On September 11, 2001, Amanda Ripley was a 27-year-old reporter living on the Upper West Side in New York City and working for Time magazine. “I got a call after the first plane,” she says. “and then after the attacks, I covered the victims, the survivors, the city. It became a kind of endless story. Every single family member had a story that you would not believe, every one. Conversations with their husband trapped on the 90th floor on his cell phone realising he’s gonna die.” There came a point, she says, when she had to stop doing 9/11 stories for a while. “You can’t write well unless you feel it, and at a certain point I had to stop feeling it.”

But she didn’t stop thinking about it. She became interested in a study of human behaviour in the twin towers. That led her to further reporting on other disasters, notably the effect of Hurricane Katrina on New Orleans in 2005, and that led to research into tragedies: fires; explosions; airplane crashes; hostage dramas. The end result is a fascinating book, The Unthinkable: who survives when disaster strikes – and why.

Ripley, now 34, has not directly been involved in a disaster herself. “I’ve led a very charmed life. I tend to show up after things have happened. But I did a lot of stories about terrible things happening to people for no good reason, and as a coping mechanism, I began to deconstruct what happened. I found behaviour was very similar in lots of disasters. And in many ways it was better than we expect.”

The Unthinkable is part study of the science of reaction to extreme fear, part indictment of the US government’s response to the terrorist threat, part call to arms. The call in question is for the individual to make an assessment of the risks we face and then to develop a strategy to help ourselves, rather than rely on others, should the need arise. “If you speak to disaster experts they only talk about what the government should be doing,” says Ripley. “That’s a mistake. Regular people can have an enormous impact on their own survival chances. The problem is: people think it won’t happen to them and, if it does, they’re screwed. It’s really lame if you think about it. We all take ourselves very seriously in other ways.

“We vastly underestimate our survival chances,” she says. “Most serious plane accidents are survivable, and most people don’t know that. They think if the plane goes down that’s it for them. There’s good evidence people die because of that mentality.”

Obviously, Ripley isn’t talking about a Lockerbie or any other high-altitude catastrophe. Rather, she cites a collision between two planes at Tenerife airport in 1977, when a KLM jet hit a Pan Am 747 awaiting take-off. Everyone on the KLM jet died instantly; many on the Pan Am flight had, however, survived the initial impact, and as it turned out they had a full minute in which to escape before the plane became engulfed in fire. Only 70 people got out, the other 326 on board died, and yet in tests the entire plane could be evacuated in 90 seconds. In another runway fire, at Manchester airport in 1985, 55 passengers died, despite there being a five-minute interval between the plane stopping and the fire preventing further rescue.

What happens in incidents such as the one at Tenerife, argues Ripley, is that some (not all) people go into a kind of paralysis, unable to speak or move, their minds curiously blank. We learn that the classic response to fear is flight or fight, but another F is equally likely: freeze. Restrained animals will often fight frantically for several seconds, and then freeze. As a strategy in the wild, this has its merits: predators will sometimes avoid limp, motionless prey, fearing disease. As a strategy in a stricken aircraft, a burning office block or a sinking ship, playing dead is no use at all.

People don’t get out even when they can, and that is partly because they think they can’t, partly because they don’t know what to do, and partly because of clusters of nuclei deep in their brain called the amygdala. The amygdala is primitive and ancient. It does not totally control an animal’s response to extreme fear (you can override it, as Ripley explains), but it has the first say.

One of the main ways the amygdala responds to a fear input is to flood the body with cortisol and adrenaline. This boosts the gross motor muscles, useful for doing battle with another caveman or scarpering from a sabre-toothed tiger, less handy when you have to strap on a fiddly lifejacket or open an unfamiliar hatch, because both hormones impair complex thinking. Given time, higher reasoning functions will assert themselves, but in an emergency, by definition, time is what you don’t have.

“Fear makes you better at some things and worse at others,” says Ripley. “One thing you’re worse at, always, is making decisions about things you are not familiar with. People have trouble unbuckling seatbelts, for instance. You lose cognitive ability.” On the flight that caught fire in Manchester, the woman nearest the door wasted seconds repeatedly yanking her arm-rest in the belief it was the door handle. Workers in the Trade Centre milled about for minutes, turning off their computers, gathering their belongings, looking for staircases they had never used.

The key to short-circuiting the paralysis response, Ripley found, is familiarity and knowledge, or even better, full-on training. Freezing has never been noted in airline crew members, as opposed to passengers. They are familiar with the plane and are trained for an emergency. “The brain’s got to have options, an alternative script,” she says. “Even a tiny amount of knowledge can make a big difference.”

She thinks airports should have simulators where you can practise evacuating a plane. And she cites the example of Tilly Smith, the English schoolgirl who recognised the signs of impending disaster in Phuket in 2004 because she’d just done tsunamis at school. The beach Tilly’s family was on was evacuated in time.

Times recommends

• A vicar’s wife: ‘There is no one to fight for us’
• Mutton dressed as lad: I’ve seen the future
• Think yourself slim with our hypnodiet podcast