Dog owners are often accused of making connections that don’t exist. “He just knows when I’m sad, somehow.” “She always wakes me up to go out before the alarm even goes off. She can tell time. She’s so smart!” But when it comes to the claim that dogs physically resemble their owners… there is actually science to back it up!
So, what turns owner/dog into a pair of doggelgängers? (Sorry, had to.) That’s exactly what Sadahiko Nakajima, a psychologist from Kwansei Gakuin University in Japan, set out to find. In his previous research, Nakajima found that people were surprisingly adept at matching pictures of owners with the correct dog. What’s more, they could also sniff out the pairs of dogs/owners that researchers had randomly put together as fakes. It was impressive that people could correctly match pairs based on facial features alone, but Nakajima was interested in finding which particular facial features people were using to make their extraordinarily accurate assessments. Here’s how he did it…
Step One: round up a whole lotta dog owners and their pups at a dog enthusiast festival, take portraits of the pairs using techniques to remove extraneous influencing factors. The photos were all basic color portraits shot against a white background, cropped at the shoulders. The humans were asked to look straight at the camera and smile slightly (somehow, the researchers seem to have gotten the dogs to comply with the same instructions!) The resulting photos included 40 owners (an equal number of males and females) and 40 dogs of varying breeds that were randomly assigned to two test sheets. One included 20 real pairs of dogs/owners and the other showcased randomly matched pairs.
Step Two: enlist 502 Japanese undergrads, present them with the two sheets, ask them to “Choose the set of dog-owner pairs that physically resemble each other. Set A or Set B.”
Step Three: make things a little harder. Study participants were randomly assigned sheets with one of five “masking conditions”: no-mask (completely unobstructed faces of both humans and dogs); eye-mask (the humans’ eyes were covered by black rectangular bars); mouth-mask (the humans’ mouths were covered by the black bar); dog-eye-mask (the dogs’ eyes are covered the black bar); or eye-only (just thin rectangular slices of the eye area for both humans and dogs are visible to the participants). Check it out:
The Results: Photos with a “no-mask” condition (full face was visible) resulted in a whopping 80% of participants correctly identifying the real dog/owner pairs.
When those same photos were viewed with the owner’s mouths covered (“mouth-maks”), the percentage of correct matches dropped only slightly to 73%.
However, if either the humans’ or dogs’ eyes were covered in the photo (“eye-mask” and “dog-eye-mask”) participants’ rate of correct pairing dropped dramatically to statistically chance levels.
Hang on, things are about to get pawsitively fascinating! Remember the “eye-only” test (just thin rectangular slices of the eye area for both humans and dogs)? Even with that little bit of information, 74% of participants still correctly chose the real-life dog/owner pairs!
Certainly makes you think about the phrase “puppy dog eyes” differently. People have always said that eyes are the window to the soul, but it seems they are also a window into that unspoken connection between dog and owner that lets a pup “just know when I’m sad, somehow.” 🙂
Want to know more about the study? Head over to Slate for the full scoop!