Can humans recognize emotional state in pet dogs by looking at their face?

Facial expressions are regularly used by humans to detect the emotional state of other people. In 1862, Darwin, for the first time, argued that all species can express emotions and that humans could find out emotions in other animals via both their facial expression and body posture. Aims of this study were: to investigate whether dog’s emotional state can be identified by humans by only looking at pictures of a dog’s face; and to explore if experience with dogs can affect this ability. To explore humans’ ability to identify dog’s facial expressions, 27 pictures of dog’s face taken from 9 subjects (different breed, sex, age and colour) under three behaviourally defined conditions (food, alone, reunion) were collected. An on-line questionnaire was developed: participants were asked to assess the emotional state (positive/negative/don’t know) for every picture. The proportion of correct responses was calculated. Kruskal-Wallis and Mann–Whitney U tests were used to identify differences among categories of people with different experience with dogs (owners, professionals, no experience). 230 participants (females=159, males=47, NA=24) answered to the questionnaire. Participants assigned 50.1% of pictures to the correct emotional state: positive emotions seemed easier to assess than negative one, 58.1% correct answers and 36.5%, respectively. Furthermore, “reunion’s face” was easier to score than “food’s face”, 62.4% and 55.1%, respectively. Differences were found among categories (Kruskal-Wallis, P=0.019): owners (51.1%) and professionals (52.0%), recognised significantly better the emotional state compared to people with no experience with dogs (46.0%) (Mann–Whitney U, P=0.016 and P=0.009, respectively). No differences were found between owners and professionals (Mann–Whitney U, P=0.548). Our findings showed that our participants were able to identify positive emotional state of dogs, with different morphology, by only looking at their facial expression. In particular, “reunion’s face” seems to be easier to identify compared to “food’s face”. This ability seems to be affected by individual everyday experience with dogs. Negative emotional state appears to be more difficult to recognize; this was probably due to the behavioural situation chosen in the present study, as being left alone is not always linked with an intense negative emotional state. Further studies should be conducted to better address the ability of people to assess negative emotional states using pictures of dogs experiencing a more intense emotions (e.g. fear).