Introduction

Theoretical Views of Aggression

Experimental Evidence

Evaluation

References

Aggression is very much a part of the lives of everyone, whether from the viewpoint of an individual engaging in aggressive behavior or being the victim of another person who is engaging in such behavior. Because of this, it is important for a psychotherapist to have an understanding of the phenomenon and of what can be done to manipulate the tendency of an individual to engage in such behavior.

In this paper aggression is considered to be any intentional act of harming another person, including both overt behavior and covert behavior where a person is not harmed. This is not to be confused with assertiveness which involves a person standing up for his own rights. The assertive person does not allow others to be aggressive or manipulative with him. Aggressiveness infringes upon the rights of others; whereas, assertiveness prevents others from infringing upon one's own rights.

This paper deals with the consideration of the effects of engaging in aggressive behavior upon subsequent tendencies to be aggressive. This involves the psychoanalytic concept of catharsis. The basic question reviewed, therefore, will be whether engaging in aggression is cathartic. To do this, the theoretical views of aggression will be examined because these theories determine whether to expect catharsis to be achieved or not. Experimental evidence is then cited which is related to the question--both supporting and refuting catharsis theory. Finally, a tentative evaluation of the studies is attempted to show some possible interpretations that a therapist might use to aid in his practice.

Back to contents

It has been widely assumed on the basis of psychodynamic theories and energy models of personality that vicarious participation in, or the direct expression of aggression discharges the aggressive drive energy and reduces the incidence of aggressive behavior. Guided by this type of theory, many parents, educators, rehabilitation workers and child psychotherapists encourage aggressive children to express aggression in one form or another (Bandura, 1969, p, 159).

There are many points of similarity between physiological drives, such as hunger and sex, and the instigation to aggression. This similarity is important when aggression is considered to be a drive that is capable of being reduced by a goal response. When an organism makes an appropriate goal response such as eating when it is hungry, the strength of the relevant physiological drive is reduced. If aggression is a drive similar to physiological drives, and causing injury to another is considered to be the appropriate goal response, the occurrence of an act of aggression should reduce the instigation to aggression. (Berkowitz, 1958.) When aggression is considered to be a drive that can be reduced by some type of aggressive responses, those responses that reduce the drive level in this manner are considered to be cathartic.

One way of examining theories of aggression is to evaluate them from the viewpoint of whether the theory predicts that it is cathartic to engage in aggressive behavior or not. There are three basic theories of aggression that predict different outcomes as far as catharsis is concerned. Both biological-instinctual and frustration-aggression theories consider aggressive behavior to be caused by a drive condition. The primary difference between these theories is in the origin of the drive. A third theory posits no particular drive and considers aggressive behavior as being the result of social conditions in an individual's past and present situation. This theory is called the social-learning theory of aggression. Both the biological-instinctual and frustration-aggression theories would expect a catharsis effect, but the social-learning viewpoint would expect a complete lack of catharsis from aggressive behavior, and in some cases would expect an increase in subsequent aggressive behavior as a result of learning aggression while making the aggressive responses.

Before evaluating the results of research to determine whether catharsis theory is tenable or not, it will be helpful to describe the three basic theories: biological-instinctual, frustration-aggression, and social-learning.

Typical of the psychoanalytic viewpoint is a description by Gilula and Daniels (1969) of what they call the biological-instinctual theory. This theory holds that aggressive behavior, including violence, is an innate component of humans that has resulted from the process of natural selection. According to this theory man is naturally aggressive. This theory holds that aggression includes a wide variety of behaviors, many of which are constructive and essential to an active existence.

According to Gilula and Daniels, one source of evidence supporting this theory is derived from psychiatric and psychoanalytic ease studies. Psychiatrists with this viewpoint describe many disordered behaviors as stemming from ramifications and distortion of the aggressive drive.

Examples range from individuals with destructive antisocial behavior who express violent aggression directly and often impulsively, to cases of depression and suicide in which violent aggression is turned against the self. Exceptions that show a low level or complete lack of aggressive behavior are explained by describing those individuals as persons who are seriously inhibited and for whom the expression of aggression, even in the form of assertion, is blocked almost entirely.

This biological-instinctual theory suggests that since aggression is inevitable, effective controls upon its expression are necessary, and reduction of violence depends upon providing constructive channels for expressing aggression.

Sigmund Freud was probably the originator of the idea that aggressive impulses are innate or instinctual. Freud described two basic instincts--sexual and aggressive. "Our hypothesis is that there are two essentially different classes of instincts: the sexual instincts, understood in the widest sense--Eros, if you prefer that name--and the aggressive instincts, whose aim is destruction." (Freud, 1966, p. 567.)

Freud was led to his theory of a special aggressive and destructive instinct in men on the basis of his experiences with the phenomena of sadism and masochism. Both phenomena presented a problem to his concept of libido theory and this problem was what prompted him to revise his theory to include the aggressive instincts. (Freud, 1966, p. 568.)

Freud considered masochism to be a more basic instinct than sadism. He believed that the basic aggressive instinct was against the self and was a form of a death wish. Sadism, therefore, was self-aggression turned outwards. This aggressive instinct can be perceived under only two conditions: One, if it is combined with erotic instincts into masochism, and the other, if it is directed against the external world as aggressiveness. If aggressiveness is not able to find satisfaction in the external world because it comes up against real obstacles, it will retreat and increase the amount of self-destructiveness holding sway in the interior. Impeded aggressiveness, therefore, seems to involve a grave injury. "It really seems as though it is necessary for us to destroy some other thing or person in order not to destroy ourselves, in order to guard against the impulsion to self-destruction." (Freud, 1966, p. 569.)

In all cathartic theory the instigation to aggression is seen as an ever-active energy source constantly impelling aggressive responses. Whether these responses appear in overt behavior depends on the presence of inhibitions or other avenues of expression. Once his aggressive drive has been aroused, whether by appropriate biological processes or some frustration, a person presumably is inclined to attack anyone unless he can displace his urge onto some inanimate object, or divert the drive into some socially acceptable channel (Berkowitz, 1964).

Hartmann, Kris, and Loewenstein (1949) illustrate this type of thinking by positing an innate biological drive based on unspecified internal sources. This drive is considered to be a force comparable to that of the libido. Like libido, this aggressive energy supposedly must find some outlet. The urge can be displaced, sublimated, or neutralized, but after expression it builds up again, and must seek release. Berkowitz (1964) states that writers influenced by the orthodox Freudian conception of aggression have envisioned a free-floating aggressive energy that may be channeled into many different activities or against a wide variety of objects. Carrying out any of the substitute actions drains the reservoir of aggressive energy.

The theory that release of aggressive impulses provides catharsis is loosely derived from Aristotle's view of the functions of tragedy. This theory contends that violence which is indulged in vicariously drains a reservoir of accumulated hostility and releases tensions that might otherwise explode into actual violent behavior. Konrad Lorenz (an ethologist) believes that behavior results from the spontaneous accumulation of some excitation or substance in neural centers. Lorenz believes that present-day civilized man suffers from insufficient discharge of his aggressive drive and he recommends that society provide people with safe ways of venting their aggressive urge. (Berkowitz, 1968.)

Gilula and Daniels (1969) describe the frustration theory of the origin of an aggressive drive state. According to this theory, aggression is a drive condition that comes from interference with ongoing purposeful activity. A person feels frustrated when a violation of his hopes or expectations occurs, and then he tries to solve the problem by behaving aggressively. Frustrations can come in various forms such as threats to life, thwarting of basic needs, and personal insults. Major factors influencing aggressive responses to frustration are in the nature of the frustration, previous experience, available alternatives for reaction, the person's maturity, and the preceding events or feelings. The response to frustration is viewed as learned behavior used to reduce the aggressive drive and to remove the obstacle that causes frustration. This theory suggests that control or reduction of violence requires reducing existing frustrations as well as encouraging constructive redirection of aggressive responses

According to Berkowitz, Green, and Macaulay (1962) aggression is an innate reaction to frustration that has one important property in common with other biological drives. Generally, when an organism makes the goal response appropriate to a biological drive the drive strength is reduced. For example, eating reduces hunger drive strength. Causing injury to another is the aggressive drive's goal response, and the occurrence of an act of aggression supposedly reduces the aggressive drive.

According to this argument, the expression of hostility should decrease the likelihood of any further aggression if there is no further frustration. This view is consistent with the catharsis theory of aggression. Many people believe that aggressive acting-out behavior reduces aggression and hostility, and most theories of play therapy for children are still based on this notion. With this type of therapy, the frustrated, angry, hostile child behaves aggressively, and supposedly this aggressive behavior reduces his level of hostility and aggression (Mallick & McCandless, 1966).

The social-learning theory of aggression described by Gilula and Daniels (1969) implies that not only can aggressive behavior be noncathartic, but that aggressive behavior will tend to increase the probability of later aggression. This theory is based on the assumption that aggressive behavior results from child-rearing practices and other forms of socialization and not from some inner drives whether instinctual or frustration-produced. Aggressive behavior can be acquired merely by watching and learning, often by imitation, and does not require frustration.

Aggressive behaviors rewarded by a society usually reflect the basic values and adaptive behaviors of the group. The social-learning theory of aggression suggests that control and reduction of violence requires changes in cultural traditions, child-rearing practices, and parental examples. Aggression cannot be reduced by catharsis according to this theory. Aggression engaged in to let off steam would only be increased due to emotional activity and new learning experiences.

Bandura (1969, p. 382) explains how participating in aggressive behavior vicariously may increase the tendency to aggress rather than to decrease it as catharsis theory supposes. He believes that the persistence of emotional frustration effects may be interpreted in terms of a self-arousal mechanism rather than as a lingering aggressive drive requiring discharge through aggression. After a person has been insulted, unjustly treated, or otherwise frustrated, the resulting emotional arousal is repeatedly recalled through symbolic reinstatement of the anger-provoking incidents. The persistence of elevated emotional arousal is attributed to the self-generated symbolic stimulation rather than the existence of an undischarged reservoir of aggressive drive.

According to Skinner (1969, p. 56) aggression is not due to a fundamental drive in humans to hurt one another or to a death instinct. Rather, it is due to an environment in which humans are reinforced when they hurt one another. To say that there is something suicidal in man that makes him enjoy war is to reverse the causal order; man's capacity to enjoy war leads to a form of suicide. In a world where a child seldom if ever successfully attacks others, aggressive behavior is not strong. The real world is now quite different. Either through simple neglect or the belief that innate needs must be expressed, children are allowed and even encouraged to attack each other in various ways. Expanding the theoretical argument outside of psychology, an experimental biologist (Scott, 1958, p. 62) has stated there is no physiological evidence of any spontaneous stimulation for fighting, either aggressive or defensive, arising from within the body.

Back to contents

Experimental evidence is available that both supports and refutes the theory that engaging in aggressive behavior is cathartic and tends to reduce the drive or internal pressure that motivates a person to be aggressive. Studies that evaluate this concept use three basic types of aggressive activity as cathartic: (1) actual aggressive behavior, (2) vicarious aggressive behavior where the person merely observes aggression and assumedly vicariously participates, and (3) fantasy aggression where a person imagines aggression or talks about aggression. These three basic methods of engaging in aggression can be used both as independent variables and dependent variables, and can involve many forms of activity.

Depending on the theoretical bias of the experimenter, a typical study would have three basic components. First, an aggressive drive is required; next, some aggressive activity is engaged in by the subjects; and last, a dependent measure is taken of resulting aggressive drive level. The instinctual theorists presuppose an already existing drive due to the biological make-up of the individual. The frustration-aggression theorists produce an aggressive drive by frustrating or insulting the subjects. The social-learning theorists look for increases in aggressive behavior and are interested in initial aggressive activity merely as a baseline from which to measure the resulting change. These latter theorists assume that due to a person's history, he will have certain aggressive tendencies that are unique to that individual.

The following studies are grouped in relation to the theoretical bias of the experimenters, the type of aggressive activities involved, and whether catharsis theory is supported or refuted. First, studies that support catharsis theory are discussed. Most of these studies involve projective techniques or subjective ratings as dependent measures.

Stone (1950) obtained indications of inhibited hostility following a socially sanctioned athletic event. He administered the Thematic Apperception Test to football players both during and after the athletic season and also to a matched group of control subjects. He found no difference in fantasy aggression during the football season, but football players showed significantly less manifest aggression on the TAT following completion of the season. This study is an example of a posited instinctual drive being reduced by the opportunity to engage in actual aggressive behavior.

In a study along the same lines, Appel (1942) also obtained results consistent with the catharsis hypothesis. In his study, children who had been in a nursery school that permitted fighting showed less hostility later in kindergarten than did children coming from a nursery school that discouraged fighting. Teacher ratings of aggression were used as the dependent measure.

Hussman(1955) used projective tests as a dependent measure of aggressive drive and compared different groups of athletes and nonathletes. He found that collegiate boxers had significantly less aggression using the Thematic Apperception Test as a measuring instrument than did wrestlers, cross-country runners, or the nonathletes. According to catharsis theory this is what would be expected since boxing is considered a much more aggressive activity than the other sports, and boxers showed less aggression which may indicate that they have more opportunity to release their drive by striking another person. A possible alternative explanation was proposed by Hussman based on his additional finding that the boxers had a stronger superego based on their responses to the Rosensweig P-F Studyadministered shortly after a boxing match. The boxers presumably felt guilty about their aggressive activity even though it was socially sanctioned and this caused their aggressive responses to be reduced.

Johnson and Hutton (1955) administered the House-Tree-Person test to eight collegiate wrestlers before the wrestling season, four to five hours before the first wrestling match of the season, and the morning after the match. The authors reported that aggressive feelings increased before the wrestling match, and then afterwards decreased considerably in most cases even below the level obtained before the wrestling season. However, this experiment lacked proper controls and there was no statistical data given so the conclusions may have little significance even if the dependent measure was accurate and reliable. No assessment of reliability is possible since the authors gave no statistical data in their report, and since it was not stated whether the scorer of the House-Tree-Person test knew the conditions under which they were obtained.

The preceding studies used actual aggressive activity for the independent variable, and were based on the instinctual origin of aggressive drive. Feshbach (1961) studied the effects of participating in a vicarious aggressive activity on subsequent aggressive behavior, and found results generally supporting the catharsis hypothesis.

This study was based on the frustration-aggression hypothesis. He found that under certain conditions vicarious participation in aggression tends to reduce subsequent aggressive activity, but only when the person who is experiencing vicarious aggression has previously been made angry. However, when anger has not been previously been aroused, he found a tendency for subsequent aggressive behavior to increase. For his study he used college students as subjects and then angered one group by insulting them. Both groups then witnessed either an aggressive boxing film or a neutral film concerning the consequence of the spread of rumors in a factory. Half of each group saw the aggressive film, and half saw the neutral film. This provided four conditions: (1) insult-aggression, (2) insult-neutral, (3) noninsult-aggression, and (4) noninsult-neutral. The dependent measures consisted of word association tests where the number of aggressive associations were measured, and a questionnaire about the subjects' attitudes toward the experimenter and the experiment.

A significant interaction was obtained with the insult-aggression group responding with fewer aggressive word associations than the insult-neutral group, and the noninsult-aggression group responding with more aggressive associations than the noninsult-neutral group. Since the interaction was disordinal, the main effect was not interpreted.

In addition to the consideration that aggressive behavior may reduce subsequent aggression by a cathartic effect, it is also possible that fantasy behavior may also affect any aggressive drive or tendency. Feshbach (1955) tested the hypothesis that the expression of aggression in fantasy will serve to reduce aggressive drive. Working within the frustration-aggression hypothesis, he experimentally induced aggressive drive in his subjects by insulting them. He then interpolated a fantasy or nonfantasy activity and subsequently measured the strength of the aggressive drive by a questionnaire and a sentence completion test.

The results were consistent with the cathartic hypothesis. The insulted group that had an opportunity to express aggression in fantasy subsequently displayed significantly less aggression than did the control group which engaged in nonfantasy activities on both measures of aggression.

Additionally, he found a significant negative correlation (-.25, p <.01) between the amount of aggression expressed in fantasy and subsequent aggression for the insulted group. The insulted subjects expressed significantly more aggression in their fantasies than did the noninsulted subjects. In this study, aggressive drive was induced by the experimenter assuming an insulting attitude toward a class of college students. For example he made such comments as, "Now I realize that you ***** college students, or should I say ***** college grinds have few academic interests outside your concern for grades... If you will try to look beyond your limited horizons, your cooperation will be useful. In other words, I'd like you to act like adults rather than adolescents." The fantasy activities consisted of responding to TAT cards and the nonfantasy activities consisted of various objective psychological tests.

Evidence that overt aggression has physiological tension reducing qualities has been provided by Hokanson and Shetler (1961). In their study, 56 college students were exposed to the following conditions in a three-way factorial experiment: high or low frustration by a high- or low-status experimenter with a subsequent opportunity or no opportunity to aggress physically (by electric shocks) towards the frustrator. The dependent measure obtained was systolic blood pressure taken before and after the frustration manipulation and the expression of aggression.

The main findings of the experiment were: (1) Frustration led to significantly greater systolic increases than the no-frustration condition. (2) When subjects were frustrated by the low-status experimenter, a significantly higher systolic pressure was maintained at the conclusion of the experiment by those subjects who were not allowed to aggress against the experimenter than the subjects who were allowed to aggress against him. For these latter subjects, the systolic pressure returned to prefrustration levels after aggression occurred. (3) Subjects frustrated by a high-status experimenter showed a return of systolic pressure to prefrustration levels whether the subjects were allowed to aggress against the experimenter or not.

The results of this experiment indicate that under some conditions, overt aggression has some physiological tension reducing qualities but that other variables may also reduce tension if they are appropriate to the situation.

The foregoing studies all had one thing in common--they obtained evidence that supported the catharsis hypothesis. A possible criticism of these studies is whether the dependent measures were actually measuring aggressive drive. Most of the dependent measures involved interpretation of projective tests rather than actual measurement of aggressive behavior. Since aggressive drive is a hypothetical construct, this may be the only way to measure it. However, validity of the measures used is open to question. The following studies either fail to support the catharsis hypothesis or show opposite results where interpolated aggressive activity increases subsequent aggression. These studies include some subjective measurements, but most use actual aggressive behavior as a dependent measure.

Several studies have attempted to assess the strength of the instigation to aggression in children following a series of experiences in which aggressive behavior is permitted or encouraged. The following study is directly related to the use of play therapy as a therapeutic technique. In a test of instinctual theory, Kenny (1953) provided an experimental group of 25 first-grade children with two catharsis situations using a play-therapy technique. A matched group of 15 children in the control condition spent an equal time playing on the swings or working on a jigsaw puzzle. Both groups were administered the first five episodes of the Korner Incomplete Story Test for a measure of initial aggressive drive, and the last five episodes after the experimental procedures as a measure of final aggressive drive. Although both groups showed a decrease in final aggressive drive, the control group showed a significantly greater decrease than did the experimental group. (Berkowitz, 1958.)

Very little controlled research has been done on the problem of aggression reduction by catharsis using autonomic measures as dependent variables rather than overt behavior. It would seem to be a logical step to measuring the construct of aggressive drive rather than having to make inferences about drive from overt behavior patterns.

Kahn (1966) has done such a study based on the frustration-aggression theory of aggression. In his study he angered 36 male college students and then randomly assigned each to either a catharsis or non-catharsis condition. Physiological measures were taken on heart rates, psychogalvanic response, skin conductance, finger temperature, muscle tension, and systolic and diastolic blood pressure. A baseline of recovery patterns was obtained for each subject by placing his foot in 4 degree centigrade water for 80 seconds and then obtaining his particular pattern of recovery from stress. After recovery from this procedure each subject was angered by an ill-mannered technician who acted very crude and vulgar but never hostile or threatening.

After a 5-minute period of anger arousal, the experimenter replaced the technician and in the control condition merely asked the subject to sit quietly for 20 minutes. In the experimental condition (catharsis), the experimenter expressed surprise that the physiological measures were so high and questioned the subject about what had taken place. He kept probing until the subject had revealed a substantial part of the arousal procedure and had expressed his feelings of anger or annoyance. He offered the subject vicarious counter-aggression by telling him he would have the technician's supervisor reprimand him. This procedure took about 5 minutes and the subject was then asked to wait quietly for 20 minutes.

Physiological measures were taken continuously for all subjects and all subjects were given a questionnaire at the end of the experiment to determine how they then felt about the technician. Results showed that the catharsis subjects disliked their annoyer significantly more than control subjects; the catharsis subjects showed a significantly slower rate of physiological recovery in all measures except blood pressure which showed a faster rate of recovery. In general, these findings do not support the catharsis hypothesis.

Also working with the frustration-aggression theory of aggression, Mallick and McCandless (1966) hypothesized that: (l) Angry aggression directed toward an inanimate object is not cathartic. (2) Aggression, unmotivated by anger or hostility, has no cathartic effect, but may, instead, lead to an increase in aggressive responses, particularly in a socially permissive atmosphere. (3) Positive and reasonable verbal interpretation of a frustrating situation to the subject who has been frustrated has cathartic value in that it reduces hostility toward the frustrator. (4) Verbal aggression against a frustrator of the same sex does not reduce the hostility toward him. (5) United States girls, presumably because of cultural forces, will show less open aggression than boys; however, in a permissive situation where privacy is assured, sex differences in open expression of aggression will be reduced.

These authors conducted three studies to test these hypotheses. In all studies third-grade children of both sexes were used as subjects. The experimental designs were 2 x 3 x 2 factorial designs with two treatments (frustration and nonfrustration), three types of interpolated activities (shooting a play gun at different targets on which were drawn figures either of a boy, girl, man, woman, cat, or dog; shooting at targets blank except for a bull's-eye; and solving simple arithmetic problems), and sex forming the third dimension of the design. The subjects were either frustrated by another child or not frustrated; then they were provided the opportunity to participate in one of the three interpolated activities; then they were shown the child who frustrated them (or the child who helped them in the nonfrustration condition) and allowed to shock him. All subjects were assured that no one would know who was shocking the child. The dependent measure was the number of shocks each subject gave the child. The results of this study were significant only for the frustration dimension (p<.01).

A second study by these authors used the same basic situation except that the interpolated activities were changed. The three activities consisted of: (1) Shooting guns at a target on which was placed a picture of an 11-year-old child of the same sex. (2) Engaging in social talk (moderately standard for all subjects) with the experimenter for 8 minutes. (3) Engaging in social talk as in condition 2, but including interpretation to the effect that the frustrator was sleepy, upset, and would probably have been more cooperative if the subject had offered him two of five nickels that were given to the subject.

The results of this study were significant only for the dimension involving the three activities. Multiple comparisons showed significance at the .005 level for differences between frustration and nonfrustration followed by social talk and nonfrustration followed by aggressive play. Significant differences at the .001 level were found between frustration followed by aggressive play and nonfrustration followed by social talk; between frustration followed by aggressive play and frustration followed by interpretation; between frustration followed by social talk and nonfrustration followed by social talk; and between frustration followed by social talk and frustration followed by reinterpretation.

Based on these results the authors concluded that expression of aggression as a catharsis serving to reduce aggression or hostility is not effective, at least when expression of aggression is toward inanimate objects. They also concluded that aggressive play in the presence of a permissive adult may lead to increased aggression. The authors propose that reasonable, positive interpretation of the frustrating situation has a cathartic effect because when the experimenter interpreted the reason for the frustration to the subjects of this study, the subjects showed less aggression than subjects who did not receive the interpretation.

The authors also found that verbal expression of aggression in the form of rating scales made by the subjects about how much they liked or disliked the frustrating child tended to increase the aggression of the subjects. The condition of being asked to rate their feelings toward the frustrator caused a significant difference in aggression scores with more aggression being exhibited by those who were asked to rate than those who were not asked to rate.

The authors concluded that verbal expressions of hostility may lead to an increase of aggressive behavior toward the subject of the hostile expression. They posed the question: Does expression of hostile feelings in therapy lead to aggressive behavior in real life?

Zillman, Katcher, and Milavsky (1972) reasoned that according to catharsis theory, strenuous exercise may be expected to result in a catharsis of aggressive inclinations. These authors note that Freud and other psychoanalysts have explicitly implicated motor activities of the skeletal musculature in discharge of aggressive energy. According to this reasoning it would be expected that aggressiveness will be low following strenuous exercise and relatively more reduced under severe instigation than under negligible or mild provocation.

These authors undertook a study based on the frustration-aggression theory to test this hypothesis. Their subjects were 38 male undergraduates who were recruited by a written announcement and who were paid for participation. A 2 x 2 factorial design was used with low versus high initial aggressive instigation as one independent variable and with low versus high degrees of physiological excitation deriving from physical activities as the other.

The low versus high degree of initial aggressive instigation was provided by ostensibly having a second subject administer either three shocks (low instigation) or nine shocks (high instigation) to the subject supposedly in response to the subject's opinions about 12 controversial issues. The second subject was later the recipient of shocks as a dependent measure of the first subject's aggression. The second subject was a confederate of the experimenter. The low versus high degree of physiological excitation was achieved by having the subject either place discs on a string (low excitation) or by having him exercise on a stationary bicycle (high excitation).

Subsequent to high or low instigation, and then either high or low excitation, the subject was instructed to deliver shocks to another subject (confederate) who was acting as a learner in a teaching situation whenever the learner made an error. As programmed, the learner made errors on 12 out of 20 items and the subjects acting as teachers delivered shocks on each of these trials. The shock intensity was determined by the subject and was the dependent measure of aggression: the higher the shock intensity, the higher the aggression.

The findings of this experiment are counter to the expectation that strenuous physical exercise can serve to drain aggressive tensions and thus induce catharsis. Subjects who were given the opportunity to drain off aggressive tensions caused by receiving shocks from another person by engaging in strenuous exercise showed greater aggressive responses than subjects who did not exercise. In all cases, the increased aggression by the subjects was associated with high instigation or provocation and there was no indication that aggressive responses were reduced by exercise being interpolated between instigation and opportunity to aggress.

Consistent with the argument against cathartic aggression is the relationship between viewing violence on television and aggressive behavior found by Eron, Huesmann, Lefkowitz, and Walder (1972) in a longitudinal study of children that lasted from the third grade until the thirteenth grade.

Working with the social-learning theory of aggression, these authors collected data on 427 teenagers of an original group of 875 children who had participated in a study of third-grade children in 1960. The original 875 constituted the entire third-grade population of a semi-rural county in the Hudson River Valley of New York. The 427 teenagers were the only ones of the original group who could be located and interviewed 10 years after the original study.

The authors took peer ratings of aggressive behavior as the third-grade measure. For measures of aggression in the thirteenth grade they used peer ratings, the personal report of the subject, and the sum of scales 4 and 9 on the MMPI. The authors compared these measures of aggression with 17 variables for the third-grade children and with four variables when the same children were in the thirteenth grade. Three variables were common for both levels: (1) Father's occupational status, (2) subject's hours of watching television, and (3) subject's preference for watching violent television programs.

The data were analyzed separately for males and females, and the authors found a highly significant relationship between a preference for violent television in the third grade and aggressive behavior in the thirteenth grade for boys but not for girls. For both, the relationship between preference for violent programs in the third and thirteenth grades was nonsignificant. The authors found that boys at the age of third-graders who preferred violent television programs were rated as more aggressive by their peers than boys who preferred less violent programs. In the 10-year follow-up study, they found that the violence of programs preferred by the male subjects in the third grade was even more strongly related to aggression 10 years later.

By the use of cross-lagged correlations, partial correlations, and multiple regression, they concluded that with boys there is a probably causative influence of watching violent television programs in early formative years on later aggression. The authors do not claim that television violence is the only causative factor on later aggression, but a preference for violence on television was relatively independent of the other factors studied and explains a larger portion of the variance than does any other single factor studied, including IQ, social status, mobility aspirations, religious practice ethnicity, and parental disharmony. The authors concluded that there may be a critical period in a boy's development, when regular viewing and liking of violent television will lead to the formation of a more aggressive life style. It should be noted that these results were obtained only for boys in the study; therefore, girls were not included in the final analysis.

Also using the social-learning theory, Feshbach (1956) found no evidence of a cathartic reduction of hostility as a function of play activity. He found that boys initially low in aggressive behavior increased significantly in overt aggression after a series of permissive free play experiences, but there was not a significant effect for girls. He also found that play with aggressive toys seemed to promote more inappropriate later aggression than did play with neutral toys.

In his study, Feshbach hypothesized that if catharsis theory is correct, one would expect to find less inappropriate aggression among children playing with aggressive toys; however, if the stimulating properties of aggressive toys are productive of more aggression, one would expect to find more inappropriate aggression among those same children. Inappropriate aggression refers to aggressive behavior not directly determined by the play object such as quarreling, fighting, and destruction which does not arise in the context of object-oriented play.

For the study, children varying in age from five to eight years were designated as high aggressive and low aggressive, depending on teachers' ratings. The children were then randomly assigned to one of three groups: an aggressive-toy group, a neutral-toy group, and a control group. Each of the two experimental groups was subdivided into four play groups of five children that met for one 50-minute play session each week for four weeks. The aggressive-toy group was provided with toys that lent themselves to thematic aggressive play, while the neutral-toy group was provided with nonaggressive toys. The results of the experiment showed that the aggressive toys elicited significantly more inappropriate aggression than did the neutral toys, with the effect more pronounced for the upper-grade groups.

Back to contents

Although there is some evidence that aggressive behavior can be cathartic, much of this evidence involves fantasied aggression rather than overt aggression. When a physiological measure is used as an indication of aggressive drive, fairly consistent results obtain across studies. Blood pressure appears to reflect a catharsis effect whereas other physiological measures do not. Whether physiological arousal indicates aggressive drive is a problem open to argument.

Most studies supporting a catharsis effect have used dependent measures rather loosely. Measures of aggressive drive obtained from projective tests are open to considerable doubt, and this is the primary type of dependent measure used to measure an instinctual drive. Based on these kinds of measures, fantasied aggression does appear to decrease after some aggressive activity, particularly when that activity is socially sanctioned as in a sporting event.

Vicarious aggression such as can be achieved by viewing violence on television or movies tends to have several effects, but generally the tendency is for this type of aggressive activity to cause an increase in overt aggression rather than a decrease. If a person is angered and then allowed to experience vicarious aggression, however, he may tend to achieve some catharsis which will reduce the tendency toward later overt aggression.

Young children who view violence on television probably are learning aggression rather than achieving catharsis, and children who are exposed to violence while they are young will probably grow into more violent adults than children who are protected from violence in their early years. This conclusion should be viewed cautiously, however, because the studies cited found these types of results with boys and not with girls.

Physical activity with or without fantasied aggression tends to increase aggressive behavior. It may not be practical to tell someone to go for a bicycle ride or some other innocuous activity to reduce physiological tension due to aggressive tendencies.

Play therapy with children, particularly when the therapist has a highly permissive attitude toward aggressive behavior, will probably serve to increase aggressive tendencies in the child. If the goal of therapy is to make a child more aggressive, then permissive, aggressive play activity with aggressive toys will probably help to achieve that goal. If the goal is catharsis which is intended to reduce aggressive activity in the child, aggression in the play room may make matters worse. It would be better to encourage nonaggressive play and also to use explanations and interpretations of why frustrations occur to reduce overt aggression in the child.

In general therapy situations, talking about hostile feelings may make those feelings more intense rather than reducing them by catharsis. This may also lead to more aggressive behavior in real life. This implies that the goal of therapy should be considered. If the goal is to make a person more aggressive, then it may be best to encourage talk of hostile feelings, but if the goal is to reduce hostility, then talk of such feelings should be discouraged, and reasons for frustrations discussed instead.

Back to contents

Back to contents