Advertisement

SKIP ADVERTISEMENT

Eureka

Crowds Are Not People, My Friend

A dense crowd in London in 1908.Credit...Hulton Archive/Getty Images

We have all been to a church or a concert without merging with the rest of the audience into some sort of hive mind. Likewise, we all know from experience that Internet message boards aren’t Borg-like mind-melding machines that multiply users’ brain power. So why, then, do we insist on treating crowds, real or virtual, like sentient beings? We’ve long believed that physical crowds are emotional, irrational and prone to violence. Over the last decade, we’ve come to think of virtual crowds as sources of wisdom that can’t be found in individuals. Both these ideas treat crowds as entities, rather than groups of people — an idea that has its origins in 19th-century sociology, which, according to scientists studying crowd behavior today, is deeply flawed.

Clark McPhail, an emeritus professor of sociology at the University of Illinois at Urbana-Champaign, and one of the first people to actually document and study how people behave when they come together in large gatherings, doesn’t even like to use the word “crowd.” It’s too weighed down by inaccurate stereotypes. For years, sociologists thought a crowd behaved like a herd of animals: at some point, it reaches a critical mass and the will of the crowd overrides individual intelligence and individual decision making.

But that’s not what happens. Groups of people are still made up of people. They can behave in helpful and intelligent ways, or they can behave in dumb and dangerous ways. But in either case, a crowd’s behavior depends on what individuals are thinking and how they interact with one another — not some overpowering collective consciousness. “Crowds don’t have central nervous systems,” McPhail said. And that is true whether the crowds you’re talking about are physical or virtual.

Gustave Le Bon was one of the first people to write about crowds as entities separate from the people in them. His 1895 book, “The Crowd: A Study of the Popular Mind,” shaped academic discussions of human gatherings for half a century and encouraged 20th-century fascist dictators, including Benito Mussolini, to treat crowds as emotional organisms — something to be manipulated and controlled. (Perhaps a Le Bonian understanding of crowds makes us feel more comfortable about the atrocities of the 20th century.) But “The Crowd” was more a work of philosophy than of science, McPhail told me. Le Bon’s ideas were based on armchair analysis of past events, not on carefully documented studies of crowds in action. In the 1960s, sociologists began to study protests and public gatherings, and they realized that the things they believed about crowd behavior didn’t align with what took place in the real world.

Take, for example, the effect fear has on a crowd. Common sense — which is to say, the Le Bon-influenced myths you’ve been steeped in since high school — would suggest that a panicked crowd loses all semblance of rationality, charging madly and trampling anyone who doesn’t keep up. But despite individual instances that come to mind — the tragic Who concert in Cincinnati in 1979, say — studies since the early 1980s have shown that groups of people generally don’t move as a collective front, and they aren’t all crazed with terror, even in terrifying situations. On Sept. 11, for instance, large numbers of people organized themselves into a quick, careful and efficient evacuation of the World Trade Center towers. They knew one another, so they discussed plans, they made decisions, they behaved rationally and independently.

In 1997, McPhail and a team of researchers documented the behavior of individuals and small groups that made up the crowd of 500,000 at a Promise Keepers rally in Washington. The Promise Keepers are an evangelical Christian men’s organization, and as such, the event was highly structured, with performers and preachers explicitly asking the audience to do certain things: pray, sing, etc. But at no point during the entire rally were more than 80 percent of the participants doing the same thing simultaneously. Most of the time when the audience acted in unison, less than 55 percent participated.

Scientists who focus on virtual groups see similar patterns. Conor Mayo-Wilson is a researcher of mathematical philosophy at Carnegie Mellon University who studies how people learn and solve problems by sharing information — how scientists in a given field reach a consensus, for instance, or even how African farmers choose which crops to plant. These groups, however different, take advantage of a diverse range of experiences and knowledge, so it’s reasonable to think that collective intelligence might come to a more accurate conclusion than any one individual. But research done by Mayo-Wilson and others shows that this isn’t exactly the case.

For instance, we know today that stomach ulcers are caused by bacteria, Mayo-Wilson said. Scientists were making connections between bacteria and stomach ulcers as early as 1889. But in 1954, Edward Palmer published a paper that claimed to find no bacteria whatsoever in 1,000 human stomachs. Palmer’s study was flawed, but knowledge of that paper spread faster and more widely than the earlier work. Soon everybody knew that bacteria couldn’t live in the stomach, but what they knew was completely false. Linking people into virtual groups enables the sharing of knowledge, but when that information isn’t accurate, it can lead the group consensus astray. “When information comes from a common source, that can cause problems with individual decision making, because it can eradicate minority viewpoints,” Mayo-Wilson said.

This assumption that crowds have some non-fragmented consciousness leads us to the false dichotomy we draw between physical and virtual crowds: one is dumb, the other is smart. But in both cases, we’re placing too much emphasis on the crowd as distinct from the people involved in it. “The thing we’re trying to emphasize is that it’s the individuals and how they interact with one another,” Mayo-Wilson told me. “How those people receive information can influence whether or not they make a good decision.”

This has real-world consequences. When police officers show up at a protest or political rally, they tend to think of the crowd in Le Bonian terms, McPhail told me. That can be dangerous. If the police assume the crowd is acting as one, it becomes easier for a handful of people to provoke a violent reaction from law enforcement — and vice versa. McPhail uses what he has learned from 40 years of studying groups of people to advise law enforcement on better, safer ways to deal with crowds. He told me that 150 years of records from Europe and the United States show violence happens at less than 15 percent of political gatherings. So he instructs officers to never respond categorically to a crowd. If one person is breaking the law, address that person in an unobtrusive way. “If you are blatant and violent, you affect people who weren’t doing anything, and that . . . turns them against you,” he said.

At the same time, knowing that virtual crowds are merely human helps us better predict when one is likely to be smart and when it’s likely to be stupid. Reddit can help someone understand a medical diagnosis just as easily as it can foster a men’s rights movement. Scientists, working as a virtual group, are capable of sharing diverse research to reach a consensus on climate change, but they’re also capable of passing down the received wisdom that crowds have minds. The group itself isn’t what matters. What matters is who they are, what they know and how they interact.

Maggie Koerth-Baker is science editor at BoingBoing.net and author of “Before the Lights Go Out,” on the future of energy production and consumption.

A version of this article appears in print on  , Page 12 of the Sunday Magazine with the headline: Head in the Crowds. Order Reprints | Today’s Paper | Subscribe

Advertisement

SKIP ADVERTISEMENT