Canebrake wren
© Karla Rivera-CaceresCanebrake wrens are known for their highly coordinated, operatic duets. This adult male wren was photographed mid-song
On a dewy summer morning, Karla Rivera-Cáceres, an ornithology researcher at the University of Miami, crouched in her usual workspace -- the tall grasses of Costa Rica's woodland -- and heard something unusual.

Rivera-Cáceres studies bird song, and that day she was listening to the canebrake wren, a brown bird whose bland appearance (it was once named the "plain wren") belies an unusual and extremely complex call.

Canebrake wrens are songbirds, the subset of species whose calls develop beyond the standard tweet or chirp into full-fledged ballads -- and within that group they are part of a somewhat exclusive club: duetting birds.

When two of these wrens communicate, they weave their songs into an elaborate, Sonny and Cher-style duet. They warble back and forth, literally finishing each other's phrases, with such high coordination that, to an outsider, they sound like a single voice.

But as Rivera-Cáceres sat listening that morning in 2011, she noticed something odd about this pair's effort: their duet was really bad.

The birds were sloppy. They chirped over each other. They sang the wrong responses and screwed up the timing. They were young, still inexperienced at singing, and it showed. As birdsong goes, their act was like a five-year-old belting opera. In the world of bird science, her observation proved huge.

Namely: it demonstrated that birdsong works even more like language than we think -- that in order to achieve their Pavarotti-esque exchanges, wrens need to learn a specific set of social rules which are similar to what humans might call manners.

Rivera-Cáceres' discovery launched her on a years-long research project at the University of Miami. This year, her ​​findings​ were ​published​ in the journal Royal Society Open Science, and are already altering the way scientists conceive of birdsong.

"People think humans are so unique," Rivera-Cáceres said, "But these birds are talking to each other, developing intimate codes. We're not the only ones who communicate or hold complex thoughts."

Humans have been fascinated by the music of birds for centuries. Expert bird watchers learn to identify species they might not even be able to see simply from their distinctive calls. Here, for example, is the call of the mangrove cuckoo, a small bird found mainly in coastal South Florida forests. For scientists, calls and songs can tell different stories. Most recently, according to a recent piece in the New York Times, researchers are increasingly finding that some birds are forced to alter songs to adapt to the surrounding din of humanity.

Comparisons between birdsong and human language date back to Darwin, Rivera-Cáceres said ("The sounds uttered by birds offer in several aspects the nearest analogy to language," he wrote in The Descent of Man), and scientists have long studied how baby birds learn the tunes of their songs --just like infants learns new words. One recent study found that genes in song birds might shed light on human speech disorders.

But Rivera-Cáceres' focus was more like an avian Emily Post - she was interested in how those birds learned to use those songs and how their interactions reflected a set of shared social guidelines, much like the subtle norms that govern human conversation with friends, family, and romantic partners.

If you aren't a bird expert, the wrens' song (loud, high-pitched and alternating between notes sort of like a see-saw) might sound like something off the Rainforest setting of a white noise machine. But to the trained ear, their calls are highly specific and follow a strict set of rules.

"The rules are called 'duet codes,'" said Rivera-Cáceres. "Every time a male sings one song type, the female will answer with a certain other song type." People do this too, Rivera-Cáceres said: when someone asks about weather, they expect an on-topic answer, not a monologue on television or cooking.

The wrens' tunes are also carefully timed to avoid pauses or interruption. "If they overlap, the birds stop singing," said Rivera-Cáceres. Aversion to pauses and interruptions is also a convention of human conversation. That's why, for example, people prefer not to speak over each other, or why delays between TV hosts and their correspondents can seem so awkward.

Human manners, as anyone with children will know, are learned with practice - often after many reminders not to interrupt. But when Rivera-Cáceres first proposed her project, most researchers believed wrens were born with instinctive codes and rhythms, the way a chimp automatically knows how to grasp or grunt.

Few had challenged this theory, in part because duetting wrens are so difficult to study, with their young hard to find in the wild. Even Rivera-Cáceres' advisor, Professor William Searcy at the University of Miami, doubted she could break new ground.

"He said it would be very difficult. It's hard to study a system that few-to-no people have studied." Rivera-Cáceres said. "But when I showed him these duets where the juveniles would do a terrible job at dueting with adults, he was excited."

"I was excited because nobody had demonstrated that juveniles were worse at duetting than adults," Searcy said. "Nobody had shown that they improve over time."

Rivera-Cáceres' discovery demonstrated that wren-world-etiquette was more like our own: that it was learned by trial-and-error.

It's not clear if Rivera-Cáceres' findings are necessarily applicable to other creatures, according to her longtime collaborator Christopher Templeton, an assistant professor at Pacific University in Oregon -- but they do provide food for thought.

"It's difficult to extrapolate to other species," he said, "but I suspect that the patterns that we're finding here are broadly applicable to other animals."

Rivera-Cáceres' first breakthrough opened the door for more findings.

The canebrake wrens use their elaborate duets to defend territory and attract mates. Before Rivera-Cáceres' research, a 1990's paper had argued that, unlike humans, wrens' repertoire of songs was fixed from an early age. In order to mate, in other words, they had to find another bird who sang just like them.

Rivera-Cáceres found otherwise. Mates frequently sing songs that were not used by their parents. In fact, it can be hard to find two pairs of birds, even within the same population, that follow the same exact set of rules.

She suspected that wrens developed their songbooks with their companions--just like friends form inside jokes, or couples create their own modes of communication.

To prove it, Rivera-Cáceres captured several wrens in Costa Rica, and separated them from their mate, facilitating a kind of avian divorce. It wasn't as cruel as it sounds, she said. "They keep singing--like, where are you, where have you gone? But as soon as someone else arrives, they're fine. They mate with the new bird".

She set the birds up with new partners, matching them with wrens from other areas. Then, she watched the birds as they got to know each other; she heard them determine their duets.

In these new couples, Rivera-Cáceres found that the wrens were like kids again: bad at timing, bad at responses. Some chirped when they should have trilled. Others interrupted and overlapped. They underwent--like many new couples--some problems with communication.

"Pairs that were together for less time made mistakes more often," Rivera- Cáceres said. "Instead of always answering with a certain song type, they would mess up and sing something else."

But over time, that changed. After a while, the new pairs mastered their rhythms and responses. What's more, they sang duet codes that they had not used with their previous mates. With their second partner, in other words, the wrens had developed a whole new way of relating.

"As they were together for a longer period of time, they figured it out," she said.

According to Searcy, Rivera- Cáceres' advisor, her work has paved the way for more insights.

"Karla has opened up a new area of investigation in song development -- and that is the development of interaction rules," Searcy said. "I hope this becomes an active area of research in ornithology. "