Does your dog know the names of its toys? Lots of dog owners may think their precious pooch can identify their favorite toys by name, but the phenomenon is actually pretty rare. In fact, many dogs who are believed to have this ability have usually had it trained into them and they can only identify a few objects. But then there are spontaneous learners, the special cases, the truly “gifted”.
Dogs are popular subjects for studying social cognition because they have evolved alongside humans. So aside from being adorable, their ability to identify object labels (the names of things), and to associate labels with specific objects could offer insights into various cognitive mechanisms.
For instance, this ability could help us understand how object labels are acquired and how they work in reference communicative signals in non-human species. That is, how animals understand labels as references for things. Moreover, a dog’s knowledge of object labels could indicate whether they have bias towards certain features on an object. And the influence of these verbal labels – what is essentially language – on mental representation can be explored by comparing dogs with knowledge to those who lack it.
This sounds great, but not all dogs can do it. Most of the very few studies into this phenomenon have only included a few animals – one or two – because, contrary to what their owners may think, the dogs themselves do not demonstrate evidence of knowing objects on an individual level. And efforts to teach dogs have usually failed. As such, what are called Gifted Word Learner (GWL) dogs, those who have this ability, are rare.
Given their apparent rareness, how can we explore GWL dogs? Well, a team of researchers turned to the citizen science approach to gather information.
“As GWL dogs are rare, the search for them and the documentation of their behaviour can most efficiently be done with the help of educated dog owners”, Dr Shany Dror, Eötvös Loránd University, Hungary, and colleagues wrote.
The authors used this method to create the first study conducted on a relatively large sample size of dogs that possess a vocabulary of multiple object names.
Citizens to the rescue
Dror and colleagues operated on the hypothesis that an owner’s skills in dog training or experience with dogs could contribute to the ability of their dogs to learn object names.
“In this case, we would expect that most owners of GWL dogs are professional trainers or experienced dog owners and that most of them deliberately taught the toy names. Additionally, we aimed to explore whether most GWL dog owners had many dogs in the past and whether their previous dogs or dogs adopted after the GWL dogs, also showed this exceptional ability,” they explained.
“Moreover, we expect that households with multiple dogs would likely include more than one GWL dog, assuming that owners have a consistent approach and lifestyle towards their dogs.”
Through this citizen-led approach, the team were able to identify 41 GWL dogs. They identified them through a social media campaign and press articles. The owners who took part were first instructed on how to test their dog’s knowledge of toy names by themselves. Then the team performed a controlled Vocabulary Assessment Test (VAT) whereby researchers observed the dogs over the internet (not in a creepy way).
Owners of the dogs that performed significantly better than chance were then asked to carry out a questionnaire about the dogs’ life history, the owner’s experience as owners, and the process through which their dog acquired their knowledge.
“In addition, we tested for correlation between the number of toys the dogs confirmed to know during the VAT, and the number of toys they were estimated to know approximately two years after the VAT, as well as between the estimated number of toys and the dogs’ Accuracy in the VAT,” the team added.
The results, which took five years to gather, are certainly something to wag your tail over.
“During the tests, we found that most of the dogs had already learned over 15 object labels, and their owners reported that they continued to acquire more labels, with many dogs now being able to recognize over 100”, the team explained.
Interestingly, out of the 35 owners who took part in the study, only three of them reported having a professional background in dog training. This suggests dog owner’s skills are not a significant factor behind GWL dogs’ abilities.
“Obviously, owners play some role in facilitating GWL dogs’ ability to learn object labels”, the team added, “because the dogs would not be able to learn toy names without owners providing them with toys, and dedicating time to play with them.”
“And yet, the results presented here suggest that the owners’ training skills are likely not the primary driver behind these dogs’ skill.”
Crucially, around 74 percent of the owners who participated reported that they did not intentionally train their dogs to learn toy names. Instead, they noticed that their dog had learned the names spontaneously through play interactions. This then prompted the owners to introduce more toys. As such, the owners did not necessarily recognize that their doggo was doing anything special.
It is possible that “playfulness” may be a factor in this ability, but even among naturally playful breeds – such as Border collies – the presence of true GWL dogs is rare.
“Thus, we suggest that playfulness, by itself, is likely not enough for the development of GWL dogs’ vocabulary. Here we also present the first evidence that, although the ability to learn object labels is most frequently found in Border collies, it also exists in a variety of other breeds.”
Obviously, with a citizen science study, there are limitations about bias and, in this case, there is a small sample size. Nevertheless, the study offers new insights using this novel approach.
“While our findings on GWL dogs should not be generalized to the wider population of typical family dogs, they support previous findings on GWL dogs, increasing their validity and suggesting that what has been found with a specific GWL dog, may be extended to other GWL dogs”, the team concluded.
The study is published in Scientific Reports.