How The Brain’s Reward System Works
The reward system is an integral part of our perceptual system. When you are awake, you are constantly moving around and doing things. You are both producing behaviors and receiving inputs.
As you engage the world, your brain does a preliminary assessment of importance. It works on the principle of triage. The most important things are dealt with first, things that can wait are put on hold, and things that stay the same are ignored.
If something is important, a small dose of dopamine, norepinephrine and other neurotransmitters are released. Dopamine is not so much a pleasure maker as an importance marker. Since we think pleasure is important, there is quite a bit of overlap between them. But dopamine is also released when you feel you are in danger.
If something really important happens, large doses of dopamine are released. Obviously, threats and novel stimuli are most important. Moving objects, strange noises, loud sounds and unusual smells are also very important to us.
Rewards can be classified as lures or reinforcements. Lures are attractive things you dangle just out of reach. This is the carrot of carrot-stick fame. You get the donkey to go by keeping the carrot dangled just in front of the donkey’s nose. In human terms, this is an inducement. It is simple to use, comes naturally do us, and effective in the short run.
Imagine you are trying to train a dog to jump through a hoop. One way is to take a treat and, keeping it just out of reach, move the treat and dog through the hoop. You can get a dog to walk beside you or sit by moving the food in your hand along the path you want the dog’s body to follow.
The downside is that the dog’s focus will be on the treat. The dog will not look at you during or after the exercise. Nor will they be focused on doing the task. So lures only work for simple tasks. Even proponents of its use recommend limiting it to 3-4 times, and they usually combine it with giving the dog the treat (reinforcement).
Reinforcements teach the dog the rule. It is a contract. You do this and I’ll give you that. On a grand scale, it’s not a very complicated rule but, unfortunately, dogs aren’t really good at rule acquisition. Animal trainers point out that dolphins, whales and elephants are superfast at learning rules. Dogs? Not so much.
The advantage of reinforcement is that it is long lasting, portable and fair. Reinforcement looks at patterns of behavior over time. It is inconvenient to carry fish food, apples and bones in your pocket just in case you have to lure your goldfish, horse or dog. Reinforcements maintain behavior patterns over long periods of time. You still do some things you were reinforced for as a child.
Reinforcement is portable. Luring tends to be environmentally specific. The dog learns to do it only under certain circumstances, the most obvious being that you have a treat in your hand.
Compared to luring, reinforcement is fair. It a contract that says it A happens then B will happen. If you never give the dog his bone, and only use it as a lure, you’re not being fair. If you give it to the dog after getting the dog to do what you want, it is called reinforcement. Skip the lure and start with reinforcement.
People like luring because they want to be in control. I know someone who wants their dog to “respect” them. What they mean is compliance. I’ve seen people do the same thing with their children. “When I walk into a room, my children should stop playing and immediately come and hug me. It’s a matter of respect.”
It is hard to forget that we are not in control of others’ lives. Being a tyrant is a bad thing. Working with others, humans or animals, is a cooperative process. Just because you can pick them up, doesn’t mean you have the right to force them.
I always recommend people begin animal training with chickens or whales. Neither animal will do what it doesn’t want to do. Whales, elephants and Great Danes are great because they are too big to bully. You must learn to interact with them. You must learn to negotiate.
Chickens, goldfish and hamsters can be physically bullied but they are so dumb that again you must learn to negotiate. They will only do what you and they agree to do.
Contract negotiation produces cooperative interaction. If you want instant compliance, buy a robot.
Another contrast between luring and reinforcement is the focus. In luring, the focus in on the food. In reinforcement, the focus is on performance. The perceptual centers of the brain are centered on different sources. It makes a difference in what the animal tracks.
Reinforcement also allows your dog to learn more complex patterns of behavior. Luring only works on simple tasks.
Which portion of the brain being activated is different too. Being lured into activity provides a little of spike of dopamine. This is anticipatory reward. It is marking it as something to look forward to. It is a simple reward and activates the general reward centers. This is what happens when you enjoy a snack. Your brain registers that it tastes good. You “good” meter spikes.
In reinforcement, more brain centers are activated. In humans, it means that the prefrontal cortex of the frontal lobe is involved, in addition to the amygdala’s emotional “good” response. Whenever we are involved in rule making, more processing is required. But the rule is more portable. It will work in more situations.
Reinforcement requires practice in other situations too. You must teach your dog that the rule applies here, and here, and here. This is similar to teaching your children that this rule applies here but not there. It is okay to yell at a baseball game but not in the library. Every rule needs generalization but reinforcement rules work better than luring.
Notice that dopamine works as a marker. Its release signifies that THIS is something to remember in the future.
In training, it helps to have a marker you can use to specific what THIS is. A marker can be a word, gesture or sound. You can say “Good” or “Hai.” Your dog isn’t a native speaker of English or Japanese. Any word you choose could work. But shorter is better.
Sound works even better as a marker. A toy clicker is a great marker. It is a bent piece of metal and clicks when you push is down. And clicks again when you release it. The click a clear cue, has a quick onset and is distinctive. The quick onset helps the animal know what is being rewarded. The distinctiveness lets it stand out against all the background noises. It can even be used from far away (something a lure can never do). A clicker is good. It would be even better if it didn’t have two sounds.
I should point out that the clicker is used only for training. You have to have it with you during training but not afterward. By the way, having training sessions is a good thing. It helps differentiate between school and home.
Clickers, like words, don’t mean anything to the dog until you pair it to a reward. Remember Pavlov’s dogs? They began to salivate to the sound of a bell because it had been repeatedly paired with their food.
Basically, it is a reflex being triggered by a previously neutral stimulus. This classical conditioning occurs with all of the stimuli around us. We learn to find peace at the ocean or excitement at Disneyland. We associate Christmas with snow, even if we live in Southern California.
These environment cues not only serve to reward us but to protect us. When we were children, our family went out for Chinese food for special occasions, including (as we got older) birthdays. My brother loves prawns. He could eat a million of them, he said. And what he really wanted for his birthday at age 14 or 15 was a whole platter for himself.
My parents complied. My brother ate them all (or close to it) and never wanted them again. For several years, just the word prawns disturbed him.
All the smells, lighting, decorations, and ambiance provide cues we use to store important information, such as don’t let me be that stupid again.
My story is as an adult. I took my family to a fancy burger joint, got sick and never wanted to go back to that place. When you get food poisoning, your brain keeps track. In fact, it makes a BIG mental note. Positive circumstances may take dozens of pairings for us to make associations. It took Pavlov’s dogs 50-75 trial to learn that the bell meant food was coming. But one dose of food poisoning is enough. Your reward system is very sensitive to adverse events.
One other thing before we talk about how the system can break. We not only learn single activities but chains of behaviors.
Think of your morning or evening routine. You get up, rub your eyes, go the bathroom, wash your face, go to the kitchen, take a shower, brush your teeth, get dressed and drink some coffee. Maybe not in that order but there is a chain of behavior you follow every day.
We learn forward chain but learning one item and then adding to it. This is how we typically learn a song. The good news is that works. The bad news is that we remember the first part well but less and less as we go. We remember “O Canada! Our home and native land!” Then we hum along until we get to “We stand on guard for thee.”
Next we will take a look at how the reward system gets broken or misappropriated.
Click for more topics.
