Let’s say you have a dog. Or maybe you’re watching a friend’s pup. You have a treat in your hand, so you have the dog’s attention. Then you put the treat into a doggy-proof box and let them figure it out. What do you think the dog will do?

If you’re a dog owner, you probably know. Spike or Spot or Ms. Fluffy will probably try in vain for a while. But soon they’ll probably turn back to you with those puppy dog eyes with a look that’s both hurt (why would you make it so hard for them after all?) and pleading.

[CLIP: Dog sounds]

That, it turns out, is a sophisticated cognitive phenomenon called referential communication. It’s an attempt to shift the human’s attention toward the problem. And dogs, as it also turns out, aren’t the only species who can do this.

In 2003 a Hungarian biologist named Ádám Miklósi tested this ability with an experiment. He gathered a handful of dogs and a group of wolves. Both groups of canines had been hand raised from birth by people in much the same way.

He put a piece of food inside of that doggy-proof box. The wolves kept trying and failing to get the food hidden inside. We already know what the dogs did.

Alan McElligott, an animal cognition researcher at the City University of Hong Kong, wanted to push this test further. He wanted to move beyond dogs versus wolves.

“The discussions of those papers, and the discussion in the press, was all about them being domesticated closely by humans.”

He thinks that there’s more to this story. What if animals don’t need to be domesticated but instead simply need to be socialized? Maybe it’s not that dogs are especially social but instead that wolves are just especially stubborn.

“So we came along in 2016 and did the work on goats and showed that, actually, it’s probably a general domestic animal thing. It’s nothing to do with being domesticated as a companion animal.”

In a new study, he’s gone even further. Enter—kangaroos?

“And kangaroos generally haven’t been used for any of that sort of research. I think most people, when I e-mailed them and suggested, ‘Oh, you know, this problem-solving task with kangaroos,’ they thought I was kind of a bit mad or some really eccentric scientist.”

McElligott rounded up a group of marsupials from three different zoos and sanctuaries. In all, he tested three different species: Kangaroo Island kangaroos, eastern grey kangaroos and red kangaroos.

Of 11 animals tested, 10 of them passed, turning their gaze back to McElligott. And of the 10 who turned to look at him, nine kangaroos actively shifted their gaze between the box and the human.The results were published in December in the journal Biology Letters. [Alan G. McElligott et al., Kangaroos display gazing and gaze alternations during an unsolvable problem task]

“This domestication hypothesis for all these behaviors, we’ve over-relied on it. I’m not saying that kangaroos are just as good at dogs or anything like that, but I’m saying the cognitive capabilities of wild animals that are habituated to humans have probably been underestimated.”

McElligott says that it’s not enough to study domesticates and their wild counterparts, like dogs and wolves. That’s why he decided to try kangaroos, which were never domesticated.

“You can’t deduce the effects of domestication on animal behavior, animal cognition, by only studying domesticated animals. It sounds pretty simple, but it’s important. So you need to see what the other ones are capable of doing if they’re used to being around people.”

McElligott says that the pouched marsupials offer up a whole new set of possibilities for understanding the evolution of social cognition. And he’s now turned his attention to yet another unexpected species: the Asian water buffalo.

—Jason G. Goldman

[The above text is a transcript of this podcast.]