Through contact with humans, dogs have developed socio-cognitive skills that allow them to communicate with us. For example, they are able to distinguish different emotions on our faces. A recent study carried out by the University of Bari (Italy) tried to determine whether dogs were more “sensitive” to particular emotions. Are dogs more “receptive” to a face that demonstrates fear, surprise, anger or sadness? Read on to find out.

Research has already demonstrated that canines process human faces in the same way we do: by analyzing facial features (mostly the eyes, nose, and mouth). This is how they are able to recognize familiar faces, distinguish neutral or emotional facial expressions and even differentiate happy or angry faces. Moreover, when they "analyze" human faces, like us, dogs tend to favor information coming from their left visual field. Keep in mind that, while they are symmetrical, the two halves of our face differ in emotional expressiveness. Research has shown that people perceive the left hemiface as displaying more intense emotions than the right. Since the muscles of the left side of our face are largely controlled by the right hemisphere of the brain, this hemisphere plays a dominant role in both emotional expression and processing. This asymmetrical processing of human faces has already been demonstrated in dogs.

To learn more about the dogs’ mechanism for perceiving human faces (and its similarity to our own), the researchers decided to use the six interculturally recognized basic emotions identified by Ekman: fear, anger, happiness, sadness, surprise, and disgust. To do so, two men and two women were photographed expressing each of these emotions plus a neutral expression. Then, 21 domestic dogs of different breeds were recruited (12 males and 9 females, ages 1 to 13 years.) Each animal entered a room containing a bowl of food. Two monitors were placed to the left and right of the bowl (at a distance of 1.9 meters). Once the animal was in the central position, a photograph was shown on the monitors for a duration of 4 seconds. The protocol consisted of two weekly trials during which each dog was shown a maximum of two dyads with different facial emotions.

For their study, the researchers analyzed the orientation of the dog’s head (was it turned to the right, left, or not at all?), taking into account the latency between the photographs appearing on the screen and the moment when the dog turned its head, as well as the time necessary to resume feeding at the bowl. The animal’s overall behavior (26 features were recorded, for example: pricked up ears, folded tail, raised legs, moans, trembling, etc.) was also studied. Finally, cardiac activity was recorded for at least 10 seconds following the presentation of the stimuli or until the dog began feeding again.

The results indicate that the dogs turned their head to the left when the two screens showed faces expressing anger, fear, and happiness. In these cases, they also took longer to resume eating. Their stress levels were higher, and their cardiac activity increased. Conversely, the dogs turned their head to the right when they saw faces expressing surprise. This research supports the existence of an asymmetrical emotional modulation in the canine brain for processing basic human emotions (as is the case for us). However, the results seen for “happy” faces don’t agree with those obtained in previous experiments. Dogs generally process negative emotions (fear, anger) in the right hemisphere, while positive emotions are processed by the left hemisphere. According to the scientists, the absence of auditory stimuli to accompany the "happy" face could trigger a behavioral alarm reaction (activity in the right hemisphere) in the animal (who is only seeing bared teeth and lifted lips). Likewise, relying only on visual information, dogs would interpret the “surprised” face as showing a relaxed expression (processed by the left hemisphere).

Note that no bias was noted for “disgust” (no preference for the right or left). This is in agreement with previous studies where this emotion proved to be far less distinguishable for dogs. The authors of the study put forth the following hypothesis: “in daily life, the same object or situation can provoke different emotional and motivational states in humans and dogs. For example, excrement can provoke an emotional state of disgust in the human owner, while it can be considered attractive for the animal.” An important clarification, indeed!