A four-year residential college is what sociologists call a total institution: it controls the conditions under which students eat, sleep, work, and party. “You can just imagine all these contextual dimensions in college that could be tinkered with to create a less stressful, less hard-drinking, more respectful environment,” Hirsch said. The assumption is that some college students are committing sexual assault when they don’t intend to, and that many are more vulnerable to sexual harm than they ought to be. Either idea can be controversial, and focussing on contributing factors, such as drinking, rather than just on the bad acts of perpetrators, can seem beside the point. But Hirsch and Mellins insist that their approach to prevention does not ignore personal responsibility; rather, it aims to nudge students toward responsible behavior on a collective scale. The first time we met, on Columbia’s main campus, Hirsch put it to me more plainly: “We have to stop working one penis at a time!”

SHIFT was born out of a crisis. In the past several years, as students all over the country became more vocal about the problem of rape in college, the press seized on a series of dramatic incidents, including one at Columbia. A rare combination of academic talent and initiative was then unleashed by the university, which may have felt the need to demonstrate its commitment to the cause, and this produced, after two years of sunup-to-sundown effort, the most rigorous, nuanced, and wide-ranging examination of the problem that has ever been carried out on a college campus. “It’s better for universities if sexual assault is positioned as a matter of sexual health, rather than as a scary threat,” the journalist Vanessa Grigoriadis, who published a book last year, “Blurred Lines,” about sexual assault on campus, told me. She added, “We’re in a new phase of the movement.”

You can trace that movement back at least four decades, to 1977, when a senior at Yale named Ann Olivarius—along with another student, three graduates, and a male professor—sued the school, citing quid-pro-quo sexual harassment by professors, a hostile environment, and a lack of reporting procedures. The plaintiffs, advised by a recent Yale Law graduate named Catharine MacKinnon, argued that this was a violation of Title IX—the federal statute, passed five years before, that prohibits gender discrimination in educational institutions. The women lost the case, but the district court ruled that it was “perfectly reasonable to maintain that academic advancement conditioned upon submission to sexual demands constitutes sex discrimination in education.” Two years later, MacKinnon published her landmark book, “Sexual Harassment of Working Women,” which argued that “economic power is to sexual harassment as physical force is to rape.”

The proper scope of Title IX was argued in court over and over in the years that followed; rulings narrowed its application, then expanded it again. Meanwhile, anti-rape activism progressed on campuses across the country. Take Back the Night marches, which had begun in the seventies, became a feature of college life in the eighties; Columbia’s first Take Back the Night march was held, in front of the Barnard gates, in 1988. The Columbia University Senate passed the first school-wide sexual-assault policy in 1995—it required that complaints be handled through an alternative form of the school’s standard disciplinary procedure. Student activists weren’t satisfied: they wanted the deans who handled sexual-assault cases to receive additional training, and they wanted to know how many incidents were being reported. They staged a prolonged campaign that culminated, in 1999, in a twenty-three-hour vigil, during which hundreds of students marched through campus shouting, “Red tape can’t cover up rape!”

Seven years ago, the Office of Civil Rights, under President Obama, issued a “Dear Colleague” letter, reasserting that sexual violence on campus was a violation of Title IX, and pushing universities to handle sexual-assault cases in a timely, transparent, accuser-friendly manner. A year later, the Department of Justice expanded its definition of rape to include male victims and multiple types of violations. (The previous definition—“the carnal knowledge of a female, forcibly and against her will”—had been in place since 1927.) Today, the D.O.J. defines sexual assault as unwanted sexual contact, which means that groping counts, as does attempted assault. The crime hinges on intention, and there are often no witnesses, which makes it uniquely difficult to adjudicate in any legal system, let alone one made up of college administrators. Campus judiciary systems don’t have a criminal court’s investigative powers or evidentiary procedures, but they do have many of a criminal court’s responsibilities. To complicate matters further, everyone involved in the process—accuser, accused, administrator—essentially works under the same roof. Betsy DeVos, Trump’s Secretary of Education, has called the current approach a “failed system,” and said that she would seek to replace it.

It might seem simpler to let the criminal-justice system handle things, but universities have a responsibility to insure that women have equal access to education. And, in addition, many students prefer to address these matters outside that system—they don’t necessarily want to send their assaulters to prison, and they may not be able to prove their cases beyond a reasonable doubt. Columbia now has twenty-three staffers with Title IX responsibilities, including case managers, investigators, and administrators, and provides free legal services to accusers and accused. The school’s gender-based misconduct policy is thirty-one pages long.

Freshman year at Columbia, as at any college, can be overwhelming: awkward encounters at parties in the “social dorm,” where the long wooden doors can be taken down to serve as beer-pong tables; a rush to find a home base in extracurriculars and clubs. Juliana Kaplan, a Barnard junior, told me, “On the one hand, you have kids at Columbia who come from kings of Wall Street—you have a secret society based completely on wealth. On the other, you have a demographic of first-generation, low-income students of color. People come in through very different contexts.” When I asked her about the tenor of student conversation on sexual assault, she told me, “I try to remember that some people have been super aware of these issues for their whole life, due to any number of factors, and then there are some people, such as men, who have to actively learn about it while they’re here.”

Five years ago, a Columbia sophomore named Emma Sulkowicz filed a complaint with the university, accusing another student of rape. (Sulkowicz, who has been working as an artist since graduation, identifies as non-binary, and uses the gender-neutral pronouns “they” and “them.”) A consensual encounter on the first day of the school year had turned violent, Sulkowicz alleged: in the midst of sex, the student anally penetrated and choked them while they struggled and told him to stop. (He has consistently maintained that the entire encounter was consensual.) Sulkowicz decided to report the incident after another student said that she’d had a similar experience with the same man. Columbia held a series of hearings and found the man “not responsible,” and Sulkowicz was subsequently denied an appeal. The following April, twenty-three students and alumni, each with a story of assault, filed a hundred-page federal complaint against the university. Student activists formed a group called No Red Tape, evoking the protests of the nineties. When the fall semester began, Sulkowicz, an art major, started carrying a fifty-pound, twin XL mattress around campus. It was a performance project: they would stop carrying it, they said, when the student who had raped them was expelled. (Sulkowicz carried it until graduation; the man they accused later sued Columbia, arguing that the university’s support of the project, for which Sulkowicz had received academic credit, constituted gender discrimination, negligence, and intentional infliction of emotional distress. The university settled with him out of court.) Soon after Sulkowicz began carrying the mattress, dozens of other Columbia students brought mattresses to the steps of Low Library and told their own stories of sexual assault.

It was around this time that Jennifer Hirsch attended a meeting of Columbia’s Women’s, Gender, and Sexuality Studies Council, where faculty members gathered in a conference room and picked over a catered breakfast. She sat next to Suzanne Goldberg, who at the time was a special adviser to Columbia’s president, Lee Bollinger, on the subject of sexual-assault prevention and response. The debates concerning rape on campus and what to do about it have been waged primarily between students and administrators, with professors off to the side. Hirsch had become frustrated by the focus, in those debates, on adjudication and punishment, rather than on the ways in which the environment of college makes students vulnerable. As the meeting ended, and people began collecting their things, Hirsch turned to Goldberg and spontaneously proposed conducting an ethnography: she would interview students, learn the everyday context of their sex lives, document the stories that the university couldn’t see. Goldberg said that sounded terrific, and told Hirsch to write up a few pages pitching the project.

A couple of weeks later, Hirsch popped into Mellins’s office, two floors down from hers in the Mailman building. The two professors have been friends since 2005, when Hirsch, who teaches in the sociomedical-sciences department, began doing work at the H.I.V. center at Columbia, which Mellins co-directs. Hirsch handed Mellins the paper she’d drafted, and began peppering her with questions. Mellins was the lead author of a 2011 study into the mental health, drug use, and sexual behavior of adolescents who had been infected with H.I.V. in the womb or as infants. She knew something about discussing uncomfortable matters with young people, and quantifying those conversations for research purposes. She answered Hirsch’s questions, and started asking her own.

Hirsch looked at her closely. “Do you want to do this with me?” she asked.

They spent the next few weeks brainstorming—on the phone, over e-mail, in each other’s offices, on whiteboards. They thought about the relevant expertise of their colleagues. Who really knew about interpersonal violence? Who really knew about epidemiology? Statistics? Trauma in young adults? As the fall turned crisp, they tracked down the faculty members whose help they wanted, and asked them if they would join SHIFT. In November, 2014, they submitted their proposal, and Goldberg quickly secured the university’s approval.

Goldberg, who is in her mid-fifties and speaks with a flat, equanimous affect, became Columbia’s first executive vice-president for university life in 2015. She has a long career of progressive advocacy—she was a co-counsel for the defendants in Lawrence v. Texas, which nullified Texas’s sodomy law. She leads Columbia Law School’s Center for Gender and Sexuality Law, and she was integral to the development of SHIFT. (“We had many breakfasts,” Hirsch explained.) But she has become a maligned figure among student activists. Amelia Roskin-Frazee, a senior involved with No Red Tape, spoke to me dismissively about the Sexual Respect Initiative, a consent-education requirement, instituted by Goldberg in 2014, that included an arts option: students could write a poem, submit a drawing, or perform a dance. When I asked Goldberg about this criticism, she said, “The initiative meets students where they are.”

Roskin-Frazee is a queer activist who, at fourteen, founded a nonprofit that provides schools and shelters with L.G.B.T.Q.-themed books. She is currently suing Columbia. She says that, two months after arriving on campus, she was violently raped in her dorm room by a stranger, and that, a few months later, she was raped again, by an assailant she suspects to be the same person. (She told me more than once that she knew this was not a typical campus assault; last year, she wrote a piece for HuffPost criticizing the notion that “true stranger rapes” are any more serious than those committed by people who know their victims.) She asked to move out of her dorm room, and alleges that Columbia violated Title IX by requiring her to do so within twenty-four hours, and telling her it would cost five hundred dollars. Columbia has moved to dismiss Roskin-Frazee’s lawsuit, arguing that she obstructed her own investigation by waiting months to file an official report. In October of last year, with a group of protesters, Roskin-Frazee barged into one of Goldberg’s law classes to publicly accuse her of endangering student survivors.

The creation of SHIFT was announced to the university at the end of February, 2015, in an e-mail from President Bollinger. Hirsch and Mellins began soliciting applications for a paid undergraduate advisory board, ultimately selecting a dozen or so students, including members of the Greek system, student-government leaders, a ballerina in Columbia’s General Studies program, anti-sexual-assault activists, a sex educator, a Barnard student, and an R.A. For the next two years, when school was in session, the group met over bagels at 8 A.M. every Monday. The board created a typology of Columbia students—the hyper-involved, the completely disinterested, the kids who find their thing and stick to it—and corrected the researchers in their sometimes fumbling attempts to classify student identities. (Each time they pointed out such a mistake, one student-board member told me, the researchers’ eyes would pop in surprise, and then they’d come back the next week saying, “We had seventeen meetings since the last time we saw you, and we’re going to do what you say.”) The students planned promotional events, setting up SHIFT tables outside the dining hall and the gym. They brought the researchers, who answered questions for students, and made sure they always had snacks. “Snacks, we learned, were a really big thing,” Hirsch said.

Meanwhile, Hirsch and the Columbia sociologist Shamus Khan prepared a team of ethnographers—current and recent grad students, who were close to their subjects in age—to talk with undergraduates about intimate subjects. These interviews would be the first big component of SHIFT’s information-gathering. The ethnographers began, that fall, with “participant observation”—i.e., hanging around football games and drinking club soda at student bars. Shortly afterward, a story appeared in the student newspaper the Daily Spectator, in which an unnamed sophomore said that Khan had been spotted taking notes at 1020, a popular bar near campus. (The plans for the ethnographic research had been announced in the Daily Spectator months before.) The story was picked up by the Post, which reported that “Columbia University researchers are spying on the school’s students at bars and campus parties as part of a new study about sexual health and violence—and the students say it’s creeping them out.”

In fact, by all accounts, the process went pretty smoothly. Some students, after talking to the researchers for a while, invited them to parties, or to kick it in dorm rooms. Many university employees are required to report sexual assault to the Title IX coördinator, but the researchers received a waiver so that they could promise students confidentiality while engaged in SHIFT research. Alexander Wamboldt, an affable, bearded Princeton Ph.D. who worked as a SHIFT ethnographer, told me that it was important, in these encounters, to “model good, consensual research behavior”—he announced his name and his purpose, along with a disclaimer about confidentiality, before entering a conversation. He and the other researchers conducted one-on-one interviews with a hundred and fifty-one students about their sex lives and their experiences at Columbia. (Students were paid for the time they spent in these interviews.) Hirsch and Khan sorted through the data and adjusted their approach when they weren’t getting all the information they needed. Wamboldt was hired to focus on so-called high-status men, such as those involved in athletics and fraternities, a group of students who hadn’t, up to that point, spoken much with the ethnographers, perhaps wary of the possibility that they’d be portrayed badly in whatever the researchers wrote up.

The interviews were bracing. Talking about sex brings a lot to the surface—students discussed loss, family, trauma, hardship, fear. Some of the men Wamboldt spoke to cracked offhand jokes about having been raped. The members of the ethnography team soon decided that they needed to do a mental-health check-in at their weekly meetings: they would go around the room, and everyone would relate how he or she was coping with the work.

In one advisory-board meeting, Mellins and Hirsch shared preliminary observations, and Mellins brought up affirmative consent—the practice of actively, mutually soliciting enthusiasm throughout a sexual encounter, which is now the legal standard for universities in New York and California. Most college students learn about it in orientation seminars or from online modules that they are required to complete. Mellins told the administrators that affirmative consent rarely factored into the experiences that students were describing.

“One of our institutional advisers pretty much fell off her chair,” Mellins told me. “She said, ‘How can it not be a thing? We’re working so hard to teach them.’ And our point was: there’s a really broad disjuncture between what students learn and what they actually practice.” The researchers found that the practice is much simpler to understand than its detractors, who tend to picture a stack of paperwork accompanying every make-out session, seem to think—and also less common than its proponents would like to believe. (SHIFT plans to publish a paper on affirmative consent later this year.)

Hirsch and Mellins launched the second phase of the study, an enormous daily-diary project, in October. Four hundred and twelve students were asked to fill out a short online questionnaire every day for sixty days. (The student board convinced the researchers that the only way to maintain subject participation through midterms was to pay: diarists got a dollar an entry.) The idea was that researchers would be able to quickly scan each twenty-four-hour period for mood, sleep, sexual activity, substance use, and unusual experiences. The pool of data could then be parsed for patterns and fine-grained interactions. Researchers might find, for example, that unwanted sexual contact is more likely to occur in the midst of other crises, or after a person has experienced unwanted sexual attention in another setting.

In January, 2016, the SHIFT team recruited students for Part 3: a sweeping, onetime survey. The student board roped in peers with the promise of gift cards, and by talking to them about how important the project was, how it could show that Columbia took sexual assault more seriously than other universities, and how, if they participated, they’d get snacks. (Students who took the survey in SHIFT’s temporary office got fruit, candy, pizza, and chips.)

The survey contains hundreds of questions, many of them startlingly intimate. It seems likely that no previous survey has so accurately reflected how sexual assault actually occurs in college—as an event embedded within the fabric of everyday life, which both perpetrator and victim understand based on their background, their habits, their state of mind. The survey asks students about sleep, exercise, eating habits, mental health, where they get alcohol, what sort of dorm room they live in, where they party and how. It asks about money, family, friends, their sexual experiences before college, their sense of agency and of self-worth. It asks about gender identity and attraction, about the moments just before an incident—who was around, what was happening—and what followed, immediately and in the long term. It asks about consent: if students expect their partners to ask, if they think it’s a matter of body language, if they think that asking once at the beginning of a hookup is fine. It asks about attitudes regarding sex and gender, sussing out common cultural biases: To what degree do they think that women lie to get ahead? Do they think that men should reveal vulnerability? Do they believe that it can’t be rape if both people are drunk? Are they not at all sure, a little sure, somewhat sure, pretty sure, or very sure that they could say no to having sex with someone if they want to date that person? What if they want the person to fall in love with them, or if the person won’t use a condom? What if they’ve had sex with the person before?

Twenty-five hundred Columbia and Barnard undergraduates were invited to participate in SHIFT’s survey, and sixty-seven per cent of them did so. I took the survey myself one day at the end of December—answering in the present, as a twenty-nine-year-old, and thinking about how I would have answered at eighteen. In the course of a half hour, I felt nauseated, and then oddly comforted, by how well the questions were outlining my life. A detailed constellation emerged of all the things that had protected me in college: a chemically stable disposition, satisfying relationships, a sense of control over my experience at school, a lack of confusion about what I wanted sex to be. My vulnerabilities—a certain recklessness, a freshman-year social life that depended on spaces and substances provided by men—were just as clear. I could see the desires and the habits, sexual and otherwise, that traced the path between then and now. I started to wonder if the research that SHIFT is producing might start closing the gap between two seemingly contradictory realities. Sexual assault on campus is frequently portrayed as lurid and dark and complex. But the experiences that live in our heads are often obvious and ordinary, sometimes heartbreakingly so. SHIFT is, in a sense, a reporting project of unprecedented scale, a map that genuinely reflects the size of the territory. It could be one of the first endeavors to show the magnitude and the texture of the problem at the same time.

SHIFT’s research concluded in the fall of 2017. Since then, the team has been analyzing the data and preparing to publish a slew of papers about the results in peer-reviewed journals. (In December, Hirsch and Khan sold a book about SHIFT, tentatively titled “The Sexual Project,” to Norton, to be published in 2019.) The first paper, which appeared in the open-access online journal PLOS ONE in November, laid out what the team learned about the frequency of sexual assault at Columbia. Sexual-assault research is notoriously contested and spotty—many regularly cited statistics come from studies with big design flaws, such as small sample sizes, or loose definitions of “college student.” The record-setting response rate for the SHIFT survey makes its data unusually comprehensive and reliable. In certain important respects, its numbers are in keeping with previous findings: a little more than one in five respondents said they had experienced sexual assault since starting college—twenty-eight per cent of women, twelve per cent of men, and nearly forty per cent of gender-nonconforming students. (The survey did not use the term “sexual assault”; it asked about “unwanted sexual contact.”) But there were also surprises. It’s long been established that women and L.G.B.T.Q. students are especially vulnerable to assault; SHIFT found that students who are struggling to pay for basic necessities are, too. Men in fraternities are, in fact, more likely than other male students to be perpetrators; SHIFT found that they were more likely than other men to be victims as well. A culture that doesn’t teach men to ask for consent often doesn’t teach them that they can withhold it, either.

Hirsch and Mellins avoid the term “rape culture” when discussing their work. I’ve never liked that phrase, not because it doesn’t name something real but because it emphasizes the way that the world is already prepared to hurt me, rather than emphasizing my personal, and not entirely predictable, relationship to the world. (As Jennifer Doyle, an English professor at the University of California, Riverside, puts it in her book “Campus Sex, Campus Security,” the term distances sexual violence from “the force of the ordinary.”) Hirsch and Mellins often talk about “sexual citizenship,” which they define as a “person’s understanding of his or her right, and other people’s equivalent right, to sexual self-determination.” In the conference room at the Mailman building, Hirsch told me, “Part of what I see our work doing is disrupting these scripts that women give consent and men secure it—that men are sexual agents and women are gatekeepers, which is affirmed by consent education that frames men exclusively as potential perpetrators.”

“Of course, you don’t want to minimize the fact that women are still holding the burden on this, in terms of absolute numbers,” Mellins said. She hesitated. “But you want to work in a way where there isn’t a single story.” A trans student who is assaulted at a party is experiencing something different from a freshman girl whose hookup is ignoring her protests in a dorm room. Both of them are experiencing something different from a boy who has never imagined that he would ever give or receive a no. Mellins pointed to an article about a Brown University student who’d been assaulted in a bathroom by another man, and then, later that day, attended a standard prevention workshop, where he felt entirely alone. “If you don’t give someone permission to be at risk, then they can’t seek help,” Mellins said.

The researchers discussed their findings with the student board—they’re all still in a group chat together—and also with administrators. Certain fixes, they’ve realized, are impossible to implement. All college students would benefit from drinking alcohol in a gentler manner: often with food, rarely in basements. But colleges can’t encourage that among underage students without breaking federal law. When I was talking with Hirsch and Mellins, I thought about my own experience with the Greek system. The National Panhellenic Conference, which adheres to rather antiquated gender norms, forbids sororities from holding parties where alcohol is served, which means that, at many schools, the most accessible parties for freshmen take place on fraternity terms, and on fraternity turf.

Every school’s environment is different—where students drink, how they get home from parties, the geographies and the conditions of their vulnerability—and the nudges and interventions have to vary accordingly. But Hirsch and Mellins hope that their research can serve as the beginning of a network of innovative cross-campus studies. In the meantime, they’re talking to administrators about the interrelationship of mental health, substance abuse, and sexual assault, and about how different types of incidents and different types of students require different types of prevention and response. Many of these conversations have echoed long-standing conclusions in public-health research, and also what some students are already asking for: more crisis support, more consideration for specific populations, more access to spaces on campus that feel like their own. “I’m grateful the SHIFT team chose to do this,” Roskin-Frazee told me. “I hope they are persuasive to administrators who are not easily persuaded.”

One night in January, I called Emma Sulkowicz to talk about Hirsch and Mellins’s project. Sulkowicz was disarming and philosophical, despite having spent five hours in the dentist’s chair earlier that day. Sulkowicz had not heard about SHIFT before, and was politely resistant to the idea: “My view in this whole thing is that, the more that Columbia can retreat behind ‘Here’s a program, here’s a study, here’s a process,’ the less that any human that finds themselves in this machine will ever be incentivized to act based on their moral compass.”

What if, I asked, the idea behind the study was tinkering with the machine, figuring out how to reorient that moral compass?

“That makes me think of asking someone to wash the dishes, and they tell you, ‘I’ll try,’ ” Sulkowicz said. “I think that’s the difference between spending two million dollars to try to understand the conditions that create a community that’s conducive to sexual assault versus just doing the right thing—expelling people who sexually assault other students.”

Sulkowicz wants to change behavior, too, but thinks that punishment is more efficacious than tweaks to campus life. When Columbia settled the lawsuit filed by the man Sulkowicz accused of rape, it put out a statement, noting that his “remaining time at Columbia became very difficult for him and not what Columbia would want any of its students to experience.” But Sulkowicz believes that what he went through had a salutary effect. “He’s been scared shitless,” they said. (The man’s lawyer called this statement “preposterous,” and said that he had done nothing wrong.)

Sulkowicz also said something that I kept hearing from Columbia students: “It’s about finding a way to make your institution, and the people who run it, more human.” Earlier that week, I’d spoken to a former SHIFT student-board member named Morgan Hughes, a laid-back twenty-three-year-old hip-hop musician. She called me from a coffee shop in Cleveland, where she’d moved after graduation. She had been a disengaged student, by her own account, mainly focussed on her music. Her friends at school, most of whom were people of color, had found it difficult to secure space and permission from Columbia to hold their own events, she told me. “Everything is so regulated, so limited, everything’s super uptight,” she said. “Columbia always says they’re listening, taking students into account, and then they turn around and make a decision that doesn’t acknowledge any of that conversation. But SHIFT did listen. They changed their agenda based on what we talked about. It didn’t feel like we were just wasting our breath.”

Would SHIFT make things different at Columbia? “Every four years, there’s a new student body, and I think Columbia is used to just waiting it out,” she said. “But this time there are professors involved. Shamus Khan is going to be there, Jennifer Hirsch is going to be there. It’s up to Columbia if they want to shoot themselves in the foot and ignore it, but people are actually paying attention to this.” She paused, and coffee-shop noises tinkled in the background. “I mean, Columbia, you should want to solve the problem, so you don’t keep having to solve the problem, you know what I mean?”