Most behavioral psychologists advise against the use of punitive methods, and advocate positive methods of behavior control. Punishment usually results in only a short-term behavior change, and it is often accompanied by undesirable and unpredictable side effects. As was illustrated in the case of Joe and his critical boss, the supervisor who uses punishment can become an antecedent that signals pending punishment. Not only does this usually result in many avoidance and escape behaviors such as absenteeism, nonproductive work, and obsessive agreement with the supervisor; it makes it almost impossible for the supervisor to function as an administrator of positive reinforcement. That is, it becomes physically impossible for the supervisor to deliver a positive reinforcement to someone who is avoiding him or her. When the punitive boss attempts to be positive, it is likely to arouse the suspicions of subordinates and to discredit any potentially positive reinforcement delivered.
Because the dynamics of emotional responses are so complex and are not completely understood, it is difficult to predict the emotional response to a punitive experience. Thus, punitive experiences frequently elicit dysfunctional emotional responses such as angry out bursts, breaking of equipment, threats, and sabotage. Related to avoidance and escape behavior is the problem of a temporary suppression of the undesired behavior. The undesired behavior, such as Joe's loafing, is suppressed only as long as the punitive boss is present; but as soon as he leaves, the undesired behavior usually returns. Another problem is the probability of behavioral inflexibility. A behavior that is viewed as undesirable at one time may at another time be considered highly desirable. Yet as a result of past punishment, the behavior may be inhibited or permanently suppressed. For example, the suggestions of a management trainee may be suppressed by derogatory remarks from the supervisor, so that later on, when the supervisor actively solicits suggestions, the trainee may be unwilling or unable to generate any ideas.
A very serious problem with the use of punishment is that punishment begets punishment. The more people use punishment, the more likely they will be to use it again. Punishment is usually followed by the immediate cessation of an undesired behavior {negative reinforcement) and possibly by the appearance of a desired behavior (positive reinforcement). In short, the consequences of punishment tend to reinforce the punitive behavior of the supervisor who delivered the punishment. Once again we have a vicious cycle, one that results in an increase in the frequency of punishment.
Suppose a foreman's job is threatened by a drop in production (antecedent) and that in response to this drop, the foreman upbraids the subordinates and threatens that heads will roll if there isn't an immediate increase (punishing behavior). In response, subordinates work faster (consequence) and production increases. As would be expected, after several days the subordinates again slow down. The foreman will again most likely engage in punishing behavior because it has been both positively and negatively reinforced. The foreman's threats have been positively reinforced by the increase in production and negatively reinforced by the temporary removal of the threat to his or her own termination. The final problem with the use of punishment is that, although it suppresses undesirable behaviors, it does not teach desired behaviors. An awkward teenager's invitation for a date may be punished by rejection. The rejection will tend to suppress socially inept behavior, but it does not teach the teenager how to extend an invitation that will be accepted.
Extinction
Extinction is the process by which conditioning or learning is reversed. In general, conditioned or learned responses established through classical conditioning can be extinguished by continued presentation of the conditioned stimulus in the absence of the unconditioned stimulus. Thus, if the conditioned and unconditioned stimuli are never again paired, the conditioned stimulus will lose its power to elicit the conditioned response. For example, if the dog in Pavlov's experiment had heard the bell ring hundreds of times yet did not get any meat, the bell would have eventually lost its power to make the dog salivate.
There are two exceptions to this general rule. First, if the conditioned response is itself aversive, the conditioning will tend to remain intact. Responses such as anxiety and fear are inherently painful.
Consequently, these uncomfortable sensations themselves become associated with the conditioned stimulus (such as the rat's causing Little Albert to cry) and tend to maintain the learned response (crying). The second exception is the case in which the conditioned response is followed by reinforcement. Each time Little Albert cries, his mother gives him a lollipop. The lollipop is a positive consequence that reinforces and maintains crying behavior. In this situation we would not expect the crying behavior in response to a white rat to extinguish even though the rat is never again paired with a loud noise. Avoidance and escape behaviors are difficult to extinguish because they are negatively reinforced. For example, a dog who has learned to jump out of a compartment in response to a flashing light and an electric shock will probably continue to do so when the light flashes even though the light is never again paired with the shock. Escape behavior is negatively reinforced by the very act of escaping.
In contrast, behaviors established through operant conditioning can be extinguished by withholding any consequence. For example, when gossip behavior is ignored and is no longer followed by a rein forcing consequence such as attentive listening, it will probably stop.
In some cases the process of extinction is functionally similar to that of negative punishment. Negative punishment is the withholding or removal of something positive. When the gossip who is accustomed to receiving attentive listening is ignored, that lack of attention can be experienced as negative punishment. Thus, it is best to simultaneously reinforce a behavior that is incompatible with the behavior to be extinguished. For instance, talking about work-related topics is incompatible with gossiping, because one cannot do both at once. To extinguish gossip behavior most efficiently, a manager should ignore all gossip talk and listen attentively to work-related talk. This principle of reinforcing desired incompatible behavior rather than punishing undesired behavior can be applied to all areas of the work environment. (Examples of incompatible behaviors include standing-sitting, tension relaxation, loafing-working, laughing-crying, frowning-smiling, and assertiveness-passivity.)
The Premack Principle
As was stated earlier, reinforcement is functionally defined as a consequence that makes the performance of a certain behavior more likely. Many people mistakenly assume that reinforcement is synonymous with reward; a reward is probably a reinforcement, but a reinforcement is not necessarily a reward. For example, the quality control inspector's behavior is probably reinforced by identifying a defective item; although one can hardly consider spotting a defective item a reward. The Premack principle of reinforcement states that a high-probability behavior (a behavior that is enacted very frequently) can reinforce a low-probability behavior (a behavior that is performed infrequently). This principle is very important for managers, because when utilized contingently it can help managers use work to reinforce work.
Suppose that a salesperson calls on a lot of old clients, but has very few new clients. The number of new clients called upon can be increased if calling on old clients is made contingent on calling upon a specified number of new clients. Basically, it is a matter of doing the least liked work first and the most liked work second. Generally, any behavior (that is not painful or unpleasant for that person) that a person enacts very frequently can be used as a reinforcer. If a secretary uses the copy machine several times a day, then going to the copy machine can be used to reinforce a low-probability behavior such as filing. In order to be effective as a reinforcer, going to the copy machine must be contingent on a specified amount of filing. Common high-probability behaviors include looking at one's watch, drinking coffee, smoking cigarettes, and talking on the telephone.
Vicarious Learning
To learn that a particular behavior will lead to a specific consequence, a person does not have to engage in that behavior and be personally reinforced for doing so; that person can learn vicariously by watching a model's behavior and its consequences. This phenomenon is known as modeling. For example, if you observe your friend petting a neighborhood dog and then being bitten by it, you will probably avoid petting that dog even though you were not bitten by it.
You learned through modeling that the probability of being bitten is increased by petting the dog.
Modeling is used extensively in all types of training. When you learn to play tennis the instructor usually first models the different kinds of swings you can use. You observe that one type of swing usually results in hitting the ball over the net and another does not, A sales trainee can learn to give a sales pitch by watching the trainer model a sales pitch. (Trainees will learn more rapidly, of course, if after watching the model they then practice giving a sales pitch themselves and are reinforced for doing so.)
A person does not necessarily even have to observe a behavior/ consequence to learn the probable consequences of that behavior. For example, your friend could tell you that she was bitten by the dog after petting it. Based on this information, you would know that you’ve likely to be bitten if you pet the dog. This is one form of symbolic learning. Through the use of words (symbols) and abstract concepts you can learn contingency relationships. Being instructed or reading instructions is another form of symbolic learning.
Superstitious Behavior
Superstitious behavior is designed to effect a consequence when there is actually no causal relationship between that consequence and the behavior. Blowing on dice before rolling them probably does not influence the roll, and knocking on wood probably has little influence over future events. The ace pilot who while flying must carry a talisman is engaging in superstitious behavior. The person who pushes the elevator button (provided it is not defective) several times is engaging in superstitious behavior, because pushing a button more than once does not bring the elevator faster.
Superstitious behavior is established through chance. Suppose that the wiring in your car is defective so that you must turn the key several times in order to start the car. If you turn the key in vain, then bang on the dashboard and the car starts on the next key turn, you might mistakenly assume that banging on the dashboard resulted in the car's starting; in fact, the association between the banging and the starting of the car was an accident. Yet the car did start (consequence) after you banged on the dashboard (behavior), and essentially this is how superstitious behavior is established, Superstitious behavior can become very resistant to extinction because it is usually maintained by a variable schedule of reinforcement: Because there is no real cause-effect (contingency) relationship, the desired consequence occurs occasionally by chance. Superstitious behaviors that are maintained by negative reinforcement are extremely resistant to extinction. Ace pilots carry charms so that they will not crash. Each time they fly while carrying the talisman and don't crash (a negative stimulus that is withheld) their charm-carrying behavior is negatively reinforced. Only if a pilot were to crash several times while carrying the charm would that particular behavior be extinguished.
Cognitive-Symbolic Processes
Individual behavior patterns are largely influenced by cognitive mediation. Thus, although rats may respond in a very predictable fashion to the presence of an antecedent cue and to the subsequent consequences, people's behavior is affected by their perception of the cue and by their anticipation of the consequences. Feelings, physical sensations, thoughts, fantasies, or visual images can function as antecedent cues, as behaviors, or as consequences in the same manner as overt actions do. For example, a sexual fantasy may function as an antecedent cue that elicits sexual behavior, as a positive reinforcer for the behavior of looking at attractive people, or as a behavior itself. Cognitive-symbolic processes play an important role in self-instruction and in self-reinforcement. One person may respond to a particular antecedent cue such as an insult by instructing himself to count to ten; another person may reinforce herself by thinking positive thoughts about herself. Self-reinforcement is an important component of self-esteem, and it plays an instrumental role in maintaining what could be called self-directed behavior. The department self-starter probably uses a lot of self-instruction and self-reinforcement.
Cognitive-symbolic processes can influence a person's perceptions of an antecedent cue. If you offer to take an employee who is a vegetarian for a steak dinner, he or she may respond with hostility rather than with the expected appreciation. The employee may have used various thoughts and images to change the invitation from a positive to a negative event. Such internal processes can also influence people's perceptions of consequences. You may intend to reinforce an employee by using praise, but if that person thinks, "I know he's just saying that to get me to write the report," then the reinforcing power of the praise is negated.
Because of the existence of cognitive-symbolic processes, modification of complex work behavior can be difficult; these internal processes can be modified, however. Research has demonstrated that when they are defined as behaviors, they do function in accord with the laws of learning.
Response Frequency
No doubt you have noticed that throughout this chapter there have been numerous references to probability (i.e., a behavior is more or less likely) and frequency. Measurement is at the heart of behavior modification: What people do or how they do it is not as important as how often they do it. Only when you measure the frequency of response can you determine if a behavior is increasing, decreasing, or remaining the same. Knowing this is essential to determining whether a consequence is functioning as a reinforcer or as a punisher (recall that these are defined in terms of whether appearance of the behavior is more or less likely), or whether a behavior is in the process of extinction. Likewise, counting behaviors helps to determine if a particular antecedent cue has stimulus control over a particular behavior. Finally, measuring response frequency is an essential step in both monitoring and evaluating a behavior change strategy.
