Reinforcement: What it is, and what it isn’t
Dog trainers love to talk about reinforcement, “reinforcing our dogs” and “the best” reinforcers. Reinforcement in psychological or behavior analysis terms, however, is really just one of two labels: positive reinforcement or negative reinforcement. “Ah ha!” you may be thinking. “Easy. Positive reinforcement is the good one, and negative reinforcement is the bad one.”
Or, you’ve heard someone say, “I tried positive reinforcement with my dog, but it just didn’t work. So I had to switch to negative reinforcement. I feel bad, but it’s just what had to be done.”
Hmm. In both cases, we’re misunderstanding the behavior analyst’s definition of reinforcement. Let’s take a minute to address reinforcement, in behavioral terms.
How does reinforcement work?
Behavior is lawful — it follows certain rules, as demonstrated by the historical work of John Watson, Ivan Pavlov, Edward Thorndike, B.F. Skinner, and others.
A stimulus is anything an animal can perceive. An aversive stimulus is a stimulus an organism tries to avoid. An appetitive stimulus is a stimulus an organism tries to get, or to get more of. Dog trainers often use the shorthand “aversive” for aversive stimulus when referring to the use of something a dog doesn’t like in training, and “reward” for appetitive stimulus, when referring to the use of something the dog wants.
Reinforcement is anything that causes behavior to increase. To know whether a behavior has been reinforced, we must ask, “Did the behavior increase?” If the behavior did not increase, it was not reinforced! This is why we must never say, “Reinforcement didn’t work.”
Reinforcement, by definition, always works! If you provide the animal with an appetitive stimulus (piece of liver) for a particular behavior (standing still), but instead of standing still, the animal jumps and barks, we can say with confidence that the liver did not act as a reinforcer for the target behavior, standing still, at that time. The dog was not reinforced for standing still.
Let’s have an example. Say you accept a job stuffing envelopes. You stuff 3,000 envelopes and at the end of the week, you are given $50 for stuffing 3,000 envelopes. When you are given another opportunity to stuff 3,000 envelopes in one week, will you accept? If you do not accept, the $50 may have been a reward, but it was not a reinforcer. Why not? Your target behavior, envelope-stuffing, did not increase.
Is feeding a treat reinforcing?
To know what is reinforcing a behavior, we must look at two things: 1) What happened immediately after the behavior (the consequence); and 2) Did the behavior increase? (Was it reinforced?)
Here we come to two more tricky terms. Positive reinforcement is the addition of something after a behavior occurs, followed by the target behavior increasing. Negative reinforcement is the removal of something after a behavior occurs, also followed by the target behavior increasing. The easiest way to keep these straight is to think of “positive” as adding something, and “negative” as subtracting something, and “reinforcement” as increasing the behavior. For either positive or negative reinforcement to occur, the behavior must increase. Whether the stimulus we add or remove to increase the behavior is appetitive or aversive, is up to the animal’s way of thinking at that time — not our intentions!
Myths about reinforcement in dog training
There’s one more misconception about reinforcement: that using stuff the dog wants in training is a “type” of training. It’s all training! There are only two ways to motivate dogs. The first is by using aversive stimuli. The second is by using appetitive stimuli. Even the nicest or harshest training method you can imagine still relies on the principles of behavior.
If you training using a mix of aversive and appetitive stimuli, you are following the laws of behavior. Every trainer is working within the confines of operant and associative learning: There is nothing else. The choice for the trainer is not whether or not to use the principles of behavior, but to learn to use them effectively and in a way that is respectful of the animal, or not.
Learn more
If you’re interested in these concepts and want to dig deeper, grab an introductory psychology textbook, or pick up a copy of Pamela Reid’s dog training book, “Excel-erated Learning.”
Resources:
Reid, Pamela J. Excel-erated learning : explaining in plain English how dogs learn and how best to teach them. Berkeley, Calif: James & Kenneth, 1996.
McLeod, S. (2007). Skinner – Operant Conditioning. Retrieved June 05, 2016, from www.simplypsychology.org/operant-conditioning.html
</div