A Lovely Stroll

Imagine you're strolling through an open field, it's a beautiful day, and your friend is with you. You and your friend don't speak the same language, but you know they care about you and enjoy your companionship. But you're friend is a bit bossy. As you walk together, your friend prods you with a stick, ouch, you move away. Then every couple steps your friend prods you again. First left, then right, you don't know where to go to escape the potential jab or slap from the stick. You slow down and take every step carefully, watching for the stick to raise up to let you know you chose wrong. Quick! Jump the other way, ouch, too fast. You tip toe trying to guess which invisible path is laid out in front of you that doesn't result in pain or the looming threat of it. But your friend is annoyed you're taking so long, so whap! Hurry up! Make a choice! Wrong choice!

This pleasant stroll through the field with your friend isn't so fun anymore, and your friend, probably isn't your friend anymore.

Negative reinforcement training is all about telling the horse which answers are wrong. The wrong answers are met with aversives until they find the answer you were looking for. All problem solving is scary and threatening until they find the path to avoid it. They learn to see and predict the threats of the aversive, to guess and avoid, through fear and discomfort. By the end, its not easy to stay friends with the one holding the stick.

Positive reinforcement is the opposite. It communicates through "yes". The learner feels their way through the invisible maze by being told when they are on the right path, not the wrong path. The learner remains encouraged, because there is no threat of pain or fear or discomfort for making mistakes. They also enjoy the problem solving puzzle because it results in good things. And if the friend guiding the puzzle is generous, supportive and kind, they will likely become better and better friends as time goes on and they solve more puzzles together.

Can we use "yes" and "no"? Well, sure, but the more "no" in the equation the more risk the learner will give up, not want to risk trying, and your learning becomes unpredictable and stressful, as does your relationship. Rather than using more "no" use puzzles you've solved in the past as communication tools, to help make challenging puzzles just a collection of well known answers.

Comic by Fed Up Fred