Imagine two people on a blind date. Who sets the tone? Who picks the topics? You might think it’s the more attractive person—but new research suggests that the factors influencing conversational dominance are not what you might expect. 

In a new paper presented at the annual conference of the American Anthropological Association in Washington, D.C. last week, Joseph Manson, a professor of anthropology, and colleagues at UCLA, analyze what happens when strangers are told to make conversation. Manson recruited 105 college students, organized them into three-person, same-sex groups, and told them to talk about whatever they wanted. The topics they ended up covering were fairly predictable; most groups discussed their majors, their hometowns, their class year. Manson videotaped the conversations and used a few measures to figure out who was dominating the conversation: How many words did each person contribute? Who determined the topic? Who did more of the interrupting?

Then, they looked at how several factors affected conversational dominance. Age had a small effect: Older students were slightly more likely to determine the topic of conversation. People from wealthier zip codes were less dominant, which Manson attributes to a kind of cool indifference. “It’s almost like they’re disdainful of new acquaintances,” he said. “They kind of disengage. They already have friends.”

Surprisingly, strength, fighting ability (as judged by independent evaluators looking at photos and videos of the conversation), social status (as judged by various signifiers like hairstyle, clothing, and the “prestige” of the subject’s major), and facial attractiveness had no significant effect on conversational dominance. Similar experiments have shown that people will accommodate or even mimic a more attractive stranger’s conversational preferences and habits. If, for example, two women are paired for conversation, the less attractive woman often starts to mimic the vocal pattern of the prettier one. (In money-allocation scenarios, where people have to decide how much of a pot of money they’re willing to share with a stranger, people are usually more generous toward the more attractive person.)

Manson discovered one factor that was by far the best predictor of conversational dominance: students’ scores on a psychopathy questionnaire. There are two facets of psychopathy, Manson explained. One, “primary psychopathy,” has to do with being callous and manipulative; the other, “secondary psychopathy,” is marked by impulsive and anti-social behavior. They’re “somewhat correlated,” he told me, but they lead to opposite effects: “People who are high on primary psychopathy dominate the conversation—they tend to be glib and charming.” People high on secondary psychopathy, meanwhile, are less likely to dominate.

“We think the function of the glib charm is to extract information, which the psychopath can then use,” he explained. “People high in psychopathy are judging the usefulness of the acquaintance.” If they’re likely to run into their conversational partner in the future—if they have the same major, for instance, have mutual friends, or come from the same area—they’re going to behave differently, making more of an effort to assert themselves and establish a rapport. Similarly, people high in psychopathy are much more likely to give money to a stranger if they discover they have something in common.