Thinking in Pictures: My Life with Autism

by Temple Grandin

  Dawkins 's work begs the deeper question of whether a child with autism who is unable to generalize can think. For example, a person with classic Kanner autism can be taught not to run out into the busy street in front of his house because it is dangerous. Unfortunately, he often fails to generalize this knowledge to a street at somebody else 's house. In another scenario, the autistic person may learn the procedure for buying a candy bar at Safeway but have difficulty figuring out how to buy a candy bar at Wal-green 's. Such people are not able to comprehend any deviations from the pictures in their memory.

  According to Dawkins 's criteria, then, savant autistics are not capable of true thought. Autistic people like myself are able to satisfy her criteria for thinking, but I would be denied the ability to think by scientists who maintain that language is essential for thinking.

  When a well-respected animal scientist told me that animals do not think, I replied that if this were true, then I would have to conclude that I was unable to think. He could not imagine thinking in pictures, nor assign it the validity of real thought. Mine is a world of thinking that many language-based thinkers do not comprehend. I have observed that the people who are most likely to deny animals thought are often highly verbal thinkers who have poor visualization skills. They excel at verbal or sequential thinking activities but are unable to read blueprints.

  It is very likely that animals think in pictures and memories of smell, light, and sound patterns. In fact, my visual thinking patterns probably resemble animal thinking more closely than those of verbal thinkers. It seems silly to me to debate whether or not animals can think. To me it has always been obvious that they do. I have always pictured in my mind how the animal responds to the visual images in his head. Since I have pictures in my imagination, I assume that animals have similar pictures. Differences between language-based thought and picture-based thought may explain why artists and accountants fail to understand each other. They are like apples and oranges.

  Studies by Jane Goodall, Dian Fossey, and many other researchers have shown very clearly that primates such as chimpanzees and gorillas can think, though few scientists would also concede that farm animals have thinking abilities. Yet anyone who has spent any time working with cattle knows that they are able to recognize familiar objects when they see them in a new location. My experience suggests that these animals think in discrete visual images. They are able to make an association between a visual image stored in their memory and what they are seeing in the present. During an experiment on the farm at Colorado State University, for example, cattle were handled in a squeeze chute for blood testing once a month for five months. Most cattle willingly reentered the squeeze chute during each blood test after the first one, but a few refused to enter. These animals were very discriminating as to which part of the squeeze chute they disliked, often refusing to put their heads in the stanchion though voluntarily entering the body-squeezing part.

  Apparently when the person operating the lever closed the stanchion too quickly, the animal got banged on the head. Animals that had been accidentally struck were more likely to balk at the head stanchion. Most of them marched right up to the squeeze chute and willingly walked into the body squeeze section, but they stopped short of the stanchion because they feared getting banged on the head. Some animals poked their head toward the stanchion and then quickly jerked it back before the operator could close the stanchion around their neck. They acted like sissy swimmers who put a toe in the cold water and then jerk it out.

  Over the five-month period the animals grew too large for the manually operated chute, so they were taken to a hydraulically operated squeeze chute for the fifth and final blood test. The hydraulic chute was painted a different color and looked somewhat different from the manually powered squeeze chute. Likewise, the alleys and corrals leading up to the hydraulic chute were totally different. When the cattle approached the hydraulic squeeze chute, many of them balked and refused to put their heads into the stanchion. They recognized the squeeze chute in spite of the different design and new location. They had generalized their knowledge of squeeze chutes and stanchions to a new place.

  Cattle I have worked with have had the ability to apply previously learned skills to new situations, which also indicates a capacity for thought. Cattle with large horns, such as Texas long-horns, have good spatial sense and will turn their heads to walk up a thirty-inch-wide truck loading ramp. But young cattle that have had no prior experience with narrow chutes and ramps will hit their horns on the entrance and be unable to enter. Turning the head to pass through a narrow place is not governed by instinct. Experienced animals learn to turn their heads. After they have learned, they will turn their heads before they enter a chute they have never seen before. When an experienced animal approaches the chute entrance, he turns his head and enters effortlessly.

  Some very elegant research with birds has shown that even our feathered friends can think. Herb Terrace, the famous chimpanzee trainer, trained pigeons to peck at a series of lighted buttons in a specific order to obtain food. The task was designed to make it impossible for the pigeon to use a simple rule of thumb such as “red light equals food. ” All of the experiments were conducted in an enclosed box and controlled by a computer to insure that the pigeons did not receive cues from the trainer. (Whenever animal thinking is being evaluated, the “Clever Hans effect ” must be taken into account. Hans was a famous horse that had been trained to count by tapping his hoof. Many people were very impressed and thought the horse really could count. Hans did not know how to count, but he was a very perceptive horse who picked up subtle cues from his trainer.) Terrace designed a whole series of trials to show that the pigeons could apply previously learned knowledge about the button order to new button-pushing problems.

  Irene Pepperberg has slowly and laboriously taught an African gray parrot named Alex to use language beyond mere repetition by having him watch two people talking to each other. One person would hold up an object such as a cork and ask, “What is this? ” If the second person gave the correct name for the cork, she would be praised by the first person and given the cork. However, if the second person gave the wrong name for the object, she was told “no ” very firmly After Alex watched many of these conversations, he started to use words in appropriate ways. Each small step was mastered before he went on to the next step.

  For a reward, the parrot would be given the object. He had to learn that the correct word could get him things he wanted. People teaching language to severely autistic children use similar methods. The Lovaas language-teaching method requires seeing the object, hearing the word, and pairing the word with both the object and reward. After a child learns the objects, he is given pictures of objects. For some children severely afflicted with autism, relating to such pictures is difficult.

  More evidence to support the idea of animal thought can be found in Benjamin Beck 's extensive review of the published scientific literature. While it is well known that monkeys and chimpanzees can use tools, Beck found many reports of tool use in birds and nonprimate mammals. Tool use is another sign that animals can really think. Elephants will push uprooted trees onto electric fences to break them, and one elephant even used a bamboo stake to scrape off a leech. Eskimo lore is full of accounts of polar bears throwing chunks of ice at seals. I have watched seagulls carry shellfish up over the roof of a steel boathouse and then drop them to break them open. The gulls also dropped clams on the road and waited for cars to run over them, exposing the tasty morsels. Beck 's review of the literature indicated that birds can learn tool use by observation. When one bluejay in a captive colony had learned to use reaching tools, five other jays also learned. A Galapagos finch that does not usually use sticks for probing learned to use them after observing another species of bird using this tool.

  At the University of Illinois farm where I worked as a graduate student, the pigs in one pen learned to unscrew the bolts that held the fence to the wall. As fast as I could screw the bolts back
in, their little tongues were unscrewing them. All five pigs in that pen learned to unscrew bolts. My aunt had a horse that learned to put its head through a gate to lift it off the hinges; and at every large cattle feedlot, there are always one or two cattle who rival the techniques of the great escape artists among us. One time I witnessed a twelve-hundred-pound crossbred Brahman steer jump six six-foot gates. He just levitated over them. A horse has to run to jump a gate, but this big Brahman rose up like a leaping whale and effortlessly cleared the top of the gates. The vast majority of cattle are content to stay in the pens and don 't try to get out, but a bull that has learned how to break barbed-wire fences is impossible to keep in, because he has learned that he will not get cut if he presses against the posts. Fences only work because cattle do not know that they can break them.

  Dolphins at the University of Hawaii are being taught to understand symbolic sign language. Initial training is conducted by a person who makes hand signals that represent a simple sequence of commands. After the dolphin learns how to do a series of these tasks with a person, the next step is to have it look at a videotape of the person. This helps to prevent the Clever Hans effect. The simple command sentences are rearranged into hundreds of different combinations so the dolphins cannot memorize a set routine. Dolphins can easily transfer instructions from a real person to a videotape of the person. A third step further prevents possible cuing from the trainer. The trainer is now dressed in black and videotaped against a black curtain. The only thing the dolphin can see is the trainer 's white gloves making the signs against a black backdrop. The dolphins are able to understand the videotaped hand signals, too. At this point, the images are more abstract, and the dolphins are taking the first steps toward understanding symbolic representation of words.

  My experience as a visual thinker with autism makes it clear to me that thought does not have to be verbal or sequential to be real. I considered my thoughts to be real long before I learned that there was a difference between visual and verbal thinkers. I am not saying that animals and normal humans and autistics think alike. But I do believe that recognizing different capacities and kinds of thought and expression can lead to greater connectedness and understanding. Science is just beginning to prove what little old ladies in tennis shoes have always known: little Fifi really does think.

  Bird Savants

  The ability of birds to migrate is based on capabilities that resemble savant skills. It is possible that savant skills are part of an older memory-imaging system that is masked by higher thinking skills. Professor Floriano Papi, in Italy, has written an important book, titled Animal Homing, on the abilities of animals and birds to migrate and home. Since the ancient Romans, carrier pigeons have been used to carry messages. How does a pigeon find its way home after it has been taken far away in a cage?

  Birds navigate by using a combination of an innate sense that enables them to detect the earth 's magnetic field and memories they have acquired. In some birds, the innate magnetic detection system is coupled with genetic programming that forms the basis of an instinct to migrate. This will get the bird headed in the right general direction, but information from memory is also essential for accurate homing and migration. If a young bird migrates with its flockmates, it simply learns visual landmarks and other information, such as constellations and orientation of the sun. Some birds, such as the European teal, can distinguish and memorize the constellations. Papi reports that some birds can make visual calibrations of constellations, correcting for the earth 's rotation during different times of the year, which doesn 't seem all that different from the intense savantlike visual memory.

  Clara Parks, whose autistic daughter has great artistic talents, noted that when her daughter painted a picture of their house, the constellations she included were very accurate. Mrs. Parks has commented that her daughter 's eye is like a camera. Possibly, her visual skill and birds' navigational skills have similarities. This explains migration, but it fails to explain how a carrier pigeon can find its way home over a landscape it has never seen before. The pigeons rely on visual landmarks when they fly over familiar territory, but when they fly over unknown territory, they rely on smell. When a pigeon is transported from its home loft to the release point, it remembers smells along the way, and it uses these smell cues to get back home. Pigeons deprived of their sense of smell will become lost. Those with their sense of smell intact will also get lost if they are transported in a container that blocks smell. It appears that visual landmarks are the preferred method of homing, but a bird will switch gears and use olfactory cues when it finds itself over strange territory where familiar visual landmarks are absent. It may be using “smell pictures. ”

  A fairly high percentage of people with autism have a very acute sense of smell and become overwhelmed by strong odors. I am embarrassed to admit it, but when I was a young child, I liked to sniff people like a dog. The scents of different people were interesting. Some animals have highly developed senses which are more acute than ours. Bloodhounds can track a fugitive for miles by smell, and predatory birds have greater visual acuity than humans. Many animals have very sensitive hearing and can hear high-frequency noises that are out of the range of human hearing. Many people with autism share these hyperacute senses. They are unable to concentrate in the classroom because they can hear talking in three other rooms. I have often observed that the senses of some people with autism resemble the acute senses of animals.

  Emotions in Farm Animals

  The manager of a very large swine farm once asked me in all seriousness, “Do pigs have emotions? ” To him, pigs were simply pork-producing entities. We have seen that their ability to think and learn exceeds conditioned stimulus response, but do they experience true emotions? Are the feelings of a sow defending her piglets or an antelope running in fear from a lion similar to feelings in people under similar circumstances? Even a chicken can be highly motivated; Ian Duncan, at the University of Guelph, found that a hen would push open a very heavy door to reach a nest box, though she was not motivated to push open a lightweight door to reach a rooster. Is this behavior driven by emotion?

  Early in my career I befriended two pet steers at the Kelly feedlot in Maricopa, Arizona, while I was doing a photography assignment for a company that made meat packaging equipment. The advertising agency wanted a photo of a great majestic Angus steer against the blue Arizona sky. To get the picture I had to lie down on the ground and wait for the cattle to come up to me. Cattle are less afraid of people when they reduce their size by kneeling or lying down. These two black steers let me touch them, and by the end of the afternoon they would allow me to pet them. At first they seemed to be afraid, but then they started to like it. They stretched out their necks to get stroked under the chin.

  About two weeks later I returned to the feedlot, and I wanted to see if the steers would remember me. I stopped my truck in front of the pen, and the black steers immediately ran over to the fence and stuck their heads out to be petted. They wanted to be petted even though I did not offer them food. They simply wanted to be stroked.

  There are many other examples of both farm animals and wild animals seeking pleasurable contact with people. Sows that have become pets will turn their bellies toward people so the people will scratch them. At one farm, a pet sow would squeal and become agitated if people walked by and failed to stop and rub her belly. When they stopped and rubbed, she would lie down, stretch out, and appear to be in bliss. Rhinos in a game park in Texas also solicited petting. When people walked up to their enclosure, one fellow would push his body up against the fence so that visitors could rub a soft spot where his rear leg joined his body. After he was petted and fed a few oranges, he would run along the fence and jump up and down like a calf on a spring day. To me, he appeared to be happy.

  To the scientist who wants objective data, these anecdotes do not prove that animals have emotions. But scientists have proved that laboratory rats are capable of recognizing a familiar person and seeking him o
ut. Psychologist Hank Davis found that lab rats will bond with a person who has petted, handled, and fed them. When a rat is placed on a table between a familiar caretaker and a stranger, it will investigate both of them and choose the familiar person most of the time. In most mammals and birds, the young will become very upset when they are separated from their mother. When calves are weaned, both the cows and the calves bellow for about twenty-four hours. Some calves bellow until they are hoarse.

  Cattle will also bellow for departed penmates. This is most likely to occur with Holsteins, which are very calm cattle. Their social behavior is easy to observe because the presence of an observer is not likely to disturb them. I have seen Holstein steers bellowing to penmates that were departing in a truck. The cattle that were left behind watched as their fat penmates walked up the ramp to get on the truck that would take them to Burgerland. Two steers stared at the truck as it turned out of the parking lot. One stretched out his neck and bellowed at the truck, and his penmate on the truck bellowed back. The nice feedlot manager was worried that his cattle knew they were going to die. They had no way of knowing this; they just didn 't like being separated from their buddies. Research by Joe Stookey and his colleagues at the University of Saskatchewan confirms that cattle do not like being alone; the cattle in their study would stand more quietly during weighing on a scale if they could see another animal in front of them.