For the Love of a Dog (45 page)

In one experiment, it looked as though dogs weren’t able to use the concept of “different” to solve a problem because they failed what’s called the “delayed nonmatching to sample test.” In this exercise the dogs are first shown a tray with just one item on it, and under that item is some yummy food. The dog is allowed to push it over and eat the food. The tray is then taken away, and after a delay of ten seconds the dog is presented with a tray again, this time with two objects on it. One object is the same as before, the other is different. The “right” choice is the
different
object, the one that doesn’t match the object that was on the first tray. After hundreds of trials, dogs were unable to learn to choose the different object, although rhesus monkeys could figure it out pretty quickly.

But in a testament to good science, the researchers varied the task, by asking the dogs to choose an object in a different
location
than it was the first time. Dogs not only figured that out, they got it right 90 percent of the time, even when they had to wait twenty seconds between the trays. Studies like this support anecdotal observations suggesting that dogs can learn abstractions. Ken Ramirez, the head animal trainer at the Shedd Aquarium in Chicago, says it’s relatively easy to teach dogs the concept of larger versus smaller, same versus different, and right versus left.

Dogs also seem to have some type of concept of number, which is also an abstraction. In one study originally done on children, dogs watched items being placed behind a screen, one at a time. After the last object was placed, the screen is removed. A subject who understands
something about number or quantity should expect to see the same number of objects sitting together as were lowered, one by one, behind the screen in the first place. It’s these tests that taught us that children about five months old have some concept of quantity—if the “wrong” number of objects is revealed behind the screen, they stare at them for much longer than they do if they see the number they expected. Dogs do the same thing, appearing to be surprised by the “wrong” number of objects once they’re revealed.

The concept of number, although we take it for granted every day, is relatively complicated. It includes the understanding that number is unrelated to size or shape (five Saint Bernards are the same number as five Miniature Poodles) and what’s called the property of cardinality, which states that the last number in a count is the total number of things counted. Children don’t understand this concept until they’re around three and a half or four years of age; before that they don’t seem to realize that if they count “One, two, three,” there is a total of three things in front of them. Studies asking whether at least some animals can understand the basic concept of number are on the rise, and they overwhelmingly suggest that it’s a concept many nonhuman animals can grasp.

Although we need much more research, it appears that animals of many species can keep up to about seven items in their working memory, but after that their performance begins to degrade. That’s another benefit of our big, linguistically oriented brains: language and our large working memory allow us to keep far more facts or concepts in our brains at the same time than the members of any other species can. However, the research suggests that the difference between us and other animals is one of degree. It’s not as simple as “Humans can think, and animals can’t,” but rather a matter of how much and in what ways animals think in comparison to humans.

Once you start looking, you can find an infinite number of stories about dogs solving problems that suggest far more than mechanical reactions to the world around them. A dog named Izzie figured out how to get around the mousetraps that his owners had placed to prevent
him from sleeping on the furniture. The owners couldn’t figure out how Izzie was managing to avoid the traps—until they set up a video camera, and watched in awe as Izzie picked a blanket up in his mouth, awkwardly dragged it across the traps until they all snapped, and then cozied up on the couch with the blanket as a pillow.

Stories of dogs teaching their humans to fetch are a dime a dozen. Some special clients of mine had a Border Collie, Bear, who had learned he wasn’t allowed to go off the grass into the road. His owners began noticing that when they played fetch, somehow the ball kept ending up in the road, no matter which way they threw it. They finally decided that Bear was purposely moving the ball off the grass, and then laying back, tail wagging, and watching them do the fetching. The same dog, like many I’ve known, played fetch with himself by trotting up the stairs with the ball in his mouth, dropping it at the top of the stairs, dashing down the stairs to catch it, and then carrying it back up again.

Another story comes from an article in
Time
magazine titled “Can Animals Think?” The article includes a story about the author’s experience with a Golden Retriever named Newton, who, like Bear, created his own version of fetch. After his human friend Michael tossed the ball, Newton appeared to search high and low for it, even looking up into the trees. When Michael gave up on Newton’s finding the ball and came to fetch it himself, Newton would wait until he was about ten feet away and then make a beeline to the ball and pounce on it. Michael wrote that Newton’s expression as he dashed away with his prize looked “suspiciously like a grin.”

Of course, often the dog’s behavior in stories like these could be explained by simple trial and error, not by the dog doing what is strictly defined as thinking. In some ways, the value in these stories is that you can use them as litmus tests: tell them to other people and you’ll get a good idea of how they evaluate the mental capabilities of dogs. Just be ready to hear some strongly held opinions. It’s an interesting and useful exercise to ask yourself what your dog would have to do to prove to you that he’s thinking. Don’t hesitate to be skeptical; it’s no disservice to your dog to be a critical thinker, dismissing examples that could be explained by simple trial-and-error learning. One objective way to ask if your example is a good one is
to put another species in your dog’s place. Your dog may come running every time she hears the can opener, but you could teach any animal to do that, even a grasshopper. Does that mean grasshoppers can think?

That’s why the research I’ve described is so important, because it helps us sort out in a more organized fashion what animals can and can’t do. An article from 2004 in the journal
Science
is a good example of how research can help us learn more about the mental experiences of our dogs. This experiment tested the ability of a Border Collie named Rico to link an unfamiliar word with an unfamiliar object. After being taught the names of two hundred different objects by his owners, Rico was asked to go into another room and fetch an object he’d never heard of before. Rico heard the phrase “Fetch the spoon” for the first time, having never been taught the word “spoon” before, and never having been asked to fetch one.
¹⁰
When Rico entered the room, he’d find seven familiar objects he knew the names of, and one unfamiliar one. In seven of ten sessions Rico was able to link the unfamiliar word with the unfamiliar object and bring back what was asked. By doing this, he not only made clear that he understood the abstract concept “unfamiliar,” he also demonstrated a skill called “fast mapping” by inferring that the unfamiliar word referred to the unfamiliar object. Until this study was conducted, it was believed that only human children could fast map. The authors concluded that the “building blocks” of language acquisition may exist in other species.

The experiment on Rico brings up the second and often strongly voiced argument that dogs can’t think because they don’t “have language.” It is true that the use of language in all its complexity is one of the defining characteristics of our species. It allows us to think in multiple levels of abstraction, to use words and thoughts to consider our actions rather than responding without reflection, and to communicate with one another on a level unimaginable in other species. But what of our dogs, and other animals like them? Is it possible to think, as we understand thinking, without being as adept at language as we are? How does our ability to use language affect our emotions,
and how might the lack of that ability affect our dogs? We’ll address these questions in the next section.

A philosopher, R. G. Frey, is on record as arguing that animals can’t have desires if they have no language. While that statement is illogical and has no basis in biology, it is understandable that we find it difficult to imagine thinking without language. Thinking and language are so closely associated in our minds that a standard psychology text links them together in the same chapter.
¹¹
But does one really have to be able to use human language to form abstractions in the brain, to understand concepts and use them to solve problems? Perhaps the question is, How much is language an expression of what goes on inside the brain, and how much of what happens inside the brain does it actually create? That’s the chicken-and-egg question of the ages, and it’s keeping a lot of people busy thinking, talking, and writing in pursuit of an answer.

In one famous example of research on cognition and language, researchers attempted to teach African gray parrots to use English words by rewarding the bird with a food treat for making a sound like the one played from a tape recorder. The birds failed miserably, and it was concluded that parrots were unable to learn or use the English language functionally. That is, until the biologist Irene Pepperberg came along and taught a parrot named Alex to learn words in ways more similar to the ways we teach our children. She held up an object that Alex wanted, and asked a human assistant what it was. If the person named it by saying “cork” or “nail,” Pepperberg handed her the object. All the while Alex was simply observing. When Alex finally made a noise himself that sounded something like “cork” or “nail,” he was handed the object. That was enough to keep his attention, since what parrots want more than anything is to play with objects using their beaks and feet. Several years and hundreds of hours later, the world’s most famous bird is able to use words to identify an object’s composition (cork, wood), its shape (round, triangular), its color, and its size. Alex can tell you
which yellow, round, wooden object is “bigger” on a tray of other objects that are yellow, round, and made out of wood. He can tell you what category is “different” on a tray of five blue keys and one red one. Most amazingly, without being taught to ask questions himself, Alex asked one of his own. During the process of learning his colors, he asked “What color?” while looking at his image in the mirror. “Gray,” he was told, and after that he was able to point out anything gray, no matter its size, shape, or composition.

Another study, this time on chimpanzees, looked at language from an entirely different perspective. It was done by the primatologist Sally Boyson, and its result emphasizes the importance of language on one’s ability to think in the abstract. In her study, chimpanzees were first taught to associate written numbers (“1,” “2,” “3,” etc.), with actual quantities. The chimps could look at a basket with four peaches in it and point to the number “4.” They could look at a basket of two peaches and then one with three peaches, add them up in their minds, and then point to the number “5.”

This finding enabled Boyson to proceed to a second study, in which one chimp got to decide which of two bowls of candy he got, and which bowl was given to a second chimp. Each bowl had a different quantity of M&M’s (which chimps like as much as we do). The active chimp’s job was to point to the bowl that the
other
chimp would get; the remaining bowl of candy was his to eat. Time and time again, the active chimp would point to the bowl with more candy, and look disheartened when Boyson handed the larger amount of candy to the other ape. The active chimp seemed unable to point to the bowl with fewer M&M’s, even after hundreds of opportunities to learn that by pointing to the largest number of candies he’d end up with fewer himself. However, when Boyson substituted numbers in the bowls, to symbolize how much candy each bowl contained, the chimp quickly learned to point to the lower number, thereby getting more candy for himself. In other words, once the chimp was able to add a layer of abstraction between himself and the candy through the use of language-like symbols, he was able to make a smarter choice.

I can’t imagine a study that makes a more compelling argument for the importance of language in our ability to think in abstractions, and to use those abstractions to our advantage. The study suggests that our
linguistic abilities might often act to temper our emotions, allowing us to pause and reflect before we react. However, studies like this do not support the belief that individuals are unable to think
at all
unless they have language. If thinking is the ability to form and manipulate mental abstractions, then the chimps showed that they could think: they were able to use numbers as abstract symbols of quantity. To look at it from a wholly different perspective, that of our own species, just because a person grew up without the concept of language doesn’t mean she can’t think. Could Helen Keller have had no thoughts in her mind before she realized that the sign for water referred to the liquid that was splashing over her hands?

I hope we see much more research on this issue in the years to come, especially on our dogs. However, we know enough already to make a reasonable argument that although language enables us to use abstractions in ways unimaginable by other animals, it is not a prerequisite for thinking. The evidence suggests that, as we’ve defined thought, dogs can indeed think, although their thought processes are undoubtedly much simpler than our own.
¹²
The scientific evidence that some nonhuman animals can understand abstract concepts keeps getting stronger and stronger. Look at what the dolphins can do at Kewalo Basin Marine Mammal Laboratory: not only have they learned that they’ll be rewarded if they create a “new” behavior, one they’ve never performed before, but they understand a cue called “Tandem create.” This cue asks two dolphins to come up with a behavior they’ve never done before, communicate it to each other, plan how to execute it, and then perform it simultaneously. If that’s not thinking, I don’t know what is.