Chaser (21 page)

All was well in our world. In her own distinctive way Chaser was doing what she was born to do. And I was a proud and happy daddy watching her.

When the sheep were ten or fifteen feet away from me I said, “Chaser, that'll do.” She trotted forward out of the little flock and came to my side, and I praised her for her good work. Without Chaser to push and accompany them in my direction, the sheep stopped and milled around where they were. It was a little after two o'clock now, and I thought we'd move the sheep in another direction for a bit before heading home.

A black and white shape streaked by me on the right. It was the unfamiliar dog from the kennel. Somehow she had gotten loose and then jumped the pasture fence.

I called out, “There,” the command to stop, but the dog ignored me. She got behind the sheep and began rushing them toward Chaser and me in the corner. “There,” I called out again in a firm voice, but the loose dog kept moving the sheep toward us in a corner. She too was doing what she was born to do, move the sheep to the farmer, even if she wasn't going to give her ear to that farmer.

The kennel escapee nipped one of the sheep in the heel, and they all rushed away from her a little faster. That would have been fine with me, except for Chaser's being by my side. If the sheep crowded in on us, I would be okay. But Chaser might get knocked down and trampled. She could wind up with a broken leg or worse, much worse.

The sheep were almost on top of us. There was no point in giving the other dog more commands she wouldn't heed. I took off along the fence line for the gate, running as fast as I could. Chaser stayed right with me, although she could have raced ahead or peeled off into the center of the field.

The mostly black dog responded by pushing the sheep even harder behind us. It was a miniature stampede. But its consequences could be full-size.

I kept running, and Chaser kept matching my pace. There was the gate, thirty feet ahead.

I glanced back. The sheep were gaining on us.

I sprinted harder, my lungs beginning to burn. Ten feet from the gate I slipped and almost went down. But I managed to hold my balance and keep going.

The sheep were only a few feet behind us now.

Finally here was the gate. I flung up the latch, opened the gate a few inches, and squeezed through it with Chaser. I shut the gate just as the sheep galloped past.

Perplexed, the mostly black dog stopped running at the gate. She clearly couldn't understand why I wasn't holding the gate open for her to drive the sheep through.

I leaned on the gate and caught my breath.

Chaser was fine. She was ready to work the sheep again, in fact. But I figured our close call was enough for one day.

For her part, the other dog was now lying on her belly on the other side of the little flock, holding them by the gate as she assumed the shepherd—me—wanted. I opened the gate and went into the pen, leaving Chaser outside it. The other dog still wasn't willing to heed my commands, but she let me come up to her, and I took off my belt to use as a leash. With that she walked dutifully beside me out of the pen and back to the kennel.

Up close I could see that she was a young dog, perhaps around Chaser's age. Although the pattern wasn't identical, the kennel escapee's mostly black coat with white patches struck me as a reverse image of Chaser's mostly white coat with black patches. I thought they would make a great picture if they were posed side by side.

There was nothing wrong that I could see with the fence around the kennel yard. The dog must have jumped it, driven by the frustration of seeing Chaser go into the sheep pasture to do what she was longing to do. When I put her back inside the kennel yard, she seemed willing to stay there. I was about to put her into the kennel proper, where she couldn't possibly get out, when Wayne drove up in his truck.

Hearing about the dog's behavior, Wayne shook his head and then apologized. With a wry smile he added, “I reckon we could chalk this up to sibling rivalry.”

To my astonishment he explained that the escapee, Kate, was Chaser's littermate. One of Wayne's daughters, Sandy, had taken Kate as an eight-week-old puppy, only to find that she was a nipper. With a young baby in her family, Sandy wasn't able to work with Kate. Wayne had worked with Kate himself, and he had recently sent her to David Johnson for additional training. Kate promised to be a good herding dog, and Wayne was looking for the right situation for her.

Sally was upset when I told her what had almost happened, concerned for me as well as Chaser. As we talked over the incident she realized that I had never been in any real danger. And we both laughed about Kate's being Chaser's sister, recalling some of the more raucous moments between Robin and Debbie when they were young.

The question was what to do about Chaser. Should I stop taking her to Wayne's? By dinnertime Sally agreed that was too extreme a reaction. Kate's jumping the kennel fence was just one of those things that can happen now and then in the best of circumstances. But Sally made me promise to make sure all of Wayne's dogs were secure in their kennel and keep an extra sharp eye out for any escapees. That reminded us both of those great escape artists Blue and Timber, and we shared a few stories of their misadventures over dinner.

Three days later Chaser and I went back out to Wayne's. Kate stayed in the kennel, and Wayne soon found her a good home on a farm with horses and cattle. Since then Chaser and I have continued our regular visits to work Wayne's sheep without any mishaps. Thanks to Wayne, I have been able to keep my promise to Chaser when she was an eight-week-old puppy, that I would strive to help her fulfill herself as a Border collie. I love that she can herd sheep the way she was bred and born to do while continuing her progress in herding words.

In Chaser's second and third years, that meant a greater focus on common nouns, words that stand for categories, and matching another benchmark from the Rico study, learning by exclusion.

11

I
N CHASER'S SECOND
year I put more effort into teaching common nouns. My goal was to teach her two more common noun categories in addition to “toy”: “ball” and “Frisbee.”

To teach Chaser what a ball is I started with eight balls of different sizes, colors, and materials on the floor: a tennis ball, a racquetball, a baseball, a lacrosse ball, a golf ball, and big and small balls made of foam and spongelike stuff. The balls had many different characteristics, but they were all round and, to different degrees, bouncy. To enable Chaser to start off with errorless learning, there were no other objects on the floor. So when I said, “Chaser, find a ball,” there was no way she could choose an incorrect object. I used the 8-of-8 test procedure I described earlier, in which I did not replace objects in the group after she picked them correctly. Instead I kept saying, “Chaser, find a ball,” until there were no balls left on the floor.

We did this several times off and on through the day. And then we went through the same procedure with another eight balls of different types, followed by another eight balls, and so on. If this process was successful, Chaser would learn to generalize that balls are round and bouncy. She would acquire an abstract concept of what a ball is, based on the common physical characteristics of all the balls she encountered.

After many repetitions of this over several weeks, I made the task more challenging by putting out eight balls and eight non-balls. She not only had to generalize what characterized a ball, but also had to discriminate a ball from a non-ball. I randomly asked her to retrieve balls and non-balls. If she retrieved a non-ball when she was asked to retrieve a ball, I softly told her, “No, Chaser. That is not a ball.” Informal tests soon demonstrated Chaser's ability to bring a ball, and only a ball, when I asked her to do so.

I used the same procedure for teaching her what a Frisbee is, using that brand name to apply to any throwable, catchable spinning disk or ring in her flock of objects, no matter what it was made of. First I put eight Frisbees on the floor, with no other objects in sight. After errorless retrieval of all eight, I tested her on other sets of eight Frisbees so that gradually she generalized the characteristics of a Frisbee across all the throwable, catchable disks in her flock. And then I put eight Frisbees on the floor with eight non-Frisbees, so that she would have to discriminate successfully among the Frisbees and non-Frisbees.

As I mentioned earlier, Chaser learned the concept of “toy” not on physical characteristics, but based on an abstract functional characteristic. Toys were objects she knew by their individual names and could play with. Everything else was a non-toy. It was fascinating that she grasped the higher-level, more abstract concept of “toy” first.

As Chaser's learning continued through the fall of 2006, I was of two minds. All the experiments Chaser and I had conducted so far—proper noun learning, independent meanings for two elements of syntax, and common noun learning—showed that her language abilities reached far beyond those documented for Rico. The evidence was piling up for a major scientific paper in a peer-reviewed journal.

Part of me was itching to document Chaser's achievement for its own sake, as a contribution to the understanding of learning, and as a stimulus to other scholars. The other part of me didn't want to slow down research in order to write such a paper, which would limit the time I could spend with Chaser day in and day out. Instead, I wanted to press ahead and extend Chaser's language training. Her ability to take an object in her mouth, or nose or paw it, on command showed that she could respond correctly to sentences with two elements of grammar: a verb and a direct object. In the early 1980s Louis Herman and his colleagues at the University of Hawaii studied the ability of two bottlenose dolphins, Phoenix and Akeakamai, to respond correctly to sentences containing two, three, four, and five elements of grammar, including verbs, direct objects, indirect objects, prepositions, and adjectives and adverbs. Understanding three elements of grammar sentences was the next step for Chaser in my mind.

Working alone, I had to make a choice. If I didn't write a paper now, who knew what might prevent me from doing so in the future? Ultimately I couldn't take the chance of not documenting Chaser's learning as soon as I reasonably could. It would be a terrible waste if her learning wasn't shared with others.

More important than any concern I might have about being able to write a paper later, however, was my desire to honor Chaser's achievements. By this point she had achieved nearly all of my initial research goals for her, matching or exceeding Rico's learning and doing so in a way that met the objections that the Yale research psychologist Paul Bloom and two other childhood language learning researchers, Stanford's Ellen Markman and Maxim Abelev, had raised about the Rico study. These objections all involved language understanding that one-to-three-year-old toddlers display but that had never been demonstrated or was subject to question in animals.

In assessing the Rico study, Bloom emphasized that children learn words by overhearing them, the implication being that animals never did this. Yet it seemed plain to me that Chaser had learned “sheep” and quite a few other words by overhearing them. All of my experiences with dogs over the years made me suspect that many dogs commonly understood the meanings of a dozen or more words they heard frequently, such as the obedience commands or the words “walk” and “treat.”

In addition to “Let's go to the sheep,” Chaser responded enthusiastically to “Let's go for a hike,” or “a walk” or “play Frisbee.” If Sally said, “I'm going over to Sue's” or “I'm going over to Nora's,” Chaser headed straight to the door. As soon as Sally opened the door, Chaser ran to the edge of the road and waited there for her.

There were also words she learned by overhearing that she preferred not to hear. If we said, “Chaser, let's go to the store,” she looked peeved and didn't want to come. Her ears and tail went down, and her usual grin turned into what looked like a slight frown. She came along reluctantly only if we pressed her. She had learned that going to the store meant sitting in the car waiting for one or both of us to return from shopping. Where was the fun in that?

Another thing Chaser didn't like was taking her monthly heartworm pill. If Sally said, “Chaser, come get your pill,” she slunk away to the bedroom and tried to avoid it. Sally got around that by saying, “Chaser, come get your pill and I'll give you a treat.” That brought Chaser quickstepping to swallow her pill dutifully and then enjoy a treat.

The real problem with overheard word learning by an animal was demonstrating it empirically. To show that an animal's exposure to a word was only by overhearing it, with no explicit teaching, would require a 24/7 visual and audio record of the animal's life to that point. I put that aspect of Chaser's learning to the side in my mind, unsure whether I should include it in a peer-reviewed paper.

Bloom also said that children are able to learn words by being shown an object and hearing a person name it, again with the implication that animals could not do this. Showing Chaser an object and naming it became my basic method for teaching her words. Once Chaser learned the concept that objects have names, as I described in chapter 9, my tests showed that she learned the names of objects on one trial when she was shown an object and told its name. When I tested retention after ten minutes and twenty-four hours, I found that she needed additional rehearsal to process the learning into long-term memory. But toddlers also need rehearsal to lodge new words into long-term memory.

Another of Bloom's reservations about Rico's learning was that “it is always Rico's owner who is communicating with him. . . . Yet, if Rico really is learning sound-meaning relations . . . it should not matter who the speaker is.” That seemed reasonable to me. As Chaser's word learning advanced, I brought other people in as trainers. Sally, Robin, and two Wofford undergraduates who assisted me at different times, Caroline Reid and Katie Grainger, all taught Chaser proper noun object names. I was also planning to do a set of formal blind and double-blind tests of Chaser's learning, in which others beside myself gave her the commands to perform language tasks.