Is it possible to learn about what dogs see?
Dogs are wonderful human companions. They are the first species that humans domesticated and they have a long history of close relationships with people.
If you spend any time with dogs (I have two wonderful adopted dogs who shower me with affection and fur), then you see first-hand the many ways that dogs can respond to their owners. A dog may treat a stranger with suspicion, but dances with joy when its owner returns. Indeed, when one of our grown children returns after a long absence, the dogs respond with excitement.
It is easy to give complex interpretations to your pet’s behavior. But what does your pet really know? It turns out that there is a lot of work on dog cognition (or as I like to call it, dognition). Some of this work was summarized in a special issue of the journal Current Directions in Psychological Science in 2016.
One interesting paper in this issue by Ludwig Huber looked at what dogs see when they look at people.
It is hard to answer these kinds of questions, because dogs cannot tell you what they know. Instead, you have to find experimental procedures that allow you to figure out what dogs must have seen based on their behavior.
As this paper points out, the visual system of dogs evolved to function well in low light. That is one reason why dogs have more limited color vision than humans. Dark places make it hard to discriminate colors, so the ancestors of dogs de-emphasized color vision in favor of being able to see in low light and to perceive the motion of objects.
One interesting question is whether dogs associate what a familiar person looks like with what they sound like. To explore this question, researchers played the voice of the dog’s owner and then showed either a picture of the owner or a picture of an unfamiliar person. Dogs looked longer at the picture if it was the picture of an unfamiliar person than of the owner. Control conditions ruled out the explanation that dogs just look longer at unfamiliar people in general. This result suggests that hearing a familiar voice paired with an unfamiliar face violated the dog’s expectations and so the dog looked longer in this situation.
Dogs seem to pay particular attention to the eyes in a face. One way to test what dogs pay attention to is to use discrimination tasks in which dogs see pictures of pairs of faces on a touch screen. They are trained to press the touch screen to get a reward. Then, they are only rewarded when they touch the proper face. If the dog can discriminate between the faces, then it can learn to get the reward reliably on each trial of the study.
Dogs differ in their ability in this task overall, but most learn to discriminate between pictures of different faces. They are best when they see a whole face. Dogs (like people) are good at using information about the configuration of the eyes, nose, and mouth. If dogs are shown only the eyes, the nose, or the mouth, they are best at learning to discriminate between faces when shown the eyes rather than the nose or mouth. This suggests that dogs are paying most attention to the eyes. That makes sense, of course, dogs are good at following people’s gaze to look at what they are looking at.
Finally, dogs seem to be able to recognize differences between at least some human facial expressions for emotions (though it is not always obvious what they know about these facial expressions).
It turns out to be hard to study what dogs learn about facial expressions, because there is a lot of information that dogs could be paying attention to if you just ask them to discriminate between pictures of the same person expressing different emotions. Most people show their teeth when smiling, but not when they are angry. So, if you just show the faces, it is possible that dogs are paying attention to the presence of teeth. (This is not a crazy idea, because dogs tend to bare their teeth at each other when angry, so teeth might be something important for dogs to pay attention to.)
But, one clever experiment trained dogs to discriminate between a happy and an angry face using only the lower half of the face (nose, mouth and chin). After learning to reliably discriminate these faces, the dogs were then shown the top half of the same face (the eyes and eyebrows) as well as the bottom and top half of new faces they were not trained on. The dogs were able to discriminate between the happy and angry faces for all three of these types of faces. That is, when trained on a happy mouth, dogs could find happy mouths on other faces, but also knew what kinds of eyes and eyebrows went with a happy face rather than an angry face.
These studies were just focused on the visual abilities of dogs. This doesn’t tell us anything about how dogs react to these expressions or what faces or expressions mean to dogs, but it does give us information about what they are able to see. Over the next few blog entries, I will talk about other interesting things that psychologists have learned about dog behavior.
Huber, L. (2016). How dogs perceive and understand us. Current directions in psychological science, 25(5), 339-344.
Get the help you need from a therapist near you–a FREE service from Psychology Today. City or Zip
Dogs notice when you’re sad, mad, or suspicious. They can even detect cancer. Turns out, dogs know more about your emotions and health than you ever suspected.
Why shouldn’t you look a dog in the eyes?
In their wolf ancestors, staring is considered threatening and rude. Some dogs still retain that attitude. That’s why you should never stare down strange dogs or hold dogs still to stare into their eyes. If a dog gives you a hard stare, with unblinking eyes and a stiff posture, back away and don’t make eye contact.