Copyright © 2007-2017 Russ Dewey
If a behavior occurs because of a reinforcing stimulus, then removing the stimulus will make the behavior gradually disappear. That is exinction.
Extinction happens if a behavior is extrinsically motivated: reinforced from outside. Some behaviors are intrinsically reinforcing, which means the reinforcement comes from inside. Then a behavior will continue without external reinforcements.
What is intrinsic vs extrinsic motivation?
Extinction of an externally-motivated behavior occurs when the reinforcing stimulus is removed. The animal usually tries it again at a later time to see if maybe the reinforcer is available again.
What is spontaneous recovery, in extinction, and how is it overcome?
The re-appearance of an extinguished behavior is called spontaneous recovery in Skinnerian conditioning just as it is in Pavlovian conditioning. An extinction procedure must be carried out several times to completely eliminate a behavior, because of spontaneous recovery.
One of the useful principles discovered by behavioral psychologists is that intermittent reinforcement causes resistance to extinction. The word intermittent means not every time.
Intermittent reinforcement is the opposite of continuous reinforcement. Under conditions of continuous reinforcement, every time the rat hits the bar, it receives a food pellet. Under intermittent reinforcement, the rat might be required to hit the bar many times to get a pellet.
What is intermittent reinforcement?
Intermittent reinforcement can be based on schedules. For example, in the pattern of reinforcement called a fixed interval 5 schedule (FI5) a rat is reinforced for the first bar-press after five minutes. This produces a predictable pattern: faster behavior (more bar-presses) as the reinforcement time grows closer.
B.F. Skinner and his colleagues studied schedules of reinforcement in detail, analyzing the distinctive patterns of behavior caused by each. The fastest responding is produced by variable ratio schedules, in which the rat must press the bar many times to get a pellet, but it never knows how many times (the ratio of bar-presses to pellet deliveries is variable).
Variable ratio schedules produce steady, fast behavior, along with resistance to extinction. The behavior is fast because the number of reinforcers depends on how fast the animal responds. The behavior resists extinction because the animal never knows when a reinforcement will occur.
Gambling is an example of a human behavior that is on a variable ratio scale. The ratio of behaviors to reinforcements (the number of bets required to get a win) is variable and cannot be predicted.
However, the more times a gambler places bets, the more often a winning bet will occur, so this encourages a high rate of responding (many bets). The payoffs are irregular, so there is resistance to extinction.
What schedule of reinforcement produces the fastest and most durable rates of response?
The "resistance to extinction" phenomenon is relevant to parents dealing with toddlers who throw tantrums. Experts on parenting advise parents never to reinforce a tantrum.
The worst thing a parent can do (if the goal is to eliminate tantrums) is to give in to a tantrum occasionally. That is an intermittent reinforcement schedule, and it teaches persistence. Maybe this tantrum did not work, but the next one might! So the tantrums continue.
During an extinction period, animals usually show a great variety of behaviors. The animal appears to test variations of behavior to see if anything will make the reinforcers start again.
Epstein (1985) called this "extinction-induced resurgence." Others call it an extinction burst.
Extinction-induced resurgence is a handy phenomenon for animal trainers. If they want animals to learn new tricks, they put the animals temporarily on an extinction schedule (stop delivering reinforcers).
This results in an increase in the activity level and variability of behavior. Then the trainer can pick out a new behavior to be reinforced.
What is extinction-induced resurgence? How is it useful to animal trainers?
In 1992 two trained porpoises escaped from an enclosure and swam out to sea. A few weeks later they turned up in a waterway next to a golf course, performing tricks for amazed golfers.
The porpoises had been trained to perform tricks for fish, but by escaping they put themselves on an extinction schedule (except for catching their own fish). Their behavior at the golf course was lively and variable...and it worked.
Before long some golfers bought a load of fresh fish, and the porpoises were feasting. Porpoises are social animals, and they make eye contact with humans before and after a trick. Attention is a powerful reinforcer for most social animals including humans.
Suppose you were on a golf course by the ocean, and you saw a porpoise close to the shore, doing what appeared to be a show trick (such as swimming on its tail, with its head high out of the water) while looking you straight in the eye. Wouldn't you feel the urge to go buy some fish?
How did some porpoises trains some humans?
In short, it was not only the porpoises being reinforced! This brings to mind an old cartoon where a rat is sitting in a Skinner Box telling another rat, "Boy, have I got that human trained...every time I hit this bar, I get a food pellet!" Humans think they are training the animals, but the animals also train the humans.
Epstein, R. (1985) Extinction-induced resurgence: Preliminary investigations and possible applications. Psychological Record, 35, 143-153.
Write to Dr. Dewey at firstname.lastname@example.org.
Copyright © 2007-2017 Russ Dewey