Researchers from University of California, San Diego, have just released the results of a new study that used the increasingly-popular button dog mats to scientifically test dogs’ understanding of words. Button dog mats are, yes, mats with buttons, in which a dog’s person has recorded one word for each button. The idea is that with training, the dog can communicate in the human’s language by pressing on specific buttons to “speak.” The UC San Diego scientists compared dogs’ reactions to words spoken by humans and buttons, as a way to gauge whether dogs were responding to the words themselves or to people’s behavioral cues. Turns out dogs understand words related to walking/playing/going outside, regardless of whether it was spoken by man or machine, yet food-related words did not elicit the same responses. Exercise more motivating than food? That is so totally the opposite of how things go down in my house…
This is a test: The first experiment was conducted in person, with researchers visiting 30 dogs’ homes across the country to test their responses to soundboard buttons. The voice on the buttons was that of the owners. An experimenter pressed the buttons while the owner was in another room, then watched the dog for 60 seconds. The second experiment involved “citizen science,” with 29 dog owners conducting the trials themselves at home under remote guidance. The owner would either push the same buttons or say the words out loud to the dog without pushing buttons and then observe the dog’s reaction for 60 seconds.
Dogs understood links between certain words and resulting activities: The dogs were more likely to exhibit “play-related” or “outside-related” behaviors after hearing the relevant words, whether spoken by a human or emitted by the soundboard. “This demonstrates that dogs are, at the very least, capable of learning an association between these words or buttons and their outcomes in the world,” the researchers wrote in the study. Dogs didn’t show “food-related” behaviors in response to food-related words, suggesting the dogs either weren’t hungry or didn’t expect food outside of their usual feeding times, according to the study.
Comprehension confirmed: The research suggests that dogs can grasp the meaning of specific words and respond appropriately, regardless of whether they hear them from people or whether the words are triggered by someone pressing a button on a prerecorded soundboard. The researchers said the findings show that the dogs understand the words and aren’t reacting to body-language cues from humans. “We know that dogs understand quite a few words,” said Federico Rossano, associate professor in the department of cognitive science at the University of California at San Diego. … The study shows that the dogs “are paying attention to the words and produce appropriate responses when they hear them,” Rossano said. He added that the team has two more publications under review that look at how dogs are using the buttons.
And then a horse trots into the story: The study shows that the dogs’ responses were not a “Clever Hans effect,” a term originating with an early 20th-century horse in Berlin named Hans. Hans was known for tapping numbers or letters with his hoof in answering questions. It was later revealed that the horse was picking up microscopic facial signals from his handler.
Some experts are still skeptics: “Dogs act similarly when the button says ‘play’ or ‘out’ as when their person says ‘play’ or ‘out,’ but it’s the owner’s voice on the button, so that is expected,” said Alexandra Horowitz, who heads the Dog Cognition Lab at Barnard College. “Dogs easily learn that my asking ‘Where’s your ball?’ or a visiting friend asking ‘Where’s your ball?’ and my voice on speakerphone asking ‘Where’s your ball?’ mean the same thing.”
I, for one, have never met a dog who didn’t know the word “walk.” Even across different humans saying it in varying ways or goosing it up in the baby talk dogs love, “walk” is always understood. There’s the instant freezing of motion, then the eye contact with the person, like the dog is saying, “Are we really doing it? Is it happening? Can I burst into exuberant joy now?” And then an exhale and they go bananas. So I can buy that there’s comprehension of certain words, regardless of who or what says them.
What I’m still hung up on is why food-related words seemed to fall flat. Because… it’s FOOD! I guess this is just another area where my dearly-departed My Girl and I were perfectly suited. I guarantee you, if these researchers had run the tests with My Girl, she would have thrown the results into chaos! None of this “she just wasn’t hungry” or didn’t expect food at “non-feeding times” nonsense. My Girl expected, wanted, and sought out food at every availability opportunity. So if you would, Prof. Rossano, I’d love for some of the next studies to examine this stark contrast in reactions.
And lastly, because I’m a bit of a researcher myself, a show of hands please, and BE HONEST: who here knew about Clever Hans before today?
