What Scientists Learned by Eavesdropping on Thousands of People
.jpeg)
Twenty years ago, a team of psychologists reviewed audio from almost four hundred university students who had been carrying “electronically activated recorders”—EARs—for several days in a row. Five times an hour, the devices automatically turned on for thirty seconds. The psychologists then analyzed the audio to test a popular claim. A recent book, “The Female Brain,” had asserted that women were intrinsically chattier than men—blabbing roughly twenty thousand words a day, while men uttered only seven thousand. The study found no evidence of that. On a typical day, both men and women spoke an estimated sixteen thousand words a day—roughly the word count of “A Midsummer Night’s Dream.”
Matthias Mehl, a social psychologist at the University of Arizona, who helped conduct the study, recently set out to replicate his findings with a larger data set: audio from more than two thousand people between the ages of ten and ninety-four, recorded between 2005 and 2019. Once again, Mehl concluded that men and women were equally talkative. But, strangely, when he and a co-author further analyzed the results, they found that participants had spoken an estimated twelve thousand seven hundred words a day—twenty per cent less than in the earlier study. “We thought we must have made a mistake,” Mehl said. Each year, he went on, the number of words spoken daily seemed to decline by about three hundred and thirty-eight. That translates into around a hundred and twenty thousand words a year—about the length of “Sense and Sensibility.” And Mehl noticed that the decline was even steeper for people under twenty-five, who lost an average of four hundred and fifty-one daily words a year. “This is many, many minutes of conversations,” Mehl told me.
A paper about the lost words, published earlier this year, has inspired intriguing theories. Valeria Pfeifer, one of its co-authors, speculated that spoken words have been supplanted by digital communication. This isn’t necessarily a bad thing; I text with a good friend in London almost every day, which keeps us connected when we go months without seeing each other. “Maybe I don’t have to talk to people about trivial matters anymore, and I’m judiciously pruning my communications to be focussed on those things that are more important,” Joshua Smyth, a health psychologist at the Ohio State University who was not affiliated with the study, told me. Mehl raised more concerning possibilities. “We don’t ask people for directions anymore,” he said. “We take the self-checkout line at the grocery store, and we don’t talk to our neighbors.” He worries that the EAR logs reflect something that other psychologists have observed: the erosion of small, everyday interactions. In a new book, “Once Upon a Stranger,” the psychologist Gillian Sandstrom notes that speaking to people whom we don’t know has been shown to teach us new things, to put us in a better mood, and to make our lives feel more satisfying. Failing to do so could make us isolated or distrustful of others.
The EAR can’t prove any of these theories—it can only listen. Mehl co-created it, in the nineties, to gauge who people talked to, how often they talked to them, and the kind of language they used. “We basically get an unobtrusive record of a person’s day as it naturally unfolds,” he said. The first version was a modified microcassette recorder; Mehl had to call participants in the morning to remind them to flip the tape. Now the EAR comes in the form of a smartphone app. As long as participants remember to keep their phones on them, it will passively document sounds in their lives. Researchers have used the EAR to listen in on postpartum women, couples, Americans after 9/11, and young athletes.
In April, I loaded the EAR app onto a borrowed Android phone. I wondered if the soundscape of my own life was quieting down—would I wind up recording a lot of silence? In the span of a week, the phone spent time on my desk, in my kitchen, at dinner parties, and in meetings with a contractor. Even though I record several hours of journalistic interviews each week, and my husband and I have a tradition of recording recaps of our travels when the memories are fresh, I felt slightly nervous about replaying the audio. The EAR exists because we aren’t fully aware of how we sound. What would it reveal?
In the twentieth century, many social scientists advanced the argument that measuring what happened in “real life” could unveil more than highly controlled laboratory experiments. In 1949, eight observers worked in shifts to generate data for “One Boy’s Day,” a report that documented every activity of a seven-year-old in Kansas. In the seventies, Mihaly Csikszentmihalyi, a psychologist recognized for naming the “flow state,” helped develop what came to be known as the experience-sampling method, in which research participants filled out questionnaires whenever a wristwatch or pager beeped. In the nineties, two other psychologists, Arthur Stone and Saul Shiffman, introduced a similar approach, the ecological momentary assessment (E.M.A.), which asked people to share how they felt or what they were doing throughout the day. Stone, the director of U.S.C.’s Center for Self-Report Science, told me that the resulting snapshots gave researchers an “understanding of what that person is really like.”
It’s not that people lie. Rather, in Smyth’s telling, it’s that we can’t say accurately what we did last month, or how we felt this morning. “When I’m asked to summarize my experiences, it tends not necessarily to reflect my actual experience, but more how I think my experiences are,” Smyth told me. Different psychological tools can document different aspects of life. The E.M.A. captures momentary feelings; other methods passively track physical movements or social-media use. (Most people say that they move much more than they really do.) The EAR listens. Susan Wenze, an associate professor of psychology at Lafayette College, emphasized that it does not capture inner experiences or feelings. It can’t say whether you’re lying or putting on a brave face. But it often leads to surprising findings.
In 2016, a study that used the EAR found that children’s asthma symptoms, such as wheezing, increased when they fought with their family members—an association that wasn’t as noticeable when families filled out daily diaries. Another study found that romantic partners rarely discussed breast cancer after one of them was diagnosed, except to work through logistical considerations such as appointment times and doctors’ credentials. A third one found that when people with rheumatoid arthritis or cancer swore around others they received less emotional support from friends, and their depression symptoms tended to worsen. Yet another review of EAR audio found that people who sighed frequently were not necessarily burdened by negative emotions.
People quickly forget they’re being recorded, according to Charles Raison, a professor of human ecology and psychiatry at the University of Wisconsin-Madison who has used the EAR in several studies. Participants usually wear pins disclosing to others that they may be recorded, but the audio rarely includes discussions of the research, and people don’t seem to change their behavior as a result. “After a while, they’d relax,” Raison told me. “They’d have sex with it in the room. They would go to the bathroom with it on.” In one study, on whether morality is a consistent personality trait, the EAR overheard sarcastic remarks (“You want my lemonade? Sure! After you drank the whole thing!”) and bragging (“My dad’s car collection is so large he had to build a new garage.”). Raison, who had previously studied meditation, used the EAR to evaluate whether two hundred people who participated in meditation subsequently became kinder, swore less, or engaged in deeper conversations. He saw no effect. “I’d probably still be a meditation researcher if that hadn’t been such a negative study,” he said. A project by other researchers listened in on people who had recently separated. More frequent contact did not seem to cause immediate distress, but it seemed to be a predictor of emotional struggles two months later—suggesting that going “no contact” might help people with moving on.
The week that I recorded wasn’t a dramatic one. I wasn’t in the midst of a separation or raising a child with asthma. I expected a neutral recap of a somewhat boring period, and when I clicked through the thirty-second snippets from April 30th I did hear my voice conversing on routine work calls and over dinners, talking to my sister, and discussing a basement leak. Several clips were devoid of any sound at all, except rapid typing and the occasional clearing of the throat. I noticed myself feeling self-conscious about solitary days when I spoke little, and proud when I’d had a chatty day. Unfortunately, I encountered some technical difficulties mid-week; I was sad when I realized I had no snippets from a long dinner at a friend’s house.
I really did hear things that I didn’t know about myself. I learned that I should probably take more breaks that don’t involve work on my house, and that a cough from a recent cold had lingered more than I’d realized. During work calls, I found myself to be friendly, but much more formal than I expected. Listening back to a conversation with my mom, I cringed at the impatience that rapidly showed up in my voice. In a few dinnertime recordings, I was taken aback by the sound of soft, relaxed voices over the sounds of forks and knives. My husband and I were using tones that I didn’t hear myself use with anyone else; we were quiet and completely unguarded. Mehl, the co-creator of EAR, told me that his study participants often share similar micro-insights. “One person said, ‘I didn’t know I was so quiet,’ ” he told me. Another told him, “I don’t think of myself as arrogant, but I can see why people say that.” When Mehl recorded himself, he was surprised by the share of his time that he spends alone. “We typically see our lives from the inside out,” he said. “This gives you a chance to hear your life from the outside in.” Lately, in response to the study about lost words, he’s been consciously trying to chat with strangers.
From the data that I gathered, I couldn’t figure out if I was speaking fewer words than I might have in the past. Yet there was almost no small talk in my recordings. I’d expected to hear myself spending more time in my neighborhood—chatting at my local bakery, say, or visiting the community garden. According to Pfeifer, the researcher who co-wrote the lost-words paper, in-person interactions are worth preserving because they’re fundamentally different from other ways that we communicate. We use gestures and facial expressions that aren’t available on the phone or in writing; there’s no chance to save a response for later, or to agonize over what to say. I was struck by how fragmentary spoken language can be. We interrupt. We finish one another’s sentences. “What we say is fleeting,” Pfeifer said, and there’s usually no record of it. “Once said, it is lost.” ♦