18 to 29 years using evidence-based methods to enhance response rates and sample representativeness (Dillman, Smyth, and Christian 2014); after our sample was drawn, the survey was conducted between March 2016 and May 2016.Of the 2,500 students invited via email to participate in a web-based survey, 1,671 (67 percent) consented to participate; our sample was representative of the available demographic information about the student body (for more on our sample and procedures, see Mellins et al. 2017).Incident rates of sexual assault in our random population sample were 28 percent for women and 12.5 percent for men, with the cumulative risk being 36.4 percent for women and 15.6 percent for men by the time they graduate college (Mellins et al. 2017).Ethnographic interviews slightly overrepresent the experiences of students who were assaulted.This is not surprising given our object of study and the fact that some interview subjects self-selected into the interview process because they had "a story to tell."As we gathered information about all kinds of sexual experiences, it is important to note that the majority of those we interviewed never experienced any sexual assault.

Findings: The Social Risks of Labeling, Telling, and Reporting
Students described a variety of rationales and concerns that shaped how they thought about, whether or not they discussed, and to whom they disclosed their experiences of assault.Table 2 outlines the distribution of labeling, telling, and reporting by year in school and gender identity.It is important to recall two things about our findings.First, we only report the experiences of students whose descriptions meet the behavioral definition of being assaulted.Absent from this article are accounts of (alleged) assailants.Although research shows a very low rate of false reports (typically between 2 and 5 percent; Spohn, White, and Tellis 2014), our accounts are of how one person experienced and reflected upon an interaction.The account of the other party may well be different, and we have no data to speak to this difference; this has important implications when we discuss the challenges of adjudication.
We categorize social risks in three ways: identity risks, interpersonal risks, and college project-related risks.These different social risks affect how students interpret what has happened to them and whether and how they act upon these interpretations, pointing to aspects of the college social environment that make reporting or discussing assault difficult and, as a result, pose challenges for ensuring that students get the help they need and report if they would like to do so.In what follows, we present these three categories of social risks, providing examples of the particular concerns students had in categorizing or discussing their experiences.

Identity Risks
Both students who did not label experiences of sexual assault as assault and those who did not tell their friends that they had been assaulted (regardless of how they labeled it) expressed concerns about negative impacts on their identities, encompassing both their internal perceptions of themselves and their ideas about how others might perceive them.One heterosexual woman, a freshman, referred to unwanted sex as "the scariest thing"; she described that she had had "nonconsensual" sex but insisted that it was not assault, refusing to imagine herself in what she perceived to be a disempowered state, unable to express her desires or unable to realize them and thereby be subjected to the power of another.She noted that  a One participant identified as both queer and female.
labeling the experience as an assault would be "giving him the power to say. . . he did that to me, and I just want to feel like it didn't affect me."For others, a more explicit rejection of the survivor identity (see examples below), which they viewed as abject, politicized, and necessarily deeply traumatized, motivated their refusal to label an assault as such or their silence to their peers.Although we spoke to students who found the survivor identity empowering, for others, refusing to define something as an assault, and hence refusing the identity of a sexual-assault survivor, allowed them to hold fast to the identity of someone who was in control of their life, someone who had not been harmed. 1 Concerns about identity also surfaced in the stories of students who did in fact label an experience as assault but who only shared that information within a finely circumscribed friend group.They experienced sexual assault as, in part, a social failure.By "social failure," we mean thinking of oneself as not having it "together" or being able to navigate a social terrain in a way that one expects one should be able to.Students expressed feelings of guilt and shame about "putting themselves in that situation," whether that meant being at a particular party, having had too much to drink, being too flirtatious, or not recognizing the "reasonable" or "expected" consequences of certain actions.Like most people, students fear being seen as ridiculous and inexperienced.The lack of experience that most concerned them, particularly when first at college (when they were most at risk for assault), were those with substance use (mostly alcohol) and sex.
Students framed this inexperience as the product of their "bad choices" rather than the social context.Both women and men described instances in which such bad choices had, in their opinion, led to unwanted sexual experiences.A male student who was groped by three senior women his freshman year noted that he "would have said no to them if they asked."He told us that he should have known that "this is how people interact; this is how people get down" in college.As a result of having his genitals repeatedly grabbed by women, he quickly left the party, overcome by a feeling of embarrassment.He did not consider the groping to be sexual assault.
Regardless of how they labeled their experience, talking about it with their friends provided students an opportunity to discuss their feelings; this seemed, in the vast majority of instances, to be helpful to students emotionally.But when we asked about expanding such conversations beyond their friendship groups, students suggested that their feelings of social failure and shame would be exacerbated by sharing the incident with anyone outside their friend group.For these students, "survivor" was not the way they wanted to be publicly (or more widely) known.In the words of one student, "I didn't want to be that girl."The identity risk of being "that girl" or "that guy" figured among students' reasons not to report.In the words of one, "It involves way too many people. . . .It means that a ton of people that I could see on a regular basis would know." The negative identity association of being "that person" was not the only motivation; students also mobilized more positive identities.These students expressed not wanting the potentially severe consequences of the investigation of the person who assaulted them on their conscience, nor did they want to manage the socially difficult challenge of confronting this person within the investigation.Countering those who argue that silence about a sexual assault perpetuates community risk by allowing someone who committed an assault to "go free," those who did not report their experience thought of themselves in different moral terms: as a person who is able to turn the other cheek and allow a second chance for growth and transformation.The social risk, in this case, is that of being perceived-or perceiving oneself-as someone who would act without regard to the consequences for a peer.

Interpersonal Risks
The interpersonal risks of labeling an interaction as assault, sharing the information with friends, or reporting revealed how students prioritized relationship preservation.One woman, who was assaulted in her freshman year by a friend of her roommate while her roommate and her roommate's boyfriend slept in the other bed, worried that if her roommate found out, it would lead to conflict between her and the roommate: she "would just hate me even more and be even . . .more horrible to me."For these students, the social risk in identifying an event as sexual assault was the conflict it might cause with a peer or friend rather than with the person who assaulted them.We found that students had real challenges making and maintaining friendships, particularly in their first year at college, when they were most at risk of assault (Mellins et al. 2017).In contexts in which friendships are valuable, rare, and a limited social resource-as was the case in our field site-the interpersonal social risks to reporting are that much greater.A refusal to label unwanted sex as assault can also be related to a desire to continue the relationship with the person who assaulted them.For example, one participant was forced to have penetrative vaginal intercourse while pushed up against a car and repeatedly saying "no."In recounting this story, she clarified that he must have understood her objection not to the sex but to the place in which it was occurring, as when she said "no," he moved her to the ground.In several interviews, participants voiced that the person who assaulted them was "a nice guy."In these cases, people insist that their friends or partners are respectful, good human beings; these sentiments appeared to be sincerely felt.The assault was an aberration.Not labeling the event as an assault allowed subjects to feel comfortable continuing with the relationship and help preserve the dignity and goodness of the person who assaulted them.The woman who was raped against a parked car explained how attracted she was to her friend and former partner and how much she liked him; she reasoned that he must have interpreted her "no's" as to being about the car rather than the sex.She conveyed the experience to researchers as more amusing than disturbing, laughing as she recounted finding dirt in her vagina afterwards.
This representation of those who committed assault frequently took place in contexts in which victims were socially embedded with and interested in preserving the reputation of the person who assaulted them, such as when both belong to a tightly knit friendship group or a club or when the two people are in an intimate relationship.The framing of those who perpetrate assault as sociopaths has consequences for the majority of victims who are assaulted by people they know or like.When the victim's many other experiences with that person do not suggest that the person who assaulted them is a sociopath, labeling the experience as assault is a challenge.In this context, labeling and telling are not just about defining one's own identity, they also have implications for the identity of another, creating a kind of interpersonal risk.To say that someone else "committed assault" is identity transforming or imposing for that person, and it makes the accused a kind of person: a perpetrator.In a context in which most assaults are committed within embedded relationships and in which such assaults have been constructed as "the worst thing ever," labeling is not just a definitional act for the person assaulted, it constructs the identity of the person who committed that assault as a sociopath or someone with deep moral failings.In this sense, not labeling an assault may well be understood as a kind of positive identity work not for oneself but for another whom you care about (Besbris and Khan 2018).For some, the challenge of labeling assault is not only making sense of what happened to you, it is imposing a morally polluted identity upon another who is a friend, partner, and/or, most likely, embedded within one's community.
These social ties augment the concerns of jeopardizing friendships and relationships with people they share in common with the person who assaulted them.Students managed the risk of surfacing conflict within a friend group or student organization either by not labeling something as assault or by closely restricting the number of peers with whom they shared the story.Sometimes, what feels to be at risk are the relationships with the peer group itself, as in the case of the small number of students who labeled an experience as assault but did not tell anyone; their words suggested a fear that sharing the news would place a burden on their social network or result in negative reputational consequences for the group as a whole.One woman, during her first week on campus, left a party while intoxicated and went to another first-year student's room.They were becoming friendly, and he had, on a previous occasion, invited her to come by to listen to some music.Once there, he began to grope her against her will.She left, ashamed that she had not accurately assessed his intentions-that for him "listening to some music" was code for "hooking up."She did not tell anyone this story, explaining that by referring to her concern that because "the campus is really small," other students would judge her to be too foolish, naïve, and inexperienced.She imagined that this negative impression would have lasting negative social repercussions.Instead of talking to anyone else about her experience with this man and her lack of consent, she decided to stay silent because, in her assessment, the fault was her own.She thought that she "shouldn't have gone to that room.That's kinda stupid."Her biggest concern was not the emotional consequence for herself but instead that others would not want to be her friend if she told the story of what happened to her, particularly because her friends knew the man, and people thought of him as a "nice guy."These relationship-related risks are keenly felt: in a residential higher-education setting, a student's friends are the people with whom they live, eat, socialize, study, and engage in extracurricular activities.A social break in one's group presents the risk of enormous affective losses in multiple arenas both by creating conflict in the group and through the fear that group members would, if forced to choose, side with the person who committed the assault.Peer groups provide a set of individuals with whom to socialize, access to otherwise inaccessible leisure time spaces and social events, a sense of belonging and identity, and, in many cases (as we discuss below), both social prestige and labor market opportunity structures.

College Project-Related Risks
Students considering labeling and reporting an experience as assault also needed to navigate risks that related to their particular college projects. 2By "college projects," we are referring to the numerous social goals students possess within the institutional setting of a university, including but not limited to academic goals, career goals, developing an identity, and extracurricular interests.In the case of formal student organizations (e.g., sororities, athletic teams, and registered student organizations), even more is at stake than potential affective losses.Students devote enormous effort and many hours to group activities and perceive them to be critical prestige structures in their day-to-day lives as well as opportunity structures for securing postcollege employment.Identifying or discussing an experience as assault, in some students' eyes, jeopardizes their very goals for what they want to get out of college.Some students explained their reluctance to make an official report by talking about the stress it would add to their already stressful lives and thus how it would do more harm than good.They often doubted that they could effectively make a case for themselves-as a result of internalized shame or responsibility for the incident-and thus decided that the costs in time, stress, and affective labor outweighed whatever social support they might secure by telling others.These students considered reporting an incident to be one of many activities competing for their limited time and emotional resources, which they judged to be better suited to pursuing other goals within the university.One female student explained her decisions not to report by stating that "dealing with public safety, dealing with the police, that would just add too much to my plate."Although, in general, our application of the social risks framework emphasizes the culturally variable elements of rationality, these examples illustrate students as in some ways consummate rational actors, submerging (or at least attempting to submerge) potentially very painful and upsetting experiences in order to optimize college time and avoid the opportunity costs they imagine to be involved in making an official report.
Some of these college project-related risks were tied to inequalities within the student body.As university settings have diversified, students are acutely aware of the different resources students can mobilize.Such inequalities in material and symbolic resources mean that some students critically evaluate the investment it would take for them to "present their side" against another, more richly resourced student and decide that reporting simply is not worth it because the outcome is unlikely to bring satisfaction.These students decided that the bureaucratic processes of reporting itself presented a risk to their goals as students.
In terms of material resources, some who were assaulted reported that the high social status of the person who had assaulted them rendered it unlikely that a convincing case could be mounted against that person.Students recounted concerns that others could hire lawyers or rely upon their long-term familial relationships with the university to protect them.In these cases, the risk was that reporting upon particularly privileged students-of whom there are many-would involve considerable personal burden and with an outcome that was unlikely to end in the reporter's favor.
Men expressed the concern that the process of reporting inherently meant that they would need to tell the university that they had had sex with someone while incapacitated and without clear memory of the incident.By "symbolic resources," we mean the ways in which assault is often not legible for some victims or is more easily challenged by some of those who perpetrate it.For example, officially reporting would allow the woman who assaulted a man to respond with accusations of her own.This would be particularly consequential in a context in which the man had already admitted to sexual contact and incapacitation.One male student who was incapacitated during his assault characterized his female partner's possible response succinctly and pessimistically, "All she has to say is, 'He was drunk.He doesn't remember.He raped me,' right?"In other instances, black men expressed concerns about whether they would be believed, noting that nonblack sexual partners had the capacity to rely upon racial privileges wherein their telling of the story would be more believable (see Hirsch et al. 2018 for an elaboration of this example).

Confidential and Mandated Reporting
In nine of the 89 incidents, students used confidential university resources (nonmandated reporters, such as religious clergy, mental health counselors, and sexual assault advocates) to manage their experiences.Their explanations for seeking professional help but not formally reporting reflect the influence of a variety of social risks and those particular students' clear understanding (which was not universal) of the difference between a mandated reporter and a confidential resource. 3Students sought confidential resources because they offered a type of support that their friends could not.Some acknowledged that an incident had been jarring or disorienting enough to require professional help in thinking through it but denied that the incident had been morally wrong enough for someone to be held accountable.In this type of situation, the framing of sexual assault as a significant moral violation lowers the likelihood of official reporting, as many don't think of their experience as "serious enough."For example, the only male student in our sample who reported either confidentially or to a mandated reporter noted that he went to Columbia's Sexual Violence Response program (an on-campus confidential reporting, advocacy, and other services group) because he had experienced something that he was "uncomfortable with. . .something I needed to deal with" but did not make a formal report because "I didn't really want for it to go any further.I didn't think that anyone had done anything that wrong."Others who accessed confidential resources emphasized the social risks they perceived in nonconfidentially reporting to the police or the university.One individual clearly laid out her reasons for not formally reporting: "One, it's ridiculous that, like, it's an investigative process because that means that someone can literally get a lawyer and argue against my experiences.Two, it's traumatic."The trauma and problems this student anticipated were juxtaposed with other, "more important" parts of her life that she prioritized over reporting this particular incident.
Given the substantial risks that students perceive in making a formal report to the university, it is unsurprising that relatively few students whom we spoke with officially reported their experiences.Only eight of the 89 incidents (experienced by five participants) were reported nonconfidentially either to an on-campus mandated reporter or directly to the police.The mandated reporters involved included a supervisor of a preorientation program, a resident advisor, and an academic advisor.One participant reported an assault that occurred off campus directly to the police, and one participant reported an on-campus assault directly to campus safety.The experiences of the very few who engaged with the mandated reporting process largely confirm their peers' fears about the social costs one might face in the process. 4 The woman who reported to the police was assaulted off campus.She immediately labeled the incident as assault, told her close friends and family, went to the hospital for medical care, and filed a report with the city police department.When discussing reporting, she says she filed a report without hesitation because she thought it was what she was supposed to do.Ultimately, she found the police to be unhelpful and even coercive.The detectives assigned to her case wanted her to call her assailant in an attempt to get him to incriminate himself as they listened in on the call.This was something she was unwilling to do, but she felt enormous pressure from the police to do it anyway.As she recounted the story, "And I was, like, 'No, I'm not doing [it],' and she's [the police officer] just, like, pressuring me over and over and over again."She later worked with Sexual Violence Response, through which she claimed to have finally received effective advocacy and extensive support.Although she initially reported without hesitation, she now says she understands why people do not, saying, "Now I understand because my first instinct was go to the police; that makes sense.Oh, it's horrible." The other four engaged in a more protracted process that included figuring out how to label what happened, talking with peers about it, seeking support from nonmandated reporters, and, ultimately, speaking with a mandated reporter or campus safety.None of these four viewed a formal report as their first or best option (indeed, as evidenced by the total eight events, some participants only reported after experiencing multiple assaults).Two did not understand that speaking with a mandated reporter meant that a formal report was necessarily going to be filed, and none wanted to complete the on-campus adjudication process once it was initiated.However, they ended up going through with it only because it did not feel as if there was a choice to do otherwise.Consistent with what is broadly found within the literature across a wide range of contexts, they described the actual adjudication process as exacerbating mental health challenges and providing little relief.Given that the broader literature both within and outside of the college context consistently presents negative experiences of reporting and relatively low rates of reporting compared to population-based studies on experiences of assault (Fisher et al. 2003;Frazier and Haney 1996;Sable et al. 2006;Zinzow and Thompson 2011), we assert that our empirical material should not be interpreted as unique to our site.Instead, it is likely a consequence of sexual assault investigations themselves.Given the potential consequences of a report (e.g., being expelled from the university), institutions have a responsibility to those accused to conduct an investigation and to not assume guilt.But the act of investigating-asking both the reporter and the reported what happened, asking others who were present before (typically friends), reinterviewing parties multiple times given competing accounts or confusing information (particularly when incapacitation is involved, as it frequently is)-although procedurally responsible is almost necessarily traumatic as it involves repeated questioning of stories and interventions into valuable but potentially fragile social networks.Though there are models available beyond an investigatory one, particularly the restorative justice model (Koss 2014), for a variety of reasons (most importantly, adherence to federal guidelines), the investigatory model is overwhelming used by institutions, and we assert that our case reflects some of the negative consequences of that model rather than particularities of our site.
The four individuals in our data who discussed their experiences with reporting to the university found it to be an enormously complex and challenging process; this meant struggling with how to process what happened and, ultimately, others either convincing them to report or making the decision for them.Following a similar trajectory as that of the woman who reported to the police, another participant labeled her assault as such immediately.However, in telling a friend about it, she received little social support and even became worried that her friends (whose busy lives featured pains and struggles of their own) had little room to care about what had happened to her.She tried to "move on," but the man who assaulted her was a fellow student leader in a campus group, and after what she describes as ongoing manipulation from him after the assault, she told another friend in the group.This time, she received strong support.Together, they tried to find a way to hold the man who assaulted her accountable in order to make the group safer for her, seeking help first from a confidential resource who knew both students and ultimately going to the university staff member who was the advisor of the program.Although she knew the advisor was a mandated reporter, she felt this was her only option.She believed that the advisor would be able to take steps within the group to keep the man who assaulted her separate from her and that she would be able to stop the Office of Gender-Based Misconduct (OGBM) investigation after it was initiated.In order to tell the advisor a story she found too difficult to verbalize, she wrote down her account of the assault and gave it to her advisor.She did not realize that the advisor, acting in accordance with the university's interpretation of federal law, would submit this written report to the university; subsequently, she was told by the OGBM that the investigation had to continue because of the evidence provided in her write-up.She experienced this as a further violation of her autonomy: ". . .my agency was totally stripped out of my hands.Like, I had no choice.There was going to be an investigation launched whether I liked it or not."Part of what this indicates is that even in a context in which the university has invested heavily in improving adjudication-Columbia has hired several sex crimes investigators and/or lawyers to help investigate and manage cases and guaranteed free representation for both the reporting and responding student-the process remains relentlessly difficult for students.Our evidence suggests that this difficulty will persist, regardless of steps taken to improve the reporting and adjudication processes, because of the "collateral damage" to students' identities, friendships, networks, and college and/or life projects.
The participant who reported to campus safety experienced ambivalence both about how to label her experience and whether to report it.When she told her boyfriend about what had happened to her-friends walked her home from a night of drinking, and she went in and out of consciousness to discover she was "having sex" with her friends-her boyfriend immediately labeled her experience as a rape.He strongly encouraged her to seek medical care and file a report, which she did.However, another student, who was friends with both her and those who assaulted her, suggested that she was not, in fact, raped.This produced an ongoing conflict around how to label what happened and how to proceed.It led her to attempt to recategorize the event as something other than sexual assault once already in the hospital.Even when she felt more confident in her decision about labeling and reporting the assault, she worried about the impact that reporting would have on those who raped her and the social impact of reporting for all involved, as she had classes and social activities with one of the men who had assaulted her.She also expressed a desire to just move forward with her life and a fear that the reporting process would prevent her from moving on, as it would force her to continually think about the assault.She ultimately decided to go forward with the investigation in large part because she was informed it would go on with or without her because it was already initiated: "Every step of the way, I've been like, 'Oh, I wish I wasn't doing this.But, like, I have to be.' There's no alternative 'cause . . .I can't just pretend it didn't happen.That's unrealistic.So, I was like, 'I have to deal with it no matter what, and the process really sucks, but, like, it's the only thing that they have.So, I just have to do it."'Although she acknowledged receiving some benefits from reporting, such as not having to go to class with a man who raped her, she noted that if she had known how long and difficult the reporting process would actually be, she likely wouldn't have participated.She experienced the reporting process itself as challenginglonger than she wanted, with lots of questions about her account of what happened.Research has noted the difficulties of conducting investigations in a way that is sensitive to the needs of those who have experienced sexual assault (Frazier and Haney 1996).Less discussed, however, is how students experience that process.In the most recent story, the social impacts were pronounced: the woman who was raped experienced the identity risks of being seen as "out of control" with her drinking; she experienced interpersonal risks with the crumbling of her friendship group; and she experienced college project-related risks, as she was unable to fully participate in activities either because her assailants were associated with those activities or because of the time and emotional energy she was devoting to the reporting process.
Two other participants eventually engaged in formal reporting of their assault, but neither did so intentionally.One genderqueer participant, who uses the pronoun "they," continued talking with the person who assaulted them right after the assault, as they had been trying to avoid labeling what happened as an assault.In their estimation, although speaking to the person who assaulted them was unpleasant, identifying as an assault victim or even survivor seemed worse at the time."I didn't want to talk to them. . .but I also thought that if I didn't talk to them, like, things wouldn't be normal. . .and it would be like admitting that something had happened to me."A few months later, they labeled it as assault and talked with non-Columbia friends about it.They later told peers at Columbia, with the goal of ensuring that the assailant would not be living in the same building and wanting that person to be held accountable within their friend community.Finally, they also sought emotional support from the confidential Counseling and Psychological Services.They only sought assistance from a mandated reporter, an academic advisor, after their mental health deteriorated over the course of processing the assault, and they began to feel the need to drop a class.They did not know that the advisor was a mandated reporter and first learned that a report had been filed after receiving an email from the Office of Gender-Based Misconduct.They had no way to achieve their academic goals, they felt, without talking to this person about what had happened.
Summarizing the experience, the student recounted: "You know, I tried the community justice model. . . .I didn't want to go to the university.And I probably wouldn't have reached out to them on my own.But they already knew. . .because my advisor had told them."This participant followed through with the adjudication process but experienced the process as harmful to their mental health, and too sociological science | www.sociologicalscience.com 27 PUBMONTH PUBYEAR | Volume 3 prolonged."[I]t was supposed to be done before the summer started. . .but it wasn't done until like halfway through the fall semester of sophomore year. . .which just, like, made my mental state so much worse because, like, this whole situation was so protracted so far beyond how, like, it was supposed to go, you know?"A range of desires led people to report formally, including the desire to drop a class as a way of managing the mental health consequences of the assault, to safely participate in a group, to move on with their lives, or to hold the person who assaulted them accountable outside of a formal adjudication process.In order to try to access these things, people navigated a difficult process of labeling and telling, facing social risks as they moved along a trajectory that led them to formally report.Sometimes, they encountered exactly what they had feared, and sometimes, they found unanticipated support, but a common thread for all of these students was that the perceived social risks of not reporting were greater than the perceived risks of reporting.However, in none of the cases we observed was adjudication perceived to be the mechanism that would be most helpful to the people in getting what they felt they needed, and afterwards, those involved felt it did little to help them heal or to hold their assailants accountable.
Our explanations run into important sampling limitations of ethnographic research; although the descriptions we provide may illustrate a breadth of student experiences, they do not necessarily illustrate their distribution.The random population-based sample survey portion of our work helps validate some of the ethnographic findings.In the survey data, only a very small proportion of students who experienced sexual assault formally reported it to university officials (2.2 percent).Of those who indicated experiencing assault on the survey questionnaire, 81 percent talked to someone about the incident, typically a close friend or roommate.Of those who talked to someone, they typically talked to a close friend or roommate; only 13 percent told mental health personnel, and 3 percent talked to a campus sexual assault advocate.The reasons people gave for not telling (people could have multiple reasons) were that they feared the reactions of others (26 percent), they were embarrassed (29 percent), they didn't think what happened was serious enough (84 percent), they didn't want others to worry about them (38 percent), or they just wanted to forget about what happened (34 percent).Those who did tell someone did so quickly, with most telling someone within 24 hours (67 percent) and 80 percent telling someone within a week.Irrespective of telling or reporting, those who experienced assault found that their academic lives (16 percent), their mental health (38 percent), their extracurricular activities (11 percent), and their social lives (36 percent) were affected.

Limit of the Social Risks Framework: Community Mental Health Burden
Our intention here has been to demonstrate the utility of social risk as an explanatory framework rather than to suggest it is the only factor shaping how students respond to unwanted, nonconsensual sexual experiences.The most important other factor for our current discussion is the community-level mental health burden of sexual assault.Students realized that their friends might be too exhausted from other incidents to be willing or able to contain and help process an experience and its aftermath.This should not be surprising given that in our context, we find that 32 percent of women and almost 16 percent of men will be assaulted by the time they finish college.On average, those who experience an assault are assaulted multiple times; we find that, on average, those who report being assaulted experience three assaults (Mellins et al. 2017).One participant spoke of telling her close friend about her assault only to hear that her friend had been assaulted in a far more violent way: I told one of my female friends about it....And she just, like, had a war story.She had a story that was, like, more. . .she just had, like, essentially, like, a more violent story. . . .She just kind of missed mine because hers was worse....So, I was, I, I felt, like, less able to, like, I don't know.I stopped talking to my female friends about it because they just, like, all had their own stuff going on. . . .And it didn't seem important to them.She then decided to stop discussing her assault with friends.This community-level burden was not something we anticipated.Some friend groups and social networks appear to have been too saturated with other emotionally heavy incidents to assist their friends with processing an incident or recovering from an assault.And in general, even within groups that were not so saturated, we noted the mental health ripple effects of students holding such painful experiences for each other.

Discussion and Conclusion
Students have strongly felt reasons for labeling an experience as something other than sexual assault, for not telling anyone, for telling only their friends, and for seeking professional help but not filing a formal report.And although these dynamics are true of many other types of traumatic experiences, there are some common features across all of the responses examined in SHIFT.First, students see their time at college as valuable and limited and do not want to be derailed in their commitment to accomplishing other vital projects.Whether it is holding fast to a friend group or extracurricular activity or securing an identity as a cool girl, competent partier, or successful socializer, there are social and identity-building reasons for remaining in close control of one's story.Overwhelmingly, students' post-assault strategies reflected a desire to balance sometimes competing priorities; they wanted to secure whatever support they needed to get back to whatever their "normal" self was, but they also wanted to minimize the affective, identity, time, and social collateral damage.Not reporting is one of the most effective ways to balance these two somewhat contradictory ends; keeping an experience ambiguous allows for social continuation rather than social rupture.
For men, our findings suggest that thinking of sex as unwanted can be a challenge because gendered sexual scripts suggest that men always want sex (Hust, Rodgers, and Bayly 2017); it may be productive to consider the ways in which this category of assault can be identity invalidating in gendered ways for men as well as to consider more generally the ways in which labeling an experience as assault invalidates or affirms gender or sexual identities.For students from marginalized groups, such as in the example of the genderqueer student, labeling an experience as an assault can be politically laden insofar as it suggests a kind of pathology within one's community.This political quality of describing someone within your community as a perpetrator or embracing the survivor identity, however, is a landscape all of those who are trying to make sense of their experiences must navigate.Some activism has successfully used the strategy of framing assault as "the worst thing ever" in order to demand more attention to and resources for the issue.But an unintended impact of sexual assault advocacy is that it emphasizes how bad assault is; some chose not to label their experiences as assault because they do not view their experience as that bad (this is the case for many men, and recall that 84 percent of our survey respondents did not report because they did not think what happened was that serious), they do not want to embrace the somewhat totalizing identity of being a survivor, or they do not want to impose the identity of a perpetrator upon someone they care about or are socially embedded with.Labeling has dramatic identity consequences not just for the person who was assaulted but also for the accused and even for the group within which they are typically embedded, potentially implying a pathology of the group or a relational dynamic that can include sexual assault.We find that some students avoid the survivor identity because it feels like failure, trauma, and an admission of not being able to do important sexual and college projects "properly."It is important for us to recognize that advocacy creates categories of experience.The gay rights movement created opportunities for sexual selves; abortion advocacy gives meaning to an experience that differs widely in stigma around the world.Such advocacy happens as part of the terrain of the symbolic world, where these (new) constructed categories of experience and identification have unintended consequences.
Such categorical formulation does not reveal an objectively "true" experience but instead gives us modes of understanding.In this article, we have avoided, when possible, the terms "survivor" and "perpetrator" other than to discuss critically how students may eschew them, in part because these words suggest a kind of person rather than an act or experience.This point is more than academic; it suggests that an understanding of assault organized around identity categories-because they are so encompassing-may serve as impediments to people making sense of their experiences or getting the kinds of help that they need.Approaching sexual assault from a sexuality studies perspective rather than one grounded in violence and trauma or criminology renders this observation not entirely unsurprising; more than two decades ago, human immunodeficiency virus (HIV) activists began using the term "men who have sex with men" to index the fact that sexual behavior does not always map neatly onto identity categories.We are making essentially the same point about sexual assault: having experienced an assault does not automatically mean that one wants to take up a survivor identity, nor does having committed an assault mean that one's peers will apply the label of perpetrator to them.
These identity processes are gendered; researching the aspects of such gendering is an important avenue for future research.Labeling or even talking about being assaulted can be identity invalidating for men, as it runs against conceptualizations sociological science | www.sociologicalscience.com 30 PUBMONTH PUBYEAR | Volume 3 of masculinity and the presumption of gendered social agency.We speculate that the gendered social scripts that assume that men are almost always seeking sex make the category of unwanted sex less legible for men (including gay and queeridentified men).For women, the gendered work is about being a particular kind of woman, a woman who can manage her sexuality.As we have shown, this sexuality management works both for and against labeling experiences as assault.
On the one hand, women do not want to suggest that they are "out of control," inexpert in modern sexual situations, or subject to the will of another.On the other, given the ways in which social activism has created an affirming survivor identity, suggesting resilience and a commitment to seeking justice not just for oneself but to protect others, labeling an experience as assault, we suggest, offers gender-affirming possibilities for women that do not exist currently for heterosexual men.For queer students, the identity dynamics at play are somewhat different.When a lesbian student is groped by a man at a party or a gay man is compelled by a woman friend to have intercourse, it is the experience itself rather than the labeling that is identity invalidating.Labeling an experience as assault then becomes a way of reaffirming one's sexual identity, pushing back against the experience that can be subjectively experienced as a denial of one's identity.Similarly, when a student in the middle of transitioning to being a woman is verbally coerced into using their penis for penetrative sex, labeling that experience as unwanted has social force, repositioning the student with their desired gender identity.These different experiences of affirmation and denial may, in part, explain why some of the highest rates of assault are reported by lesbian, bisexual, gay, transgender, queer/questioning plus (LBGTQ+)-identified students and why heterosexual men report some of the lowest rates of assault.
Our research also highlights how intellectually productive it can be to examine the instability in an outcome of interest.There is a great deal to be learned about the labor of meaning making, identity building, and relationship construction by looking at the deployment of categories, even in their avoidance. 5Sexual assault researchers have expended a great deal of effort to develop behaviorally specific measures of sexual assault (Koss and Gidycz 1985;Koss and Oros 1982), and those measures have laid the groundwork for a generation of research that more accurately estimates the prevalence of assault among college students (Mellins et al. 2017).But efforts to estimate accurately the prevalence of an experience by using measures that avoid naming that experience, although crucially important for demonstrating the prevalence of unwanted nonconsensual sex, sidestep or perhaps even naturalize people's reluctance to label something as an assault even in the relative anonymity of an online survey.Research might take that avoidance as a central question of interest.
This work may help explain why those who label an experience as assault are more likely to report having experienced multiple assaults or, as the literature has shown, why having been assaulted is predictive of future assaults.Once the label has been embraced, its deployment is more easily available for future experiences (and for interpretations of past ones).Finally, we understand this problem of categorization within a framework of how young people are doing gender and sexuality, in short, in terms of how such gender and sexual projects fit within their college and life projects.Our social risk framework points to how making sense of an experience of assault requires embedding within aspects of individual identity, within dynamics of social networks, and within broader culturally gendered and sexual structures and scripts.In short, making sense of sexual assault, both scholastically and for those who experience it, requires a broad ecological approach.

Policy Implications
Given the heightened interest in sexual assault, we end with four brief policy implications.First, the social costs of labeling are widely acknowledged by survey researchers as the reason that behaviorally specific measures are important, and yet, services for students are frequently labeled explicitly as being for those who experienced sexual assault or sexual violence.Although we do not recommend a wholesale rejection of these terms, at least some outreach and communication should be directed towards students who have experienced upsetting, confusing, or even just disagreeable sexual experiences.Second, in this era of heightened concern for undergraduate mental health (Byrd and McKinney 2012), we call attention to the community-level mental health impact of students being the primary ones working to contain and respond to their peers' confusing and sometimes traumatic experiences.The policy implication is not that students should receive another form of mandatory training in how to listen with empathy, set boundaries, and respond effectively to a story of sexual assault; rather, it is that effective primary prevention of sexual assault should be considered as a means to reduce stress and improve mental health for the broader student population.Third, the social risk of disruption within a friend group reflects, at least to some degree, a modifiable social factor; students' reluctance to close the door on a friend group, even if it means enduring years of contact with someone with whom they had a "weird" experience, reflects their assessment of the probable difficulty of replacing that friend group once social networks have settled.There is the broader context as well; students' time and labor investment in extracurricular activities is part of their career paths, and so, they see walking away from a shot on the executive board as having a cost in terms of their competitiveness in the job market.This reflects a general finding from our project about the unrecognized potential for multisectoral approaches to sexual assault prevention, including prevention approaches that are not necessarily related to social or sexual practices; instead, we might think of a range of community-level interventions, such as those that might help build more robust safety nets in which the barriers to exiting social groups (and the power that any individual wields within those groups) are lowered.
The final implication is about using increases in sexual assault incident reports as a metric of improvements in the campus climate.There is a tension, perhaps, between a singular focus on adjudication and punishment and a broader focus on ensuring access to supportive services for those who have experienced unwanted nonconsensual sex of any kind.Unquestionably, it is vital to continue to improve processes of reporting and adjudication so that they are sensitive and fair to all involved.At the same time, however, we have shown that students have good reasons for choosing not to engage with those processes and that parts of the investi-gatory model can be experienced as traumatizing in themselves and not particularly helpful to the victim.Somewhat ironically, after our extended discussion of the value to students of ambiguity, this leaves us with the final recommendation about the importance of a greater scientific, evidence-based approach: repeated surveys, population-based sampling, approaches that achieve response rates that are high enough to generalize from them with confidence, behaviorally specific measures, network (relational) data, and institutional and/or contextual information as central elements of any strategy to measure changes over time in the campus climate.
Notes 1 This is not to say that those who identify as survivors are not in control of themselves or necessarily any more or less harmed by nonconsensual unwanted sexual experiences than those who do not identify as such; it is merely to argue that some eschew the label of survivor because they perceive it to be a strategy to preserve a particular sense of self.
2 In our broader research, we think of sex as a project.The concept of a sexual project (Hirsch 2015) catches the many reasons why all of us engage in the kinds of sex that we do; pleasure is an obvious one, but a sexual project can also be to develop and maintain a particular kind of a relationship, or it can be a project of not having sex, of having sex for comfort or to have a kind of experience, or because sex can advance our position within a group or increase the status of groups we are a part of.Sexual projects are embedded within other projects-such as the college project-which together make up people's life projects.The structure of a society, particularly its organizations, influences the kinds of sexual projects people undertake.
3 Participants have institutional spaces in which to confidentially report their sexual assault, though they do not always know the differences between those they can report to confidentially and those who are mandatory reporters, meaning these people are required to make an official report of the incident to the Title IX office.Most individuals who reported confidentially also reported to friends, which typically happened first.
4 Of these data on mandatory reporting, one subject was recruited in the field (in the course of ethnographic observations) without knowing that the subject experienced an incident of mandated reporting.The other subjects self-selected into the study; this selection might explain some of the negative experiences we observed with mandated reporting because these subjects may have wanted to tell their negative experiences to people from the university community.There was, however, no meaningful difference in the account of mandatory reporting between the subject who did not self-select into the study in order to tell a story and the others who did.

Table 2 :
Labeling, telling, and reporting sexual assault by year in school and gender.a. Year in school by reporting.