Want to encourage cooperation? Try exchanging names.

Dale Carnegie famously called one's own name the “sweetest, most important sound in any language.” And according to new research, knowing each other's names might also help bring out the best in us. 

A study published earlier this month in the journal Science Advances examines the effects of onymity – that is, the opposite of anonymity – on Chinese students in a classic two-player social experiment in which the most rational choice is betrayal. What researchers found, however, seems to defy rationality: Participants who learned each other’s names opted for cooperation over treachery.

In an age marked by xenophobia and political polarization, studying onymity may offer insight into practical ways of helping strangers get along. This particular study suggests that even small steps toward getting to know one another can bring big benefits for society as a whole, whether it's in a town hall meeting, on a jammed roadway, or in an online discussion forum.

“Since the spirit of cooperation that social cohesion is based upon is crumbling away in some places, be it on Facebook or in societies that are about to be torn apart about issues such as immigration, we sought insight into what enhances cooperation,” said co-author Jürgen Kurths from the Potsdam Institute for Climate Impact Research, Germany, who contributed statistical analyses, in a statement.

Humans have been engaging with the concept of anonymity since the dawn of civilization, with the oldest known masks dating back 9,000 years. The idea of a link between anonymity and immorality goes back at least to the 4th century BC, when Plato discussed the potentially corrupting effects of a magical ring of Gyges that would render its wearer invisible.

More recently, psychologists have examined the role that anonymity plays in promoting impulsivity and a disregard for social norms and reducing one's ability to accurately weigh risks. In 2004, John Suler coined the term "online disinhibition effect" to describe how anonymity, when combined with the absence of a recognized authority and face-to-face real-time interaction, results in people behaving in ways that they would not in the real world.

"Our focus was on onymity as the opposite of anonymity," says Marko Jusup, an assistant professor of mathematics at Hokkaido University in Sapporo, Japan. Rather than explore how anonymity erodes civil discourse, Professor Jusup and his colleagues sought to understand how onymity might promote cooperation by adapting a classic social experiment known as the Prisoner's Dilemma.

Developed in the 1950s by researchers at the RAND Corp., the classical formulation of the Prisoner's Dilemma supposes that two criminals are under arrest and held separately. Each prisoner can choose either to testify against the other or to remain silent. If both prisoners testify, each serves two years in prison. If the one testifies and the other remains silent, the silent one gets three years and the stool pigeon walks free. If both keep mum, they each serve just one year on a lesser charge.

Because snitching on your partner results in a sentence of either zero years or two years, and keeping quiet results in a sentence of one year or three years, from a purely self-interested standpoint, you should always betray your partner, even though you would both be better off by cooperating.

“A purely rational perspective then dictates that onymity should have been disregarded, and people should have kept playing the game as if they were anonymous,” says Jusup. “Rationality, however, is just a part of human psyche, which is nicely emphasized by the results of our experiment.”

Jusup and his colleagues found that, when the participants did not know each other’s names, they cooperated about 25 percent of the time. When they did know each other’s names, they cooperated between 50 percent and 75 percent of the time. In this particular game, onymity offers no additional benefit, suggesting that this willingness to cooperate may be driven by deeper psychological mechanisms that can potentially be leveraged, for good or ill, in the real world.

The researchers added a few twists to the classic game, including a 75 percent chance of there being an additional round and the option for one player to "punish" another player by incurring a small cost to make that player pay an even greater one.

But "punishment did nothing to promote cooperation," says Jusup. Rather, it prompted players to betray their partners again or to retaliate with a counter-punishment. Overall, the participants in the experiment acted on short-term thinking, mostly just responding to the previous move.

"This would suggest that one bad (or good) move may outweigh a series of good (or bad) prior moves," says Jusup. "We must acknowledge that human decision-making is a combination of rational thinking and quirky cognitive biases," he says.

But Justin Grana, a postdoctoral fellow at New Mexico's Santa Fe Institute who specializes in game theory and who was not part of this study, is hesitant to describe the player's cooperation under onymity as a cognitive bias, instead seeing the players as incorporating their anticipated feelings into their cost-benefit calculations. 

"These people are actually incurring a cost," says Dr. Grana. "They don't like it when other people see them as selfish. They care about their self image." Grana also speculates that familiarity may increase sympathy between the players, and that acting against feelings of sympathy may make them feel bad.

The idea that something as simple as a name encourages people to be more sympathetic is particularly intriguing for researchers and developers who study interactions in online communities.

"People act differently when their identity is intact. If you know my name, if you know my face then I'm apt to be a more humane person," says Arthur Santana an assistant professor at San Diego State University's School of Journalism and Media Studies who was not part of this study. In 2014 Professor Santana published a study of newspaper comment boards that found that more than half of anonymous comments included language that was vulgar, racist, profane, or hateful, compared with less than a third of non-anonymous comments. "People's inhibitions drop when they are anonymous," he says.

Grana interprets the online disinhibition effect in traditional economic terms, noting that anonymity greatly reduces the cost of leaving a comment: "If costs go down to do something, more people will do it, especially the people who possibly value it less or maybe put less time into it," he says.

That said, eradicating anonymity online entirely might not be desirable. "Before we all jump on this bandwagon that we say anonymity is terrible, it's important to recognize the counterargument, that there might be some value in anonymity in the sense that you get the raw, unfiltered comment," Santana says.

"Of course, there's no value in hate speech online," he adds, "but I think we're making a mistake in completely removing a person's ability to speak openly and freely."

For his part, Jusup was wary of extrapolating these findings to address real-world problems, pointing out how marketers have used behavioral economics exploit and manipulate our cognitive biases. "How far we should go in using these biases to achieve certain goals – even if these goals are geared towards improving the society – is first and foremost a matter of an ethical debate," he says.

[Editor's note: An earlier version misstated the location of the Santa Fe Institute.]