Emotions in dogs: How can we measure them?

Tiziano Travain, Elisa Silvia Colombo, Laura Clara Grandi, Eugenio Heinzl, Annalisa Pelosi, Emanuela Prato Previde, Paola Valsecchi (2016) How good is this food? A study on dogs’ emotional responses to a potentially pleasant event using infrared thermography, Physiology & Behavior, 159, 80-87. 

For centuries, the importance of understanding animal’s emotional states was not recognised. However, with concern for animal welfare increasing, animal’s emotions are becoming a frequently studied subject. Until recently the focus of these studies were primarily on negative emotions. Animal welfare scientists now recognise that welfare is not only about the absence of negative emotions, but also includes the experience of positive emotions. This study investigated if eye temperature in domestic dogs is a reliable indicator of their emotional state. The findings show that, while eye temperature is a good indicator of arousal, it does not necessarily differentiate between a negative and positive emotional state.

Even though it is now widely accepted that non-human animals experience different emotions just like humans do, measuring them remains difficult. Current indicators that are used to measure emotions, such as heart rate and behaviour do not always account for differentiation between positive and negative emotional states. In addition to arousal, valence (the intrinsic positivity or negativity of an event or situation) is the other dimension used to categorise emotions.

To find out if measuring dogs’ eye temperature is a reliable indicator for their emotional state, the researchers used Infrared Thermology (IRT). This parameter was measured alongside more commonly used indicators, namely heart rate (HR), heart rate variability (HRV) and behavioural signals. The team hypothesized that a treat received from their owners should cause arousal in the dogs, visible by an increased HR and variable HRV parameters. An increase in tail wagging would indicate their positive emotional state.

Nineteen healthy pet dogs were part of the experiment, with different breeds, ages and sizes represented. The experiment was divided into three phases; the baseline phase; the feeding phase, during which the owner offered the dog treats over a 10-minute period at approximately 20-second intervals; and the post-feeding phase. To uncover any potential habituation effect during the baseline phase, the researchers compared the first and last five minutes from this part of the experiment. They found a significant decrease in HR values during phase one, and therefore only included the last five minutes in their analyses.

Positive or negative arousal?

The results of the experiment indicate that, while the mean eye temperature of the dogs increased significantly during phase two, this parameter does not always differentiate between experiencing positive and negative emotions. The behavioural signals that were recorded during the experiment indicate that the dogs experienced getting the treat from their owners as a positive event; other than some nose and lip licking in phase one, they did not show any signs of stress and instead showed an increase in tail wagging during the feeding phase. A previous study found that stress (from a veterinary visit) could also cause an increased eye temperature in dogs. During this experiment, dogs did show significantly more stress signals, and looked away from the IRT camera in many instances.

Interestingly, HRV did not change across the different phases. It appears that the autonomic system remained stable during the entire experiment, which could mean that receiving a food treat is not a sufficient stimulant to cause long-term changes in dog’s autonomic system. A reason for this could be that the dogs in the study did not have to do anything to get their treat, or that the pet dogs might receive treats on a regular basis.

Concluding, IRT is an effective way of measuring the level of arousal in domestic dogs, but cannot give definitive indications regarding valence. Further research with laboratory dogs would be a valuable addition to this study. This would eliminate any potential impact that the dog-owner relationship has on the outcome of the experiment.

World Animal Protection’s view

While a considerable number of studies have conducted research into negative emotions experienced by animals, this study is one of the few that aims to find better ways of identifying positive emotional states in animals. Research aiming to identify measure of emotional states in animals is an important area of research, which can help aid animal welfare assesments.