Can Dogs Recognize Emotions Just by Looking at a Human Face?

Source: Clipart.com

When you and I look at a photograph of a human face, it is easy for us to recognize the emotional state of the person shown in it just by their expression. Can dogs do the same? I first became interested in this problem several years ago when I visited the laboratory of a colleague who was a well-known researcher in the area of human emotions. He was preparing a display and had prepared several oversized photographs of one person displaying a variety of emotional expressions. He had them resting on the floor, propped up against the wall so that he could select the ones that he wanted and mount them on a larger board. I had my young Beagle, Darby, with me that day, and in the way that dogs do, Darby walked along the line of pictures sniffing and glancing at each. That was until he reached the one which had an angry expression, at which point he stopped, looked at the photo, and then backed away and did not approach it again. I wondered as to whether he was reading the emotion in that angry expression and avoiding the picture for that reason. However, at that time there was no data I knew of that showed that dogs had that degree of emotional sensitivity to the human face. Fortunately, science progresses and now there is some data that confirms that dogs, like humans, can discriminate emotions from photographs of faces.

There is no doubt that dogs watch humans, read their emotional reactions, and modify their behaviors on the basis of what they see (see here, or here, or here for examples). There is even evidence that the brain mechanisms used to interpret emotions are similar in dogs and humans (see here). However, in most of these studies researchers used a lot more than simply the photograph of a human face. Not only did the people in these studies make facial expressions, but they also used their voices, and often their body language, to convey positive or negative emotions to the dog. To pick up emotions simply from a static, unmoving, photographic representation requires a fairly subtle level of interpretation.

A new study that has been accepted for publication in the journal Current Biology*, explores this issue. The team of researchers was headed by Corsin Müller from the Messerli Research Institute at the University of Veterinary Medicine in Vienna. The team trained a group of 11 dogs to distinguish between images of the same person making either a happy or an angry face. The first thing to note is that the dogs did not get to see the full face. Each dog was shown only the upper half or the lower half of the person's face. For each dog, one expression, happy or angry, was chosen to be the correct one, and the dog had to indicate that face by touching a picture of it shown on a computer display screen with their noses. Tapping the correct face gave the dog a food reward.

Source: Anjuli Barber, Messerli Research Institute

Now some of you are probably thinking that the fact that dogs could learn this discrimination does not mean that dogs automatically read and interpret the emotional expressions of people, but rather that they can be taught this skill. However, the test section of the experiment goes much further. The most interesting finding comes after the dogs had proved that they could distinguish happy versus angry seeing only the upper or lower half of the face. When they were next shown either the other half of the face used in the training stage, or the other halves of people's faces not used in training, the dogs could still discriminate between the happy and angry expressions.

Müller summarizes his conclusions by saying “Our results suggest that the successful dogs realized that a smiling mouth means the same thing as smiling eyes." Furthermore, he suggests that the dog's ability to do this is based on their prior knowledge of the meaning of human emotional expressions when he says, “We think the dogs in our study could have solved the task only by applying their knowledge of emotional expressions in humans to the unfamiliar pictures we presented to them."

Ludwig Huber, a senior author on the team, added some additional interpretation, "Our study demonstrates that dogs can distinguish angry and happy expressions in humans. They can tell that these two expressions have different meanings, and they can do this not only for people they know well, but even for faces and they have never seen before." What exactly those different meanings are for the dogs is not directly addressed by the study however Huber suggests that "it appears likely to us that the dogs associate a smiling face with a positive meaning and an angry facial expression with a negative meaning."

What is particularly interesting to me in reviewing this new set of data is that the researchers report that the dogs were much slower to learn to associate an angry face with a reward during the training portion of the study. This certainly seems to suggest that the dogs already had an idea, based on prior experience, that it's best to stay away from people when they look angry. Therefore, asking them to approach and touch the angry face in order to get her reward conflicted with their natural avoidance tendencies. This brings me back to my young Beagle who looked at the photograph of an angry face that was propped up on the floor, seemed to read its expression, and decided that that was one photo he really didn't want to be near.

Stanley Coren is the author of many books including: The Wisdom of Dogs; Do Dogs Dream? Born to Bark; The Modern Dog; Why Do Dogs Have Wet Noses? The Pawprints of History; How Dogs Think; How To Speak Dog; Why We Love the Dogs We Do; What Do Dogs Know? The Intelligence of Dogs; Why Does My Dog Act That Way? Understanding Dogs for Dummies; Sleep Thieves; The Left-hander Syndrome

Copyright SC Psychological Enterprises Ltd. May not be reprinted or reposted without permission

* Data From: Corsin A. Müller, Kira Schmitt, Anjuli L.A. Barber, Ludwig Huber, (2014). Dogs Can Discriminate Emotional Expressions of Human Faces. Current Biology, DOI: http://dx.doi.org/10.1016/j.cub.2014.12.055