Si quieres ver este blog en versión española, pincha aquí.

Jan 19, 2013

Premack´s Principle (AKA "Finish your vegetables and I´ll give you an ice cream")

First of all, let me apologize for not having updated this blog as soon as I would have wanted.

Today I will talk about a very useful principle in the field of behavior and learning, of application, as usual, in animals of any species, including humans, dogs, dolphins and of course, parrots.

American psychologist David Premack (1925) enunciated, after various experiments with rats and apes, that "
high-probability behaviors (those performed frequently under conditions of free choice) can be used to reinforce low-probability behaviors " thus leading to Premack´s Principle, of paramount importance in animal training and of application, for example, in advanced techniques such as back-chaining. Another way of looking at it, is understanding that reinforcements might be food, form example, or a behavior itself (if it is perceived as something nice, such as playing).

If we transfer this principle to humans, we´ll find it everywhere. When our mother told us to do our homework for an hour so we could go to the park for playing, she was simply applying the Premack Principle. In this example, the access to the desired behavior (playing in the park) served as a stimulus for the less desired behavior (studying), reinforcing the latter. Exactly the same principle is behind the phrase "if you finish the vegetables, I´ll  give you an ice cream."

Although some of the behaviors we want to teach our parrots are simple responses (eg stepping-up on our hand or waving), most of which will be discussed in this blog are more complex (not in the sense of difficulty, but meaning that are behaviors composed of many simple responses, such as to walk to different rings, and put them on different pegs using color discrimination) and are approached by chaining successive responses. These "chains" sometimes are worked "forward" (forward-chaining) and sometimes “backward” (back-chaining), ie from the last response of the complete behavior toward the first one.

Let´s consider an easy example:

Imagine that we want Rufo to take an object and bring it back to our hand (retrieve). Even it sounds as an easy to run behavior, in reality it is a very complex chain of responses, as follows:

1. - Rufo looks at the ball that is on the table.

2. - Rufo walks one step towards the ball.

3. - Rufo walks two steps toward the ball.

4. - Rufo walks up to ten steps and reaches the ball.

5. - Rufo stretches his body and beaks the ball.

6. - Rufo holds the ball with his beak for a second.

7. - Rufo holds the ball with his beak for two seconds.

8. - Rufo turns, holding the ball with his beak, and walk one step towards us.

9. - Rufo walks, holding the ball with his beak, two steps towards us.

10. - Rufo walks, holding the ball, up to ten steps and comes to us.

11. - Rufo opens his beak and places the ball on our hand.


As we see, the desired behavior has been divided into many tiny steps to make learning possible.

Probably (depends on the individual) this retrieve behavior is easier to learn if we first teach Rufo the last responses of the chain (back-chaining), because doing so Rufo will always be moving from an unknown response (and thus less desirable) to an already known (most desirable) one, so that the preferred response will enhance the less familiar one, reinforcing it and facilitating and speeding up the learning process.

Thus, we would begin to teach him to hold the ball and put it in our hand, then we would step back a little and teach him to walk toward us holding the ball, and finally we would teach him to walk away from us to hold the ball off the table. This way, Rufo will perceive as a reinforcement the possibility of doing the first response he learned (last one in the chain, placing the ball on our hand), learning faster the previous responses of the complete behavior.

Returning to humans, it may surprise you to know that many musicians start by learning the last part of the score so its easier for them to learn the begining of it ... so when, being at school, I had to memorize a poem and I used to begin by the last lines of it, I was simply implementing, unwittingly, the Premack Principle.


No comments:

Post a Comment