Many positive reinforcement dog trainers use a hand-held noise-making device called a clicker in order to pinpoint the exact moment that the dog does whatever is that you want him to do. The click noise is always followed by some kind of reward–usually something tiny and edible. This is classical conditioning, also known as Pavlovian Conditioning: a neutral stimulus is repeatedly paired with something meaningful until the neutral stimulus itself becomes meaningful. Some people classically condition a special word or a light or a whistle in place of the click. This method of training is used in zoos, aquariums, and service animal training facilities around the country. The point is that some kind of meaningful and precise signal can indicate exactly what was good faster and more effectively than trying to get a fish into the mouth of a dolphin at the exact moment that he jumps through the hoop.
Clicker training operates in a binary way: If there’s a click (or another classically conditioned stimulus), the dog knows that it has done the correct thing. If there is no click, no reward. Clicker training isn’t just for dogs. There’s a burgeoning world of educators who use this kind of “conditioned reinforcer” (usually a clicker) to teach both children and adults everything from playing the violin to tying a shoe to learning a language. When used with people, clicker training is usually referred to as TAG teaching.
But it’s also possible to use the clicker method to point out what is incorrect — to make it a conditioned punisher instead of a conditioned reinforcer. In a Boston Globe piece that ran earlier this week, reporter Billy Baker interviews Marjorie Feinstein-Whittaker, an accent coach who helps Bostonians learn how to pronounce their “R’s.” In the story, Baker points out that Whittaker uses a clicker, clicking whenever her students make an error. The clicker is a more precise way of pinpointing the mistake than verbally trying to correct it. I heard about this story on NPR, which played Baker using the clicker to try to rid his own father of his accent. My dog Amos was lying in the room with me and immediately perked up and went to the speaker where the clicking nose was coming from. He loves the clicker! But to Baker’s father, the click had come to mean a correction. He quickly got pissed at being clicked and told his son to shove off. It was an interesting reminder of how a neutral stimulus can come to mean such different things.
It all made me think of a famous Bostonian: Cliff Clavin. As a kid I recall this episode of Cheers where Cliff, the annoying postman, receives an electric shock every time he says something dumb. Same idea. Only difference is that the shock isn’t paired with any kind of conditioned punisher. The jolt is delivered precisely and quickly enough to make that sort of pairing unnecessary.
Because we have the ability to communicate with each other using language, it wasn’t necessary for Baker to condition the click so that his father learned it meant he’d made a mistake. He could just explain that much. If you were going to use this system with a dog, however, you’d need to do some classical conditioning with the stimulus (in this case, the click) to insure that he learn that the sound–heard during the precise moment when he lifts his leg on the couch, for example–meant that something bad is about to happen. When a dog approaches an electric fence, his shock collar usually makes a beeping noise which encourages the dog to back away. The noise comes to have meaning because it has been continually paired with a shock.
In positive reinforcement training, this kind of thing is completely frowned upon. The clicker should ALWAYS be a sign of something good, never anything bad. In my experience, once people see how much can be done with a dog by simply rewarding the wanted behaviors and ignoring the rest, there’s little incentive to use punishment.
On the Karen Pryor Academy for Animal Behavior and Training message board, Karen Pryor, one of the world’s leaders in spreading the gospel of positive reinforcement, wrote that she’d contacted both the reporter and the speech coach to explain them how to use the clicker. Her approach to dealing with clicker misuse? Positive reinforcement, of course:
You can believe that clicker trainers and TAG teachers weighed in from all over the universe. Both the speech coach and the reporter were attacked from all sides.
Hey, that’s not very clickerly! Educate, but don’t get mad!
I emailed the speech coach, explained the more powerful way
of using the clicker, she responded positively and said she is going to try it and keep me posted. I emailed the reporter, who said I was the only person who had given him any positive reinforcement, and who wants to do a story about me
and the ‘real’ clicker training, in the fall.
Moral: thank you, everyone, for being so protective of our wonderful technology; and remember, you catch more flies with honey than with vinegar. Keep clickerly tools in mind even when someone else’s behavior is not what you want.