Why dogs imitate irrelevant actions

Many dogs copy human actions even though that is actually irrelevant for achieving their goals. At the Messerli Research Institute in Vienna, canine over-imitation is being tested in numerous experiments. In a Science publication, Messerli researcher Shany Dror recently demonstrated that some dogs can learn words merely by overhearing conversations – just like toddlers.

Biology and Medicine

by Alois Pumhösel

February 16, 2026

Humans and dogs share a unique bond. For tens of thousands of years dogs and humans have been interacting, complementing each other, and helping each other. Better than probably any other animal, dogs can understand humans, interpret their actions, and learn from them. An important mechanism behind these abilities lies in what is known as overimitation. Similar to children, dogs imitate human behaviors that are of no immediate benefit to them – from seemingly singing along when music is played to opening their mouths when humans yawn.

Whereas the complex aspects of overimitation in children have already been well investigated, systematic research into this phenomenon in dogs only started relatively recently. A pioneering institution in this field is the Clever Dog Lab at the Messerli Research Institute for Human-Animal Relations, a joint institution of the University of Vienna and the University of Veterinary Medicine Vienna (Vetmeduni Wien). In the current FWF-funded project “Overimitation in Dogs,” researchers are working to test key hypotheses from previous research through experiments. In a publication in Science, project team member Shany Dror recently demonstrated that the learning abilities of some dogs go much further than previously thought.

Research shows that some dogs have exceptional word comprehension. They even can learn new words simply by overhearing conversations, performing as well as or even better than 18-month-old children as researchers at the University of Veterinary Medicine in Vienna have discovered.

Animals, humans, and culture

The issue of overimitation has a far-reaching background. It extends well beyond behavioral research and touches on fundamental questions of culture and social learning. “In children, the imitation of actions that are not relevant for achieving a goal is often understood in a cultural context. It is an attempt to follow social rules, even if one does not understand them. There is no doubt that the technique of -overimitation has helped to bring about complex human culture,” notes Shany Dror, who is conducting research together with principal investigator Ludwig Huber and other colleagues. “This is interesting in relation to dogs because it is often assumed that culture is a factor that distinguishes humans from animals.”

So far, there has been no evidence of overimitation in other animals – including great apes. Dogs, which developed in the same environment as humans in the course of their long domestication, are the only animal species in which this ability has been verifiably proven to date. “It is likely that those dogs that best understood humans' communication and best integrated into their community were able to reproduce most frequently,” explains Dror. “Dogs were thus subject to similar evolutionary pressure as humans – an aspect that makes canine social behavior particularly interesting for research.”

(Not) a question of relationship

In the course of the project, Dror and her colleagues are investigating three specific questions. The first concerns the relationship between the animals and their owners: does the quality of the relationship influence the tendency to engage in overimitation? A study by Dror's colleague Karoline Gerwisch, who examined the behavior of the animals in their own environment in their owners' homes, has already produced surprising insights. It showed that assistance dogs that, according to the survey, have a closer bond with their caregiver do not exhibit significantly higher overimitation than family dogs. In fact, their tendency to overimitate was even lower – which could be a consequence of the special training these assistance dogs receive.

Dror herself is trying to explore the influence of the bond with the owner in a completely different approach. Her focus is on the hormone oxytocin, which influences social bonds in both dogs and humans. “In my study, I am testing whether higher levels of oxytocin in dogs, which usually increases attention to social signals, also promotes a tendency toward overimitation,” Dror explains. The findings are expected to provide further important insights into the significance of the human-animal relationship as regards overimitation.

Study proves dogs' special learning abilities

Some dogs eavesdrop on their owners: Shany Dror et al.

Just a misunderstanding?

Another set of questions in the project deals with the motivation for overimitation. Do dogs understand that there is no actual need to perform acts of overimitation to achieve their goals? Or is it merely a misunderstanding on their part, and they do not even know that their behavior has no functional benefit? The research strategy involves finding imitation tasks that are not only irrelevant, but also make the animals aware of this irrelevance. For example, a human can perform an obviously unnecessary gesture in reaching for a treat in a box – will the dog copy it?

The researchers also pursue the question of whether overimitation is related not only to evolution and domestication, but also to the individual development and socialization of the respective animal. On the one hand, the experiments involve working with puppies in order to compare their behavior with that of adult animals. On the other hand, the focus is on the comparison between dogs and wolves. “Although they are closely related genetically, there are enormous differences in the behavior of dogs and wolves that could furnish insights into the development of overimitation,” notes Dror. “Even though working with wolves is challenging, we still hope to obtain reliable results.”

The amazing abilities of dog prodigies

In her dissertation at Eötvös Loránd University (ELTE) in Budapest, Shany Dror already worked with a very rare group of dogs that have special abilities: so-called Gifted Word Learner dogs, which exhibit an extreme talent for learning the names of objects. It is known that they can memorize hundreds of object names through interaction with their owners. Dror and colleagues at ELTE have now presented a highly acclaimed new finding in a publication in the journal Science. “We show that these dogs can learn new words even when overhearing two people talking,” explains Dror. “And that’s not all: this learning process was just as good and effective as that achieved through direct interaction.”

Surprisingly, these Gifted Word Learner dogs display abilities comparable to those of 18-month-old toddlers. This said, the underlying mechanisms and cognitive abilities are likely to be different, says Dror. “I think future research will find differences in the extent to which children and dogs can assimilate the perspective of those involved and understand the underlying mental processes.” Nevertheless, the more scientists like Dror explore the behavior and cognitive abilities of dogs, the clearer it becomes that the boundaries between humans and animals, once believed immovable, are starting to become more fluid.

About the researchers

Shany Dror is a postdoctoral researcher at the Clever Dog Lab at the Messerli Research Institute for Human-Animal Relations, an interuniversity institution of the University of Veterinary Medicine Vienna (Vetmeduni) with the University of Vienna and the Medical University of Vienna. She earned her doctorate at the Department of Behavioral Biology at Eötvös Loránd University (ELTE) in Budapest. Her work with the rare Gifted Word Learner dogs and the associated live broadcasts of online experiments involving a “Genius Dog Challenge” have already attracted worldwide media attention.

Ludwig Huber is Professor of the Scientific Foundations of Animal Ethics and Human-Animal Relations, Head of the Department of Cognition and Applied Ethology at the Messerli Research Institute, which also includes the Clever Dog Lab, and currently also Head of the overarching Department of Interdisciplinary Life Sciences at Vetmeduni Vienna. His broad, comparative research approach focuses on animal cognition. Set to run from 2024 to 2027, the project “Overimitation in Dogs” received EUR 449,000 in funding from the Austrian Science Fund FWF.

Science Publication