Rewards and Punishment Reinforcement Theory

Rewards and Punishment

Reinforcement Theory holds that consequences either increase or decrease a behavior. Positive reinforcement is the "presentation of a reinforcer (satisfying stimulus) contingent upon a response that results in the strengthening of that response" (Driscoll, 1994). Examples include praise, a reward, or a gift used to encourage behavior. Punishment, on the other hand, is used to weaken a response, or decrease an inappropriate behavior. Although critics of positive reinforcement and rewards present valid criticisms, their objections are mainly because they have extended the scope of the application of what positive reinforcement and punishment are intended to accomplish.

Arguments against punishment have been around for some time. Skinner is perhaps the most noted early anti-punishment researcher. He opposed punishment for five reasons (Skinner 1953):

Its ineffectiveness in decreasing inappropriate behavior

Considerations that the negative behavior being punished is not operant, but caused by some other process

The production of side effects

The ethics of using punishment when other techniques are viable

The quality of life that aversives adversely affect Skinner

In his discussion of punishment and children, Skinner (1953) states, "Punishment informs the child that a particular behavior is unacceptable but punishment does not necessarily inform the child of what alternative behaviors are acceptable...In addition, punishment, if not carefully used, might cause emotional reactions in the child which may make learning quite difficult."

More recently, rewards have also been attacked as a means for modifying behavior. Kohn (1993) questions the advisability of reinforcement, even the use of extrinsic rewards, offering the following disadvantages:

1. Rewards punish because they are controlling, and also because some people do not get the rewards they deserve -- which essentially makes them indistinguishable from punishment.

2. Rewards create or exacerbate imbalances of power, and disrupt relationships. Knowing that someone is sitting in judgment contributes to a level of anxiety that interferes with performance, and a feeling of being evaluated rather than supported.

3. Rewards do not attend to the reasons that the trouble developed.

4. Rewards discourage risk-taking. When working for a reward, people only do what is necessary.

5. Rewards smother peoples' enthusiasm for activities they may otherwise enjoy, and undermine the intrinsic motivation that promotes optimal performance.

A well noted critic of Skinner's view of punishment is Richard Solomon. He claims that there was no scientific basis for Skinner's conclusions, citing conflicting data at the time of his research. Even more than a decade later, Solomon claims,

Our laboratory knowledge of the effects of punishment on instrumental and emotional behaviour is still rudimentary -- much too rudimentary to make an intelligent choice among conflicting ideas about it. The polarized doctrines are probably inadequate and in error. The popularized Skinnerian position concerning the inadequacy of punishment in suppressing "instrumental" behaviour is, if correct at all, only conditionally correct."

Still other researchers such as Baron (1977) state that punishment can work under certain conditions: "(a) if you can punish almost every time, (b) punish immediately, - punish in socially acceptable ways, and (d) do not punish harshly or become overly angry." Regardless of scientific evidence or suspected measures that would make punishment more effective, the intent of punishment in Reinforcement Theory is to suppress inappropriate behavior, not to solve underlying problems. Considering the scope of the application of punishment, it is, in many instances, neither practical nor desirable to replace punishment with a rewards system. Crime is an obvious example. True, suppression may only be short-term with punishment as the only redress, but this is why the length of the prison sentence is designed to match the severity of the crime.

Several behavioral researches have contributed to understanding the benefits of rewards. For example, Thorndike defined the Law of Effect which says: "behavior that is followed by satisfying consequences will be more likely to be repeated and behavior that is followed by unsatisfying consequences will be less likely to be repeated." (cited in Ormrod, 1999) Skinner would add to this understanding with his Law of Conditioning and his Law of Extinction (cited in Ormrod, 1999):