Waiting your turn in a conversation is not just polite—it’s an extreme exercise in timekeeping. It typically takes only a fifth of a second, or about half as long as a blink, for people to respond to another speaker. Anything shorter, you are talking over someone. Anything longer, you risk creating an awkward pause. “If you ever want to experience it for yourself, try waiting a full second in between conversational turns. It is extremely painful,” says Michael A. Long, PhD, associate professor of neuroscience and physiology at NYU Langone Health’s Neuroscience Institute.

However, awkward silences rarely happen, and in fact, we have often formulated our response before the other speaker finishes. Given the tight window for switching turns, it’s a wonder that people don’t interrupt each other more often. To find out what enables such precise turn-taking, Dr. Long has spent a long time listening to loud chirps of Alston’s singing mice. These musical creatures, native to Costa Rica’s cloud forest, sing to each other to announce themselves, entice mates, and negotiate territories. Barring occasional blunders, the mice are greatly skilled at waiting for each other to finish before immediately responding with their own song. “They’re quite considerate,” Dr. Long says.

In a new study published in Science, Dr. Long and colleagues found that the mice divide the task load between two brain regions. One area produces the songs unique to each mouse. “It’s what allows them to get up on their hind legs and sing out who they are to the world: ‘I am Ralph the mouse,’” Dr. Long says. Another area, called the orofacial motor cortex, or OMC, is the maestro orchestrating the mice’s conversational duets. When the researchers slowed neural activity by mildly cooling the OMC, the mice produced a longer song. When the OMC was turned off entirely, the animals started to sing over each other.

“This is the first demonstration of the brain mechanisms behind coordinated vocal turn-taking in any mammal,” says Dr. Long, who is now looking for similar mechanisms in humans. What he finds may offer a new way of understanding communication problems in autism or stroke-related language deficits. Precise timing is a central component to speech, and not just in conversations. To produce just one word, the brain needs to orchestrate more than 100 speech-involved muscles on a 20-millisecond timescale.

Before studying social communication in singing mice, Dr. Long’s laboratory had focused on another musical prodigy, the zebra finch. In 2008, Dr. Long and his colleagues showed that the pace of the birds’ songs relies on a small area in their brains; when the area is cooled, the birds sing in slow motion. In a collaboration with neurosurgeons that was published in 2016, Dr. Long found humans, too, have an internal metronome: an area that, when cooled in people undergoing brain surgery, slows speech. “Instead of saying ‘Monday,’ they’d say ‘Mooondaaaay,’” Dr. Long says.

Carrying on a conversation adds another layer of complexity to this sophisticated clockwork. It’s the leap between learning to hit a tennis ball over the net and competing against another player, or learning to play the violin and performing with others in an orchestra. Although we practice conversation every day, “it’s an exceptionally impressive act when we take a step back and see just how much skill is involved,” Dr. Long says. Just like any complex skill, conversational turn-taking is probably something humans hone through practice. “Even these mice can get in sync much better when they have known each other for a while,” Dr. Long says.