Punished by Rewards

Has anyone ever heard of this guy, Alfie Kohn? He has written a book called Punished by Rewards, and I recently read an interview with him at this Web site.

So let's set the backbone for discussion. Behavior, by definition, is a "muscle, glandular, or neuro-electrical activity" ^[1]. And to go along with that, in order for a behavior to occur, there must be some sort of reinforcement behind that behavior. Whether it is stimuli arising from moving your hand to your face, or dropping a piece of candy into your mouth, there is a reinforcement contingency behind EVERYTHING you do.

But of course, regular, normal, natural reinforcement is a lot different than added, performance management, reinforcement. And that's what Mr. Kohn is discussing, at least in his interview.

"They do work in the short term, but at a great cost. Rewards, like punishments, are useful for getting exactly one thing: temporary compliance. By bribing or threatening kids, you can get them to do what you want as long as the reward or the punishment keeps coming. You'll never get anything more substantial than that."

I think he's half right: rewards get compliance. But you have to look at it this way: if a behavior isn't occurring, then there isn't an effective naturally occurring reinforcement contingency behind that behavior. So the point of performance management is to add additional components to the environment in order for a desired behavior to occur.

Bribes are different than rewards. Bribes come before the behavior, and often involve a sort of unethical deal. For example, "I'll give you $100 now if you blow the big game on Friday." That's a bribe. "You'll get your allowance when you clean your room." is not a bribe, because the outcome comes after the behavior, and it's an ethical behavior that the person is asked to engage in; it's a rule describing an outcome. Mr. Kohn is confused about this distinction.

I will admit, that there's a problem with performance management. The moment you take the added contingencies away, the behavior is most likely to fall back to where it was before. However, any great performance management contingency will account for this, and slowly fade out the system, so the children will be able to go off their natural environment, more and more, instead of an added incentive program.

"In the course of a day, we tell people things over and over instead of bringing them in and getting their perspective and solutions. For example, a child who comes up with her own remedy for getting out the door on time, every morning, is a child who has been helped to take charge of her own life. She's been encouraged to think about how to solve problems."

I think he's on to something here. However, you have to understand that "problem solving" is a complex skill set, just like organizing your room or writing a paper. If a child does not have these skills, they need to be taught. We don't just develop "problem solving skills" out of thin air.

(continuing from the last quote) "...The child has taken ownership of the solution, and that makes it more likely to succeed. Regardless of the specific result, the process itself is also terribly important. It creates a very different dynamic in a family that's based on respect, not on treating kids like pets."

Yes, we need to focus on the process. However, although I hate to say it, the same principles of reinforcement and punishment that work for your pets, also work for you and I. This is chilling for some people to grasp, and I understand that. But the only thing that separates you from Fido is that you a have a complex set of verbal skills. In the end, we're all just products of our environment.

Question: "If rewards are out, then what should we do when kids refuse to cooperate?"

Kohn: "I think we have to look at the deeper reasons why the child is not interested. In many cases involving school, the trouble lies not with the child's motivation, but with the task and the way it's being taught"(emphasis added).

"With the way it is being taught." If there's a motivational issue, and if the behavior is not occurring at a desired frequency, rate, whatever... then there's not an effective reinforcement contingency behind that behavior, or in other words, rewards.

"The skills and intellectual exploration nested in the tasks we ask kids to perform must be connected to their own questions and interests. Many schools don't do this, so they "have to" manipulate kids with grades, stickers, and the like. It's a reflection of a problem with their teaching strategy, and parents should know this."

I think Mr. Kohn should start a school where you can come and go as you please, there's no incentive to do anything, and there's no competition because "you're all hear to advance as individuals." And see how many people show up. Get back to me with those numbers, but I can already guess that it will be close to zero.

I must show my respect to Mr. Kohn. He seems like a very well read, very successful individual. But I think he's misled about how rewards work in our environment, as subtle or as substantial as they may seem. If it weren't for rewards, he wouldn't have written the many fine books he has published, and he wouldn't have raised his kids in a probably well rounded manor. If we took the rewards out of Mr. Kohn's life, he'd lay in bed, all day long, staring at the ceiling, with nothing to live for.

As always, if anyone has anything to add to this, please submit a comment. I'd be pleased if we can get a discussion going on any one of these posts/ideas.

^[1]Malott 2008