Amy and I had signed up for sessions 2 and 3 back-to-back. This was 12 days of class with a 3-day break. All this learning is tiring, and session 3 is the longest of the sessions at 7 days. Still, it was fun and stands out in my memory.
Our task was to create a fixed behavior chain which, at the end of the week, our hens would perform with no external cues or reinforcements until she crossed her finish line. The behavior chain was built around the hen proceeding over a gymnastic apparatus specially built for this session. The apparatus consisted of two towers, each with a platform around its post at about a foot above the table. The two towers were placed about three feet and a “balance beam” or catwalk connected them. Finally, a ladder led diagonally to each tower. The hen would climb one ladder, proceed around the *outside* of the post on the platform (instead of jumping straight to the catwalk), move across the catwalk, circle the outside of the other tower, and then descend the second ladder. This sounds complicated enough, but in addition to this sequence, each hen had to perform some other behaviors during her journey on the apparatus. We had a choice of several behaviors. Between the two birds, we had to pick five behaviors: One hen would do two and the other would perform three. One behavior had to involve a cue, and one had to be indiscrete.
I assigned the cueing behavior and one other to my novice hen. Unfortunately, apparently as a result of prior training mistakes by other students, she would not stop pecking the target without a cue. Her extinction bursts for this behavior become more and more extreme as the days passed and it became apparent that the behavior was unusable. In the end, she did her “racecourse” with only one appended behavior.
My experienced hen did not run into such a snag. The three behaviors I settled on for her were to pull a rubber hand which was attached to the edge of the table a certain distance before releasing it; to peck a specific colored target; and to pick up a small stuffed toy from a dish and move it to the table. The rubber band pulling behavior is harder than it sounds because it closely resembles a normal chicken hunting behavior of pulling a worm or grub from the ground. Hens instinctively want to keep pulling until the worm comes loose, so letting it go can be tough. Also, while you are trying to build the length of the pull, she might get snapped while letting go and become nervous of the rubber band. The goal length of the pull was 11 inches, which involved quite a hard, prolonged effort. Teaching the color target selection wasn’t too hard since it was something I had taught before, and the hen had learned (from other students) before. Finally, it seemed that my hen had never before been exposed to the behavior of moving the little “Piglet” stuffed figure from the dish to the table. Teaching this behavior took some time.
Getting my hen doing all these behaviors reliably was not my only task. I also had to get her to move through the racecourse apparatus in the proper order. The hardest part for most of us was convincing our hens to circle to the outside of the posts instead of hopping straight to the catwalk (for the first post) or back to the ladder from the catwalk (for the second post). We worked on this by chaining both forwards and backwards. I’d start my hen at the bottom of the up ladder and click her for climbing it. Then I’d use the principle of feeding to promote your training goals and feed her at the outside edge of the first platform. If she associated the outside of the platform with the food delivery, she’d be more likely to proceed in that direction instead of taking the shortcut. Getting her across the balance beam, again, was not difficult, and then the second platform demanded more careful fine timing and feeding. I’d also start her on the second platform to build up a smooth finish, and so on. (This is backward chaining.)
I had to decide where to have my hen perform each added task. I ended up putting the rubber band pull first. This was risky as hens can get really stuck on pulling rubber bands and if she got stuck, she would blow the entire sequence! However, I needed the other available spots for other behaviors. One behavior had to take place on a platform so I put my targets on the first platform. Finally, I added the Piglet-in-a-dish behavior at the finish line, past the end of the down ladder. This was the hardest of the three added tasks for my hen and I wanted the rest of the course under her belt so she would not get stuck if she failed the Piglet toss.
The day of the test came and my turn took my by surprise. My hen had worked on each of the three behaviors in place… but never at the same time! I had *no* idea if she actually knew and could perform the whole sequence. My job was to stand at the end of the table with the finish line, holding the food cup, and not moving or speaking at all until I could click for the last behavior. My partner placed my hen next to her rubber band and off she went.
I honestly don’t remember much about the next 40 seconds as she progressed through the course. I am pretty sure she didn’t pull the rubber band to 11 inches, but she probably made it to 8 or 9. She went around her tower, pecked the correct target, crossed the catwalk, circled the far target, walked down the second ladder, picked up Piglet, and deposited him on the table. Click! Treat! I was pretty stunned. She had done it just about perfectly.
One of Bob Bailey’s “Chicken Camp Shirts” has the word “Believe” emblazoned on the back of it. Though this sounds somewhat religious, it’s not. It’s just a reminder that the techniques we were learning are very well known principles of learning, and they work, as has been shown in laboratories and real life situations over and over. Watching my hen correctly perform this long sequence was pretty convincing! I will never again dismiss chickens as stupid.