Dogs recognize our emotions, and they don’t like it when they see angry

Please welcome today's guest contributor, Natalia de Souza Albuquerque, a PhD Student at the University of São Paulo, Brazil and the University of Lincoln, UK.


Natalia's wonderful dog, Polly.
Hello Dog Believers!

The title of this post might seem obvious to dog owners, but it turns out there’s a lot more to the emotional world of dogs than most people expected. That’s what I want to share here today.

You will probably agree with me that the relationship between dogs and people is quite unique. In fact, dogs seem especially connected to human beings, in a way that no other two animal species are. And the secret of this fascinating relationship may rely on a very important ability: to read and respond to our emotions. 

Emotions are a very interesting (and complex!) research topic. They encompass the mechanisms we have to assess our physical and social surroundings, and they are also linked to how we perceive and respond to different stimuli. 

Recent studies have shown dogs are very sensitive to our emotional expressions: they can discriminate between happy, neutral and angry faces. They look at facial expressions in different ways depending on the content of the image, and they can link together different parts of a face that are expressing the same emotion (e.g. happy mouth with happy eyes).


Happy.

But do dogs actually recognise the information conveyed in certain facial expressions or vocalisations? Do dogs understand that angry facial expressions mean ‘angry’ and respond to them accordingly? Aiming to answer these questions, we (Dr. Briseida Resende from the University of Sao Paulo, Brazil, and Prof. Daniel Mills, Dr. Kun Guo, and Dr. Anna Wilkinson from the University of Lincoln, UK and I) decided to run a broad, non-invasive study.

We showed domestic dogs pairs of facial expressions on a screen: one angry and one happy (images were of the same individual, which could be a dog or a human—either female or male). At the same time they saw the facial expressions, dogs also heard a happy, angry, or neutral sound. In this type of set up, if an individual recognises the emotional content of the faces and voices, they will look longer towards the positive face when listening to the positive sound and look longer towards the negative face when listening to the negative sound. Essentially, recognition is indicated by "matching" what they see with what they hear.


Examples of stimuli used in the study: faces (human happy vs angry, dog playful vs aggressive) & correspondent vocalizations.

The first step was to do a thorough analysis of the looking behaviour of each dog. What we found was fascinating! Dogs were really good at linking sound and image of the same emotion, regardless of species (dog or human), gender (female or male), content (positive or negative), or side of presentation (on the left or on the right-side of the screen). This means that dogs have a cognitive representation of positive and negative emotions. 


Dog looking at screen and hearing auditory stimuli.
To clarify what happened, picture this: let’s say you are in a room all by yourself, and you hear someone laughing outside. What would you expect to see? Someone happy or someone angry? Happy, right? This is because we have stored in our memory several features of a “happy emotion” (visual, auditory, etc.) and we use this in our day-to-day lives. The ability to recognise emotions of one’s own species had previously only been shown in humans and other primates, and the ability to recognise emotions of another species was thought to be unique to humans… until dogs showed us they are way more complex than we imagined!

The second step was to undertake a detailed examination of dogs’ mouth-licking behaviour (the behaviour to lick around one’s own “mouth area”). We were particularly interested whether dogs in the study mouth-licked when they saw the different facial expressions and heard the different sounds. Although there is a quite extensive body of literature that uses mouth-licking as a stress response in dogs, no study had systematically investigated its association with the actual perception of negative emotions in dogs. And what we found was that this display has a lot more to tell us than we thought.


The behavior of interest.
Mouth-licking in dogs is more than the expression of a desire to be fed or a simple response to uncertainties and general discomfort. In fact, the occurrence of this display was dependent on (a) the emotion: dogs licked more often when they saw negative faces; (b) the sensory modality: dogs mouth-licked more often when seeing negative emotions, but not when hearing negative emotions and (c) the species of the stimulus: dogs licked more often when they saw angry humans in comparison with angry dogs. 

In other words, dogs seem to have perceived our angry faces as unpleasant, which changed their own emotional state and triggered mouth-licking. Since we found that dogs responded to angry human faces especially, and that only the visual cues influenced the occurrence of the display, we believe that mouth-licking in dogs may be a cue that signals a dog’s perception of negative information. These abilities may have been selected for (probably unintentionally) during domestication, as they facilitate dog-human communication. Want to see the study in action? Here is a short video clip:

 
Dogs are multi-faceted animals and they possess very complex cognitive abilities. Our research findings lead to the idea that dogs are not only able to recognise and respond to emotions of humans, but also may be capable of understanding them at some level. 


Natalia de Souza Albuquerque
~ Stay in touch with Natalia on Twitter, Facebook, & ResearchGate

References
Albuquerque N., Guo K., Wilkinson A., Savalli C., Otta E., Mills D. (2016). Dogs recognize dog and humans emotions. Biology Letters, 12.

Albuquerque N., Guo K., Wilkinson A., Resende B., Mills D.S. (2017). Mouth-licking by dogs as a response to emotional stimuli. Behavioural Processes.


All images copyright Albuquerque.