While at the grocery store, you hear a toddler beginning to protest. You see him straining to reach a colorful box of cookies. His grunts become piercing squeals as he looks at his father. Both of them know what he wants. Both also know that a tantrum is forthcoming if the child doesn’t get what he wants.
When is the best time to give the toddler a treat to avoid a tantrum?
You may already be familiar with “classical conditioning”. This is what people are talking about when they talk about Pavlov’s dogs. Dogs cannot help but drool when they smell food.
Pavlov rang bells when he presented dogs with food, so dogs soon learned that a ringing bell meant food was coming.
Pavlov then only rang bells without food. The dogs drooled anyway. To the dogs, the sound of the ringing bell was a signal that food was coming. Food was the original stimulus, but the bell took over that role. Drooling was the response.
“Operant conditioning” is a different form of behavior conditioning. It focuses on consequences, not on pairing involuntary responses with arbitrary stimuli.
In operant conditioning, the word “reinforcement” refers to increasing the likelihood that a behavior will happen. The “positive” in “positive reinforcement” refers to adding something to increase the chances a behavior will happen. It does not refer to something “nice” or “pleasant”.
The “negative” in “negative reinforcement” means taking something away to make it more likely a behavior will happen. Some examples will be useful here.
Say I meet one of Pavlov’s dogs and I want him to sit. One strategy I can try is to give the dog a small cube of cheese every single time he sits down. Because I know dogs love cheese, I am intentionally using cheese (stimulus) to increase the likelihood that the dog will sit down (behavior). The dog, in turn, learns that this person will offer FREE CHEESE just for sitting down! This delightful consequence shapes the dog’s behavior. The delivery of cheese to the dog is the addition of a stimulus, so this example is “positive” reinforcement.
“Negative” reinforcement is usually not as effective (or as fun) as “positive” reinforcement, but we can still use it to make it more likely this dog will sit. When the dog is doing anything but sitting, I yell at the dog. When the dog happens to sit down, I stop yelling. The moment it stands back up, I resume yelling. If the dog values quiet, then it will learn to sit down (behavior) to shut me up. The withdrawal of my yelling is the removal of a stimulus, so this is “negative” reinforcement. The goal is to get the dog to sit down.
So let’s return to the example above of the toddler who is on the verge of screaming in the grocery store. If the father wants to increase the likelihood that the child will be serene, when is the best time to give the little one a treat?
Sometimes positive reinforcers stop happening. Let’s revisit the dog who was the lucky recipient of cheese for sitting.
Maybe I run out of cheese, or I’m bored with giving the dog cheese for sitting down. The dog keeps sitting down around me because it is expecting the consequence of free cheese. However, I’m not giving the dog any cheese. What do we expect the dog to do?
This is a bit of a silly example, but chances are good that the dog will try harder to sit more to receive FREE CHEESE. Maybe the dog will sit closer to me. Maybe the dog will even sit ON me. It’s almost as if the dog is saying, “Hey, I’m sitting here! You’re supposed to give me cheese when I sit. Why aren’t you giving me cheese? Can’t you see that I’m sitting?”
This surge in behavior to receive the positive reinforcer is called “extinction”. After heroic efforts at sitting, the dog will eventually learn that no FREE CHEESE is coming. The dog will give up on sitting around me. The cheese (stimulus) has gone away, so the sitting (behavior) also goes away.
If the father wants to the toddler to stay quiet and serene, the best time to give the child his preferred cookies is when the kid is peaceful and chill. The child will likely experience the cookies as a reward and will learn that he gets cookies when he is NOT yelling or screaming.
If the father tends to give the child cookies when the toddler is about to have a tantrum, then the child learns that the best way to get cookies is to yell and scream.
If the father happens to read this blog post and realizes that he is reinforcing an undesired behavior, he may soon see “extinction” behavior from the child. The child may have a meltdown of epic proportions, the likes of which no one has seen, before learning that these tantrums won’t give him the consequences that he wants (cookies).
I’ve been writing about dogs and cheese and children and cookies, but this is actually about American politics and the results from the recent midterm elections. We should expect surges in behavior from the federal administration. Maybe these surges in behavior represent extinction behavior; maybe they represent retribution. (They could be both.)
If the model of operant conditioning is accurate here, then we can expect the behaviors to be more intense and frequent than what we witnessed before. This might mean even more outlandish comments about body parts, more cruel commentary about different groups of people, and more heinous actions that will harm US residents and others around the globe. The behaviors in extinction bursts always end, but they will be hard to tolerate while they persist. We must not give in, and we must not reinforce any bad behavior.
