Speak, Boy, Speak!

Dog owners know their pets are vocal creatures. Rover doesn’t just bark—he howls, yips, and growls. The same holds true for coyotes and wolves in the wild. Man’s best friends (and their cousins) are definitely out there communicating, but the human ear isn’t so good at picking up the subtleties of what they’re saying. Despite this aural shortcoming, untrained people from across the globe are tuning into canine communiqué, hoping to make heads and tails of it.

For the Canid Howl Project, scientists are recording members of the dog family all over the world: jackals in India, hyenas in Kenya, coyotes in North America, wolves from just about everywhere. On the domestic dog front, owners are submitting the sounds of their loyal pets to the database (donate your hound’s howl here). Over the last year, hundreds of volunteers have been analyzing thousands of recordings. What they find could provide insight into the pooch parked on the couch—and eventually help guide conservation efforts for wolves.

Here’s how it works: Volunteers log onto the project website. They listen to a call and trace the precise acoustic features of the howl on a spectrogram (basically a picture of the sound). “Nothing works quite as well as having a human look at a spectrogram,” says Arik Kershenbaum, an animal-vocalization expert at the University of Cambridge who helped launch the project. “Computers still aren’t as good at pattern recognition as the human brain.”

Running the project is Jessica Owens, an assistant professor at Union College in Kentucky (who has three rescue pit bulls at home). Owens has a deep understanding of what it takes to scrutinize calls. “For my dissertation, I spent 2,000 hours analyzing bird calls,” she says of her work on titmice. “It was painful.”

Therein lies the beauty of having the public pitch in. “One person can do about 100 calls in 30 minutes,” says Owens.

To increase the accuracy, multiple people view each spectrogram. For a presentation at the conference of the International Society of Behavioral Ecology last year, Owens investigated whether volunteers were coming up with different results for the same spectrograms. They weren’t. “Whether it was just a person online or a student from academia”—she offers her students extra credit to take part—“whether you had a scored 1,000 wolf calls or 10, put in 10 hours or 10 minutes, it didn’t matter,” she says. “That’s really cool. It means anyone from the public can go in and help.” In the future, Owens plans to further look into how the public’s hearing compares to the experts’.

In the meantime, the project is recruiting more volunteers. A new partnership with the International Wolf Center should help; the nonprofit is encouraging its nearly 70,000 Facebook followers to log onto Canid Howl. “I’m really glad to see these folks doing this project,” says David Mech, IWC’s founder and a wolf researcher for more than 50 years. “It takes a lot to gather all these howls, a lot of work to compare them, and it will give us more insights into all these animals.”

The scientists hope to learn about how canine communication evolved and the purpose of particular howls—for instance, to warn invaders off their territory or to get pack members ready for a hunt. Mech imagines conservation applications, too—assuming the effort is successful and researchers can interpret the calls of different species.

Kershenbaum, who is currently working on a project to remotely record wolves in Yellowstone, says one possibility for the research would be to learn how to keep wolves off ranches. Previous attempts to keep the carnivores away from cattle by blasting howls have only proven somewhat effective. “What if they’re playing cohesion calls, and they’re saying, Come here, come here?” he says. Parsing the howls to select only territorial ones might be more effective.

The red wolf, one of the world’s most highly endangered canids, could also benefit from the research. Living in southeastern United States, this wolf has a habit of hybridizing with coyotes, which threatens to the species’ genetic integrity. Usually these canine cousins wouldn’t get it on—social structures and territories prevent it—but with red wolves’ limited dating pool and an abundance of coyotes, the two do end up interbreeding. According to Kershenbaum, if the Canid Howl project could identify the howls of those hybrids, researchers would at least have a noninvasive way to measure the problem.

And finally, what every dog person longs to know: Fido’s thoughts. Owens, who is also interested in learning more about how domestic canines communicate, mentions this YouTube video of people acting like dogs at a park. They run around, sniffing, peeing, and saying “hey, hey, hey” to each other. “Obviously, their communication is more sophisticated than that,” Owens assures me. “But are we ever going to know what every single bark means? Not likely.”