With most cases of operant conditioning, the desired behavior is too complex to be emitted without first being shaped by the environment. Shaping is a procedure in which the experimenter or the environment first rewards gross approximations of the behavior, then closer approximations, and finally the desired behavior itself. Through this process of reinforcing successive approximations, the experimenter or the environment gradually shapes the final complex set of behaviors (Skinner, 1953).
Shaping can be illustrated by the example of training a severely mentally challenged boy to dress himself. The child's ultimate behavior is putting on all his own clothes. If the parent withheld reinforcement until this target behavior occurred the child would never successfully complete the chore. To train the boy, the parent must break down the complex behavior of dressing hito simple segments. First, the parent gives the child a reward say, candy, whenever the boy approxhnates the behavior of positioning his left hand near the inside of the left sleeve of his shirt. Once that behavior is sufficiently reinforced the parent withholds reward until the child places his hand into the proper sleeve. Then the parent rewards the child only for putting his left arm entirely through the sleeve. Following this, the same procedures are used with the right sleeve, the buttons, trousers, socks, and shoes. After the child leams to dress himself completely, reinforcement need not follow every successful trial. By this time, in fact, the ability to put on all his clothes will probably become a reward hi itself. Quite apparently, the child can only reach the final target behavior if the parent breaks up the complex behavior into its component parts and then rehiforces successive approximations to each response.
In this example, as hi all instances of operant conditioning, three conditions are present: the antecedent (A), the behavior (B), and the consequence (C). The antecedent (A) refers to the environment or setting hi which the behavior takes place. In our example, this environment would be the home or any other place the child might be putting on clothes. The second essential condition in this example is the boy's behavior (B) of dressing himself. This response must be within the boy's repertoire and must not be hiterfered with by competing or antagonistic behaviors, such as distractions from siblings or television. The consequence is the reward (C), that is, the candy.
If reinforcement increases the probability that a given response will recur, then how can behavior be shaped from the relatively undifferentiated into the highly complex? In other words, why doesn't the organism simply repeat the old reinforced response? Why does it emit new responses that have never been reinforced but that gradually move it toward the target behavior? The answer is that behavior is not discrete but continuous; that is, the organism usually moves slightly beyond the previously reinforced response. If behavior were discrete, shaping could not occur because the organism would become locked into simply emitting previously rehiforced responses. Because behavior is continuous, the organism moves slightly beyond the previously rehiforced response, and this slightly exceptional value can then be used as the new minimum standard for reinforcement. (The organism may also move slightly backward or slightly sideways, but only movements toward the desired target are reinforced.) Skinner (1953) compared shaping behavior to a sculptor molding a statue from a large lump of clay. In both cases, the final product seems to be different from the original form, but the history of the transformation reveals continuous behavior and not a set of discrete
Operant behavior always takes place hi some environment, and the environment has a selective role hi shaping and maintaining behavior. Each of us has a history of being rehiforced by reacting to some elements in our environment but not to others. Tins history of differential reinforcement results in operant discrimination. Skinner claimed that discrimination is not an ability that we possess but a consequence of our reinforcement history. We do not come to the dhuier table because we discern that the food is ready; we come because our previous experiences of reacting hi a shnilar way have been mostly rehiforced. This distinction may seem to be splitting hairs, but Skinner felt that it had important theoretical and practical implications. Advocates of the first explanation see discrimination as a cognitive function,
existing within the person, whereas Skinner accounted for this behavior by environmental differences and by the individuals history of reinforcement. The first explanation is beyond the scope of empirical observation; the second can be scientifically studied.
A response to a similar environment in the absence of previous reinforcement is called stimulus generalization. An example of stimulus generalization is provided by a college students purchase of a ticket to a rock concert performed by a group she has neither seen nor heard but one she lias been told is similar to her favorite rock group. Technically, people do not generalize from one situation to another, but rather they react to a new situation hi the same maimer that they reacted to an earlier one because the two situations possess some identical elements; that is, buying a ticket to one rock concert contains elements identical to buying a ticket to a different rock concert. Skinner (1953) put it this way: "The reinforcement of a response hicreases the probability of all responses containing the same elements" (p. 94).
Was this article helpful?