Hitomi Chijiiwa and Yasuhiro Kanakogi
Animal Behaviour, volume 240 (October 2026), article 123710. First published online September 9, 2026. DOI: 10.1016/j.anbehav.2026.123710.
Original abstract
When we hear a person's voice, we can often infer attributes such as age or identity, even without seeing the speaker. This ability suggests the presence of a cross-modal representation that integrates information from both the visual and auditory modalities. Such representations are not limited to interactions within the same species but also occur between different species. Previous studies have shown that dogs, Canis familiaris, can match human voices with faces based on familiarity, sex and emotional expression. However, it remains unclear whether they can distinguish humans by age across modalities. Such ability could be important for dogs to adapt their behaviour appropriately towards children versus adults in everyday interactions. This study investigated whether dogs can form cross-modal representations of human adults and children using an expectancy-violation paradigm. We hypothesized that if dogs possess such representations, they would exhibit longer looking times during voice–face mismatched (incongruent) trials compared to voice–face matched (congruent) trials for both adult and child stimuli. The experiment involved seating dogs in front of a monitor. Immediately after hearing the voice of an adult woman or a girl played through a speaker, the dogs were shown a photograph of a face that either matched or mismatched the voice. We then measured gaze duration towards the monitor. Contrary to our prediction, dogs tended to look longer in congruent than in incongruent conditions, although this effect was only marginally significant. This pattern suggests that dogs may discriminate between congruent and incongruent voice–face pairings, but not in the predicted direction. Exploratory analyses further suggested that this tendency was more pronounced among dogs who had lived with children. Overall, these findings provide tentative evidence that dogs may associate human female voices and faces across age classes, and that experience with children could influence this process.
Download the full article
Read the complete six-page paper, including its methods, results, discussion and references.
Download the full article (PDF)
Source and reuse license
Chijiiwa, H., & Kanakogi, Y. (2026). Dogs might associate the voices and faces of human girls and adult women. Animal Behaviour, 240, 123710. Original publisher article and supplementary material.
© 2026 The Authors. Published by Elsevier Ltd on behalf of The Association for the Study of Animal Behaviour. Licensed under Creative Commons Attribution 4.0 International (CC BY 4.0).
The abstract is reproduced verbatim. The downloadable publisher PDF is provided unchanged, with original credits and notices intact. Hosting does not imply endorsement of DogTraining.World by the authors or publisher. Supplementary files remain available from the publisher.