Sculpting Behavior: Positive Reinforcement

Animals are very efficient in their behavior. If a behavior is inherently pleasurable (eating, playing, chasing), or if it gets something pleasurable for the animal (like food, attention or interaction), the animal will display that behavior more and more often. If a behavior is not pleasurable, or if it does not work to obtain something pleasurable, the animal will use that behavior less and less. Whenever you interact with an animal, you’re constantly giving her feedback about what works to get pleasurable things and what doesn’t work. If a dog jumps up and gets attention, even if the attention is that you push him, then he knows that jumping “works” – that is, it gets him attention. If a cat play bites and you don’t end the game, then he knows that play biting “works” – the fun continues.

The good news about this is that we can easily use the way animals learn to “sculpt” their behavior, by consistently rewarding the desirable behaviors we see and ignoring or interrupting the undesirable behaviors. Gradually, you will see the animal behaving more and more in desirable ways, and less and less in undesirable ways.

But what about, for example, dogs who jump all the time? Well, that’s just it: no dog ever jumps literally all the time. Even with a dog that jumps a lot, there’s a moment when she isn’t jumping, so reinforce that moment with attention and some food! If you don’t like what she’s doing, show her what you would like her to do (how about sitting down, instead?).

Repetition and patience are the key in animal training. There’s never a magic moment when the animal understands the meaning of our requests. Animals gradually become conditioned through lots of repetition that certain behaviors in certain situations will or will not “pay off.”

We use these principles – rewarding desirable behaviors and ignoring undesirable behavior or removing rewards when the animal behaves in an undesirable way – in Open Paw training, and do not use physical punishment. Animals make associations with you and with the situation every time you interact with them. Thus, an unfortunate side effect of using punishment to try to train animals is that, while they may learn to respond to cues, they may also form negative associations to you, to the situation, to people, or to training. Furthermore, often you don’t get the result you wanted from trying to use punishment to train. Take for example a dog jumping on people. It’s not a desirable behavior to people, but in the dog-dog world it is an appeasing, friendly greeting gesture. If you use punishment to try to get the dog to stop jumping, you have to use a severe enough punishment the first time that it effectively outweighs the positive associations of the friendly greeting gesture. If the punishment is not severe enough, then, you are not effectively damping that behavior. You may even unintentionally be rewarding it. Furthermore, some dogs may try to stop the punishment by offering an appeasement gesture rather than by stopping the undesirable behavior– so the result might be more rather than less jumping.

So, using punishment to train is pretty inefficient, difficult to do correctly, and, in order to be effective, must be severe. A much more efficient, friendlier way to train is to teach the dog a desirable, incompatible behavior; ask yourself, “if this is ‘wrong,’ what is ‘right’? In this case, we might train the dog to sit to greet people.

Click here for a printable PDF version of this article.


<< Back to Pet Basics • Items You Will Need to Buy for Your Dog >>