Joan Orr: That leads me to another question I had. This is based on a discussion I had with Anne, my daughter who is working on her BCBA right now, and is on the call here, along with two other students who signed up for this. This question is about shaping.

Karen and I talked about this yesterday. The description in the textbook says that shaping includes extinction, so you put one behavior on extinction and then you move to the next behavior. This is an old-fashioned description of shaping and it’s really not what we do anymore. Karen, I wondered if you would just address that, for the record here.

Karen Pryor: Absolutely. I’m sure that is not what Skinner meant when he first discovered shaping. He discovered it on an afternoon in a laboratory in Minnesota somewhere. Playing around with the pigeon, he and a couple of his students discovered that by clicking the feeding mechanism you could mark any number of behaviors, and he taught the pigeon to knock a ping-pong ball around – in about five minutes. He was using shaping as we understand it, and there was never a point in which the bird did not get reinforced for touching the ball. It just got reinforced for more and more active touching the ball.

I was amazed to discover a year ago when Ken Ramirez and I were working on a survey paper of modern marker-based training, covering history and best practices, for a chapter in a text book. I was amazed to discover from the editor, from his standpoint, that this kind of plunge into the despair of extinction and then come out of an extinction burst and find out, oh yes, some people did this over and over again, was the only way he knew of doing shaping. I think it’s very cruel. There’s nothing worse than extinction. I hate it myself. We all do, you know. The computer stops working and you just get so frustrated and angry. It has nothing to do with shaping. It’s a training procedure but it is not shaping. And it’s not a good training procedure, in my opinion, not even a safe one.

As we were talking earlier, Joan, you and Theresa were talking about observing the child and observing the behavior. You want to stop before the frustration and anger and extinction burst. You want to stop way before that, and change the circumstances so that they are not so aversive for the child. I think it’s a crazy idea myself. They put this in the textbooks, and I don’t understand how people could do that for so long without taking a look at the effect on the learner.

Joan Orr: I think what’s really going on when we are shaping we start reinforcing an increase in criteria that we’ve already seen starting to happen. And that behavior is reinforced, and that behavior reinforces the behavior before that. So we are building a chain.

So, when he is going down the stairs (boy in pool video), going down two steps, then he goes down three steps; going down the third step reinforces going down the second step. Then he goes to the fourth step, which reinforces going down the third step. Then jump (into pool) and that reinforces going down the whole set of stairs. So we are building reinforcers on top of reinforcers on top of reinforcers, and not going into extinction at all.

Karen:  Absolutely. For example, how unhealthy would it be if you had to extinguish stepping on the second step before you could reinforce stepping on the third step? That’s not what we do at all. You’ve got it right with the chain. We build these little chains, and everything that previously happened becomes reinforced by what’s happening now.