I’m always curious about how much my dogs understand when I talk to them. I do know one thing: They know their names. Call either “Watson” or “Raylan,” and one looks up, even in a noisy room. They know who they are.
They also know that the hearing and attention skills of dogs are sharper than that of babies, while language cognition — how they learn and understand certain words — is similar in both. (To be sure, dogs cannot speak, but just try saying “cookie” followed by “sit” to my dogs and watch what happens.)
Because babies develop these multiple skills simultaneously, scientists cannot tell which ones are at work when they react — or don’t react — to speakers in a noisy environment, they say.
“Dogs are like infants in some ways . . . and by seeing how they behave when doing listening and noise tasks can tell us which of the skills is likely involved in infants,” says Rochelle Newman, associate director of the Maryland Language Science Center and chair of the University of Maryland’s Department of Hearing and Speech Sciences.
“Babies are bad at listening in noise, and we don’t know why,” she says. “Figuring out why babies have trouble might tell us what we need to do to make it easier for them.”
Babies seem to tune out when there is noise, not just to hearing their names but also to other tasks.
“They no longer show recognition of what is being said,” Newman says. “So if there are two objects to look at, if a voice tells them to look at one in quiet, they will look appropriately, but in noise they will look randomly back and forth, suggesting they no longer understand what is being said.”
Because dogs typically respond better to speech-in-noise than infants — and researchers are aware of their skills differences — dog experiments using their names can provide important clues about what may be going on in babies. That is why Watson, my black Labrador retriever, and Raylan, my yellow German shepherd-beagle mix, and I found ourselves recently in a small, three-cornered test booth at U-Md.’s Canine Language Perception Lab, eager to make our contribution to science.
While there, the scientists tested both dogs as part of their ongoing research into background noise response.
The researchers had prepared several recordings in advance to see how the dogs respond to hearing their names amid different levels of background chatter, an experiment that allows them to make comparisons to infants.
“Dogs have really good hearing, much better than babies, but they don’t have any more language than babies, which suggests that if the reason babies were failing was because they didn’t have language, then dogs would be failing the same way,” Newman says.
“Dogs know their names, and so do babies, so if language was the problem — and if you needed language to listen in noise — then you would expect dogs and babies to have similar responses,” she adds. “But dogs are better, which tells us that something else is going on.”
My dogs and I sat in a small booth with pegboard lining the walls. The booth was equipped with a video camera in front and audio speakers mounted on each side. Each dog was tested separately and heard several recordings that used either his name or a fake name (the scientists call it a “foil”), in our case, “Echo.”
They began with the dog’s name spoken without background noise. With each test, the background noise grew in intensity, although the sound of their names was always 5 decibels louder than that of the background speech. To avoid undue influence on my part, they had me wear headphones with piped-in music to block the outside sounds.
Each dog turned toward the speakers a few times upon hearing his name, so they knew the disembodied voice was talking to them. Watson actually barked upon hearing it. Raylan, on the other hand, looked at the speakers, then seemed to grow bored. Neither reacted to “Echo.”
What about that barking? He may have found hearing his name from an unknown source confusing, or “he just enjoyed hearing his name,” says Amritha Mallikarjun, the doctoral student who ran the experiment. Raylan may have become less interested as the testing progressed, she says. “To be fair, we do play the same stimuli several times, so he might have gotten bored, or he might have been tired,” she says.
The responses do not measure a dog’s intelligence.
“We’re interested in how dogs as a group perform, rather than any individual dogs,” Newman says. “What are the commonalities among these dogs when it comes to their responses?”
My dogs’ information will join a body of data from more than 300 dogs they have studied to help tease out the skills infants use — and do not use — when coping with the “cocktail party” effect.
“We think the main reason has to do with attention skills,” Newman says. “We as adults can choose what to pay attention to, but infants don’t have the same control over that — although sounds do grab their attention. We don’t know why infants show this difference. It could be because of poor language skills, poor attention skills or poor hearing skills. Because all of these things mature together, we can’t know which might be having an influence. Since dogs likewise have poor language skills, but have better hearing skills, comparing performance between infants and dogs can help us single out what is more likely to be the cause.”
The results from this study also will be used to help those who train dogs destined for service jobs — for example, as emotional support companions and for dogs who engage in drug and cadaver detection and search-and-rescue missions — where they often must work in noisy, chaotic and stressful environments.
“Like humans, dogs often find themselves in noisy environments where they are expected to respond to human speech, despite distractions,” Newman says. “Service animals are often in environments where there are a lot of people talking, and where humans give commands. We want to know what levels of noise they can deal with.”
In other research, the scholars are studying dogs’ language “comprehension,” specifically how they distinguish between certain words. Their preliminary results suggest that “dogs can tell the difference between words that differ only with a vowel, but not when you change a consonant,” Newman says. “If a dog’s name is ‘Mick,’ the dog will not respond to ‘Mack’ or ‘Mike,’ but will respond to ‘Tick.’”
This distinction is also helpful for training service dogs, she says.
“A lot of dogs have problems learning the command ‘bow’ — which is when the dog keeps his front end down and bottom end up — because they have already learned the word “down,’ and have trouble telling the difference between consonants,” Newman says. “That’s one reason you wouldn’t want to name your dogs ‘Tom’ and “Bob,’ where the vowels are the same and the consonants different.”
Dogs often participate in research, historically in disease studies that have human applications, such as cancer. But there is also a growing interest in learning more about dog cognition. They make great research subjects because of the strong connection they have developed over time with humans, experts say.
“Dogs have this amazing convergence with us,” says Brian Hare, professor of evolutionary anthropology at Duke University, a longtime dog researcher and founder of the Duke Canine Cognition Center.
“We are so distantly related to dogs and yet, in some ways, we are more similar to dogs than to great apes,” he says. “This convergence means we can study dogs to learn more about ourselves — which is extraordinary, when you think about it.”
Duke’s dog research projects seek to gain a better understanding of dog psychology, how dogs process language, develop trusting relationships and form memories.
“Dogs are more than mere learning machines,” Hare says. “They have a rich understanding of their world, which allows them to be flexible problem solvers. Some of their skills even resemble those we see in young children.”
Similar canine perception study programs — which often use fun puzzles and games to test the dogs’ cognitive abilities — have begun in recent years at other academic sites, including Yale University, Arizona State University, Columbia University, the University of Florida, the University of Kentucky and Emory University.
These studies all rely on the willingness of owners to volunteer their pets. It’s an enjoyable outing for the dogs — and their humans, as well.
It was cool to see how my dogs did.
I had been concerned about the possibility of chaos, as Watson is a high-energy pup, easily overstimulated in new environments, and Raylan occasionally is anxious with strangers. I needn’t have worried. These scientists, accustomed to working with dogs, were patient and friendly. But a few unexpected mishaps have occurred with other dogs, they say.
One dog, fearful of shiny floors, didn’t want to walk down the hallway to the testing booth. Another dog thought the voice behind the wall belonged to a real human, and “almost knocked down the facility trying to get to the ‘person,’ ” Newman recalls. Her own dog, a Scottish deerhound, “basically just laid down and didn’t pay any attention at all,” Newman says, laughing. “She failed.”
But, for the most part, the dogs have a good time, especially since each receives a treat, a squeaky toy and a “dog”torate certificate for participating.
“Most like being here,” Newman says. “There’s a lot of petting and wagging tails. It’s fun for them, and fun for us, as well.”
