Chapter 16 Skinner: Behavioral Analysis 475Extinction
Once learned, responses can be lost for at least four reasons. First, they can simply
be forgotten during the passage of time. Second, and more likely, they can be lost
due to the interference of preceding or subsequent learning. Third, they can disap-
pear due to punishment. A fourth cause of lost learning is extinction, defined as
the tendency of a previously acquired response to become progressively weakened
upon nonreinforcement.
Operant extinction takes place when an experimenter systematically with-
holds reinforcement of a previously learned response until the probability of that
response diminishes to zero. Rate of operant extinction depends largely on the
schedule of reinforcement under which learning occurred.
Compared with responses acquired on a continuous schedule, behavior
trained on an intermittent schedule is much more resistant to extinction. Skinner
(1953) observed as many as 10,000 nonreinforced responses with intermittent
schedules. Such behavior appears to be self-perpetuating and is practically indis-
tinguishable from functionally autonomous behavior, a concept suggested by
Gordon Allport and discussed in Chapter 12. In general, the higher the rate of
responses per reinforcement, the slower the rate of extinction; the fewer responses
an organism must make or the shorter the time between reinforcers, the more
quickly extinction will occur. This finding suggests that praise and other reinforc-
ers should be used sparingly in training children.
Extinction is seldom systematically applied to human behavior outside ther-
apy or behavior modification. Most of us live in relatively unpredictable environ-
ments and almost never experience the methodical withholding of reinforcement.
Thus, many of our behaviors persist over a long period of time because they are
being intermittently reinforced, even though the nature of that reinforcement may
be obscure to us.
The Human Organism
Our discussion of Skinnerian theory to this point has dealt primarily with the
technology of behavior, a technology based exclusively on the study of animals.
But do the principles of behavior gleaned from rats and pigeons apply to the human
organism? Skinner’s (1974, 1987a) view was that an understanding of the behavior
of laboratory animals can generalize to human behavior, just as physics can be
used to interpret what is observed in outer space and just as an understanding of
basic genetics can help in interpreting complex evolutionary concepts.
Skinner (1953, 1990a) agreed with John Watson (1913) that psychology must
be confined to a scientific study of observable phenomena, namely behavior. Sci-
ence must begin with the simple and move to the more complex. This sequence
might proceed from the behavior of animals to that of psychotics, to that of men-
tally challenged children, then to that of other children, and finally to the complex
behavior of adults. Skinner (1974, 1987a), therefore, made no apology for begin-
ning with the study of animals.
According to Skinner (1987a), human behavior (and human personality) is
shaped by three forces: (1) natural selection, (2) cultural practices, and (3) the
individual’s history of reinforcement, which we have just discussed. Ultimately,