Reading emotions in dog eyes and faces by Japanese observers: A replication and extension study of Burza et al (2022)

Previous studies conducted in the U.S. and Brazil have demonstrated that humans can recognize dog emotions from static images of canine eyes or full faces (Bloom et al., 2021; Burza et al., 2022). The present study aimed to replicate and extend these findings in a Japanese context, investigating whether the ability to interpret dog emotions generalizes across cultures. Japanese participants (N = 342) were shown photographs of three dog breeds previously used by Burza et al. (2022) and asked to presume the dog's emotional state. Results revealed that Japanese participants identified presumed dog emotions at levels significantly above chance, both when viewing full-face images and eye-only images. Accuracy was notably higher for full-face images, supporting the idea that broader facial context enhances accurate identification of emotion. Dog ownership had no effect on accuracy, and participants' ability to recognize human emotions did not predict their performance with dogs. This suggests that interpreting canine emotions may rely on distinct cognitive mechanisms. Whereas the overall findings replicate previous results from Western populations, some cross-cultural differences emerged in which specific emotions were most accurately identified. These differences underscore the cultural factors and perceptual biases in cross-species emotion interpretation. In sum, the study suggests that humans possess a moderate but consistent ability to interpret dog emotions from facial cues, and that this capacity, whereas partially universal, may be shaped by cultural context.