Why Did Homer Pull the String?
To help us learn how to build prosocial behavior!
Homer, The Home Run Hitting Rooster
I was visiting a tourist attraction featuring boat rides, alligator wrestling, rattlesnake milking, a water slide, and fast food in Silver Springs, Florida. I came upon a 6 ft by 8 ft by 3 ft wooden box with a glass front. Looking through the glass, I saw what appeared to be a miniature baseball field, complete with bases and an outfield with green-painted grass. At home plate, a bat and ball lay on the ground. I noticed that there was a string attached to the bat, laying limply alongside it. A sign by the side of the box reads, “Homer, The Home Run Hitting Rooster.” Beneath the sign is a slot with 25 cents written under it. I inserted a quarter into the slot. A buzzer sounded, a door opened on the side of the box and a chicken ran out. He ran to home plate, yanked the string, and the bat swung and hit the ball into the outfield. The chicken circled the bases in order, stepping on each one, stepped on home plate, and disappeared into the box to the recorded sound of a roaring crowd.
I stepped away from the box so that a woman in line behind me could take my place. She was somewhat overweight, wearing a simple print dress, and flip-flops. She put a quarter in the slot and began jumping up and down and shouting, “Roger, Roger get over here and look at this chicken. Why he knows the rules of baseball better’n you do and you graduated high school!”
Her explanation of Homer’s behavior doesn’t tell us much about why Homer performed as he did. Hers was a common misperception; instead of looking for the cause of Homer’s behavior in his environment, she placed it inside Homer’s head and inferred a learning history not usually evident in chickens. In fact, Homer’s performance was an example of exquisite behavioral engineering that involved the careful application of some basic principles of behavioral technology. Let us examine them carefully. We will do this by discussing the way in which Homer’s act was probably created.
Reinforcement
The first is reinforcement. For a response to be acquired and maintained, it must be reinforced. There are two types of reinforcers - primary and secondary (also called conditioned). Primary reinforcers are things that satisfy basic needs like food, water, reduction of pain, and access to opportunity to mate. Any other event, if paired with one of these in time, can become a secondary (or conditioned) reinforcer. It is highly probable that Homer had occasional access to food in the box to which he retired after each performance.
Shaping Behavior
The question is, how does Homer get the food? We begin to answer this by observing chicken behavior. What do chickens do? They spend a lot of time walking around and pecking at objects on the ground. Some of those objects are edible. Occasionally, they also flap their wings. Presenting chicken feed in a small hopper for a short time will give Homer access to a few grains, enough to reinforce the behavior of approaching the hopper and pecking. We can arrange for the hopper to come up through a hole in the floor or slide out from a hole in the wall. Either way, it should make a distinctive noise, and the duration (1 or 2 seconds) of each presentation be tightly controlled. The noise the apparatus makes as it operates will naturally precede the appearance of the food and will thus become a conditioned reinforcer. We will use that to reinforce a new behavior, pecking a spot on the wall above the feeder. Homer can now operate the feeder himself by pecking the spot on the wall. Now, let’s now make that happen only if a light over the feeder is on. If the light is not on, pecking the spot will not produce the feeder. We call the light a discriminative stimulus (SD); in its presence, a peck will be reinforced by the presentation of the feeder. If it is not on, the peck produces nothing.
Since the light will be paired with the food if Homer pecks the spot, it can also serve as a conditioned reinforcer (SR), reinforcing whatever response produces it. Let’s have it come on whenever Homer enters the area. It will now reinforce the response of walking through the door. In order to make it come on, Homer will have to exit the area and re-enter. At this point, Homer walks through the door, the light comes on, he pecks the spot, and the feeder operates. We have just built a two-link chain culminating in primary reinforcement. What does Homer do now? He probably stands around and waits for the light to come. Sooner or later, he will go back outside and do Homer things. He will then reenter the box, the light will come on, and the chain continues.
This time, Homer will likely go back outside and immediately reenter the box. Does the light come on? Maybe. Maybe not. If it doesn’t come on, Homer will go back outside and try again. If the light doesn’t come on this time, he will go outside and wander around a bit before trying again. Bingo! Light comes on and the chain continues. This time when Homer goes outside and wanders around, we wait for him to be near home plate when he makes a move to reenter. The light comes on and again, the chain continues. Next time out, we wait for him to be closer to home plate before reinforcing the entry sequence. We are shaping Homer’s wandering by this procedure and it may be time-consuming. We wait for Homer to step on home plate before entering the box. Now the light comes on and the chain will continue.
When Homer steps on the home plate, it sinks into the ground a few millimeters and makes a click sound. This compound stimulus (i.e, the click and the hepatic effect of the plate moving) becomes the discriminative stimulus for running into the box. Unless Homer steps on home plate, running into the box will have no effect except to irritate Homer.
We have now built a behavior chain that looks something like this:
Each of the SD→R→SR combinations is called a link and the stimuli are called linking or bridging stimuli. They each serve two functions: they reinforce the response which produces them and signals the next response in the chain. (You’ll notice that an SD can also function as an SR). To complete Homer’s training, we would simply add links to the chain. Stepping on the home plate would reinforce running from third, and stepping on third would reinforce running from second. The first response in the chain would be pulling the string that activates the bat and sends the ball into the outfield. The availability of the string reinforces the behavior of running out of the box. The SD for that might be a flashing light over the door; it happens whenever a spectator drops a quarter into the slot. The ball going into the outfield is the SD to run to first, and so forth.
Behavioral Engineering
Now here is where behavioral engineering comes into play. Remember, we are dealing with a chicken and our aim is to produce a chain that will allow patrons to exclaim, as that woman did, “He knows the rules of baseball!” If Homer runs to first and continues into the outfield, then pecks at the ground, people will not be impressed…they might even boo! How do we keep that from happening? We use a little trick called a limited hold. The reinforcer at second is only available for a short amount of time after Homer steps on the first base. Same at the second and third.
Another critical variable is Homer’s (to use a popular term) motivation. Remember, Homer is doing this for food. Too much food during the day and he will probably stop responding. Too little and he may grow too weak to perform. We need to determine how much food Homer needs to stay at his best. We then put the whole chain on a schedule; not every run has to end in food. As long as all the conditioner reinforcers along the way are delivered, Homer will continue to perform. On busy weekends and holidays, the schedule will probably be a variable ratio. That is, on average, he gets fed after every so many runs. On rainy days, when very few if any people are in the park, we put Homer on a variable time schedule. That is, every so often, the signal for Homer to run out of the box will be delivered and he will do his act, even if there is nobody watching (Remember that Homer is playing baseball for the food, not for our entertainment). The times could be determined by the intervals between food deliveries on the variable ratio schedule. This way Homer’s food is always response-dependent and will maintain his behavior.
This Doesn’t Explain Human Behavior…Does It?
Yes, it does. Most complex human behavior is composed of chains. Daily activities like eating, dressing, and driving to work can be analyzed as chains. Order is often important; you don’t put on your shoe until after you put on your sock, for example, but you can put your socks on before or after your trousers. Many teachers and therapists speak of task analysis when breaking a complex chain into its components. A special kind of linking stimulus is one in which the reinforcing function and the discriminative function can be separated in time and space. Such a stimulus is called a token. You mow the lawn today, receive money, and exchange that money for a ticket to the movies tomorrow. Tokens are almost always tangible and are often symbolic, like point entries in an account. Regardless of the form, they are very useful when it is impossible to provide reinforcing events at the same time and place the earning response occurs.
The Token Economy
It is possible to set up systems in settings like schools and therapeutic institutions wherein tokens are dispensed for desirable behavior and are later exchanged for tangible items like toys, personal grooming aids, candy, and other edibles or access to a preferred activity like playing a board game or attending a concert. The exchange takes place in a specific area, often called a store, and usually at specified times. In designing a token economy, it is important to create a menu of items and activities, each with a price in the token currency. Prices can be adjusted according to demand. Students may earn tokens by completing assignments, for example, and later exchange them in the store. It is also possible to remove tokens for miscreant behavior, analogous to a fine in civic life. One of the authors (HSP) learned the hard way that tangible objects like poker chips or play money should not be used in elementary or secondary schools. Why? Two reasons. They are easy to counterfeit and they are available to reinforce any behavior. Including robbery and extortion. Better to use a point system whereby each participant carries an account book where points can be added or subtracted under supervision. If desired, more elaborate economies can be created which involve taxation which provides salaries for class officers and patrol kids. Those salaries are, of course, tokens themselves, and can make service a truly rewarding experience.
The ultimate in complexity could occur in a school district in which several schools each had their own token economies. This would create the conditions for inter-school trade with currency markets, exchange rates, etc. Complex economic principles involving trade, deficits, surpluses, etc. would be readily observable and could be of great educational value.
Finally, a token economy in a school could be faded into the real surrounding economy. The school currency would be set at a certain dollar value and tokens earned at school could be exchanged for dollars at the local bank. This practice would simulate the emergence of cryptocurrency in the larger economy.
A conditioned reinforcer that has been paired with many other reinforcing events is called a generalized conditioned reinforcer. Money and praise are the two most commonly mentioned examples of such reinforcers. It is important to remember that the term reinforcer implies that a functional relation exists between the behavioral event and its consequence. Thus, a reinforcer increases the frequency of the behavior that produces it. Many of us have had the experience of working with a teacher or other service provider who says “Well, I tried reinforcement but it didn’t work.” He probably tried an M&M and the kid spit it out. For that kid, an M&M was NOT a reinforcer. Some generalized conditioned reinforcers lose their potency over time.
Application of Reinforcement for a More Prosocial Culture
A basic strategy for increasing prosocial behavior involves making any event that signals benefit for another person or group into a generalized conditioned reinforcer. For example, picking up litter along a highway is reinforced for many people by the sight of a litter-free stretch of road. Having an official bestow public praise for that accomplishment can multiply the reinforcing effect. The group that accomplished the cleanup “feels good” as a result of the pairing. Over time, the good feeling one gets when performing a civic deed can maintain similar good behavior in the future. Service organizations like Lions, Rotary, and Kiwanis are adept at providing supporting events like banquets, parades, and award ceremonies that maintain the prosocial behavior of individuals or groups within their ranks. Many schools emulate this process at the elementary and secondary levels, with beneficial results for the larger community.
“We are only just beginning to understand the power of love because we are just beginning to understand the weakness of force and aggression.” - B.F. Skinner
However, care must be taken to avoid making participation aversive. If a student engages in civic behavior solely to avoid aversive consequences like low or failing grades, the behavior is not likely to persist once the individual is no longer subject to such contingencies. It may be necessary to use such contingencies initially to bring the individual into contact with the more powerful contingencies described here. Ideally, they have already been exposed to them earlier in their educational history and are susceptible to influence by the reinforcers arranged by the supporting culture.
Thus, creating a culture that promotes and reinforces civic behavior is essential for encouraging individuals to engage in such behavior over the long term. As such, schools and other institutions must focus on fostering an environment where individuals are motivated to participate in prosocial behavior, rather than simply participating to avoid negative consequences. By doing so, we can help develop a generation of responsible and engaged citizens who are committed to building a better future for all.
Pass it on and see you next week.