An Intelligent Dog?

I’ve been reading John Pilley’s book about his dog Chaser, who knew the names of 1,022 different objects and who could learn the name of a new object by hearing it just once. What I’m particularly enjoying about the book is the frequent reminder that Chaser’s abilities are not all that remarkable for a Border Collie.

I’m surprised that this is the bit that I’m enjoying because the claim that “Border Collies are so intelligent” often reveals what we value as intellingence in dogs (“trainability”) rather than their innate skills.

Pilley is certainly impressed by Chaser’s ability to learn words, but he also offers examples of her independent problem-solving abilities (bouncing a ball down a flight of stairs and persuading her person to toss it back up so that she can repeat the process) or the way she guides her own learning in working sheep.

But I think what Pilley ultimately reveals is something more profound than “Collies are clever.” There’s a far deeper and more powerful reflection on the learning of working dogs that, when we get it right, can bring so much more to the learning of dogs we consider companions. Let’s take a look at Collies for now, and then we’ll see how the claim Pilley makes can be broadened out to include other dogs.

Learning: More than Just Science

Here’s the bit that made me catch my breath. Pilley thinks about what makes Chaser such a good learner and in this moment, instead of merely deferring to the scientists, this is what he writes:

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. […] the knowledge of Border collies’ intelligence that those who live and work with them every day have accumulated goes far beyond anecdotal evidence.

Pilley himself was a scientist – a behavioural psychologist—so his valuing of the experience of those who live and work with Collies is significant. Too often, we can be led to believe that science and experience are in conflict with each other; Pilley insists, in fact, that they are often complementary. And it surely follows that such a richness of “understanding and insight that complements animal science” is found among those who work with dogs in other domains.

Beyond Training

Now, I haven’t read much, I have to confess, about dogs engaged in other types of work, but over and over again in written nonfictional accounts of Border Collies I’ve encountered extraordinary feats that couldn’t have been “achieved” by training alone: leading people to trapped or injured people or animals; finding their way home across tens of miles; alerting people to dangers; identifying and fetching individual sheep from a flock of hundreds.

Here’s one of my favourite examples: the nineteenth-century poet and shepherd James Hogg writes about a Scottish butcher who frequently sent his bitch on long unaccompanied journeys to round up sheep. When the butcher returned after once sending the bitch on a five-mile wild hill route, neither dog nor sheep were where he expected. He set out in search of them, and found the bitch driving the sheep while carrying a new-born pup; she had given birth on the trail and managed to keep the sheep together. Once the sheep were safely folded, she returned to retrieve her pups one by one and bring them to their nest; all but the last pup survived.

Legends or Learning?

I had previously regarded these stories as legends, possibly exaggerated for effect. That is, until Pilley spelled out the difference between “trick training” and “learning” for me in a way that helped me see that these stories are entirely plausible events. Pilley gives the example of a circus act in which a dog is trained through shaping to jump through hoops; the outcome here, he notes, is already decided:

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 […].

Pilley refers to this kind of learning as closed-ended – like learning multiplication tables by heart (4×4=16, always). Open-ended learning, however, is that to which shepherds repeatedly expose their dogs: real situations of increasing difficulty with likely unpredictable and variable events (six sheep in a pen at 5am on a balmy Tuesday will not behave identically to those same six sheep in a pen at 7pm on a rainy Friday).

And here’s where we can expand outwards from Collies to think about dogs who have to work in an unpredictable “real world”: war dogs and assistance dogs, for example; dogs who have to process, think fast, and respond in new ways to unforeseen events.

So What Do We Get Right?

From experience, I know it’s often easier to slip into the learning-by-rote mindset and to think of a single carefully mapped process with a clear outcome. But that’s not real life, and it’s not where the greatest rewards of learning are to be found. When creating a curriculum for our dogs to learn, what we get right is offering them space to draw on that learning, their history, their own innate knowledge, and their previously acquired skills to demonstrate independent thinking and creative application. Instead of asking a dog merely to place a ring on a hook, for example, we might offer them increasingly challenging (but still achievable) opportunities to develop their skills by inviting them pick up and place a variety of objects into or onto a variety of surfaces or containers. If the challenge is at the right level, and if they have achieved ample success, then we will be able to observe their self-teaching as they test and adopt solutions to the puzzles with which they’re faced. As Kay Laurence notes in Build the Learning

Much of the learning at the micro level is internal: how we change the angle we hold a pen, our point of balance on a bicycle, the sensation of moving through water. […] It is beneficial for the learner to discover these processes and be able to call upon them as solutions for other situations.

And Where Do We Go From Here?

The key, I think, is to see our dogs’ learning as something that is ongoing and constructive. “Training” is often thought of as “teaching dog to perform x action or movement.” “Learning” is about the skills that they can acquire, the opportunities they have to develop those skills laterally, to apply them to new situations, to refine them, to combine them with other skills; this is of so much more value to the individual dog. A dog who jumps through a hoop on cue may be able to jump through a hoop on cue, but a dog who has learned perceptive distance, who has developed proprioceptive skills, who can move their body with ease and precision, who can take off and land well, and who can assess, adjust, and adapt to changes in that environment is a dog who has skills that they carry with them across learning. This, in turn, provides a foundation for so much else, and transfers into other areas of that dog’s life. And this is what we can learn from remarkable people who have worked with remarkable dogs: training and learning are not the same thing at all.