As research participants, dogs have a powerful advantage over other highly intelligent species, like parrots or chimpanzees, for the scientists studying them. “They’re ubiquitous,” says Brian Hare, Ph.D. ’04, a professor of psychology, neuroscience, and evolutionary anthropology at Duke University. Urged on by his Harvard thesis adviser, Moore professor of biological anthropology emeritus Richard Wrangham, Hare began his academic career studying primates, but he homed in on cognition in canines during graduate school. In the two decades since, he’s had the covetable task of observing and training generations of puppies as they grow up.
“I have the largest sample of animals on campus by far,” says Hare, who heads up Duke’s Puppy Kindergarten, a longitudinal, training-based study of puppy cognition dating back to 2018 with backing from the National Institutes of Health. “People bring their dogs in just like people bring kids, and it’s literally thousands and thousands of testable animals.”
The Hare laboratory’s latest research, now a preprint for the journal Animal Behavior, focuses on puppies bred to become service dogs—and the stages at which they develop certain abilities and intelligence markers that suggest whether they will be successful at such work.
The researchers tracked 101 puppies from two service dog organizations—Canine Companions and Eyes, Ears, Nose and Paws—during the period of rapid brain development that occurs between 8 weeks and 20 weeks of age.
Every two weeks, a team led by co-author Hannah Salomons tested skills including memory, problem-solving, impulse control, and social communication. In one test, puppies learned to nudge a loose lid off a Tupperware container to access a treat. Trainers then sealed the lid and waited for the puppy to look to them for help—a social communication skill.
Of the 10 abilities tracked in the study, nine typically emerged by the time the dogs were 16 weeks old. One of the earliest, the ability to understand human gestures such as pointing, appeared at around eight weeks.
“Some of what we were able to measure emerges rapidly,” Hare says, “and [the puppies] reach adult performance levels within weeks. With other skills, it’s quite gradual, and they’re improving a little bit every other week.” Bladder control, for example, usually shows up at around 14 weeks, while eye contact could emerge somewhere between the 10th and 11th week.
When he was a graduate student at Harvard, Hare studied the evolutionary differences between dogs and wolves, focusing on the ways in which domestication changes cognitive development. Effective service dogs, he notes, have certain traits that diverge significantly from their wild ancestors.
“One of the things dogs inherited from wolves, and what makes them great guards,” he says, “is they’re very xenophobic. They’re fearful of strangers—it’s why, when you have a dog at home, they bark when somebody comes to the door, or they can be aggressive out on a walk.”
But service dogs, including those from Canine Companions, have been intentionally bred for the opposite trait. “They aren’t just neutral to strangers, they’re ‘xenophilic’; they actually prefer strange people to people they know and live with,” he says. “That’s a big, big change. These dogs are crazy attracted to new people, and that’s what makes them great service dogs.”
In the Animal Behavior study, a dog’s individual aptitudes and skills generally remained consistent into adulthood—a crucial finding that suggests those dogs that have the smarts and temperament to handle the rigors of service work could be identified earlier.
“That could really be a game changer in terms of being able to have more dogs available to help more people,” Hare says, “because we could know early which dogs would be worth investing in, and which dogs might not have the capacity for it.”
Hare’s family dog, Neutron, didn’t have the capacity for it. Despite being bred at Canine Companions as a service animal (the organization has its own breed type, a mix between a Labrador and a Golden Retriever), Neutron proved too scared of the vet’s office to hack it as a working dog. He’s perfectly happy helping around the lab, but more than a year of time, money, and effort went into his training before he flamed out. Hare hopes studies like these could improve the selection process and make service dog training more efficient and less costly.
“[What’s] important for training seems to emerge in this very narrow window,” Hare says. “And as long as they have socialization and a rich, warm family life, that maturation process will happen.”
