I remember (Dr. Bandura relates) reading a story reported by Professor Mowrer about a lonesome farmer who decided to get a parrot for company. After acquiring the bird, the farmer spent many long evenings teaching the parrot the phrase, "Say Uncle." Despite the devoted tutorial atteniton, the parrot proved totally unresponsive and finally, the frustrated farmer got a stick and struck the parrot on the head after each refusal to produce the desired phrase.

But the visceral method proved no more effective than the cerebral one, so the farmer grabbed his feathered friend and tossed him in the chicken house. A short time later the farmer heard a loud commotion in the chicken house and upon investigation found that the parrot was pummeling the startled chickens on the head with a stick and shouting, "Say Uncle! Say Uncle!"

The research whose results we print here was done by Dr. Albert Bandura, professor of Psychology at the University of Stanford. Apart from the intrinsic interest of his conclusions, the material presented here should demonstrate the full power of the media in shaping our values and molding human behavior.

It is useful, however, to try to understand the process of learning-by-imitation outlined here, because radicals could well use these methods for the opposite ends to foster decent human values and attitudes. At the very least, Dr. Bandura's work should teach us to realize quite how dangerous it is to leave control of the media in the wrong hands.

The following are excerpts from a paper delivered by Dr. Bandura. Much has been omitted. People professionally interested in his research should contact Dr. Bandura in California or the CRIMSON for references to the full published write-ups of the experiments.

A note of exhortation: the writing style, high socrelese, is not impenetrable. On the contrary, should one absorb the idiom it can be remarkably effective. It is worth the effort to do so.

Learning through Tuition vs. Learning through imitation of a Model.

One can distinguish two kinds of processes by which children acquire attitudes, values, and patterns of social behavior. First, there is learning that occurs on the basis of direct tuition or instrumental training. In this form of learning, parents and other socializing agents are relatively explicit about what they wish the child to learn and attempt to shape his behavior through rewarding and punishing consequences.

Although a certain amount of socialization of a child takes place through direct training, research bearing on modeling processes demonstrates that, unlike the relatively slow process of instrumental training, when a model is provided, patterns of behavior are rapidly acquired in large segements or in their entirety.

Let us now consider a series of experiments that both illustrates the process of learning through imitation and identifies some of the factors which serve to enhance or to reduce the occurrence of imitative behavior.

Transmission of Aggression through Film and TV.

One set of experiments was designed to determine the extent to which aggression can be transmitted to children through exposure to aggressive adult models. One group of children observed an aggressive model who exhibited relatively novel forms of physical and verbal aggression toward a large inflated plastic doll; a second group viewed the same model behave in a very subdued and inhibited manner, while children in a control group had no exposure to any models. Half the children in each of the experimental conditions observed models of the same sex as themselves and the remaining children in each group witnessed opposite-sex models.

This investigation was later extended in order to compare the effects of real-life and film-mediated or televised aggressive models on children's behavior. Children in the human film-aggression group viewed a movie showing the same adults who had served as models in the earlier experiment portraying the novel aggressive acts toward the inflated doll. Children in the cartoon-aggression group saw a film projected on a glass lenscreen in a television console. In this film a female model was costumed as a cat and exhibited aggressive behavior toward a plastic doll.

The results of these experiments leave little doubt that exposure to aggressive models heightens children's aggressive responses to subsequent frustration. As shown in Fig. 1, children who observed the aggressive models exhibited approximately twice as much aggression as did subjects in the nonaggressive model group or in the control group. In addition, children who witnessed the subdued nonaggressive model displayed the inhibited behavior characteristic of their model and expressed significantly less aggression than the control children.

Some evidence that the influence of models is partly determined by the sex appropriateness of their behavior is provided by the finding that the aggressive male model was a more powerful stimulus for aggression than the aggressive female model. Some of the children, particularly the boys, commented spontaneously on the fact that the female model's behavior was out of character (e.g., "That's no way for a lady to behave. Ladies are supposed to act like ladies" . . .).

In contrast, aggression by the male model was generally viewed as appropriate and approved by both the boys ("A1's a good socker, he beat up Bobo. I want to sock like A1.") and the girls ("That man is a strong fighter. He punched and punched, and he could hit Bobo right down to the floor and if Bobo got up he said, 'Punch your nose.' He's a good fighter like Daddy.").

Furthermore, the data reval that aggressive models are highly influential not only in reducing children's inhibitions over aggression, but also in shaping the form of their behavior. Children who observed the aggressive models displayed a great number of precisely imitative aggressive acts, whereas such responses rarely occurred in either the nonaggressive group or the control group.

The finding that film-mediated models are as effective as real-life models in eliciting and transmitting aggressive responses indicates that televised models may serve as important sources of behavior and can no longer be ignored in conceptualizations of personality development. Indeed, most youngsters probably have more exposure to prestigeful televised male models than to their own fathers. With further advances in mass media and audiovisual technology, models presented pictorially, mainly through television, are likely to play an increasingly influential role in shaping personality patterns and in modifying attitudes and social norms.

It has been widely assumed on the basis of psychoanalytic theory and other hydraulic energy models of personality that children's vicarious participation in film-mediated aggression or the direct expression of aggressive behavior will serve to discharge "pentup energies" and effects. Guided by this catharsis hypothesis, many parents, educators, and mental health workers encourage hyperaggressive children to participate in aggressive recreational actavities, to view highly aggressive televised programs, and to aggress in psychotherapeutic play-rooms and other permissive settings.

In contrast to this "drainage" view, the growing body of research provides considerable evidence that vicarious participation in aggressive activities can be highly effective in modifying observers' aggressive behavior, but not in the direction predicted by the catharsis hypothesis. These findings are perhaps not too surprising. It is highly improbable that even advocates of vicarious drive reduction would recommend community programs in which sexually aroused adolescents are shown libidinous movies at drive-in theatres as a means of reducing sexual behavior, or in which famished persons are presented displays of gourmands dining on culinary treats in order to alleviate hunger pangs.

The influential role of modeling processes in personality development is further revealed by studies demonstrating that children adopt, through observation of behavior exemplified by adults and peers, relatively complex attributes including standards of achievement and self-evaluation, patterns of emotional responsivity, syntatic styles, moral judgmental orientations, and patterns of self-gratification.

The Consequences of Rewarding Aggressive Behavior.

The manner in which rewarding or punishing consequences to the model's behavior influences imitation is demonstrated in another experiment in which nursery school children observed either an aggressive model rewarded, an aggressive model punished, or had no exposure to the models. The models were two adults, and the film presented to the children was projected into a television console. In the aggression-rewarded condition, Rocky, the aggressive model, appropriates all of Johnny's attractive play possessions and tasty foodstuffs through aggressive - domineering means. The film shown to the children in the aggression-punished condition was identical with that shown to the aggression-rewarded group except for a slight rearrangement of the film sequence so that the aggression exhibited by Rocky resulted in his being severely punished by Johnny. Following exposure to the models the children were tested for the incidence of post-exposure aggressive behavior. Children who observed Rocky's aggressive behavior rewarded readily imitated his physical and verbal aggression, whereas children who saw him punished exhibited relatively little imitative behavior and did not differ from a group of control children who had no exposure to the models.

At the conclusion of the experiment each child was asked to evaluate the behavior of Rocky and Johnny and to select the character he preferred to emulate. These data yielded some interesting and surprising findings. As might be expected, children who observed Rocky's aggressive behavior punished both failed to reproduce his behavior and rejected him as a model for emulation. On the other hand, when Rocky's aggression was highly successful in amassing rewarding resources, he was chosen by most of the children as the preferred model for imitation. The surprising finding, however, is that without exception these children were highly critical of his behavior (e.g., "Rocky is harsh" ... "Rough and bossy" ... "Mean" ... "Wicked" ... "He whack people" ...).

It was evident from the children's comments that the successful payoff of aggression rather than its intrinsic desirability served as the primary basis for emulation (e.g., "Rocky beat Johnny and chase him and get all the good toys" ... "He came and snatched Johnny's toys. Get a lot of toys" ...). The children resolved the conflict by derogating the unfortunate victim, aparently as justification for Rocky's exploitive-assaultive behavior. They criticized Johnny for his inability to control Rocky ("He's a cry baby. Didn't know how to make Rocky mind"), for his miserliness ("If he'd shared right in the beginning, Rocky might have played nice"), and generally described him as "Sulky," "Mean," and "Sort of dumb."

This study clearly demonstrates the way in which rewarding consequences to the model may outweigh the value systems of observers--children readily adopted successful modeling behavior even though they had labeled it objectionable, morally reprehensible, and publicly had criticized the model for engaging in such behavior.

In many televised and other mass media presentations antisocial models amass considerable rewarding resources through devious means but are punished following the last commercial on the assumption that the punishment ending will erase or counteract the learning of the model's antisocial behavior. The findings of an experiment by Bandura in 1965 reveal that although punishment administered to a model tends to inhibit children's performance of the modeled behavior, it has virtually no influence on the occurrence of imitative learning. In this experiment children observed a film-mediated aggressive model who was severely punished in one condition, generously rewarded in a second condition, and received no response consequences to the model in the third condition.

Consistent with the findings cited earlier, a post-exposure test of imitative behavior showed that children who observed the punished model performed significantly fewer imitative responses than children in the model-rewarded and the no-consequence groups. Children in all three groups were then offered attractive incentives contingent on their reproducing the model's behavior. The introduction of the rewards completely wiped out the previously observed performance differences, revealing an equivalent amount of learning among the children in the model-rewarded, model-punished, and the no-consequence groups. Moreover, girls had acquired approximately as much imitative aggression as did the boys.

It might be concluded from these findings that exposure of children to punished antisocial or other types of models is likely to result in little overt imitative behavior. Nevertheless, the observed behavior is learned and may be exhibited on future occasions given appropriate instigation, the instruments necessary for performing the imitative acts, and the prospect of sufficiently attractive positive rewards contingent on the successful execution of the behavior.

Closing Remarks.

Man is endowed with a capacity to behave aggressively. However, the frequency with which he exhibits such behavior, the specific forms that it takes, and the targets that are selected for attack, are strongly influenced by social experience. Societal training in aggression is partly achieved through the types of examples that it provides. There now exists a large body of evidence that the attitudes, values, and behavior of children, as well as adults, can be substantially altered through observation of the actions of others and its consequences for them.

In evaluating the effects of exposure to televised violence on the behavior of viewers it is important to distinguish between learning and doing. It has been shown that if people are exposed to televised aggression they not only learn aggressive patterns of behavior, but they also retain them over a long period of time. There is no longer any need to equivocate about whether televised stimulation produces learning effects. It can serve as an effective tutor.

Although people acquire patterns of social behavior from television, they do not necessarily perform this behavior in their everyday interactions. The casual linkages of performance are more complex because other factors enter in as determinants. This same issue applies, of course, to the influence of television on consumer behavior. For example, a well-endowed blonde begs 50 million viewers to join the Dodge rebellion. Obviously 50 million people do not jump up and purchase Dodge automobiles. The televised influence increases the probability that Dodge cars will be purchased, but one would have to consider many other factors in predicting the specific individuals who will act on what they have learned. I assume that in selling commercial time the television industry argues for the potency of televised influences with the same strong conviction that they argue for its ineffectiveness in modifying social behavior.

Findings of laboratory studies show that exposure to aggressive behavior