Do you remember the scene at the end of Close Encounters of the Third Kind where a troop of red-suited astronauts file on board the alien spaceship? Imagine that you are one of those astronauts. You have been given a highly important mission. You are to learn everything possible about the aliens, including their language. You have reason to believe that the aliens are friendly and are similarly motivated. Everyone on both sides wants your mission to succeed. Now, as soon as you are on board the alien ship, they attach a small box to the back of your neck. One of the aliens speaks to you in their tongue: “Gruhackz,” he says. He waits a moment. You look at him questioningly. He pushes a little button and an electric shock courses through you, coming from the box on your neck. This, it turns out, is the aliens’ method of teaching you their language. They say something to you, and if you do not understand or respond appropriately, they shock you. Imagine how little you would enjoy your language lessons and how slowly you would learn under this system. You have a mission to perform, and you are going to do it, but it is no fun whatsoever.
Now imagine that you are not an astronaut but a hiker, minding your own business and enjoying a warm Sunday walk in the woods when you are abducted by aliens and subjected to the same training regimen. Depending on the person’s personality, I can imagine that some might fight to the death, others might go mad, and some, in order to survive, might submit to the training, but they would always be on the lookout for the chance to escape or otherwise sabotage their captors.
These two scenarios are analogies for how we humans have traditionally trained animals. The first scenario, in which you are an astronaut, doing your best to learn, would be analogous to the training of dogs, horses and other social animals. It is instinctive for them to want to be part of a group, and they will try to learn what their group leader wants even under adverse circumstances. The second scenario would apply to wild animals and animals such as cats which do not naturally live in groups.
Look at the way a dog is taught to heel in the traditional manner: A choke collar is put on the dog with a leash attached. A choke collar runs loosely through a metal ring, so that the collar is loose when there is no pressure applied to the leash, but the collar can be tightened by pulling or jerking on the leash. In the heeling maneuver, the handler wants the dog to walk close to his left side with her shoulder more or less even with the handler’s leg. The dog is not supposed to forge ahead, lag behind, or veer away from the handler. So the handler says “Spot. Heel.” (Gruhackz) and begins walking. Whenever the dog is not in the proper position (most of the time at first) the handler gives a quick, hard jerk on the leash. This both punishes the dog for being in the wrong spot and also brings her back to where you want her (if done properly). This is not much fun for the dog or the handler and explains in part why there are so few dogs who have been taught to heel. It just doesn’t seem fair to punish the dog when she doesn’t know what you want, and because the process is so unpleasant, many people give up on it quickly.
Now try to imagine teaching a cat to heel using this method. Everyone says you can’t train cats, but that isn’t true. Most of us have been to the circus and have seen the trained lions and tigers. So it is possible. But it is far more difficult.
Now imagine another scenario. You are aboard an alien spaceship. It doesn’t matter whether you are the astronaut or the abductee. You are in a room. Suddenly you hear a clicking noise, and then a few grapes appear on a table. You eat them. How nice. A few moments later, you hear the clicking noise again, and a miniature candy bar appears. Every time you hear the clicking noise, something tasty appears. Now, when you hear the click, you look around to find out what treat you will get next. Then the clicking stops. You wait by the table for a few minutes, but no more treats appear, so you begin to explore the room. You are exploring the surfaces, picking items up, moving from place to place. You have just started to move toward a strange orange object hanging on the wall, when you hear the clicking noise. You look around, and sure enough there is half a cookie on the table. You go over and munch it. You wait for a moment, but no more clicks. You once again begin exploring the room. As you are walking from one side of the room to the other, you hear the click. You walk over to the table and eat your treat. As you are munching your popcorn, it occurs to you that the first clicking noise occurred when you approached the orange object on the wall. The second time, although you weren’t headed toward the orange object, you were passing close to it when you heard the click. You look toward the orange object on the wall. There is immediately a click and a treat. !!! Maybe it has something to do with the orange object. You walk over to the orange object. Click, treat. You touch the orange object. Click, treat. The orange object disappears, but another appears on a different wall. You walk over to the new orange object, and as you do, you hear an alien say, “Gruhackz.” Click, treat. You don’t yet know whether gruhackz is the name of the object, or the command to touch it, or something else, but you are definitely looking forward to your next language lesson.
This scenario is a fairly accurate description of a training session using the method called clicker training. This method is based on the research of behavioral psychologist B. F. Skinner. One of the pioneers of this method of training is Karen Pryor. As a young woman, she did graduate work under Skinner. Later she and her husband purchased a sea park in Hawaii similar to the later franchises which would be called Sea World. They began using Skinner’s methods to train the dolphins, whales, sea lions, etc. The idea in the beginning is to reward the trainee for any behavior which even vaguely approximates the goal behavior. The problem with sea animals is that you can’t stand next to them and hand them a fish when they do what you want. So the idea was developed to use a whistle. When the dolphin leaped, you blew the whistle. Then you threw him some fish. The whistle marks the behavior that you are rewarding. The animal quickly learns that it is the behavior which was occurring when he heard the whistle which produces the fish.
Once the animal is offering the behavior on a regular basis, then you stop offering a reward on every attempt. At first you might reward him four out of five times, then three out of four. Gradually, you make the animal perform the action more times before getting the reward. Paradoxically, this actually causes the animal to work harder. Now you can begin shaping the behavior. You do this by rewarding those attempts which are closer to your goal behavior. So the dolphin jumps several times, and you whistle (mark the behavior) the time that he jumps high. So in this way you begin shaping the dolphin to jump higher and higher.
You can train animals to perform very complex behavior chains using this method. One person trained her dog to bring her a cold beer from the refrigerator. She did this by breaking the action into a number of smaller steps: hold a cold metal can in your mouth; hold a metal can without puncturing it with your teeth; bring the can to me and put it in my hand; open the refrigerator (by pulling on a knotted rope tied to the door handle); find the beer in the refrigerator; remove the beer; close the refrigerator. Typically, you teach the last link in the chain first. In this case, put the beer can in my hand. Then you add the next to the last link - bring the beer from the kitchen to me, put it in my hand. You work backward, adding link after link in the behavior chain, until finally, as you sit on the couch watching David Letterman, you look at your dog and say, “Get me a beer,” and she pads into the kitchen, opens the fridge, pulls out a cold one, sets it down, closes the refrigerator, picks the beer up again and brings it to you.
In large outdoor situations, a whistle is generally used as the marking signal, because it is loud and can be heard at a distance and when there are other noises (such as screaming crowds for dolphins in sea parks or baa-ing sheep for sheepdogs), but indoors or other places where a whistle seems unnecessarily loud, a small clicking device has been adopted by many trainers. These used to be sold as children’s toys called crickets. Now many training and pet supply houses sell them.
Clicker training courses for people who want to train their puppies are appearing in many cities, but this method is also being used in many other venues. Some zoos are now using the clicker method both for their animal acts and also to train the animals in the basics necessary for handling them. The last time I went to the San Diego Zoo, we went to see one of their small wild animal shows and the trainer held and used a clicker throughout the show. Another zoo I read about in The Clicker Journal, uses clicker training to teach the animals to do such things as hold still for veterinary examinations and basic grooming such as nail trimming, and to move about on command (such as into smaller cage areas while the larger living area is being cleaned). The clicker method is also being used to train police dogs, assistance dogs, and search and rescue dogs as well as other domestic animals such as horses and, yes, cats.
I read an amusing account in The Clicker Journal about a man who raises llamas. He uses clicker training to teach his llamas their basic handling lessons - walking quietly on halter, standing to be groomed, etc. One day, more or less for his own amusement, he was using the clicker to train one of his llamas to pick up litter in her pasture and put it in a box. A llama in an adjoining pasture who was there for breeding purposes watched with interest for some minutes, then jumped the fence into the pasture where the training was taking place, and competed with the first llama for pieces of litter.
Now let us revisit the person who is trying to teach his dog how to heel. Using the clicker method, he goes out into his back yard with his dog (not on a leash) and begins walking around. When the dog gets tired of exploring the yard and approaches the man, he clicks and treats the dog. Now the dog is eager to be near the man because she wants treats. At first he will click any time the dog is nearby. Then he begins shaping the behavior: the dog receives clicks only when she is on the left side, not too far ahead or behind, not too close or too far away, etc. This is shaped gradually. Then the length of time the dog must be in the right position before receiving a click is extended. At first the dog is clicked any time she is in the proper position. Then she must maintain this position for two steps, three, four, halfway across the yard, etc. Now, the man can put a leash on the dog and take her out to practice heeling in the park, downtown, the airport, etc. Both the dog and the man are having a good time. For the dog it is a game. She figures out how to get the man to give her treats. For the man it is also a game. He is figuring out how to make it clear to the dog that he wants her to heel, sit, search for lost children, sniff for drugs, cancer, bombs. In the old methods dogs sometimes became afraid to try anything new for fear of punishment; in this method dogs are motivated to try all kinds of new behaviors because there is no punishment for mistakes. That didn’t work? Fine, I’ll try this.
Clicker training creates eager, motivated pupils. Instead of a grim scenario where the training is something both trainer and trainee must endure in order to get to an orderly and pleasant future, training becomes a game in which trainer and trainee both have fun, learn to know and understand one another, and can enjoy the process as much as the product.
If you are interested in learning more about clicker training, there are a number of resources available:
“The Clicker Journal is a bimonthly publication dedicated to the practical and ethical use of operant conditioning. It is a publication for clicker trainers, with articles submitted by clicker trainers.” You may read past articles or subscribe at: http://www.clickertrain.com/journal.html
Web sites
Books (all available from Amazon)
- Clicker Training for Obedience by Morgan Spector
- Clicking With Your Dog: Step-By-Step in Pictures by Peggy Tillman
- Clicker Training For Your Horse by Alexandra Kurland
- Don't Shoot the Dog! : The New Art of Teaching and Training by Karen Pryor