Jackpot: The Power of Positive Reinforcement

By Beth Terrell

This is Luca, also known as His Lordship of Eternal Cuteness. He's a papillon, (French for "butterfly"). Look at the ears, and there's no mystery about how the breed got its name. See the little bit of white on his ears? It's considered a flaw in the show ring, because it's believed that darker ears most resemble butterflies. If his ears had been dark, his breeder would have kept him to show and breed, and this little ray of shining brightness would never have come to live with us. As it was, she was looking for a home for him at exactly the same time I was being consumed by puppy fever.

Oh, I had no intentions of actually getting a puppy. I was at the dog show "researching breeds." But the minute I saw him, I was in love. I don't know what it was. I had been passing up other adorable puppies for weeks. This one had my heart the second I saw him. He weighed two and a half pounds, and most of it was ears. He still had his puppy coat (not the silky tresses you see today), and he looked like a strong wind might pick him up and carry him off. He literally took my breath away. He still does.

I know I'm hardwired to love him. With his tiny face, big eyes, and round head, he evokes the hormonal rush of adoration we humans are programmed to feel when we look at babies. But there's more than that at work here. I've loved every dog who's ever owned me, but there's an almost spiritual chemistry with this one.

A few days ago, we were clicker training in the living room. If you've never clicker trained a dog before, it's a fascinating experience. Luca loves it. It's a positive training method based on traditional operant conditioning. The idea is that animals (and people) do the things that bring pleasant results (rewards) and avoid the things that don't bring rewards. A reward might be a good feeling, a good grade, praise, or--if you're a papillon--a tasty piece of liver snack. Some people call clicker training "marker training," because the click marks the desired behavior. The click is made by a little plastic device, the clicking end of an ink pen, or even the trainer's tongue. (Sometimes I just use a quick, high-pitched "yes!", but the clicker is best, because it's a very distinctive sound, and the dog hears it only during training sessions.) The clicker is paired with a reward, so that the animal learns that the click means good things happen. (Click - reward, click-reward.) The click becomes a marker you can use to tell the dog, "Yes, that thing you just did, that's what I want."

Timing is important with clicker training, because whatever the dog is doing at the moment you click is what it gets rewarded for, and what gets rewarded, gets repeated. Sometimes you lure the dog into the position or behavior you want, then click the instant you get it. Other times you capture a behavior the dog performs spontaneously. (Luca sometimes rubs his eyes with one paw; he looks like he's a shy boy covering his face. I find this adorable and want to put it on cue, so whenever I see him do it, I click and reward. As a result, he does it more and more frequently. When he is offering the behavior often, I will begin to pair it with the command, "Shy boy." Eventually, I'll reward only when he does the behavior on command. This is how he learns that "Shy boy" means, "Do that cute thing where you cover your eyes with your paw."

Once he's learned that, and once he does it reliably on command, I'll stop clicking every time and go to an intermittent reinforcement schedule. This is the most powerful type of reward of all, because it works on the same theory as Las Vegas slot machines. We know there will be a payout; we just don't know when, so we keep on playing. Same with Luca. Once he knows what he's supposed to do, he gets praise every time, but he only gets the reward every so often, and he's never sure when. Sometimes it's a "jackpot"--a whole handful of goodies instead of the usual tiny bite. Woo hoo! Because he always gets something good (praise) and sometimes gets something really good, the behavior becomes very reliable, and Luca is happy because he's figured out how to make the good things happen.

Today, I realized that being a writer is a little bit like being clicker trained. (I bet you were wondering how I was going to relate this to writing!) Writing is chock full of that powerful intermittent reinforcement.

I write a chapter of tight, compelling prose that holds up to even my sternest inner critic.
Click. I get a warm and fizzy feeling inside. I want more of that!

I struggle to reach my goal of 1,000 words. Nothing sings. The writing is flat. What happened to that warm and fizzy feeling? No reward. Hm. Better try harder.

Someone sends me an email saying they read my book and couldn't put it down. "I'm serious," she says. "I was reading at stoplights!"

Click. Hm. Maybe I really am cut out for this.

Maybe I get a bad review and I'm stuck on the chapter where the villain escapes using a toilet plunger and a pair of his grandmother's support hose. No reward. (Sigh.) The thought of my friend reading my book at stoplights carries me through. Maybe tomorrow will be the day I get...a good review, a top New York agent, a six million dollar movie deal.

No wonder writing is so addictive!