Science in Christian Perspective



Behavioral Views of Punishment:
A Critique

Psychology Department 
Western Conservative Baptist Seminary 
5511 S.E. Hawthorne Blvd.
Portland, Oregon 97215

From: JASA 34 (September 1982):

Behavioral psychologists, notably Skinner, claim that punishment is ineffective and produces a number of adverse effects. The nature and types of punishment are discussed. Evidence on the effects of reinforcement and punishment is reviewed, alternatives to punishment are presented, and biblical teachings regarding punishment are reviewed. It is shown that punishment is effective and that the effects of reinforcement andpunishment are similar in nature but opposite in direction. While biblical teachings clearly advocate punishment and thus imply that it is effective, there is clear convergence of biblical and behavioral emphases in encouraging use of alternative approaches.

The question of what role, if any, punishment should play is one of the most controversial areas in behavioral psychology. It is also an issue that raises significant concerns for those who hold a Christian perspective and who believe that the Bible advocates the use of punishment. In the discussion that follows, we examine how behaviorists use the word punishment, consider supporting data and arguments for and against the use of punishment from a behavioral perspective, discuss alternatives to punishment in dealing with problem behaviors, and explore how these compare and contrast with a biblical perspective on punishment.

Definition and Forms of Punishment

Behavioral psychologists use the word punishment in a somewhat unique way. Where our normal social definitions of punishment tend to focus on what is done to the person, a behavioral definition of punishment focuses on how the person responds to the stimulus event in question; that is behavioral definitions of punishment focus on the function of the stimulus in affecting the behavior of the person who receives it. Two basic classes of stimulus events are functional in affecting behavior: removal of a stimulus, and presentation of a stimulus. Punishment is functionally defined by a decrease in the frequency of a target response when a stimulus is either removed or presented following that response.1

 Our definition of punishment suggests that there are two broad classes of conditions that may be termed punishment; what these two classes have in common is their effect in reducing the frequency of the behavior that they follow. Each of these is described in turn, along with Time Out, another procedure that has similar behavioral effects.

One other aspect of the relationship between reinforcement and punishment is important. A stimulus that will strengthen a response when it is presented following it, will also weaken a response if removed following that response. Thus the same stimulus events can function either to weaken or strengthen responses.

Punishment by Presenting a Stimulus

A punishing stimulus is defined as a stimulus that results in the reduction in the frequency of a response if it is presented following that response. A wide variety of events may function as punishing stimuli: spanking, being yelled at, scolding, the word NO, being slapped, electric shock and so on. The same stimulus event may function in different

Portions of this material were presented at the Annual Meeting of the American Scientific Affiliation at Taylor University, Upland, Indiana in August, 1980, adapted from Behavioral Psychology: A Biblical Perspective, published by CAPS-Harper and Row (1981).

 ways for different people or organisms. Thus we must be careful, when we describe a particular stimulus event as a punishment, that we are clear that this event functions as a punishing stimulus.

If we can show that presenting a certain event after a response, say a verbal reprimand after eating with fingers' results in a decrease in the frequency of that behavior, then we may describe the process as punishment. Punishment by a stimulus is probably the most commonly used form of punishment in our society.

Punishment by Removing a Stimulus: Response Cost

The removal of a stimulus following a response may also result in a decrease in frequency of the response. When a response is decreased in frequency by the removal of a stimulus we term this punishment response cost. Traffic tickets and fines for legal offenses are the most common examples of Response Cost procedures used in our society. Of course, removal of a particular stimulus may decrease response frequency in one person, have no other effect on another, and increase response frequency in a third person.

Time Out

Punishment by presenting a stimulus and Response Cost involve stimuli following a response. Another procedure that has similar effects involves the alteration of events that precede responding. Time Out involves either of two forms of changes in events preceding responding that result in a decrease in the frequency of a response. The first form of Time Out involves removing the child, for example, from the environment in which a variety of responses may be reinforced. Isolating a child in his/her room following tantrums, with the result that tantrums decrease in frequency, is an example of a Time Out procedure.

The second form of Time Out involves the contingent removal of a discriminative stimulus in the presence of which a variety of responses are reinforced. A frown on Mom's face (or the absence of a smile) may be an example of this type of Time Out if Mom's frown signals that one or more responses in her presence such as requests for cookies, asking her to play a game, or approaches for hugs will not currently be followed by reinforcement. If the frown results in a decrease in the frequency of responses when it is present, it might be an example of Time Out in this second sense. Prison involves some of the elements of Time Out in this second sense.

The common feature of the two forms of Time Out described here is that both involve a temporary termination of the opportunity to obtain reinforcement for certain responses. Time out does not fall within our definition of punishment given above. However, because of the similarity in behavioral effects it will be useful to deal with the Time Out procedure as essentially similar to punishment as we have defined it.2

The Punishment Controversy

Does Punishment Work?

Skinner believes that the use of punishment is widespread in our society because of the immediate effects that it produces, but he claims that "In the long run, punishment, unlike reinforcement, works to the disadvantage of both the punished organism and the punishing agency."3 Skinner cites evidence from studies that both he and Estes conducted, which he interprets as support for the conclusion that punishment is not effective .4

In contrast with Skinner, Staats is a behavioral psychologist who believes that punishment has a legitimate and necessary role. "Actually, in our present state of social advancement, it is impossible to raise a socially controlled child without the use of some form of aversive stimulation. It is thus important to ... minimum its adverse effects and maximize its productive effects.'" Staats suggests several guidelines that he believes, are important in meeting this objective.

When punishment is employed, it is suggested that it be as infrequent as possible, as slight as is necessary to be definitely aversive, applied immediately but of short duration, and be paired with words so the words will later on be capable of substituting for the direct punishment.6

These words, which will come to produce negative emotional responses much like the unconditioned aversive stimuli, will later be enough to prevent the occurrence of undesirable responses. Staats advocates the use of what we have called Time Out. He also reminds us of the importance of using statements about the consequences of behavior to help the child come under control of language as well as primary punishment.

Probably the most serious objection to punishment raised by Skinner and his colleagues is the claim that punishment doesn't work. In an extensive review of research on the effects of punishment, Azrin and Holz criticize Skinner's interpretation of the data regarding the effectiveness of punishment.

Azrin and Holz conclude that in the Skinner and Estes procedures the introduction of punishment along with extinction could have served as a discriminative stimulus which indicated that reinforcement would no longer occur; termination of the punishment reinstated the original conditions and extinction proceeded in the normal fashion. The accuracy of this interpretation is supported by the results of an ingenious experiment in which Azrin and Holz used a pseudo-conditioning procedure of changing the key light from white to green (a neutral stimulus) for pigeons who had been reinforced only in the presence of a white key. They found effects similar to those previously obtained with shock, suggesting that the discriminative rather than punishing effects of shock and bar slaps had produced the effects found by Skinner and Estes. Azrin and Holz conclude that shock and bar slaps served notice that food was no longer forthcoming, rather than having a punishing effect. Consequently, the data from the Skinner and Estes studies do not bear on the question of the effectiveness of punishment.7

Azrin and Holz not only question the interpretation of the results of the Skinner and Estes studies, but go on to present ample data to support their conclusion that punishment is a highly effective method for reducing the frequency of responses. With mild punishment there is a characteristic recovery of the base rate of the behavior when punishment is discontinued. However, they note that with severe punishment it has been shown that the results are almost irreversible. "One of the most dramatic characteristics of punishment is the virtual irreversibility or permanence or the response reduction once the behavior has become completely suppressed."8

Other Effects of Punishment

A number of additional objections to punishment have been raised by behaviorists. First, punishment results in a tendency to avoid the punishing agent. When this happens with key social agents such as parents and teachers, the child loses the opportunity for important learning experiences. At the extreme, the person may become a social isolate. While this is an important concern, Staats points out that the tendency for punishment to produce avoidance may be counteracted by any reinforcement provided by the same person.9 The effects of reinforcement and punishment on social attraction are opposite in direction. If reinforcement is the principle mode of interaction, then occasional punishment will not have a serious impact on social avoidance.

A second adverse effect attributed to use of punishment is that the punishing agent is modeling aggressive behavior, and that the person receiving the punishment is likely to adopt these behaviors." To some extent this is true; however, the relationship between modeling and imitation is complex. A number of factors are known to interact with the experience of observing a model in determining whether imitation will occur, including sex and social status of the model, context, consequences to the model and consequences to the observer. Thus under appropriate conditions, the adverse effects on imitation of punishing behavior are not likely to be serious. Further, when socially appropriate punishment methods are employed, it is probably desirable that the recipient imitate the observed behavior.

Probably the most serious objection to punishment is the claim that punishment doesn't work.

A third adverse effect of punishment identified by critics is the fact that it produces a number of emotional effects. These emotional responses are essentially respondent behaviors; that is, they occur whenever certain stimulus events are present whether or not they follow another response in contingent manner. Thus they occur whenever punishment occurs, but are not limited to such occasions.11

The unpleasant emotional effects of punishment come to be associated not only with the punishing stimulus but with all stimulus events that occur at the time of the punishing event. The emotional effects are thus associated with the punishing agent, the situation in which punishment occurs and so on. These unpleasant emotional effects play a major role in the development of avoidance responses. Behavioral psychologists generally view this tendency for negative emotional effects to generalize to all aspects of the punishment context as undesirable. Certainly the tendency to develop avoidance responses to key social agents such as parents and teachers is undesirable. But as Staats accurately notes, in some ways this generalization of unpleasant emotional effects may be beneficial; learning to avoid situations or people that increase the likelihood of undesirable responses is a desirable outcome.

Fourth, in addition to affecting the response actually followed by punishment, punishment tends to affect other behaviors as well. A child who is busily doing his assignment while talking out loud to himself may cease talking out loud if this response is punished. The presentation of punishment, however, may affect his work on the assignment. The adverse effects of punishment on other ongoing responses may be limited in three ways. (1) As we hive noticed before, the consequences of a response are most effective if they follow immediately after the response. (2) Punishment is more effective if it occurs consistently after a response. (3) Ongoing reinforcement for a response will interact with any accidental effects of punishment occurring

Rodger Bufford is a graduate of the King's College and received his PhD in Clinical Psychology from the University of Illinois at Urbana. He is currently Associate Professor of Psychology at Western Conservative Baptist Seminary, and practices at the Seminary's Counseling Center. Dr. Bufford has authored a number of articles in professional journals and one book, The Human Reflex: Behavioral Psychology in Biblical Perspective (Harper and RowlCAPSI 1981).

after the response. For all of these reasons, punishment tends to act in a selective fashion to primarily influence the response that it consistently follows in a contingent fashion.

There is one way in which the tendency for punishment to affect other responses is clearly positive. Usually, other responses similar to the punished response are also undesirable. The tendency for punishment to reduce the frequency of similar responses is thus an advantage. We need to be cautious, then, about the effect of punishment on responses other than the target response, but this need not be a major concern.

Finally, several other criticisms have been raised with regard to the effects of punishment: delay of punishment leads to weakened effects in reducing the frequency of the response; failure to consider the function or effect of a particular stimulus for a particular person may result in accidentally reinforcing undesirable behavior; punishment may serve as a form of aggression in which the punishing agent acts out their anger. Each of these is a legitimate concern, but they are more relevant to considerations of how to punish effectively. The fact that sloppy punishment contingencies do not work well is no more serious a problem than the fact that careless reinforcement contingencies also are not very effective.

Effects of Reinforcement and Punishment: Parallels and Contrasts

In this section we examine in more detail the effects of reinforcement and punishment; Table I summarizes these. The first line indicates that reinforcement increases response rate, while punishment decreases response rate. These relationships should be familiar.

As we noted earlier, one of Skinner's criticisms of punishment is that it has temporary effects. The second line of Table I indicates that both punishment and reinforcement have temporary effects. In general, reinforcement is used when we wish to strengthen the frequency of a response that occurs at a low base rate. Following the introduction of reinforcement, the base rate of the response increases. If the reinforcement is then discontinued, the response decreases in frequency. Introducing reinforcement again quickly reinstates the higher base rate found when the response is reinforced. These results are summarized in the top half of Figure 1. In the lower half of Figure I the effects of punishment are portrayed in a similar fashion. Punishment is used when we wish to decrease the rate of a response; the baseline for the response to be punished is

Table 1
Effects of Reinforcement and Punishment

Behavioral Characteristic
        Reinforcement Effects      Punishment Effects

                     Target response                                    Increase                                   Decrease

                    Permanence                                         Temporary                                 Temporary

                    Emotional Effects                                 Positive                                       Negative

                   Social Effects                                Interpersonal attraction                  Interpersonal avoidance
                                                                               (love, affection)                              (hate, dislike)

                Modeling Effects                               Imitation of reinforcing                    Imitation of punishing
                                                                               interactions                                      interactions

             Generalization                                   Similar responses strengthened;           Similar responses weak-
                                                                      Co-occurring responses                      Co-occurring
                                                                      strengthened                                        responses weakened

      "Unauthorized" behavioral                        Theft, extortion, "conning"                     Avoidance behaviors
      Effects                                                     (eg. steal M&Ms, fake task                  (eg. play hooky, run
                                                                      etc.)                                                     away from home) and

usually fairly high. When punishment is introduced, the frequency of the response decreases. Stopping the punish ment results in a recovery of the base rate of the response.


Figure 1. Effects of Reinforcement and Punishment

Reinstating punishment quickly recovers the lowered base rate found in the original punishment period.12 

Although the directions of the effects are opposite, the kinds of effects produced by reinforcement and punishment are essentially the same.

Notice that the effects of punishment are a mirror image of the effects of reinforcement. In general, both reinforcement and punishment have temporary effects. One further qualification is required: under limiting conditions both reinforcement and punishment may have virtually permanent effects.13

A second criticism of punishment raised by Skinner is that it produces adverse emotional effects. Generally overlooked in the behavioral literature is the fact that reinforcement also affects emotional behavior. I I Just as punishment produces displeasure, anger, disliking and hate, reinforcement produces emotional responses such as attraction, liking and affection.15 These relationships are summarized on line three of Table 1.

Some of the emotional effects of punishment that we have just discussed are essentially social effects as well. In addition to affecting emotional responses to other persons, however, punishment may also affect a wide range of other social behaviors. The child who is often punished by parents and teachers may soon learn to avoid contact with them. Technically we would term these responses avoidance and escape responses; they are negatively reinforced by preventing or terminating the presence of these social agents. Unfortunately, the child thus misses important learning experiences in socialization and education; in this way, both social relationships and learning experiences may be affected by the use of punishment." By contrast, reinforcement has the opposite effect of producing social attraction and thus fostering important learning interactions.

Another important principle is that the consequences which follow a given response tend to affect other responses that occur at about the same time, as well as to effect other responses that are similar in form. This principle applies with punishment as well as with reinforcement: the effects of punishment influence not only the specific response that it follows, but also other responses ongoing at the same time and responses that are similar in form.

The reduction of the probability of other responses that are also undesirable by generalization effects of punishment can actually be a beneficial effect. If the punished response is desirable, however, or if other responses that are desirable are weakened along with the punished response, problems may be presented. Consequently, the generalization of the effects of punishment may be either good or bad. Further, the degree and probability of generalization effects will be influenced in important ways by other ongoing events such as the strength of behaviors that occur at about the same time as the punished response, the ongoing reinforcement support for those behaviors, the past experiences of the person with reinforcement and punishment, and so on.17

Thus, while punishment clearly does affect responses other than those specifically followed by the punishing stimulus, this phenomenon is not limited to punishment. Reinforcement also has generalization effects. Careful management of contingencies can enhance or limit generalization effects for both punishment and reinforcement.18

A final problem that has been suggested for the use of punishment is that of "unauthorized" escape. An example of this is a rat in an experimental chamber in which an electric shock is presented by means of a metal floor grid at periodic intervals. The rat can avoid shock by pressing a bar before the shock begins, or escape by pressing the bar after the onset of shock. Rather than press the bar, some rats learn to lie down on their backs with feet, nose and tail in the air; in this manner they effectively escape the shock although the floor is electrified continuously.19 The desired response of bar pressing does not occur, yet the animal is able to avoid the unpleasant experience of electric shock.

The same principle may be seen with human behavior. A child who is punished by his teacher for failure to turn in his homework may avoid punishment by doing his homework; he may also avoid punishment by playing hooky.

Another form of "unauthorized" escape is the use of counter-aggressive measures. When the neighborhood bully tells Johnny that he will beat him up if he comes to the playground again, Johnny can avoid the punishing event by staying away. He can also avoid it by beating up the bully, provided he is strong and able enough to do so. Or he may bring his older brother along for protection; in this instance, we might consider Johnny's response to be socially acceptable. If Johnny avoided punishment for not doing homework by assaulting his teacher, however, we would disapprove.

Conceptually, we may think of "unauthorized" escape responses as negatively reinforced behaviors that are socially undesirable. Almost totally neglected by the behavioral literature, but of equal social significance in my opinion, is the problem of "unauthorized" reinforcement. Behaviors that produce unauthorized reinforecement include theft, extortion, bribery, "conning" and the like. Stealing a candy bar rather than earning one is an example. These examples show that a person may obtain positive reinforcement or escape punishment in ways other than those that were intended by parents, teachers and experimenters. Although the directionality of the behaviors is different, in many ways similar problems are posed with unauthorized effects of both reinforcement and punishment.

Although the directions of the effects are opposite, the kinds of effects produced by reinforcement and punishment are essentially the same. Both affect the rate of a response; both have temporary effects except under limiting conditions; both produce emotional effects; social attraction is affected by both; generalization occurs with both; finally, unauthorized effects may occur with both.

Alternatives to Punishment

We have seen that punishment is effective, desirable and even necessary under some conditions. Our discussion would be incomplete, however, if we did not also examine alternatives to punishment and consider the circumstances in which these alternative strategies may be more desirable than punishment.

A major contribution of the behavioral approach has been the explicit description and study of the effectiveness of alternatives to the use of punishment in dealing with problem behaviors. A review of the behavioral literature suggests six strategies apart from punishment which may be used to eliminate undesirable behaviors: (1) changing the setting conditions;20 (2) removal of the discriminative stimuli for the response;21 1 (3) terminating reinforcement for the response;22 (4) developing another response which prevents the problem behavior; (5) reinforcing any other behavior which occurs (DRO);23 (6) elimination of the opportunity to respond. These procedures are presented in Table 2, along with examples of each technique. Space does not permit extended discussion of each technique here.

There are clearly limitations to each of the approaches offered as alternatives to punishment. Sometimes we are not able to control whether another person becomes tired, hungry, or sick yet we may wish them to be patient, tolerant and so on even when undergoing these unpleasant physical and emotional states. Not all stimulus events are readily controlled, thus it may not be possible to remove them. For some responses the reinforcement is intrinsic (running is reinforced by the physical sensations and by getting somewhere more quickly) and thus not amenable to extinction. It may prove difficult to devise a suitable incompatible response for some problem behaviors. And so on. Thus each specific problem response presents a challenge in identifying the most suitable technique for reducing the frequency of that response most effectively.

In practice, it is not uncommon to find two or more of these alternative techniques used together. For example, if Johnny has a tantrum problem, it might be dealt with by using a combination of approaches. First, we might arrange a specific punishment for any tantrums that occur. Second, eliminating any reinforcement for tantrums is an important additional procedure. Third, we might arrange to reinforce either a particular incompatible alternative, some other specific alternative response, or any activity other than tantrums. The ideal reinforcements would be the very things that Johnny has previously gotten by means of talatrums.
Thus on a shopping trip a mother might do the
fol lowing (1) when a tantrum occurs she puts Johnny in the car until she finishes (a punishment), and gives him no candy or gum on that trip; or (2) in the absence of tantrums she allows Johnny to select a pack of gum or a candy bar as they leave the store and then permits him to eat it. This combination of procedures would likely be more effective than any single procedure alone.

Punishment should be viewed as one of a group of techniquesfor reducing the firequency of responses, punishment is effective, and in specific behavioral contexts it is the preferred method for reducing the frequency of specific problem behaviors.

                                                                          Table 2

                                                               Alternatives to Punishment

Procedure                                                            Examples

1. Change setting conditions                                get good night's sleep take an aspirin nothing to 
                                                                          drink for 12 hours

2. Removal of a stimulus                                      turn out lights put dessert in cupboard

3. Terminate reinforcement:                                 Stop responding to requests without "please" no response 
    Extinction                                                        to cries after being placed in bed

4. Develop specific response                               reinforce being in seat (vs reprimand for 
    which prevents problem behavior                    being up without permission) 

5. Strengthen and alternative response                 reinforce running (vs self-mutilation)

6. Eliminate opportunity to respond                     lock gate to swimming pool
                                                                          take away ball

The most -obvious alternative to punishment when the goal is to develop a response rather than to eliminate a response is the use of reinforcement procedures. Punishment is one of several effective procedures when the goal is to decrease the frequency of a specific response. When the goal is to increase the frequency of a response, or to develop a new response, punishment is not an effective procedure, and indeed the other procedures discussed here as alternatives to punishment are also not particularly effective. Punishment is not an effective means for establishing a response, though many parents in our culture attempt to use it in this way. In addition to using reinforcement to establish a response, reinforcement may also be used to strengthen a response that is already present but is so weak that it does not readily occur.

Punishment: A Biblical Perspective

In attempting to develop a biblical perspective on punishment, a number of biblical teachings should be considered. First, in the Mosaic Law there is the explicit provision for a set of procedures that correspond roughly to our current civil and criminal codes. Punishment was specified for a variety of offenses, and included a range of punishment procedures." Second, in the Proverbs there are a number of references to the use of a rod for discipline of a punitive sort in the process of child-rearing." Punishment is endorsed by the Scriptures, and there is a general principle that the nature and severity of the prescribed punishment is related to the nature of the offense. Further, it is suggested that more mild forms of punishment are a social norm: "reproofs for discipline are the way of life."26

It is interesting to note possible parallels between the use of a rod for discipline and some of the principles for punishment that we have discussed. It is definitely painful, can be applied briefly, lends itself to pairing punishment with words; the frequent references to reproof suggest that the use of words is an integral part of the discipline process advocated by the Bible. Another principle that the Bible reflects is the suggestion that punishment be used as infrequently as possible.27 Finally, the suggestion that punishment be brief parallels the biblical principle that God's forgiveness is immediate and sure.28

Many examples of the use of punishment occur throughout the Bible. In some, God is the mediator of punishment, while others are carried out by social agents. When Adam and Eve ate of the fruit of the tree of knowledge of good and evil they were put out of the Garden of Eden; in addition, their relationship with God suffered an immediate disruption. Cain was punished for his failure to bring an acceptable sacrifice to God. Achan was stoned for taking forbidden plunder. David was punished for his adultery with Bathsheba. Ananias and Sapphira were slain for lying before God. Sodom and Gomorrha were destroyed. Israel and Judah were defeated in battle and carried away into captivity." In each of these circumstances God had provided verbal warning beforehand that these behaviors were not acceptable. Indeed, the whole history of Israel and Judah is a cycle of disobedience, warning by the prophets, punishment in the form of oppression by their enemies and failure of crops, repentance, then renewal of the sinful patterns and practices.30

While it is clear that the use of punishment is endorsed and recorded in the Bible, there is much teaching that emphasizes the use of more positive methods of behavior influence. Parents are instructed to teach their children God's principles throughout the day as a part of normal daily activities: "When you sit at home, when you walk along the road, when you lie down and, when you get up."31 There are also many examples in the Bible that indicate the desirability of positive reinforcement; there are frequent references to the use of encouragement and to the focus on positive behavioral attribures.32

In summary, the Bible clearly advocates and records examples of the use of punishment. We can see a number of parallels between the biblical examples and the principles of punishment that we find in behavioral psychology. It is also clear, however, that punishment is not the sole method of behavioral influence advocated in the Bible. Thus, in broad terms, it appears that biblical teachings are compatible with the data regarding the use and effectiveness of punishment.

Punishment should be viewed as one of a group of techniques for reducing the frequency of a response.

Are the Effects of Punishment Bad?

We have seen that punishment not only reduces the probability of response, but also has unpleasant emotional effects, affects other ongoing responses, contributes to social avoidance and may foster aggression. Whether these effects are good or bad is a question of values. One way to resolve the question of values is to adopt the view that pleasant effects are good, and that unpleasant effects are bad. They could then be studied scientifically by examining which events strengthen behaviors that they follow and which weaken behaviors they follow. A second approach is to measure the reaction of people regarding whether these outcomes are good or bad, and then adopt the majority opinion. A third approach is to appeal to some a priori set of values (e.g., those given in the Bible). Central to all three of these approaches is that they make a value commitment that lies outside the scope of science. Science can tell us whether people find certain outcomes pleasant or unpleasant, or whether they view them as good or bad. But it cannot tell us that the majority view is right,, that is an extrascientific issue. Deciding that the use of punishment is good, bad, or neutral is an ethical-philosophical, moral and religious issue, not a scientific one.

I Scientifically we can say that punishment produces unpleasant emotional effects. But Skinner is making a value statement when he says that punishment is, therefore, bad or undesirable. Moreover, this is a value about which there is considerable disagreement. Staats suggests that the unpleasant emotional effects of punishment contribute in a positive way to the development of a controlling vocabulary of words such as NO, STOP, and so one which actually reduces a child's exposure to unpleasant or punishing events. When the child reaches for the flame on a candle, a loud "NO" prevents a burned hand. A second way in which Staatsviews the emotional effects of punishment as desirable is through generalization of the effects of punishment to similar responses and similar stimulus conditions. A child who is punished for throwing a baseball through the neighbor's window will be less likely to throw footballs, basketballs, rocks, or other objects through that window in the future, and will also be less likely to throw objects through the windows of buildings down the street or across town. Staats views these effects as desirable, and I concur.The emotional effects of punishment are particularly important when those emotional effects influence human social relationships. Most persons have both reinforcing and punishing relationships with others around them. Thus the emotional response to a given person, say Mother, reflects a combination of both positive emotional effects associated with reinforcing experiences and negative emotional effects due to punishing experiences. The overall quality of the relationship depends on the relative frequency and impact of reinforcing and punishing events in the relationship with Mother. Thus a mother who is mostly reinforcing, but occasionally punishes will be loved. A mother who often punishes and rarely reinforces will be disliked or hated.

These same principles apply to the avoidance and aggressive behaviors that are sometimes produced by the use of punishment. Avoidance and aggressive responses can be minimized if punishment occurs in a context that involves a high frequency of positive reinforcement, thus maintaining approach and attraction at high strength (these responses are incompatible with avoidance and aggression). Furthermore, if aggressive behavior is maintained at low strength through punishment, it will be very unlikely to occur.

One other aspect of punishment needs to be clarified. Punishment is not restricted to the action of social agents. Punishment is a natural feature of our world. If I stick my fingers into the fire I experience pain. There are several choices at my disposal regarding how to deal with this experience. First, I can keep my hands out of the fire in the future; second, I can use some method to anesthetize or destroy the pain sensors; third I can continue to stick my hands into the fire and endure the pain as I am best able; fourth, I can eliminate fires from the environment. For a variety of reasons, we usually choose to live with fire and learn to minimize our personal experiences of pain from it. Something about the social aspects of punishment seems to make it more difficult for us to deal with social punishment in such a matter of fact fashion.

A fundamental factor in the behavioral objections to punishment seems to be a frank dislike of punishment. Personally, I dislike it too. I dislike administering it (except when provoked), and especially I dislike receiving it. But this distaste does not negate either the effectiveness of punishment or the biblical instructions regarding its use. It is significant, in this regard, that the Bible recognizes our distaste for discipline, and also indicates that there are two kinds of mistakes with regard to discipline: first, the mistake of loving so little we fail to do it; second, the mistake of loving so little that we enjoy carrying out discipline.34

We defined punishment in terms of the effect of a stimulus event on behavior. With this in view, then, several conclusions are possible: (1) punishment works; (2) punishment may produce a number of effects in addition to reducing the frequency of the target response; (3) reinforcement has side effects similar in nature to those associated with punishment, but opposite in direction; (4) the potential adverse effects of punishment may be minimized by careful application of punishment; (5) biblical teachings clearly support the use of punishment; (6) the issue of whether punishment is good or bad is a value issue that must be decided on an extrascientific basis. We have suggested biblical teachings on punishment which indicate that punishment is a legitimate procedure. We need to remember, however, that both the methods and goals of punishment must be examined in light of biblical teaching to establish their legitimacy.35

The Bible clearly advocates and records examples of the use ofpunishment. .. It is also clear, however, that punishment is not the sole method of behavioral influence advocated in the Bible.


We have defined punishment in terms of the effect of a stimulus event in decreasing the frequency of a response. The principal types of punishment discussed include punishment by presenting a stimulus following a response and Response Cost or punishment by removing a stimulus following a response. In addition, we noted that Time Out has features that are conceptually different from punishment but produces similar behavioral effects.

The widespread behavioral perspective that punishment is ineffective and undesirable has been discussed in terms of the interpretation of the findings of key experiments by Skinner and Estes, and in terms of the effects of punishment on the recipient. Comparisons and contrasts were drawn between the effects of reinforcement and punishment. In general reinforcement and punishment have similar kinds of effects which are opposite in direction. Both generally have temporary effects; both affect emotional responses; both affect social attraction; both have modeling effects; generalization occurs with both; unauthorized effects are possible with both.

Research clearly suggests that punishment is effective; however, there seems to be ample reason to consider alternative techniques and to minimize the frequency of punishment in view of the potential effects of punishment on social avoidance, its generally unpleasant emotional effects, the risk of modeling effects contributing to aggressive responses and the possibilities of unauthorized avoidance.

Several alternatives to punishment as techniques for eliminating undesired responses exist including: changing the setting conditions, removing the discriminative stimuli, terminating reinforcement for the response, developing a specific alternative response, strengthening any alternative response, and eliminating the opportunity to respond. One of the major contributions of the behavioral approach has been in the development and study of these alternative techniques.

Review of biblical teachings regarding punishment suggest that punishment is clearly sanctioned by the Bible, but also reveals that there is ample support for the use of alternative techniques in managing human behavior. Thus there seems to be a convergence between biblical teachings and the interpretation of the behavioral data on punishment presented here, which suggests that punishment is highly effective, but that its use should be limited to circumstances in which elimination of a specific response is the goal and in which alternatives are not suitable.

One possible factor that contributes to the widespread behavioral objections to punishment is a personal dislike for punishment among the investigators, and a tendency to interpret data in light of this pre-experimental commitment.

In conclusion, we have seen that punishment works and that it has many effects in addition to the immediate effect on the punished response, but that these effects are similar in nature and opposite in direction from those associated with reinforcement. We have also seen that there are a number of alternative approaches to dealing with problem behaviors, and that when practical these may be preferred to use of punishment. Finally, we noted that biblical teachings support the use of punishment, but are generally consistent with the precautions that emerge from the behavioral study of the effects of punishment. While we have concluded with a limited endorsement of punishment, in the context of this presentation we have not considered how to go about punishing in an effective manner. These issues remain to be addressed at another time.


1See N. H. Azrin and W. C. Holz, Punishment, in W. K. Honig (Ed.) Operant Behavior: Areas of Research and Application (New York: Appleton, Century, Crofts, 1966), p. 381. Others define punishment as a procedure which may or may not affect the response; for example, see C. B. Ferster, S. Culbertson, and M.C.P. Boren, Behavior Principles second edition (Englewood Cliffs, N.J.: Prentice-Hall, 1975), p. 199 ff.
For a discussion of Time Out, see Azrin and Holz, op. cit., p. 389 ff.
3B. F. Skinner, Science and Human Behavior (New York: MacMillan, 1953) p. 183
.4See W. K. Estes, An experimental study of punishment (Psychological Monographs, 57, Whole No. 263, 1948) and B. F. Skinner, The Behavior of Organisms (New York: Appleton-Century-Crofts, 1938).
A. W. Staat, Child Learning, Intelligence, and Personality: Principles of a Behavioral Interaction Approach (New York: Harper and Row, 197 1) p. 234.
Ibid., p. 236.
7Azrin and Holz, op. cit., p. 420 ff.
8Ibid, p. 410.
9Staats, op. cit., p. 236 1.
10See A. Bandura, Aggression: a Social Learning Analysis (Englewood Cliffs, N.J.) 1973, and A. Bandura, Principles of Behavior Modification (New York: Holt Rinehart and Winston) 1969.
For a discussion of the nature of emotional behavior from a behavioral perspective, see S.W. Bijou and D. M. Baer, Child Development P A Systematic and Empirical Theory (New York: Appleton, Century Crofts, 1961). Bijou and Baer's thesis is that emotions are respondent behavior, and that all consequences of behavior, both reinforcing and punishing also elicit emotional responses.
See G. S. Reynolds, A Primer of Operant Conditioning, revised edition (Glenview, Illinois: Scott, Foreman and Company, 1975) pp. 35-38.
Ibid, pp. 123-128.
14or a discussion of the relationships between emotional responses and reinforecement and punishment contingencies, see S.W. Bijou and D.M. Baer, Child Development P A Systematic and Empirical Theory (New York: Apleton-Century-Crofts, 1961), pp. 73-76.
15See Byrne, D., and Rhamey, R. Magnitude of positive and negative reinforcements as a determinant of attractions (Journal of Personality and Social Psychology, Vol. 2, 884-889; 1965).
Staats, op. cit., p. 233 f.
17See Staats, op. cit., p. 239 ff.
18Azrin and Holz, op. cit., pp. 416, 433-434.
19The procedure described here is called Sidman avoidance; see Azrin and Holz, op. cit., p. 383 ff.

0S. W. Bijou and D. M. Baer, Child Development 11. Universal Stage of Infdncy (New York: Appleton-Century-Crofts, 1965).

Ferster, et. al., op cit., present an extensive discussion of the ways in which the presentation and removal of stimuli affect behavior (see p. 513 ff).
Probably the earliest example of this principle in the literature is the elimination of a phobic response to white furry objects by Jones (1924); cited in L.P. Ullmann and L. Krasner, Case Studies in Behavior Modification (New York: Holt, Rinehart and Winston, 1965). A more recent application of this procedure is found in T. Ayllon and J. Michael, The psychiatric nurse as a behavior engineer (Journal of the Experimental Analysis of Behavior, Vol. 2, 323-334, 1959).
Adapted from R. W. Malott and D. L. Whaley, Elementary Principles of Behavior, Volume I (Kalamazoo, Michigan: Department of Psychology, Western Michigan University, 1968).
Many of the basic commandments regarding specific kinds of offenses and the associated punishments may be found in: Deuteromomy 19:1-21; Exodus 20:1-22:31; Leviticus 24:10-23.
See ' eg., Proverbs 13:24; 22:15; 23:13-14; 29:15, 17.
26Proverbs 6:23; also see Matthew 5:21-24, Luke 17:3; Ephesians 6:1-2.
27See Ephesians 6:4 which admonishes fathers not to provoke their children. One of the surest ways to provoke is to be constantly punishing or
28See 2 Samuel 11: 1- 12:14, especially 12:13-14; 1 John 1:9.
29See Genesis 3:1-24; Genesis 4:1-16; Joshua 7:1-26; 2 Samuel 11:1-12:14; Acts 5: 1 -11; Genesis 18:1-19:29; 2 Kings 17:1-18ff and 2 Kings 25:1-21 respectively.
This theme of disobedience, warning and punishing is woven throughout the Old and New Testaments. Some examples occur in: Deuteronomy 28:1-66; Joshua 7:1-8:3ff; Judges 3:7-14, 4:1-24, etc.; Jeremiah 7:1-26 ff, 29:10-19 ff; Matthew 12:38-45; Luke 10:1-16; Jude 3-Ilff; Hebrews 11:1-12:29, cf. especially 12:25-29. For a discussion of the Old Testament writings on this theme see Larry Richards, The Edge ofJudgement (Elgin, Illinois: David C. Cook Publishing Company, 1977).
Deuteronomy 6:4-9, New International Version.
32For example, see: Philippians 1:3-6; Colossians 1:3-8; 1 Corinthians 1:4-7; I Thessalonians 1:2-3 Instruction to practice giving thanks and praise are given in: Ephesians 5:4; Philippians 4:8. The emphasis of Scripture in this area, however, is more on the prohibition of being critical and destructive: for example, see Ephesians 4:29-3 1; Philippians 4:2-3; Galations 5:13-15.
For a discussion of some of these issues, see the section on Psychological Engineering in C. Ellison (Editor) Modifying Man: Issues and Ethics (Washington, D.C.: University Press of America, 1978).
See Hebrews 12:7-11; Proverbs 12:1, 15:5.
35See R. Bufford, Ethics for the mass application of behavior control in C. W. Ellison, Modifying Man: Implications and ethics (Washington, D.C.: University Press of America, 1978).