As a certified applied animal behaviorist, I’ve worked with a variety of species—from Komodo dragons, orangutans, camels and horses to humans. Working with species like these and more, I’ve learned the importance of having a variety of tools to train a desired behavior. Whether challenged to train a Komodo dragon to enter a crate on cue or my horse to willingly accept dewormer or ear clipping, I open my toolbox and choose the training tools that work best for the task.
In this article, I’ll introduce and define some of those tools you’re probably already using but may not be fully taking advantage of, particularly positive and negative reinforcement. A simple breakdown will help you distinguish the training method you’re using and give you alternatives if a particular method doesn’t seem effective for a certain horse or behavior.
Though many people use negative reinforcement when training animals, I’ve found the experience is usually the most fun for both species (me and the nonhuman animal I’m training) if I rely heavily on positive reinforcement. So we’ll also apply a couple positive reinforcement tools on some common behavior issues that folks ask me about in my column, “Ask Dr. Weiss,” at www.aspca.org.
Training Tools Defined
To get to the fun stuff, we first need to dig through a few learning-theory terms that can be a bit confusing—and are often used incorrectly.
- Positive in the field of learning theory simply means “give.” That is it—just give. Not good or great or yummy or anything like that … just give.
- Negative means “take away.” That’s it—not bad or painful.
- Reinforcement means increase the likelihood that a given behavior occurs again. Think of it as making something stronger—more likely.
Now let’s put those together.
Positive reinforcement occurs if, when I give something right after a behavior occurs, it makes that behavior more likely to occur again.
For example, I want to teach a dog to sit. I have a tasty treat the dog would like me to give to him. I hold the treat a bit above his head and say “sit.” He then puts his nose up, resulting in his rear rocking back toward the floor. I give him the treat. The next time the treat appears, he’s likely to try the behavior again because it caused treat delivery previously.
Negative reinforcement occurs if, when I take something away right after the behavior occurs, it makes that behavior more likely to occur again.
In this case, I need to take something away to reinforce a behavior. Let’s use the same behavior “sit” and the same species, dog, as our example. This time I’ll apply moderate pressure to Fido’s rear when I give the “sit” cue. I’ll keep the pressure on his rear until he yields to it by placing his behind on the floor. The moment he does, I’ll release the pressure. Next time I give the cue and apply the pressure, he’ll likely sit more quickly to “turn off” the pressure.
Here are a few more terms you’ll need for your training toolbox:
- Cue: A cue is often an auditory (a word, for example), visual (a hand signal) or tactile (the touch of your leg) stimulus to initiate a particular behavior.
- Primary reinforcer: This is something the animal finds inherently reinforcing, or said more simply, something the animal naturally likes. Most commonly used primary reinforcers are food, water and social interaction.
- Secondary reinforcer: When training, timing is everything. Secondary reinforcers are used to help with timing—they signal “primary reinforcer is coming!” They are sometimes called a “bridge,” as they connect the desired behavior and the reward. To develop a secondary reinforcer, it is paired with the delivery of a primary reinforcer several times until the secondary reinforcer produces a seeking of the primary reinforcer. You may have heard of clicker training. Clickers are secondary reinforcers. I’m notorious for losing clickers, so I use the word “yes” or make a whistle sound instead—I call it “tweet and treat.”
- Shaping: Shaping a behavior is something you’re already doing … now it has a name. When we shape a behavior, we reinforce approximations of the behavior we want. When we train something like lifting a hoof, the initial behavior we reinforce may be shifting the weight off the leg we’ll be lifting. For more complex behaviors, the behavior we initially reward may look quite different from the final behavior.
Now I’ll teach you how to create a secondary reinforcer, then we’ll do three exercises to show you how to put applied learning theory into practice.