A different researcher then clicked one of these buttons, recording the dog's behaviour despite not knowing which button was which or being able to hear the words they made.
The question of whether or if humans and dogs can communicate with each other through a soundboard has become quite popular among dog enthusiasts. By revealing that dogs trained to utilise these devices react to the pre-recorded phrases in the same way that they do to spoken words, researchers claim to have made the first steps towards understanding.
“Here we show that actually [dogs] do pay attention to the [soundboard] words and they produce appropriate behaviours independently of environmental cues and who produces the word,” said Prof Federico Rossano, of the University of California San Diego, who led the research.
“While this study is most certainly not mind-blowing, it is a necessary first step,” he added.
Push-button soundboard adoption has surged in the last few years, as seen by the abundance of videos on social media featuring dogs like Bunny utilising the device.
However, there is disagreement regarding whether these dogs are actually reacting to sounds coming from the device or are just following their owners' actions or nonverbal cues.
Rossano and colleagues describe how they conducted two trials with 59 dogs in total, all of whom had been trained to use a soundboard in a paper published in the journal Plos One.
In the first experiment, the researcher recorded the words "out/outside," "play/toy," and "food/eat/dinner/hungry" on a dog's soundboard and covered the buttons with coloured stickers.
A different researcher then clicked one of these buttons, recording the dog's behaviour despite not knowing which button was which or being able to hear the words they made.
Then, dog owners conducted a similar experiment, except this time, they alternated between saying the term out loud and pressing one of the buttons.
The findings show that after the play/toy button was pressed, the dogs were around seven times more likely to exhibit play-related actions than the average for the three buttons, whereas the out/outside buttons showed comparable degrees of acceptable conduct. When the relevant button was pressed, they did not, however, exhibit a higher likelihood of displaying actions associated to eating.
The researchers are now studying whether dogs can push the correct button for specific situations, work that they say could not only help probe the depth of the canines’ word comprehension but also shed light on whether such devices can be used for humans and dogs to communicate with each other.
Prof Clive Wynne, the director of the Canine Science Collaboratory at Arizona State University, who was not involved in the work, described the new study as a “nothing burger”, noting that the main finding was that dogs responded to certain verbal cues.
“There is nothing remarkable about that,” he said, adding that the team only studied responses to three familiar words – and the dogs were only successful on two of them.
Wynne said the fact that the dogs had been trained to press buttons played no role in the current study, while the research did not shed new light on what dogs understood when certain words were spoken.
Dr Mélissa Berthet, of the University of Zürich, said the study laid the groundwork for future research and showed – contrary to some suggestions – that the dogs were indeed responding to the audio from the buttons rather than cues from their owner.
“They really needed to show this,” she said. “And now I think the community of scientists is waiting for the rest that’s going to be exciting.”