If you were provided with photos of a random set of people, as well as separate photos of those people’s dogs, chances are, you could match up the right person with their pet.
How? Scientists think it has something to do with the eyes — they’re just not sure what exactly that something is.
Sadahiko Nakajima, a psychologist at Japan’s Kwansei Gakuin University, has been trying to figure out how people are able to match pets with their owners at a rate greater than chance could account for.
He considered the obvious: Maybe men are simply more likely to own large breed dogs. Perhaps women with long hair are more likely to have dogs with long, floppy ears.
However, even when experiments are designed to exclude these superficial characteristics, people are still able to match strangers with the correct pets.
In a 2009 study, Nakajima found that participants could not only match photos of pet owners with their dogs by facial appearance alone, but they could also recognize which randomly paired dog-and-owner photos were fake.
This clued Nakajima in that something about facial appearances was giving participants hints at which dog belonged to which person, so he designed an experiment to determine what part of the face was responsible.
He created test sheets that each included 20 sets of dog-human pairs. One sheet featured actual dog-human pairs while the other had randomly matched pairs.
The photos were basic headshots cropped at the shoulders, and the photos sets included an equal number of male and female owners, as well as a variety of dog breeds.
The 502 Japanese undergraduate students who participated in the study were randomly assigned to one of five different groups. Each group received test sheets that had been “masked” in a different way.
In one, the humans’ eyes were covered, and in another, the dogs’ eyes were. One also showed only both the dogs’ and humans’ eyes, while another covered their mouths.
The control group received photos that showed the full faces of the humans and the dogs.
Those who saw the unobstructed faces were quite good at determining which human-dog pairs were fake, and 80 percent were able to select the real-life human-dog pairs.
The group that received the photos with concealed mouths were also fairly good with 73 percent making the correct matches.
However, when the humans’ eyes or dogs’ eyes were covered, participants’ ability to select the right human-dog pairs fell to statistically chance levels.
But those in the eye-only group proved to be almost as good as selecting the real-life human-dog pairs as those who viewed the unobstructed faces. Seventy-four percent chose the correct pairs by looking at the eyes alone.
This finding was so surprising that Nakajima duplicated the study with an entirely new batch of participants and this time, 76 percent picked the right pairs.
As Nakajima notes in the study, because all the human models were Asian, they all had similar dark-colored eyes, so there must be something else — some shared look — that clues people in to what dog belongs to which person.
He’s not sure what it is about the eyes that communicates a shared bond, but this isn’t the first study that’s found people are able to extract psychological cues from the eyes.
In fact, a 2009 Tufts University study found that participants were able to discern a stranger’s sexual orientation simply from looking at photos of people’s eyes.
Related on MNN:
Photo: Sadahiko Nakajima, Kwansei Gakuin University