Click and create
One subject that crops up frequently in training circles is the side effects of various training techniques. We caution against the harmful fallout of punishment-based methods. We debate the relative merits of luring, shaping, and capturing. We examine studies that compare the rate of behavior acquisition using various marker signals. And, of course, we love to talk about the added benefits of clicker training—engaging the dog's mind, the respondent conditioning of a positive emotional state, the fostering of creativity in both trainer and trainee.
Valenzia bowing on her bowl.
Creativity? Is that a good thing?
Several years ago, on a large training-community mailing list, one person posed a very honest question: "Why should I want my dog to be creative?" The question merits some consideration. Aside from the giddy elation that overcomes behavior nerds (like myself) when we see a training subject acquire a new skill or thought process, how useful or desirable is it to have a creative animal? After all, the more you teach the dog to problem-solve, the more trouble the dog can get into. If I train my dog to open doors, can't he escape the house, or steal food out of the refrigerator? If he learns to manipulate objects in his environment, won't he figure out how to open the crate door or knock the lid off the trash can?
Leaving aside for the moment the question of sloppy training (good stimulus control and proactive training will prevent most of these hypothetical problems), let's look at some of the advantages of creating a creative dog.
My blue Doberman, Lucrezia, was the most clicker-savvy dog I have ever owned. She was also by far the most fearful dog I have ever owned. Among her various phobias, Lucrezia was terrified of things being over her head. If a treat rolled under a table or chair, she would duck away in fear rather than put her nose in the shadow of the object to find the treat. She would eye the furniture warily, as if the chair would collapse on her head and devour her.
Around this time I was teaching training classes in a big-box store, and Lucrezia was my demo dog. One of the things we did to kill time on slow days was a variability exercise, or the "Show Me Something New" game, in which the dog offers a completely novel behavior for each successive click.
Lucrezia loved this game, and would invent all sorts of new ways to poke, touch, grab, bite, sit on, or move around an object. Since our training area included a row of chairs, I sometimes had her "show me something new" with a chair. One day, after exhausting all manner of nose-pokes and chin-rests and paw-whacks, Lucrezia came up with wrapping her paw around the metal chair leg. Click and treat. Next, she offered the new behavior of dragging the chair a few inches with her paw. Click and treat again; move on to the next trick. When we were finished with the session, I put the chair back in the row and set up for the next class.
Some time later, I tossed Lucrezia a piece of food that hit the concrete floor and rolled, ending up against the wall behind the row of chairs. When she saw that her treat was under a chair, Lucrezia balked and backed up, looking from me to the chair helplessly, too frightened to stretch her neck under the chair to get the treat. After a few moments of conflict, suddenly Lucrezia perked up. She walked up to the chair, hooked her paw around the leg, dragged the chair out of the way, and ate her treat. Then she pranced over to me, head high: "Look, Mommy, I outsmarted that scary dog-eating chair!"
I found this astonishing at the time. Now that I have a better grasp of behavior science, it blows me away that a young, fairly novice dog (she'd been with me about a year) was able to recall a one-repetition behavior and apply it in what was clearly a stressful situation.
Back to creativity. That particular behavior, grabbing and dragging the chair, was 100%-invented by the dog. I did not teach her to move the chair. Apart from the single click and treat during our game, I had never reinforced her for moving the chair. I had not counter-conditioned her to regard chairs as non-scary objects. In fact, for the rest of her life Lucrezia remained hesitant to eat treats from beneath furniture. The original and underlying fear issue was still there. But, encouraged to develop her own creative problem-solving skills, the dog could approach and deal with something that frightened her. What Lucrezia invented was not only a means of getting to a treat. It was, in effect, a coping mechanism.
Even if confidence is not an issue, there are times when it is beneficial for a dog to find another approach, one that does not rely on the same old routine. When things go wrong (and they will!), it's nice to be able to invent another option on the fly.
A few months ago, my Doberman, Valenzia, was attacked by another dog and badly injured. To protect the hundreds of stitches and staples and drains and other insults that were embedded in her, Valenzia had to wear an Elizabethan collar, what we now call, thanks to the writers of the Pixar film Up, "the cone of shame."
Playing versatility and shaping games can develop
creativity in your training subject.
I could write an entire creativity article just about Valenzia and medical collars. She has demonstrated her creativity by inventing a way to remove every kind of protective medical collar I've ever put on her! But this anecdote involves a collar she hadn't yet removed (if only because she was too sore and swollen to twist out of it).
The door into my fenced backyard exits from a walk-out basement, which is accessible from the house only via a steep, narrow stairway. Because of the injuries to her legs, Valenzia was convalescing in a pen in the basement where she could be walked outside to relieve herself without taking any stairs. Valenzia is used to sleeping in my bedroom, and she's a very clingy dog even when she's feeling at her best. You can imagine how she felt about being confined alone downstairs! She was determined to get upstairs, and she didn't care how many stitches she ripped out doing it. I had to keep a baby gate across the stairs from the basement to prevent her from dashing up to the house.
One day I failed to block the stairs completely when I let her out of the pen, and Valenzia bolted up the steps toward the main floor. She made it about halfway before the plastic cone around her neck sagged low enough to hook on a step, jolting her to an abrupt (and probably painful) halt. She tried to push forward a couple of times, but the collar blocked her. She was stuck.
I called her from the base of the steps, but she glanced back at me with an eloquent look that distinctly translated to, "I'm not going back in that pen." Valenzia paused for a moment, looking down at the steps. Then, with astonishing athleticism and flexibility for a dog in her condition, she launched her hind end into the air and pivoted 180 degrees on the narrow stairs so that she was facing me at the bottom. Carefully, she lifted one hind foot and felt for the step behind her, then the next step. She began backing up the stairs, lifting her head to let the edge of the cone bump up each step in her wake.
I was too stunned to stop her as she backed to the top of the stairs, pivoted again, and dashed into the house to run laps and make a mess of my floors. I have never taught Valenzia to back up the stairs. Anyone who has ever trained a dog to back through ladder rungs or up steps knows that dogs aren't aware of their back feet instinctively. It usually takes time and practice to develop hind-end awareness and control. Yet, here was my dog inventing it on her own because going up the stairs front-first wasn't working very well.
Individually, all of the component behaviors make sense. With Valenzia, I've done a lot of free-shaping (a simple form of cognitive problem-solving) and enough kinetic behaviors (jumping, targeting to platforms, etc.) for her to be able to move her body consciously and with intention. The 180-degree pivot is something she can do mid-air (another behavior she invented), so doing it with her front feet on the ground is a reasonable variation. Valenzia knows how to find a platform and step up onto it, so it's not surprising that she possesses the mechanics of finding and stepping up on the stairs. What is so impressive to me is that she was able to recognize the cause of the problem (the low-hanging cone around her neck), figure out what would remove the obstacle (turning it the other way), and chain separate behaviors together as a response.
Man, clever animals are neat!
On the small scale, a dog figuring out the stairs is not necessarily a useful behavior. In this case, it caused me some additional hassle—I had to catch the dog, get her back in her pen, and scrub the blood off the floors. But apply this same thought process to a dog performing a function that is valuable to humans:
- A search dog's path is cut off by debris, and he needs to find an alternate route back to his handler.
- A service dog's harness fails, and she needs to maintain physical contact with her person.
- A police dog sent to search a building has to surmount obstacles such as fences, closed doors, and furniture to reach the suspect in hiding.
Of course animals are trained for as many contingencies as possible in situations like these, but there are many instances where something unforeseeable may go wrong. If the dog can perform the required job some other way than Plan A, it can be very helpful, or even life-saving.
As a trainer, I love building creativity in my dogs because it makes it so much easier to teach new behaviors. Not every dog follows the exact same training plan. Some dogs learn faster with capturing, others with targeting, others with adducting known behaviors, others with free shaping. But the more creativity and analytical thinking you foster in your animal, the more options you have available to train new or complex behaviors. If you have a creative dog, you may even be able to skip steps on your training plan.
When I wanted Valenzia to put her front feet on an overturned bowl and bow, I was able to shape the position in just a few minutes because she began offering me various behaviors immediately—targeting, foot movement, head movement, joint flexion, variations of one or two feet, with or without nose touch. She tried every possible combination, just to see what would work. I was presented with a wide range of behaviors to select from, instead of having to coax the dog along one muscle movement at a time to build the position I wanted. Could I have shaped a novice dog to do the exact same behavior? Certainly! But without the dog offering me everything she could think of, it would have taken more training steps and significantly more time.
Much of the speed and the offering of behavior can be attributed to experience—what we call having a clicker-savvy dog, or a dog that has learned to move until he/she hears the click and to repeat the marked behavior. But the experimentation, the offering of new variations, is a form of creativity that results from being reinforced consistently for trying new things and inventing new combinations.
The creative advantage
I am convinced that the benefits of a creative animal far outstrip the potential downsides. Sure, you may have a dog that figures out how to open the cabinets or go up the stairs when you don't want him to. You might experience a few inconveniences, at least until you use your own clicker creativity to train something more desirable. In the end, the benefits to your dog, the time you'll save in training, and the potential new behaviors your dog might invent are worth a hypothetical hassle or two.
Post new comment