Science News

Heart of a dog

A dog’s ability to assess humans and to follow human signals changes with age

Kate Telma

doi:10.1038/nindia.2017.99 Published online 1 August 2017

Adults and juvenile dogs try to make eye contact and assess humans.

© Anindita Bhadra

Human teenagers aren’t the only mammals that go through a difficult phase. Free-ranging dog researchers report that compared to adult dogs, juvenile canines are more likely to avoid interacting and to ignore the pointing gestures of reliable humans1.

Free-ranging dogs comprise nearly 80% of the world’s dog population. But studies that seek to understand the origin and evolution of canine social cognition tend to focus on pet dogs or captive wolves. “The problem is that people were [only] working with pet dogs, and trying to understand why dogs are so good with humans, but that is a circular argument,” because pets are especially bred and raised to interact with humans, explains Anindita Bhadra, an assistant professor of biological sciences at the Indian Institute of Science Education and Research in Kolkata and senior author of the study.

Free-ranging dogs maintain complex relationships with humans. They exist in close proximity to, but are not reared by humans. As scavengers, they depend on humans for food scraps and survival — but humans are also their greatest threat. The free-ranging dog population in India is more genetically diverse than various pet breeds, and also genetically closer to wolves. The centuries old population makes an ideal model system to study.

Bhadra’s team took to the streets around IISER Kolkata, observing dogs at railway stations, residential neighbourhoods, and areas adjacent to highways at every hour of the day over the course of a year.

After familiarising each dog with the experimental set up, they set out two bowls, equally distant from the dog, each covered with a piece of cardboard. One contained a piece of raw chicken, and the other had been rubbed with chicken. Blind to which had meat, the experimenter pointed to a bowl at random. The dog had thirty seconds to approach a bowl, and was allowed to eat the chicken if it found a piece. Nearly all pups approached the bowls, compared to less than half of juveniles. Once at the bowls, 78 percent of pups followed the pointing cue, whereas roughly 50 percent of adults and juveniles took the cue. 

The experimenters repeated the pointing trial three times with each dog in order to assess the role of positive reinforcement — if the dog followed the point, and was rewarded with a piece of meat, were they more likely to take a pointing cue from a human immediately in the future?

“Adults show more judgement in whether to follow pointing based on the past experience. In a way, they are more rational than the juveniles,” Bhadra said, because 14 of 16 adults followed subsequent cues after being rewarded, whereas juveniles and pups followed subsequent cues at random.

“It is the very fast adaptation of dogs [to positive reinforcement] that our study and Bhadra's study have common,” said Akiko Takaoka, a research associate at Meiji Gakuin University who studies psychology and dog cognition and who was not involved with the study. Takaoka conducted a studythat measured the ability of pet dogs to follow similar pointing cues, and the likelihood of a dog to follow cues from a reliable or unreliable experimenter.

Next, Bhadra’s team coded videos of the interactions with the dogs to assess the gaze alteration, or the amount each dog shifted gaze from the bowls to the experimenter, throughout each pointing trial. Gaze alteration can be interpreted as a measure of confidence.

Making eye contact

If unsure of the task, the dog looks to the experimenter for clues about what to do next; if the dog is trying to assess the intentions of the human, it may look to the human and try to establish eye contact. According to Takaoka, the ability to assess a human’s reliability, and to generalise about human reliability, might be even more critical for free-ranging dogs, because in contrast to pets, free-ranging dogs interact with many different and unfamiliar humans each day.

“Adults and juveniles try to make contact and assess the human,” Bhadra said, by more frequent and longer gazes, even if they were not interested in following the experimenter’s point. Communicating through gazes could be a measure of fear the dog has for the experimenter. Pups just look at what is being pointed at. “There is more blind trust in the pups,” she said.

These age-dependent results suggest behavioural plasticity — that dogs’ ability to follow human pointing, to learn with experience, and to assess human reliability, develops with age. Pups, who were studied while under the care of their mothers, likely had little previous interaction with humans. Adult dogs on a population level, however, have cumulatively more experience with humans, and have the ability to make decisions based on if following the human’s cue in the past led to a food reward.

But juvenile free-ranging dogs are just beginning to look for food on their own, and according to Bhadra often make clumsy nuisances of themselves by scattering garbage around neighbourhoods. They face the brunt of negative human interactions, between children throwing rocks and chasing them with sticks, to the high level of human-influenced mortality3 at that age. Yet there is hope for juveniles to improve their relationships and communication abilities.

“The most fascinating result for us was that adults show so much rational decision making with positive and negative reinforcement,” Bhadra said. “Really, this was almost too good to be true when we looked at the results.”


1. Bhattacharjee, D. et al. Free-ranging dogs show age related plasticity in their ability to follow human pointing. PLoS ONE (2017) doi: 10.1371/journal.pone.0180643

2. Takaoka, A. et al. Do dogs follow behavioral cues from an unreliable human? Anim. Cogn. (2014) doi: 10.1007/s10071-014-0816-2

3. Paul, M. et al. High early life mortality in free-ranging dogs is largely influenced by humans. Sci. Rep. (2016) doi: 10.1038/srep19641