Greta Kaplan, CPDT
In the first session of Chicken Camp, “Discrimination,” we taught our hens to choose and peck a colored target. Since the hens had done this before, they already had learned to peck a specific color. So, we tested them by placing the three identical targets (other than color: red, yellow and blue) in front of each hen. My hen pecked yellow, so I removed it. Then she pecked red, probably indicating that she had, at some time in the past, been reinforced for pecking yellow. Blue therefore became my new “hot target” and my job was to teach her to peck only the blue target. Our goal was to see if, eventually, she would refuse to peck the yellow and red targets *even if the blue target was not on the table, for 20 seconds.*
Certain rules applied. We could not use lures to get the behavior: No hiding a grain of food behind the blue target to get her interested in that part of the table. However, I could remove the blue target to permit her to extinguish the yellow and red pecking behaviors. If she visibly paused, withholding pecks, I would quickly slide the blue target into place so that she could peck it and earn reinforcement. This illustrates the power of the Premack Principle: A higher value behavior reinforces a lower value behavior! And by contrast, if she pecked the wrong target after a certain point, I could quickly pull away the blue target to remove the opportunity for reinforcement for a while. This is negative punishment and it reduces the likelihood of pecking the wrong target.
Once we had switched our hens to the new hot target, we performed a stimulus reversal. Until now, we had been working our hens in 30 and 60 second periods. For the stimulus reversal, we would shape nonstop until the hen had met criteria for selecting the new target color (in my case, that was red, since that had been less preferred to begin with). It took seven minutes for my advanced bird to reverse her stimuli, and that shaping session was wonderfully fun. One important lesson here: Since the old hot target was present, the first step was to click *any* behavior other than pecking the hot target — even just holding her head up away from the target was clickable, and was rewarded. Waiting for her to choose another target would have taken a long time and created “ratio strain.” The low rate of reinforcement would have reduced her interest in working and she might have wandered off or perhaps engaged in frustration behaviors. Always choose criteria at which your animal can succeed!
During the second session of Chicken Camp, “Cueing and Criteria,” among other things, we worked on attaching a cue to a behavior. Amy, my training buddy and roommate, had a beautiful success with this. She trained one of her hens to circle in place. She then taught the hen that the cue for doing this was a pen, held vertically. As long as the pen was in this vertical position, the hen was to keep circling. During her demonstration, Amy placed her hen on the table. The hen stood there for a moment, waiting for information. Then, with a very still body, Amy lifted the pen into the vertical position. The hen immediately started spinning in place. She spun in place for about 15 seconds, and then Amy lowered the pen. The hen stopped, and received a single click and treat. It was a very clean demonstration. Amy had attached this cue in just one morning’s worth of training sessions. Attaching cues is not particularly hard, once you know the trick: First you get the animal reliably doing the behavior, and then you deliver the cue when you know the animal is about to perform the behavior. You then click and treat when the animal performs. After multiple repetitions, the animal learns that the cue predicts that the specific behavior will be reinforced, and so performs that behavior when she perceives the cue. The flashy part of Amy’s demonstration was the indiscrete nature of the behavior and the cue: the behavior was to continue until the cue went away.
In my next entry, I’ll talk about what we did in our third session, “Chaining.”
Leave a Reply