Do Spoken Words Create Mental Images in Dogs?

Photo by Blue Bird / Pexels

There is no doubt that dogs can learn that words spoken by humans may contain instructions for certain behaviors that they are requested to do. You tell Lassie to sit or come and she sits or ambles to you appropriately following your instructions. However for humans, even for young children, words can represent objects, as well as actions. Many psycholinguistic researchers believe that when a person hears a spoken word naming a particular item, their brain responds by conjuring up (consciously or unconsciously) an image of that particular thing. The idea that a dog might have the mental ability to process words in the same way, invoking mental imagery of the object that the word labels, is something that many psychologists still have severe doubts about. Marianna Boros headed a team of investigators at Eötvös Loránd University in Budapest, who wanted to peek inside a dog's mind to see if canines might be calling up mental images of the items associated with words, just like people do.

Measuring Word Understanding in Dogs

Part of the problem associated with understanding how much language a dog knows comes from how we assess their linguistic knowledge. The behavioral technique typically involves speaking a word to a dog which might be a command or might name an object. If the dog obeys the command or indicates or retrieves the object, we conclude that he understands the word. Around 1993, this is exactly what I did, and in this way, I determined that dogs know about 165 words or word-like signs and signals, which put them on par with a human toddler between 2 and 2.5 years of age. Later research in other labs revealed rare canine linguistic geniuses, such as the Border Collie named Chaser, who demonstrated that she knew more than 1000 words (mostly labels assigned to dog toys and plush animals). She indicated that she understood these words by fetching the requested toy when it was named. However, later research showed that dogs, like Chaser, that could learn many (or even any) object names were few and far between.

There is a methodological problem here. The dog may form a representative image of the object associated with the word, but it has no interest or inclination to retrieve or interact with it. This is what I suspected was the case for my Beagle, Darby. I tried, over an extended time, to teach him to retrieve a few objects that I named for him. It was an abysmal failure. However, I did notice that if I put out two objects that we had practiced with, say a ball and a Frisbee, and instructed him "Darby, fetch the ball," he did not take a single step toward retrieving the appropriate object, though he did turn his head and look directly at it. I always imagined that the problem was not ignorance but obstinacy—and that he was telling me he knew the identity of the object but wasn't the least bit interested in retrieving it.

An Alternate Method of Determining Word Understanding

There is a way around this problem of getting an animal to engage in overt behaviors to indicate they understand a word's meaning. This involves measuring brain activity using some variant of an EEG. The trick here: You say a word and hold up an object corresponding to it while measuring what is happening in the brain. You now compare that same tracking of mental activity to the situation where you say a word, but now hold up an object that does not correspond to the word. If the electrical activity inside the brain is different—depending on whether the word and the object correspond—then the individual is forming a mental representation of the object and comparing it to what appears before their eyes. In other words, they are using some form of internal imagery.

Using this technique researchers have shown that in human beings—some as young as nine months of age—internal representation of objects associated with words was already beginning to form. In this new study, the Hungarian researchers decided to adapt this methodology to see the brain activity of a dog when words were spoken. They used a specially designed portable, non-invasive, EEG monitor and attached it to the heads of dogs.

Do Dogs Form Images from Words Describing Objects?

Specifically, 18 dog owners were invited to bring their pets to the laboratory along with five objects that the animals knew well, such as rubber toys, balls, slippers, and other items. During testing the owners were instructed to say a word corresponding to an object before showing the dog either the correct item or one that did not fit the label. An owner might say, "Hey, look at this ball." Sometimes he might show the ball, but other times he might show his dog a completely different item like a rubber bone.

The results were quite clear. The EEG recordings demonstrated different patterns of activity when the items presented matched the word meaning versus when the item and the word were inconsistent. The differences between consistent and inconsistent presentations were larger for words the owners believed their dogs knew best, indicating that researchers were looking at differences in levels of understanding.

The trick in science is that if something works and produces an understandable outcome in one situation, it should generalize to other situations. Thus, if we believe that these electrical patterns in brain activity are the result of comparing an internal image evoked by words to what is being seen when we are testing human adults and infants, then it seems reasonable to assume that finding the same differences in dogs means that they too are calling up some form of internal imagery when they hear a word as well. Of course, your dog may not respond appropriately by retrieving the item you have named—that's another matter entirely. You must understand that all dogs are not Border Collies.

Copyright SC Psychological Enterprises Ltd. May not be reprinted or reposted without permission