Before Cairo Wakes

Why nobody helps when everybody can

You sit in a hospital waiting room. A man near you slides off his chair. Ten people see it. Nobody moves. Each person looks at the others. Each sees calm faces. Each thinks, it must be fine, or someone else will do it. Are these ten people cold? Or is something else going on?

A murder, a headline, and a famous name

On 13 March 1964, at about 3am, a young woman named Kitty Genovese was attacked near her home in Kew Gardens, a quiet part of New York. She died. Two weeks later, the New York Times printed a story with a shocking headline: 37 people saw it and did not call the police. Inside the story the number was 38.

America asked how people could be so cold. Two New York psychologists, Bibb Latané and John Darley, did not like that answer. They asked a different question. What if the crowd itself is the problem? They ran experiments and published them in 1968. The idea got a name: the bystander effect (the more people who watch, the less likely any one of them is to help).

What the experiments showed

Their best known test used a fake emergency. A student sat alone in a room and talked to others over an intercom. One voice suddenly had a seizure and cried for help. The student believed either that he was the only one who could hear, or that several others could hear too. When he thought he was alone, most helped within a minute. When he thought four others could hear, far fewer did. I give the numbers from memory, about 85 percent against about 31 percent. I did not check them today.

A result I did check is from 1969, by Latané and Rodin. A woman in the next room seemed to fall and cry out in pain. When a person was alone, 70 percent went to help. When the person waited with a stranger, only about 40 percent did.

Why it happens

There are two simple reasons.

First, diffusion of responsibility (the duty gets spread so thin that nobody feels it). If you are the only one who can help, the whole duty is yours. If ten people can see, each feels one tenth of it. Each quietly hopes another will take it.

Second, pluralistic ignorance (everyone is unsure, so everyone watches everyone else). You do not know if the man is ill or just asleep. So you look at the others. They look calm. You take their calm as proof that all is well. But they are doing exactly the same thing to you. The room becomes a circle of people copying each other's calm face.

The part the famous story got wrong

Here is the twist. The murder story that made this idea famous is mostly false.

In 2007, three researchers, Manning, Levine and Collins, went back to the records. In American Psychologist, they reported no evidence of 38 witnesses. No evidence that witnesses saw the murder. No evidence that they all did nothing. About a dozen people heard or saw parts of the attack. None saw all of it. The police were called. A neighbour and close friend, Sophia Farrar, went to Kitty and held her until the ambulance came. The killer, Winston Moseley, was arrested six days later. The Times later called its own reporting flawed.

I read this through a Wikipedia summary, because the original paper would not open for me. The lesson is plain. A story that fits our fears spreads fast, even when the facts are weaker than the headline.

The strongest case against the idea

The lab effect is real, but it is not a law. A 2011 review of many studies, by Fischer and colleagues, found the effect gets weaker when the danger is clear, when the attacker is still there, and when helping is risky. People step forward when it is obviously serious.

Then came cameras. Philpot and colleagues studied CCTV video of over 200 real street fights in the UK, the Netherlands and South Africa. In over 90 percent of them, somebody stepped in. And more people watching made help more likely, not less. So crowds are not always lazy. The effect is strongest when the situation is unclear and the cost of being wrong feels high: a man who may be drunk, a stranger who may be fine, or a message in a busy group chat.

The part that goes against normal advice

Normal advice says there is safety in numbers. In a vague emergency, one person beside you can be worth more than ten. And silence in a crowd tells you very little about what people feel. Many of those quiet people are worried too. They are just waiting for someone to go first.

Where it shows up

Family. Your mother writes in the family group that she feels unwell. Six cousins read it. Each thinks another has already called her. Nobody does.

Money. A shared bill, a loan to a relative, a joint gift. When everyone owns it, nobody owns it, and it slips.

Work. The mistake on the shared slide. The full shared folder. Everyone saw it. Everyone assumed someone had reported it.

Three things to try

1. When something looks wrong in a crowd, go first, or say out loud, "I am calling." One voice releases everyone else.
2. If you need help in a crowd, point at one person. "You, in the blue shirt, call 123." A name turns a crowd into one person's job.
3. In a family chat, answer a request in the first hour, even if only with "I will do it."