On the basis of my successful effort to quit smoking, I concluded that as I had to quit smoking, Chaser needed a little negative reinforcement, as well as lots of positive reinforcement, to quit chasing cars. Negative reinforcement, often in an implicit form, is usually part of healthy boundary setting. The question was exactly how to provide this.
The experience of negative reinforcement depends on an individual’s temperament and personality. Some individuals are from birth thicker-skinned or more hardheaded than others. Trainers often talk about hard dogs and soft dogs. Yasha was a hard dog, and in his early obedience training he needed a fairly strong, but still gentle, hand. Grindle was a soft dog. Observation over the last month of Chaser’s equally soft temperament suggested that a little negative reinforcement should go a long way with her.
Before she chased the Jeep, I had already started to use some very mild negative reinforcement by adding a firm “No!” just before she ran to the end of her leash while chasing something. The idea was to associate the “No!” with the unpleasantness of hitting the end of the leash, so that the word would eventually stop her from pulling and lunging on the leash, and also generalize to become a stop signal for other unwanted behavior. But unfortunately the jerks she was giving herself were not sufficiently aversive to give the “No!” real strength. Trainers often suggest that when dog creates a jerk on the leash, we should intensify it enough so that the dog really doesn’t like it. Otherwise the dog will make a judgment such as, “If I lunge after the cat, I will get a tug. But it’s really not that bad, and it’s worth it because I love to see that cat run.”
I finally made the tough love decision that to enhance Chaser’s immediate responses to my commands, I would take her into the front yard on a fifteen-foot line, two and a half times as long as her normal leash. The long line would let her build up more speed in chasing something she shouldn’t, and she would then get more of a jerk when she ran to the end of the line. I also planned on adding a firm tug of my own to the long line to heighten the jerk.
When we went outside, however, I couldn’t bring myself to take that step. It seemed too abrupt. Instead, I took her into the backyard to rehearse the major obedience commands with no possible distractions from passing cars—and using only positive reinforcement. I made sure Chaser knew that I had treats in my pocket. Without positioning her, I wandered around the yard. Every minute or two, I randomly gave one of the obedience commands—there, stand, drop, crawl, here—in combination with “Chaser!”
Every time she heard her new name, Chaser instantly looked at me and I paused only a second before giving an obedience command. I smiled as she eagerly responded to my words. I lavished yummy treats, praise, and pets upon her with enthusiasm, going through several repetitions for each of the obedience commands.
A few hours later, I again took Chaser into the backyard and ran through the same exercises, again using only positive reinforcement. At the end of the session, I switched over to the obedience herding game, using a tennis ball as a surrogate for a sheep. In the course of the game I gave Chaser the obedience commands relative to the tennis ball, hoping that the herding scenario would cause the obedience commands to take on even more positive value. The success of these training and play sessions made me think that maybe we could still avoid using negative reinforcement beyond “No!”
That evening Sally and I were sitting on the front porch with Chaser lying quietly between us when a car went by, not as fast as the Jeep but still moving quickly. Chaser instantly got to her feet and barked at the car. I grabbed her collar to prevent her from scrambling off the porch in pursuit, and we took her inside.
We had to do something more. But it had to be something that wouldn’t throw her off stride in other ways. One firm jerk on the long line would surely be okay, given our good relationship with Chaser and her abundant confidence. But having to repeat that a lot worried me. Stroking Chaser, Sally said, “If the momma dog was here she’d probably get you to stop doing something she didn’t like with one little shake on the scruff of your neck.”
“That’s brilliant, Sally! That’s what we’ll do,” I said.
Puppies are biologically prepared to learn that a behavior is not allowed when the mother dog shakes them by the scruff of the neck. We knew from Wayne West that he relied on the mother dog’s doing that to teach young puppies not to defecate and urinate in the kennel. Chaser had just turned twelve weeks old and was still at her most impressionable. If I did things right, one lesson about going into the street might be enough.
The next morning I took Chaser into the front yard on the long line. Luck was with me and it wasn’t long before a car came by. Chaser immediately headed after it. As she was building up speed I said, “Chaser! Here!” She kept going. Just before she got to the end of the fifteen-foot line, I roared, “No!” When she hit the end of the line, I gave a firm tug back on it. And then I ran to her, picked her up by the scruff of the neck, and gave her one vigorous, mother-dog-like shake.
I gently put her down on the grass, moved a few feet away, and in a warm, soothing tone said, “Chaser! Here!” She came immediately to be comforted and given a treat.
Over the rest of the day’s normal routine of play, obedience training, and walks, Sally and I found that the “No” was enough to stop Chaser in her tracks if she ran after a squirrel or headed into the road for any reason. By the end of the day the “No” did not even have to be shouted, although we always made sure it was firm and clear and loud enough to be heard.
There was just one more test we needed to complete Chaser’s don’t-go-into-the-road training. The following morning we all trooped into the front yard—with no long line or leash. I told Sally I wanted her to walk into the road, luring Chaser to follow her. Sally doubted that Chaser could be tempted from my side, but she obliged me and walked toward the road. Curious, Chaser followed her. As Chaser neared the edge of our yard I said, “Chaser, here!” She kept going, and worry rose up in my head. But just before she got to the road, I said, “No!” Chaser stopped instantly and looked back to me. I called out, “Chaser, here,” and she came to me without hesitation.
I gave her all the positive reinforcement I could muster, and Sally quickly joined us and did the same. A little later we switched roles, and I went into the road while Sally gave Chaser her commands. Again we found that “No” now stopped Chaser from going into the road.
That was the turning point. Over the following week, we strengthened the power of “No” through repetition. The word had now acquired aversive meaning for Chaser. And once “No” took on a quality Chaser didn’t like, it became something she really didn’t want to hear. With consistent use, the “No” would lose its negative emotional quality, but still remain effective as a command cue and a directive piece of information. We could even use “Yes” and “No” as substitutes for “Hot” and “Cold,” when I asked Chaser to find a hidden object.
An additional tactic we used to teach Chaser not to chase the wrong things was to positively reward a response that was incompatible with chasing. When we saw or heard a car coming during a walk, we pointed to the grass and said, “Car. Go to grass.” And then we ran with her to the grass, had her lie down, and showered her with praise and pets until the car passed. At that point we softly said, “No car” or “Okay,” and continued on our walk.
Within a few days, Chaser chose to leave the road and lie down in the grass whenever she heard a car coming, usually well before we did. We gave her lavish positive reinforcement the first times she did this. Soon our positive reinforcement could be more moderate without the behavior weakening. In fact, if we forgot to say, “Okay, Chaser, c’mon,” after a car passed, she remained lying in the grass until we did. Sally was the first to experience this. Lost in conversation with the Ya-Yas, Sally suddenly realized Chaser was not with her. Spinning around, Sally saw that our bright little puppy was still patiently lying in the grass at the side of the road, waiting for the “Okay” command.
&
nbsp; Thanks to the combination of these techniques, Chaser would not go into the street even to follow her favorite ball, unless I went into the street and asked her to join me there. This learning held in relation to any road anywhere we took her after that, from Spartanburg and environs to visiting Debbie, Jay, and Aidan in Brooklyn.
Now I could concentrate on channeling Chaser’s powerful social and herding instincts into her language learning. When it comes to harnessing the power of Border collies’ instinctive drives, we can learn a lot by listening to the farmers and trainers who work with them every day.
7
Listening to the Farmer
THERE WAS A lot to be thankful for as Chaser turned four months old. The danger of the feral cat had passed, she had learned to be safe from cars, and most of all she had bonded strongly with Sally and me.
We positively reinforced Chaser’s bond with us in every way we could. But reinforcement is definitely a two-way street, and Chaser was positively reinforcing us as much or more. Just as a smile elicits a smile, Chaser’s playfulness stirred the same in Sally and me. Her wagging tail sparked joy in us to match her own.
Sally and I had been looking forward to having a dog in the family again. But we weren’t prepared for how strongly Chaser affected us. We didn’t realize how big the void in our lives had become.
We went puppy mad. We e-mailed dozens of pictures of Chaser to Debbie—Robin’s e-mail connection in the mountains was too slow for that—and our conversations with both Robin and Debbie were full of details about our new family member. Debbie joshed with us on the phone one day, “I get it. She’s cute.” But like Robin, Debbie listened patiently as we waxed on and on about Chaser.
Perhaps the impact was even greater on me than on Sally. My mind was buzzing with curiosity about Chaser’s potential to learn the meaning of words. Eight years after my retirement from Wofford I was fully a scientist again, entranced by the possibility of discovering something worth sharing with the world. Sally told me more than once during our first month with Chaser, “You’ve needed this, John.”
It went even deeper than my desire to contribute to scientific understanding. The day we brought Chaser home, science became secondary to her quality of life. Not that there was any conflict between the two. The more Chaser fulfilled herself as a dog, a Border collie, and a member of our family, the more she was likely to be able to learn. But if a conflict between Chaser’s needs and science ever came about, Chaser had priority.
Later in the conversation in which she joked that she’d seen more than enough pictures of Chaser, Debbie said, “I think she’s given you your heart back, Dad.”
I couldn’t say anything. A few seconds later, obviously concerned about how I might be reacting at the other end of the line, Debbie broke the silence and said, “It just feels like ever since Yasha you’ve been holding something in. And now that part of you is breathing again.”
I still couldn’t say much. But I managed to voice my agreement and my thanks for her, Robin’s, and Sally’s love and understanding. The black box into which I had locked my grief over Yasha had opened without my realizing it. My sorrow over his loss and my lingering sense of guilt for his suffering at the very end were gone. Yet his vital spirit was still with me, stronger than ever, now that Chaser was a member of our family. My friend who told me, “When you get a pet, sooner or later you get a broken heart,” also told me, “Your heart gets whole when you can risk its being broken again.”
I hadn’t been able to take the first step onto that path of emotional healing, but Sally had taken it for me. And then Chaser, greeting me every morning with a “What are we going to do today, Pop-Pop?” tail wag, had done the rest. Her arrival in our family was a blessing, for sure.
My scientific hopes for teaching Chaser language grew larger the more she displayed strong Border collie traits and instincts. Dogs of all breeds and mixtures of breeds can be highly intelligent, but no dogs have shown greater skill and creativity at problem solving than Border collies. The stewards of Border collie intelligence over the generations—farmers, breeders, and trainers—have amassed a body of understanding and insight that complements animal science and offers many clues for researchers to follow and test.
Scientists are trained to be skeptical about so-called anecdotal evidence, meaning observations from experience that haven’t been confirmed in rigorous experiments. However, the knowledge of Border collies’ intelligence that those who live and work with them every day have accumulated goes far beyond anecdotal evidence. One of the greatest animal scientists of recent times, John Staddon, the James B. Duke Professor of Psychology and Professor of Biology and Neuroscience Emeritus at Duke University, believes that the knowledge of farmers and shepherds provides compelling evidence of dogs’ ability to make inferences and solve complex problems without human direction. In Staddon’s view the exploits of Border collies epitomize creative learning as opposed to rote learning.
In a 2006 essay for a special issue of the International Journal of Psychology, Staddon wrote, “There are two methods to train a dog”—and by extension a person or any other animal. The method that is “the quickest, least dependent on individual aptitude . . . is called ‘shaping by successive approximations.’ It is the method used by circus trainers. . . . It is effective and reliable, especially if what is to be taught is well-defined and predictable.”
For example, a dog act in a circus might involve dogs jumping through hoops. Shaping by successive approximations begins with getting the dogs to walk through hoops that are stood up on edge on the ground or raised up only a few inches. Gradually the trainer brings the dogs to the point where they leap from one tall platform to another through hoops of blazing fire, perfectly on cue with the gestures of the trainer and any music or sound effects.
These are amazing feats of training, athletic skill, bravery, and, it should be said, intelligence. A circus dog must be a good learner and periodically be able to learn new tricks. It takes a strong bond of trust between trainer and dogs, and dedicated effort by all of them, to put on a great show. But the outcome, if all goes well, is predetermined. The dogs will never be asked to solve a problem on the spur of the moment by applying their instincts, life experiences, and reasoning powers to draw inferences. The trainer and the dogs know exactly what to expect during every microsecond of the act. There is no uncertainty or surprise, except for those in the audience who haven’t seen the performance before.
Shaping by successive approximations is essentially closed-ended. Applied to humans, Staddon says, this method “is the basis for regarding teaching [children] as training in a ‘skill,’ like a trick to be taught to an animal. It treats students like dogs, and pretty dim ones at that—Odie rather than Lassie.” This is what is happening today in schools where teachers are regimented into “teaching to the test,” and where students spend more classroom time taking practice standardized tests than they do interacting with their teachers and classmates on curriculum content.
By contrast, “creative teaching” is required for teaching things that are less well defined and do not have a predetermined outcome. In the case of humans and animals alike, Staddon argues, this involves creating an environment that challenges learners to use all their “natural propensities”—instincts, fundamental drives, emotional energy, perceptual and cognitive abilities—to solve new problems.
It’s important to realize that nature is the greatest creative teacher of all. It’s equally important for us twenty-first-century humans to realize that we are not nature’s only brilliant students and that we are not its only thinking species. Over the past decades, naturalists in the field have discovered attributes, abilities, and behaviors once thought to be “uniquely human”—tool use, “cultural” differences between groups that belong to the same species, complex emotions such as empathy, inferential reasoning, among other characteristics—in animals such as bonobos, chimpanzees, crows, dolphins, and not least of all domestic dogs.
Res
earchers are confirming and probing these characteristics both in laboratories and in animal sanctuaries that seek to replicate wild animals’ natural habitats. For example, Brian Hare and Juliane Kaminski have separately devised and conducted experiments with bonobos, chimpanzees, and dogs strongly indicating that all of these animals have an implicit theory of mind comparable to that of young children when they acquire language. Hare’s book The Genius of Dogs, written with his wife and co-investigator Vanessa Woods, offers a fascinating guide to their own work and that of many other researchers on the perceptual and cognitive traits humans share with animals. Much of this work draws on techniques first developed in behavioral and cognitive psychology experiments, with behavioral psychologists such as John Staddon pioneering a broader understanding of the role of inference in nearly all animal learning. The ability to acquire language, I am confident, will sooner or later be recognized as one more capacity that is no longer “uniquely human.”
Staddon reaches the same conclusion that I had heard from Border collie breeders and trainers: creative teaching builds on the learner’s instincts and innate tendencies. Instead of teaching single, repeatable behaviors in isolation or in a series, this method of teaching and training encourages creative learning, the ability to solve unexpected problems through spontaneous trial-and-error inference guided by accumulating experience and judgment. Creative learning is essentially open-ended. It is also highly dependent on the relationship between the learner and the teacher or trainer.
Staddon offers two examples. One is of a teacher encouraging a child’s interest in scientific discovery. The other is of a shepherd teaching a Border collie to herd sheep. Staddon quotes the early-nineteenth-century Scottish “shepherd poet” James Hogg on training the dog Sirrah by putting him in situations where his herding instincts and propensities enabled him to solve problems of increasing difficulty: “[Sirrah] would try everywhere . . . till he found out what I wanted” and in so doing would display “a great share of reasoning.”
Chaser_Unlocking the Genius of the Dog Who Knows a Thousand Words Page 9