The Murder That Wasn't Watched

On the morning of March 27, 1964, readers of The New York Times opened their paper to a four-column headline on page one: “37 Who Saw Murder Didn’t Call the Police.” Beneath it ran one of the most influential pieces of crime journalism of the twentieth century. The story described, in clinical detail, how a 28-year-old bar manager named Catherine Genovese, called Kitty, had been stabbed to death outside her apartment building in Kew Gardens, Queens, while dozens of her neighbors watched from their windows and did nothing.

The article was 1,400 words long. It quoted the assistant chief inspector of the New York Police Department as saying he had been a detective for many years but could find no explanation for what had happened on Austin Street. It quoted neighbors muttering excuses through doorways. It implied, without ever quite saying so, that something rotten had taken root in the modern American city: a willingness to watch a young woman die rather than lift a telephone.

The story horrified the country. It also did something rarer: it produced a scientific theory. Within four years, two social psychologists had taken the implied moral of the Times article and turned it into one of the most cited findings in the history of their field. The bystander effect, as they called it, would become a fixture of psychology textbooks, jury instructions, military training manuals, and Sunday sermons. It would shape how a generation of researchers thought about the moral capacities of strangers.

It was also, in most of its essential details, untrue.

The Editor and the Commissioner

The origin of the Genovese myth was not a psychologist or a reporter but a lunch. Ten days after the murder, the Times metropolitan editor, A.M. Rosenthal, sat down with Michael J. Murphy, then commissioner of the New York Police Department, at Emil’s, a restaurant near police headquarters. Rosenthal was a 41-year-old former foreign correspondent who had recently returned from postings in India, Poland, and Japan, and he was hungry, by his own later admission, for a story that would announce his arrival as a city editor of consequence. ¹

Over the meal, Murphy mentioned in passing that 38 people had watched the Genovese killing and that not one of them had called the police. The number stuck. Rosenthal returned to the newsroom and assigned the story to a reporter named Martin Gansberg. Gansberg filed within days. The headline went through revision (the body of the article said 38 witnesses, the headline said 37, an inconsistency no one corrected), and the story ran with all the weight that a Times metropolitan front page could confer.

Rosenthal himself would build a small career on the case. Within months he published a book, Thirty-Eight Witnesses, which extended the parable. He gave speeches. He saw in the killing a confirmation of something he had long believed: that the moral fabric of the city was fraying, that anonymity bred indifference, that the witnesses on Austin Street were emblems of a wider sickness. He did not, in any of this work, return to the original police files. He did not interview the witnesses again. He did not, as the journalist Jim Rasenberger would later put it, treat the story as journalism. He treated it as a parable he had been given and was responsible for delivering. ²

The Experiment in the Intercom Room

Among the readers of Rosenthal’s coverage were two young psychologists at separate New York universities. John Darley, 30, taught at NYU. Bibb Latané, 31, taught at Columbia. They had met at a party shortly after the murder and found themselves, like much of the city, unable to stop talking about it. But they were trained to convert horror into hypothesis, and the question they framed was sharper than the moral one the Times had implied. The question was not whether the witnesses were monsters. It was whether anyone, placed in the same situation, would have behaved differently.

The experiments they ran in 1968 became models of social-psychological method. In one famous version, an undergraduate was placed alone in a small room and told he would be discussing personal problems with other students via intercom, for reasons of privacy. The other students were tape recordings. After a few minutes, one of the voices began to slur and stammer, then to choke, then to plead for help, then to fall silent in what sounded unmistakably like a seizure. ³

The results were arresting. When the subject believed he was the only listener, 85 percent left the room to seek help, most within a minute. When he believed four other people were also hearing the same sounds, only 31 percent did so, and those who did took significantly longer. Darley and Latané ran variations with smoke pouring under a door, with a woman crying out from the next office, with seizures and falls and emergencies of every kind. The pattern held. The presence of other witnesses, even imagined witnesses on the other end of an intercom, suppressed the impulse to act. ⁴

They named the mechanism diffusion of responsibility. If a person believes that responsibility for acting is shared, she will feel less of it herself. They paired this with a second mechanism, pluralistic ignorance: in ambiguous situations, people look to others for cues about whether something is really wrong, and if everyone else is busy looking calm, the collective performance of calm becomes evidence that nothing is wrong. The bystander effect was born from the marriage of these two ideas, and the Genovese case was its founding myth. Darley and Latané cited the killing in their 1968 paper. Every textbook that reprinted the experiment cited it too. The murder and the theory became inseparable.

What Actually Happened on Austin Street

For decades, the foundational story went uninspected. Then, in the early 2000s, a handful of researchers and journalists began pulling on the threads. The American lawyer Joseph De May Jr., who lived in Kew Gardens and ran a local history website, started collecting court documents, trial transcripts, and police records. The British psychologists Rachel Manning, Mark Levine, and Alan Collins assembled this material into a peer-reviewed reassessment for American Psychologist in 2007. What they found dismantled the Times account almost line by line. ⁵

There were not 38 witnesses to a murder. There were perhaps a half-dozen people who saw any portion of the attack at all, and most of those saw only fragments through curtains, in darkness, from upper floors, in confusion. The killing did not take place in a single sustained scene. It happened in two attacks separated by roughly half an hour, the first on a busy street where one neighbor did shout from his window, causing the attacker to flee briefly, the second in a vestibule out of sight of most of the building. Many of the people the Times counted as witnesses had heard a single scream in the night and assumed, as people often do in cities, that it was a lover’s quarrel or a drunken argument from the bar down the block.

At least two neighbors did call the police. One call came after the first attack. The exact response, given the state of NYPD record-keeping in 1964 and the absence of the 911 system (which did not exist in New York until 1968), is difficult to reconstruct, but the calls were placed. And there was Sophia Farrar, a small woman who lived in the building and who, when a neighbor knocked on her door and said something terrible was happening downstairs, ran down to the vestibule and held Kitty Genovese in her arms until the ambulance came. Genovese died on the way to the hospital. She did not die alone, watched by 38 silent windows. She died in the arms of a friend.

Kevin Cook, in his 2014 book Kitty Genovese: The Murder, the Bystanders, the Crime That Changed America, walked through the Times errors one by one. The number 38 appears nowhere in the police files. It seems to have been an off-the-cuff figure offered by Commissioner Murphy over lunch, possibly the count of people the police had attempted to interview, not the count of people who had watched a murder unfold. By the time the discrepancy could have been caught, the story had hardened into legend. ⁶

The Inconvenient Footage

If the founding case was a fiction, the obvious question was whether the theory itself survived. For a long time, defenders of the bystander effect argued that the lab evidence stood on its own, independent of the Genovese narrative. The intercom experiments were real. The smoke-filled rooms were real. People did, demonstrably, behave more passively in groups than alone in controlled settings.

But the lab was not the street, and in the 2010s researchers gained an unprecedented way to study the street: closed-circuit television. By 2019, public space in much of the developed world was saturated with cameras. Most violent incidents in city centers were now recorded from multiple angles. For social psychologists, this was something like the invention of the telescope. They could finally observe, rather than simulate, what happens when strangers encounter a fight.

A team led by Richard Philpot, working with Mark Levine and colleagues across institutions in Denmark, the United Kingdom, and the Netherlands, assembled a dataset of 219 real public conflicts captured on CCTV in Lancaster, Amsterdam, and Cape Town. The cities were chosen to span low, medium, and high rates of violent crime. The researchers coded each video for the number of bystanders present and whether anyone intervened, either by separating combatants, blocking blows, calming aggressors, or comforting victims. ⁷

The headline finding inverted decades of textbook orthodoxy. In 91 percent of incidents, at least one bystander stepped in. And the more bystanders present, the higher the chance that someone helped. Each additional onlooker raised, not lowered, the odds of intervention. People in groups did not freeze into silent witnesses. They acted, often quickly, often in coordination, often at some physical risk to themselves.

A separate review by Peter Fischer and colleagues, published in Psychological Bulletin in 2011, had already suggested that the classical effect weakened or disappeared when the emergency was unambiguous and the danger to the victim was clear. In situations where a stranger was obviously bleeding, or obviously being attacked, the inhibitory effect of an audience shrank toward zero. The bystander effect, in other words, was real for ambiguous emergencies (the kind one might be embarrassed to misjudge) but unreliable as a general theory of human behavior in crisis. ⁸

What the Lab Got Right, and What It Missed

It would be too neat to say that Darley and Latané were simply wrong. They were not. Their experiments are reproducible. There is something that happens in a room full of strangers in the presence of an ambiguous signal: a kind of mutual reading of faces, an attempt to calibrate the right reaction. If the signal is faint enough (smoke that might be steam, a thud that might be furniture, a cry that might be laughter), then the social cost of overreacting can outweigh the perceived cost of doing nothing, and the group can talk itself into stillness.

What the lab missed was the asymmetry between ambiguity and clarity. Their experiments were designed, by necessity, to keep the emergency just plausible enough that doubt could grow. The chosen sounds were heart attacks, seizures, falls. The subject could not see the victim. He had only intercom audio and the inferred presence of other listeners. In that thin sensory environment, doubt did the work the researchers expected it to do.

The street is louder than the lab. A man swinging at a woman on a sidewalk is not an ambiguous stimulus. A teenager being kicked by three others on a tram platform is not an ambiguous stimulus. There is little to misread. And there are other forces in operation that the lab carefully excluded. There is the presence of potential allies, which can lower the perceived cost of stepping forward. There is the visibility of the act itself, which can shift the calculus from “is this happening” to “who will move first.” There is the simple animal fact that human beings, on average, do not like to watch other human beings be hurt.

Manning, Levine, and Collins made a subtler point in their 2007 paper. The Genovese parable, they argued, had not just been factually wrong. It had been pedagogically harmful. By teaching generations of students that strangers were dangerous in their indifference, the field had subtly licensed a fatalism about urban life that the data did not support. The lesson sounded scientific but functioned like folklore: a story about modern alienation that confirmed what people already half-believed.

A Woman Named Sophia Farrar

There is one figure in the Genovese case who has, in recent years, been pulled slowly out of the shadow that Rosenthal’s article cast. Sophia Farrar was 38 years old in 1964, an Italian immigrant who lived on the ground floor of the apartment building at 82-70 Austin Street with her husband. She was friends with Kitty. When another neighbor, Karl Ross, knocked on her door at around 3:30 in the morning and told her something was wrong downstairs, she did not hesitate. She did not call out from a window. She did not weigh, in any visible way, the considerations that decades of psychology would assign to her. She put on a coat. She went out into the dark hallway. She found Kitty Genovese at the foot of a stairwell, fatally wounded, and she sat with her, and she held her, and she told her she was not alone.

Farrar was alive when the Times article was published. She was alive when Rosenthal published his book. She was alive when the bystander effect became a staple of undergraduate psychology. She was never interviewed for any of it. She died in 2020, at the age of 92, by which point the Times had begun the slow and incomplete work of correcting its own story. Her name appears in the corrections, finally, as a counter-fact to the parable. ⁹

There is a temptation, in telling the revised history, to make Farrar into a moral hero who refutes the cynicism of the original tale. But she would probably have resented the elevation. By all accounts she did what she did because Kitty was her friend and someone needed to help, and the situation, viewed from the inside, was not a philosophical test of urban anonymity but a request for a coat and a flight of stairs. The point is not that she was extraordinary. The point is that she was ordinary, and that the people we have been trained to imagine as silent watchers were, in fact, ordinary in much the same way: confused, half-asleep, uncertain what they had heard, and, in at least two cases, reaching for a telephone.

The Useful Half of a Bad Story

What survives, then, of the bystander effect? Not the founding myth. Not the strong version that became cultural common sense. What survives is something narrower and more interesting: a tendency, under conditions of genuine ambiguity, for groups of strangers to read each other’s stillness as evidence that nothing is wrong. This is real. It is worth knowing. It is also, by the standards of the original parable, almost embarrassingly modest.

The larger lesson is about what happens when a discipline takes its founding case on trust. Darley and Latané did important work. They were also working downstream of a newspaper article that had skipped the most basic steps of verification, and they built a theoretical edifice on top of it without going back to the records themselves. This is not a unique sin in social science. The Stanford prison experiment, the marshmallow test, the power-pose findings, the implicit-association test in some of its strong forms: each has had its own season of unraveling. Each was built, in part, on a story too good to interrogate.

There is something almost reassuring about the revision. The world the Times described in 1964 was a world in which a young woman could die under the gaze of dozens of indifferent neighbors. The world the CCTV footage describes is one in which strangers, in the overwhelming majority of cases, do something. They block punches. They pull people apart. They run downstairs in the middle of the night to sit with a dying friend. They are not heroes. They are people who, when the situation is clear enough to read, tend to move.

This is not a triumphalist conclusion. Help is not always enough. Sophia Farrar’s arrival did not save Kitty Genovese’s life. Bystanders intervene, on the CCTV footage, in incidents that are sometimes fatal anyway. The point is narrower. It is that the picture we inherited of ourselves as a species of silent watchers was wrong, and that the truth, when finally photographed from enough angles, is gentler than the parable. We hesitate. We look around. We check each other’s faces. And most of the time, somebody moves.

Sources

1. Jim Rasenberger, Nightmare on Austin Street, American Heritage, 2006. — https://www.americanheritage.com/nightmare-austin-street
2. A.M. Rosenthal, Thirty-Eight Witnesses: The Kitty Genovese Case, McGraw-Hill, 1964 (reissued Melville House, 2008). — https://www.mhpbooks.com/books/thirty-eight-witnesses/
3. John M. Darley and Bibb Latané, Bystander Intervention in Emergencies: Diffusion of Responsibility, Journal of Personality and Social Psychology, 1968. — https://psycnet.apa.org/record/1968-08862-001
4. Bibb Latané and John M. Darley, The Unresponsive Bystander: Why Doesn’t He Help?, Appleton-Century-Crofts, 1970. — https://archive.org/details/unresponsivebyst0000lata
5. Rachel Manning, Mark Levine, and Alan Collins, The Kitty Genovese Murder and the Social Psychology of Helping: The Parable of the 38 Witnesses, American Psychologist, 2007. — https://psycnet.apa.org/record/2007-12058-009
6. Kevin Cook, Kitty Genovese: The Murder, the Bystanders, the Crime That Changed America, W.W. Norton, 2014. — https://wwnorton.com/books/9780393239287
7. Richard Philpot, Lasse Suonperä Liebst, Mark Levine et al., Would I Be Helped? Cross-National CCTV Footage Shows That Intervention Is the Norm in Public Conflicts, American Psychologist, 2019. — https://psycnet.apa.org/doiLanding?doi=10.1037%2Famp0000469
8. Peter Fischer et al., The Bystander-Effect: A Meta-Analytic Review on Bystander Intervention in Dangerous and Non-Dangerous Emergencies, Psychological Bulletin, 2011. — https://psycnet.apa.org/record/2011-07658-001
9. Sam Roberts, Sophia Farrar, Who Rushed to Kitty Genovese’s Side, Dies at 92, The New York Times, 2020. — https://www.nytimes.com/2020/09/04/nyregion/sophia-farrar-dead.html
10. Nicholas Lemann, A Call for Help: What the Kitty Genovese Story Really Means, The New Yorker, 2014. — https://www.newyorker.com/magazine/2014/03/10/a-call-for-help