It was the first day of a beginner’s dog obedience class. In front of me a woman was standing beside a Golden Retriever, and the dog’s face was showing a considerable amount of gray on it. The woman said to me, in a troubled tone, “I’m not sure that this class is going to do any good. My dog is 8 years old and she’s never had any obedience training. My husband insists that you can’t teach an old dog new tricks.” This is a common concern among dog owners who worry it’s “too late” to teach their dog new behaviors.
My own dogs seem capable of a lot of learning well into their senior years. For example, my Nova Scotia Duck Tolling Retriever, Dancer, was 12 years of age when I decided to teach him the exercises needed for Rally trials. Not only did he learn these new performance skills, but he also went into competition and earned his Rally Advanced title at the ripe old age of 14 years. However, that is just my personal experience, so it is comforting to know that now there is scientific data to back the conclusion that an old dog can, in fact, continue to learn.
A recent study has come out of the Messerli Research Institute, which is part of the University of Vienna. Lisa Wallis headed the team of researchers, and the study was published in the journal Age. It was one of a series of investigations of dog behavior that have been carried out at the “Clever Dog Lab,” which is headed by Friederike Range. This three-year-long study involved 95 Border Collies ranging in age from 5 months to 13 years, all of whom were pet dogs that were brought into the laboratory at various times for testing.
So far, the effect of aging on cognitive processes (such as learning, memory, and problem-solving), has mostly been studied in people, where it has been confirmed that older brains are often less efficient in these areas. However, it has recently been established that dogs show many of the same kind of age-related changes that humans do. In fact, there are enough similarities between dogs and people in terms of the aging brain that the University of Toronto is studying the behavior of older Beagles in order to better understand what may be going on in aging humans. The Clever Dog Lab study was an exploration of the relationship between age and learning ability, focusing specifically on dogs.
The apparatus used in this investigation involves a touch screen attached to a computer. When the dog touches the correct item displayed on the screen, the computer delivers a treat from a dispenser located below it. Range discovered the usefulness of the touch screen for working with dogs totally by accident. She had started taking her Border Collie puppy, Guinness, to work and found that he enjoyed playing with the touch screen used by some other researchers to study cognitive processes in pigeons. Guinness was quite successful at learning to solve some of the problems that had been used to study bird behavior, and Range recognized that this might prove to be a technique useful for studying the mind of dogs. The touch screen has now become a standard piece of apparatus in the Clever Dog Lab.
The basic setup of this study involves projecting two items on the screen, side-by-side. If the dog touches the correct picture with his nose, he is rewarded with a food treat. Touching the incorrect picture does not result in a treat but rather in a timeout. In each set of test items there were four “positive” pictures that were always rewarded and four “negative” pictures that were never rewarded. In the easier task the pictures were simple clipart-style images (such as a clock or a cup), while in the more complex discrimination the pictures involved underwater photographs versus abstract paintings.
The first, and most important, finding is that all of the dogs were capable of learning these discriminations. However, that is not to say that there weren’t differences in learning ability depending on the age of the dogs. The data showed that it took older dogs more training trials to reach the desired level of proficiency and the older dogs had to endure more correction trials. To give you an example, the most efficient learners were dogs in the 6-month to 1 year of age group. On the easy discrimination task it took them approximately 18 trials to reach the learning criterion set by the experimenters (which was 87 percent correct choices for this task). Compare this to older dogs who were around 10 years of age. These senior canines took approximately 39 trials to reach the same learning criterion. This means that these elderly dogs took more than twice as long to learn —however they did learn and did reach exactly the same level of proficiency eventually.
In addition to the straight discrimination tasks, the investigators looked at logical reasoning on the part of the dogs. This involves what has been called an “exclusion task.” After the dogs had reached their learning criterion, they were again shown pairs of pictures on the touch screen. This time one of the pictures was completely new (one that had never been used in any of their training sessions), while the second one was familiar from their previous training where it had been one of the “negative” or always unrewarded pictures. When presented with such a choice, the logical reasoning should be, “I know that one of these pictures has never been rewarded when I touched it before, so my best bet would be to choose the other one.”
Surprisingly, the older dogs did better than the younger dogs in this inference or reasoning task. Range summarizes their findings saying, “The older the dog, the better it performed, while younger dogs were unable to master this task. This is probably due to the fact that older dogs more stubbornly insist on what they have learned before and are less flexible than younger animals.” Hmmm… I wonder if the same thing holds for humans.
Finally, there is the question as to how well the dogs retained what they had learned. For this part of the test, the investigators allowed at least six months to lapse after the first learning tests, and then they repeated the touchscreen trials using the same pictures as before in order to evaluate the dogs’ long-term memory skills. The good news is that nearly all of the dogs remembered the correct pictures as being positive. In this long-term memory test there were no significant age-related differences with both the older and younger dogs performing equally well.
So the conclusion that can be reached from this research is that you can teach an old dog new tricks, only it’ll take longer than it takes a young dog. But once that old dog has learned, he will remember these new things over the long term.
Stanley Coren, Ph.D., is a professor in the Department of Psychology at the University of British Columbia and a writer for Psychology Today.