Researcher examines how people perceive interruptions in conversation

We all know that unpleasant feeling when we're talking about something interesting and halfway through our sentence we're interup – "Wait, what's for dinner?" – pted.

But was that really an interruption? Whether or not one person interrupted another depends on whom you ask, according to new research from Stanford's Katherine Hilton, a doctoral candidate in linguistics.

"What people perceive as an interruption varies systematically across different speakers and speech acts," said Hilton, who is also a Geballe Dissertation Prize Fellow at the Stanford Humanities Center. "Listeners' own conversational styles influence whether they interpret simultaneous, overlapping talk as interruptive or cooperative. We all have different opinions about how a good conversation is supposed to go."

Using a set of carefully controlled scripted audio clips, Hilton surveyed 5,000 American English speakers to better understand what affects people's perceptions of interruptions. She had participants listen to audio clips and then answer questions about whether the speakers seemed to be friendly and engaged, listening to one another, or trying to interrupt.

Hilton found that American English speakers have different conversational styles. She identified two distinct groups: high and low intensity speakers. High intensity speakers are generally uncomfortable with moments of silence in conversation and consider talking at the same time a sign of engagement. Low intensity speakers find simultaneous chatter to be rude and prefer people speak one at a time in conversation.

Hilton also found a gender disparity among survey participants. Male listeners were more likely to view women who interrupted another speaker in the audio clips as ruder, less friendly and less intelligent than men who interrupted.

Conversational styles and gender bias

Previous research on interruptions has primarily focused on analyzing recordings of natural conversations. But their results could not be generalizable because every study defines an interruption differently, said Penny Eckert, a Stanford professor of linguistics.

"Katherine's experiment really pulls apart the whole notion of interruption," Eckert said.

For example, when a woman asserts herself in a conversation with a man, is that speaking up or interrupting? While other studies have shown that women tend to be seen more negatively than men if they speak up or interrupt, no one has measured those perceptions quantitatively before, Eckert said.

"Gendered ways of talking and interpreting matter, and they have many consequences, including political ones," said Rob Podesva, an assistant professor of linguistics and Hilton's adviser. "Katherine's research shows that there are systematic gender disparities in how we interpret interruptions. Being aware of these disparities may be the first step in figuring out how to address them in the future."

The differences in conversational styles became evident when participants listened to audio clips in which two people spoke at the same time but were agreeing with each other and stayed on topic, Hilton said.

The high intensity group reported that conversations where people spoke simultaneously when expressing agreement were not interruptive but engaged and friendlier than the conversations with pauses in between speaking turns. In contrast, the low intensity group perceived any amount of simultaneous chitchat as a rude interruption, regardless of what the speakers were saying.

"However, these two categories represent extremes on a spectrum," Hilton said. "In other words, most people are likely to be somewhere in between the two conversational styles."

Hilton also found a significant gender bias among study participants when they listened to audio clips considered highly interruptive by almost all participants. In those clips, interrupters did not just talk at the same time as another speaker, but they also changed the subject or raised their voices.

Male listeners were more likely to view a female speaker who interrupted as ruder, less friendly and less intelligent than if the interrupter were male, although both male and female speakers were performing identical scripts in the audio clips, Hilton said.

However, female listeners did not show a significant bias in favor of female or male speakers.

"Finding this gender bias wasn't as surprising as the extent of it and the fact that it altered perceptions of a female speaker's intelligence, which we don't think of as related to interruptions," Hilton said.

Understanding human interaction

Hilton was curious about how conversations work well before she started studying linguistics. She described her family as incredibly talkative, so growing up, she got used to a high-intensity conversational style.

"During my family gatherings everyone would be talking all at once, and it could get pretty chaotic," Hilton said. "I had to figure out how to navigate that, and I was always interested in who gets to speak when. Who gets to tell their stories and opinions? And who doesn't? It fascinates me how people can manage all of this complexity without any explicit rules for conversation."

Hilton said she is excited to contribute research that helps shed light on the cognitive, social and cultural aspects of human interaction.

"People care about being interrupted, and those small interruptions can have a massive effect on overall communication down the line," Hilton said. "Breaking apart what an interruption means is essential if we want to understand how humans interact with each other."