Dogs flub problem-solving test

Humans may be a bad influence on their best friends — at least when it comes to problem solving. Dogs that have lived around people lagged behind wolves in a task that wasn’t very tough: tugging a lid off a food container.

Only one of 20 dogs got the lid off a plastic storage box and gobbled the sausage treat inside, reports Monique Udell of Oregon State University in Corvallis. Yet eight of 10 wolves gnawed, pawed and ripped their way into the container.

The social tendencies of dogs may be getting in the way of persistent, independent struggling that would have freed the treat, Udell suggests September 16 in Biology Letters. The dogs in the test (10 pets and 10 shelter dogs with some history of pethood) typically spent 10 to 15 percent of their time gazing at the nearest person and 5 percent or less of the time touching the container. Wolves raised and fed by people but living outdoors, however, barely looked at a nearby person. They typically devoted about 90 percent of the two-minute trial time to grappling with the treat box.  

Even with no person around, the dogs didn’t paw or mouth the box much more than they did in human company, Udell found. When a person hovered over the dogs and actively encouraged them to keep trying to open the box, the dogs did spend more time engaged with the problem — and a few more managed to open the box — but they still did not match the wolves.

One way of looking at the results suggests that “wolves are practical problem solvers and dogs are social problem solvers,” says Clive Wynne of Arizona State University in Tempe, who worked with Udell on earlier dog research but not on this project. But Wynne suspects more research could support an alternative explanation: “We teach our dogs to be stupid.” He suggests early and abundant exposure to human outrage at a sandwich stolen off a plate, or a kitchen cabinet ravaged, may have taught pets to proceed cautiously in helping themselves. In contrast, wolves in the test were raised by people but live in an outdoor enclosure.

A tidbit of preliminary data in the paper suggests a way to test the idea that dogs learn to be inhibited. An 8-week-old puppy also managed to open the box. If young puppies with less training in the etiquette of sharing human houses do better than adult dogs at ripping open food boxes, then learned caution may be a factor.

Udell’stest grew out of an urge to reverse an experimental scenario in a line of experiments going back to a 2003 paper from Ádám Miklósi of Eötvös Loránd University in Budapest and his colleagues. As part of a surge of interest in how dogs differ from wolves, the researchers had trained nine dogs and nine human-raised wolves to open a box and score a treat. When presented with a box that couldn’t be opened as expected —because its lid was glued on —dogs were more likely than wolves to turn their gaze toward a human face.

That gazing has been interpreted as a clever move, something Udell’s study questions. (Miklósi points out that he was more cautious in reporting the results in his paper.) His larger idea, though, is that “dogs were selected to some extent for being able to cooperate with humans.” And this deep history helps explain dog sensitivity to human social cues, a sensitivity that encourages people to tolerate canine company. “Otherwise,” he says, “we would just live with wolves.”