Perceiving Emotion in Babies and Dogs

Do babies and dogs process facial expressions in the same way, and if so what does it tell us about the evolution of emotion?

By Zazie Todd, PhD

Darwin suggested that some human emotional expressions could have their origins in the facial expressions of other animals, including primates and dogs. If so, there would be similarities in the way people process emotional faces across these different species. While most research has focussed on other primates, a paper just published in PLoS One by Annett Schirmer et al (National University of Singapore) investigates whether or not there are similarities in processing facial expressions in human infants and dogs.

The study asked sixty-four participants, half of whom were dog-owners, to take part in a series of tests. The experiment included both implicit and explicit tests of emotional processing. The explicit tests asked people to look at photographs and rate the emotion being expressed. For the implicit tests, people were shown a photograph and then a set of letters on a computer screen; they had to decide if the letters were a word or not. 

The words were either positive or negative. The idea is that even though the photograph is not related to the word task, if people process the emotion then it will affect their ratings. For example, if the photograph is of a positive emotion and then the word HAPPY pops up, it will take people slightly less time to decide this is a word than if the photograph had been of a sad face. This is known as a priming effect.

The researchers looked at positive and negative facial expressions, so they refer to it as affect, because they have not specified which emotion it is.

Photographs of dogs and babies were taken especially for the experiment. Dogs were recruited at a park. For the positive affect photo, they were given food or a toy, depending on what the owner said they would prefer. For the negative affect photo, they were put in a crate for five minutes while the owner went away.

Baby photographs were taken in the lab. In order to capture pictures of positive affect, the babies were given a toy. For photos of negative affect, the parent was asked to leave the room so the baby was alone with the experimenter. In some cases this did not upset the baby, in which case the experimenter hid from them. In all cases, for babies and dogs, the duration of the negative affect was kept short.

The photographs were touched up as necessary, e.g. to account for movement or to remove the crate bars from the image. Full-face photos were used, and cropped images of the eyes were also used for the explicit task.

Images used in the study, showing baseline (neutral), positive and negative affect. Source: PLoS One

The results showed that there was a priming effect for both the human and dog faces. In other words, people were influenced by the positive or negative affect in the photo. While this has been previously shown for human faces, it is the first time it has been shown for affective photos of dogs. It suggests that people automatically process affect in both dogs and humans.

In the explicit processing task, participants were able to recognize positive and negative affect in both humans and dogs. Ownership of a dog did not have an effect on people’s ratings of the dog images.

People responded more accurately to photos of the full face than the cropped image of the eyes, for both humans and dogs. The results also suggest a similarity in the human and canine expression of sadness via the eyes, but not happiness.

There were gender differences in both the explicit and implicit tasks. While both men and women recognized affect in a dog’s face, women were better at it than men, and women dog owners were better than those who did not own dogs. It is no surprise to find gender differences, since other studies have suggested that women are more sensitive to emotional expression than men. The fact that dog ownership makes a difference for women, but not men, suggests women dog owners are more attuned to canine expressions than their male counterparts. There are several possible reasons for this, including differences in the experience of owning a dog (e.g. perhaps women take more responsibility for the dog).

People were more sensitive to human emotional expressions than to canine ones. Dog ownership did not affect this, but experience with dogs did, since some non-dog owners had no experience with dogs at all. This is in line with a study by Michele Wan, which found that although people found it easy to recognize happiness in dogs, experience was important when it came to recognizing fear.

The authors say their results show that “humans recognize positive and negative affect in the facial behavior of dogs and that they can do so without having ever interacted with a dog. Additionally, a number of similarities were revealed between the processing of human and dog expressions.”

This would be expected for non-human primates. Extending it to dogs is a significant step. This supports the idea of an overlap in emotional expressions in primates and canines due to continuity of evolution.

Do you find it easy to recognize positive and negative emotions in your pets?

Reference

Schirmer A, Seow CS, & Penney TB (2013). Humans process dog and human facial affect in similar ways. PloS one, 8 (9) PMID: 24023954  
Wan M, Bolger N, & Champagne FA (2012). Human perception of fear in dogs varies according to experience with dogs. PloS one, 7 (12) PMID: 23284765